key: cord-0005969-ak5pq312 authors: nan title: 8th European Congress of Intensive Care Medicine Athens - Greece, October 18–22, 1995 Abstracts date: 1995 journal: Intensive Care Med DOI: 10.1007/bf02426401 sha: 7ee7984e047e41456a2fc4fedba58b3524b02878 doc_id: 5969 cord_uid: ak5pq312 nan Objectives: Evaluate the levels of TNF, IL-6 and PAI-I in different moments of the ARDS and the possible relationships among them. Methods: 23 septic patients with ARDS were studied. Also Significant differences for: TNF, PAI-I and IL-6 in septic patients and both evaluations of ARDS with control gropup; PAI-1 between septics and 2nd evaluation in ARDS, and between the ist and 2nd evaluation in ARDS; IL-6 between septics and both evaluations in ARDS; and IL-~ in both evaluations in ARDS patients in relation to mortality. Conclusions: i) Elevations of TNF, PAI-I and IL-6, with clinical signs, are suggestive of infection; 2) The persistent and progressive elevation of PAI-I with any clinical criteria may suggest evolution to ARDS; 3) Due to its own kynetics, IL-6 takes part later in the acute phase, its levels being related to the magnitude of the injury in the tissues. Objectives: The influence of long-term volume therapy with different solutions on plasma levels of circulating adhesion molecules was studied. Methods: According to a randomized sequence, 30 patients with sepsis secondary to major surgery exclusively received either hydroxyethylstarch solution (10% HES, mean molecular weight (Mw) 200,000 daltons, degree of substitution (DS) 0.5) or human albumin 20% (HA) for volume therapy for 5 days. Plasma levels of circulating (soluble) adhesion molecules (endothelial leukocyte adhesion melecule-1 [sELAM -I] , intercellular adhesion molecule-1 [sICAM -I] , vascular cell adhesion molecule-1 [sVCAM -I] , and P-selectin ) were serially measured on the day of admission to the intensive care unit (='baseline ' value) and during the next 5 days. Results: sELAM-I, sICAM-I, and sVCAM-I plasma levels were markedly higher than normal at baseline in both groups. In the HES-patients, sELAM-J decreased to normal range, whereas it further increased in the HA-group (from 89• to 106• During the study period, sICAM-I and sVCAM-I plasma levels remained unchanged in the HES-patients, but further increased in the HA-group (from 626• to 1,329• sGMP-140 increased significatly only in the HA-group (483• to 683• Only PaO2/FIO 2 was significantly correlated to plasma levels of adhesion molecules. Conclusions: Sepsis is associated with markedly elevated plasma levels of adhesion molecules indicating endothelial activation or damage. By long-term volume therapy with HES, these levels remained unchanged or even decreased, whereas volume therapy with human albumin did not have any beneficial effects on soluble adhesion. Central venous catheters are frequently used in the care of the critically ill patient. The incidence of catheter related sepsis varies in the literature. We investigated the occurrence of contamination and sepsis compared to results of the EPIC study as part of quality assesment in our intensive care unit. From january until august 1994 all removed central venous catheters were examined for microbiological culture. The patients who showed signs of sepsis were also registered. The results of the contaminated catheters and septic patients were compared with results from the EPIC study. During the 8 month period ,2059 patients were hospitalized on our intensive care unit. 230 Central venous catheters were examined for microbiological culture. 118 specimens appeared to be possitive (51%). 13 patients showed clinical signs of sepsis. The incidence of sepsis due to contaminated central venous catheters was 13/118 (11%). The incidence of sepsis due to the presence of all central venous lines was 13/230 (6%). The microorganisms responsible for the sepsis syndrom were : Stapylococcus aureus (n=5), Escherichia colt (n=7), others (n=6). In the EPIC study the percentage for sepsis on the ICU was 17.6% for the Netherlands and 17.8% for Europe. Despite a high number of positive culture from removed intravascular lines, a low percentage of sepsis was seen compared to results of the EPIC study. We recommend routine bacteriological culture of all removed central venous lines and recommend to look at colonization and sepsis due to intravascular lines as a measure of quality control in the intensive care unit. Objectives: Prognostic assessment of Simplified ACute Physiology Score (SAPS) in granulocytopenie patients with septic shock (SS). Methods: The medical records of 59 admissions to an intensive care unit (ICU) of granuloeytopenic patients with SS are reviewed. Fiftytwo patients had haematological malignancies. Seven patients had aplastie anaemia. Patients were categorised as survivors (discharged from ICL 0 and non-survivors (died in the ICU). SAPS index was calculated for patients daily during their stay in ICU. All patients were severe granulocytopenic (total white cell count less than 0,5 ]09]1). Results: Five patients (8,5%) were discharged from ICU. Fifty-four patients died in ICU. Non-survivors had SAPS On admission higher than survivors ( 20.9+4.6 and 16.5+3.0, respectively, p<0,01, Mann-Whitney U test). No patient with a SAPS greater than 20 survived. Mortality among the 27 patients with SAPS from 9 to 20 was 81,5%o. The evolution of SS was rapid. The mean stay in ICU among non-survivors was only 56 hours. An analysis of the SAPS index on admission of non-survivors showed an inverse correlation with the duration of their stay in ICU (r=-0,52, P=0.001). All survivors recovered from granulocytopenia. They had normal white cell counts at the time of discharge from ICU. There was inverse correlation in survivors between SAPS and white cell counts, when these parameters were evaluated daily. However, the SAPS index alone cannot be considered to be on individual predictor factor of mortality. Patients who had failure of the malignancy to respond to chemotherapy and who had persistent granuloeytopenia died in ICU despite SAPS index on admission and recovery from SS. Conclusion: SAPS index greater than 20, failure of the malignancy to respond to chemotherapy and persistent leueopenia all point to a poor outcome of granulocytopenie patients with SS. Introduction: Antipyretics sometimes are used for fever control in febrile neutropenic patients with hematological malignancies(HM). We observed a dramatic fall of blood pressure(BP) and development of septic shock(SS) in some of the patients who received antipyretics. Aim: To clarify can antipyretics provoke SS in neutropenic patients with infection. Methods: Retrospective review of medicat records of 52 neutropenic(WBC <0,5 109/1)patients with HM, admitted to the intensive care unit for SS, was performed. There was selected group of 8 patients receiving antipyretics shortly before a fall of BP. Results: There was a definite causal relationship between receiving antipyretics and fall of BP in 4 from 8 patients. All patients had fever due to infection and had normal level of BP before receiving antipyretics. Hypotension developed within 40 minutes up to 1,5 hours after administration of antipyretics. Three patients received 0,5 g of metamisol and one 0,5 g ofparacetamol per os. In all cases we observed dramatic diaphoresis and the temperature fall to subnormal level (35.4+0.4~ accompanied'by hypotension. But in 8-12 hours the fever was coming back without blood pressure elevation. The fluid replacement was controlled by central venous or wedge pressures. There were required 1200+350 ml colloid and cristalloid solutions for volume loading. In spite of fluid administration the hypotension persisted and all patients required inotropic therapy. Only one patient survived and is alive now. Conclusion: It seems to us that our data offer to state that antipyretics administration can initiate SS in febrile neutropeuic patients with infection. Objectives: To assess the agreement between cardiac output (CO) measured by ODM t and by 3 other methods used in ICU patients. Methods: We prospectively studied 12 adu t patients requiring hemodynamic monitoring with a pulmonary artery catheter. An esophageal doppler monitor provided measurements of CO (ODM), stroke volume and flow time (FT) used as an indirect evaluation of patient's volume status. Patient hemodynamic status was evaluated by a modified fast response pulmonary artery catheter (Baxter Health Care Corporation, Santa Ana, CA), allowing CO measurements by thermodilution 0"D) and an evaluation of right ventricular ejection fraction and end diastolic volume (RVEF and RV-EDV). In the last six patients CO was measured by transthoracic echocardiography (ECHO) and oxygen consumption was measured by a DELTATRACK II metabolic monitor (Datex) allowing CO calculation according to the Fick formula (FICK). The agreement between methods measuring CO and their reproducibility, were evaluated by Bland and Altman analysis. Results: Agreement between CO measurements is expressed as bias (d) and 95% limits of agreement (L of A = d_+2SD .15 TD-FICK -2.36 -8.06 to 3.34 FICK-ECHO 0.60 -5.92 to 7.12 There was no correlation between FT and RV-EDV. Conclusions: Although CO measurements by ODMil had the best reproducibility, the limits of agreement between the four methods tested were unacceptable for clinical purposes. Further investigation is required in order to improve the accuracy of CO measurement in the ICU. PhD, A. Paltzev, V.Bajbikov, B.Dobryakov D.Sc., A.Ostanin PhD, O.Leplifia PhD, H.Chernykh PhD Munieip. Hosp. N l, N 12; Inst. of Clin. Immunol., Novosibirsk, Russia ObJectivies: Efficiency of native cytokines used in the treatment of patients with severe surgical infections has been studied. Methods: For two years 120 patients were treated with cytokine mixture (SSP) obtained by arterio-venous perfusion of swine spleen and contained the following cytokines: IL-1, IL-2, IL-3, TNFa, IFNy, GM-CSF. Results: SSP intravenous infusions were shown to accompany with mortality decrease from 23.4% to 12.5% in patients with abscessed pneumonia and lung abscesses and from 50% to 13% if disease course was complicated with sepsis. In patients with purulent peritonitis and sepsis efficiency of SSP was decreased due to endotoxieosis. Thus, we used adoptive immunotherapy with MNC activated in vitro with SSP or recombinant IL-2. Intravenous infusions of such cells resulted in transformation of a pathologic process from destructive into productive one. Moreover, clinical manifestations of sepsis were controlled in 81% and mortality was decreased from 46% to 19%. Conclusions: The use of eytokines themselves as well as cytokine-treated lymphoeytes permits to control the disease and leads to the mortnlity decrease owing to stimulation of host defence mechanisms. Background: Although red blood cell transfusions (RBCt) are used to increase oxygen availability in septic patients, several lines of evidence suggest that RBCt may actually worsen tissue hypoxia. Thus, RBCt may negatively influence outcome of septic patients. Objectives: To determine the association of 1) RBCt ; 2) number of units transfused; and 3) mean age of the units transfused on the first day of transfusion with mortality of critically ill septic patients. Methods: We prospectively identified patients who met strict criteria for sepsis syndrome (SS) seen in the ICU of St. Paul's Hospital from 1992 to 1994 and excluded patients who died in the first 5 days after the onset of sepsis. We recorded clinical characteristics, multiple system organ failure score, and APACHE II at onset of sepsis. Then, we retrospectively recorded the total number and age of RBC units transfused during the first 5 days after onset of sepsis. Overall 30-day mortality was 22%. Results: The main results are shown in the table. The mortality of patients who received RBCt was nearly double the mortality of those who did not receive RBCt even after adjusting for severity of illness using APACHE II. Objectives: Gastric mucosal acidosis is frequently observed in patients with sepsis. The aim of this study was to determine whether volume infusion using Pentaspan| decreases abnormal gastric mucosal PCO2 (PiCO2) in patients who have sepsis syndrome (SS) who have already been resuscitated using clinical endpoints. Methods: We prospectively identified 5 patients who met strict criteria for SS, had a pulmonary artery catheter and a gastric tonometer in place, and PiCO2 > 50 mmHg. Pentaspan| (500 mL) was infused in 30 rain. Measurements of hemodynamics, hemoglobin, arterial lactate, blood gas analysis, and PiCO2 were performed before and repeated 30 miff and 2 hr after Pentaspun| infusion. We calculated the PiCO2 -arterial PCO'2 difference (PiCO2-PaCO2) and pHi (using Henderson-Hasselbach equation). ANOVA was used to assess statistical significance. Results: All patients werereceiving adrenergie drugs. MAP was 73 :1:13 mmHg and lactate 1.2:1:0.6 mmol/L. Pentaspan| increased CI by 22% (p<0.05) but did not change PiCO2 ( and increase m oxygen o* Wery were simimny achieved in both groups. Nevertheless, epinephrine was associated with a lactic acidosis and increased laetate/pyruvatemia ratio (L/P) that evoke a dysoxia rather than a metabolic effect. An higher gastric mucosal PCO2 in the ep group compared to nor-rob suggests the hypothesis of an anaerobic production of CO2 in favor of a splanchnic hypoxia. In both group, arterial ketone body ratio that reflects hepatic mitochondrial redox state, compared to a control group without shock was decreased but increased between 12 and 24 hours after restoration of arterial pressure. The association norepinephrine-dobutamine seems to be better for splanehnic circulation than epinephrine and should be used for dopamine resistant septic shock. Moreover, the increase in arterial pressure with nor-dob improved gastric mueosal pH and hepatic mitochondrial redox state and argue to reconsider arterial pressure as a significant goal for resuscitation in septic shock. Conclusion: Significantly higher malondialdehyde and ghitathione levels and glutathione-peroxidase activity in group NS at the end of ICU stay were related to mortality These findings indicate an increased generation of free oxygen radicals together with increased anfioxidant activity in this group and sapport the employment of antioxidant interventions in critically ill patients. Oblecfives: To determine the role of nitric oxide (NO) in the mechanism of septic shock induced by isolated limb perfuslen with recombinant TNFcr Methods: We have measured TNFr~ and metebo~ites of NO in 5 patients with signs ot septic shock following treatment with isolated limb perfusion for nonresectable soft tissue tumors and melanomas of a limb. Perfuslen was carried out with melphalan (Burroughs Wellcome) and recombinant TNFcr (Boehringer). TNFc~ was determined by specific radiometric assay (Medgenix Diagnostics), nitrate and nitrite were measured with a modification of the Guess reaction ~. Results: Results are shown in the Table. Conclusions: During isolated limb pedusion with recombinant TNF~ very high levels of TNFcr were measured in arterial blood in 5 patients. They all showed signs of severe sepsis syndrome with shock from vasodilafion, probably due to leak of recombinant TNFt~ from the peduslen circuit to the systemic circulation. TNFc~-induced vasodilation was not accompanied by a rise in serum NO-metsbolites. Our findings do not confirm the widely accepted theory, mainly based on animal experiments, that genera• of NO is the key pathogenefic mechanism in septic vasodilafion 2, nor that TNFrt invariably induces forreafion of NO. The precise mechanism of shock in these patients remains to be elucidated. References: 1. Moshage H, Kok B, Huizenga JR, Jansen PLM Nitrite and nitrate determinaiions in plasma: a critical evaluation. Clin Chem 1995: 41/6. 2. Moncada S, Higgs A. The L-argioine-nitrio oxide pathway. N Engl J Med 1993; 329: 2002 -2012 EC is a commonly used for prolonged, stable animal anesthesia. Noting that the hypotension after IV LPS was attenuated by EC, we hypothesized EC also protects against LPS toxicity. Sprague-Dawley rats received IP saline (S), thiobutabarbita180 mg/kg (TB), or varied doses of EC, followed 2 hours later by bolus 30 mg/kg IV LPS. 7-day survival is shown below: GROUP: S TB EC(0.1GMIKGI EC(0.SGM/KG) EC(I.2GM/KG) ALIVE (n) t 0 0 3 9 ~ TOTAL (n) 10 S S 7 10 "Signiflcant;y different from all other groups, p<0.0S 5/5 rats given LPS followed 2 hours later by EC (1.2 gm/kg) also died. Additional rats were treated with S (n=10) or 1 2 gm/kg EC (n=10) followed by 30 mg/kg LPS, then sacrificed at 4 hours. Blood glucose (BG, mg/dl),.hematocrit (HCT), leukocyte count (WSC/mm~ platelet count (PLTxl0~/mm3), bicarbonate (HCO, mg/dl), gross bowel hemorrhage (BH, 0-4 scale) and lung myeioperoxidase activity (MPO, ~VmirVgm wet lung) are shown below ( We conclude that EC reduces the lethality and multiple organ toxit;~ty of LPS. Its diverse effects suggest asite of activity upstream from the cytokine cascade. These results are important for studies of LPS which may use EC anesthesia and may have potential in the therapy of septic shock. [Zo = 0 Hz impedance (Z; {dyn.sec.cm "5 }); Zl = first harmonic Z; Zc = characteristic Z; Z1 ph. = t'trst harmonic phase angle {radians}; f, #, * at least p < 0.05 between FiO2 0.4 and 0.12, FiO2 0.4 and FiO2 0. 4&NO -0.46_+0.1 -0.26_+0.1 # -0.24+0.1 m -0.85+0.1 * -0.85+0.1 * -0.74+0.1 * -0.88_+0.1 * In hyperoxia, compared to dogs at the same Q, minipigs had a higher Ppa (26+1 rnmHg versus 16+1 mmHg; p < 0.01). Hypoxia increased (Ppa-Ppao) at all levels of Q by an average of 13 mmI-Ig in minipigs and 2 mmHg in dogs. Inhaled NO inhibited hypoxia-induced (Ppao-Ppa)/Q changes in both species. Conclusions: We conclude 1 ~ that the minipig is an animal model of elevated pulmonary vascular resistance and impedance, and 2 ~ that hypoxia-induced alterations in PVZ spectrum are due to changes of resistance in small arteries. Objectives: 1) To determine the toxicity of NG-monomethyI-Larginine (NMA) administered by intravenous bolus to patients with refractory septic shock. 2) To investigate the biologic activity of nitric oxide synthase inhibitors in septic shock. Methods: From August 1993 to January 1995, thirteen patients with vasopressor refractory septic shock received NMA intravenously in escalating doses from 1 to 40 mg/kg. Results: No hepatic, renal, gastrointestinal, or hematologic toxicity was observed at doses of NMA as high as 40 mg/kg. Significant biological activity was observed at all dose levels consisting of increased blood pressure (Systolic blood pressure from 70.9 mm Hg + 3.4 to 109.0 _+ 4.3 s.e.m., p=0.0004, systemic vascular resistance (430 + 57 to766 + 93 dyne.sec/ cm s, p=.002), and a decrease in vasopressor requirements. The magnitude and duration of these effect were dose dependent. Decreased cardiac output (8.1 _+ 0.8 to 7.0 _+ 0.9 I/min p=.003) and increased pulmonary artery pressure (35.6 _+ 2.1 to 43.5 _+ 2.0 mm Hg; p=.004) were also observed. No significant effects on heart rate, pulmonary capillary wedge pressure, or central venous pressure were observed. Four of 13 patients survived for more than 28 days, 4 patients died of cancer complications (all 5 patients had maintained blood pressure for 24 h on NMA) and 4 patients died of complication attributable to septic shock (MODS, ARDS, DIC, refractory hypotension), and 1 patient was unevaluable. Conclusions: No adverse clinical effects have been observed in patients receiving bolus doses of NMA as high as 40 mg/kg. The increased pulmonary artery pressures observed in septic shock patients is further augmented by NMA and may limit the dose which can be administered by intravenous bolus. Other schedules of drug dosing may attenuate this effect. Glucose-insulin-potassium (GIK) solutions have been shown to improve cardiac contractility and increase oxygen availability in experimental and clinical settings of septic shock. Several mechanisms have been proposed to explain these effects including a direct improvemeut of the energy balance by glucose, a direct influence of insulin on cardiac performance or an increase in intravascular volume due to the hyperosmolarity of the solution. To explore the role of hyperosmolapity, we compared the effects of GIK to those of a isoosmolar hypertonic saliue solutiou in endotoxin shock in dogs. Methods : The study included 18 mongrel dogs (25• pentobarbitalanesthetized aud mechanically ventilated with air. Thirty minutes after the intravenotls administration of 3 mg/kg of E. coli endotoxin, the dogs were randomized to receive a 2ml/kg infusion in 30 rain of a hypertonic (2895 mOsm]L) solution iucludiug either a mixture of glucose 50 % with 750 U insulin and 2000 mEq KCL/L (GlK-Group 1) or hydroxyethyl starch 7.5 % in NaCI 8.4 % (HES-Group 2). In each dog, a 0.9 % saline infi~sion was continued to maintain the puhnonary arlery occluded pressure at baseline level. Hemodynamic, blood gas aualysis and laboratory data were collecled at baseline and 30 miu, 60 rain, 120 rain, and 240 nunutes later.. Results : Eudotoxin administration was followed by a fall in mean arterial pressure (MAP) aud cardiac index (CI) and a rise in blood lactate levels. Resuscitation with either GIK or HES hypertoaic solutions resulted in similm increases in MAP, CI, oxygen delivery and left ventricular stroke index (Table 1) . We conclude that during resuscitation from endotoxic shock the use of GIK solutions is not superior to hypertouic HES solutions. The higher blood lactate levels observed in the dogs receiving GIK can be attributed to the glucose metabolism. 1, 75 for group 2, 105 for group 3) were drawn and immediately analysed at 37~ using the ABL500 Radiometer for PO2, PCO2 and pH, and the OSM3 Radiometer for HbO2%, HbCO% and MetHb%. Psost (i.e. the Ps0 at pH=7.40, PCO2=40 mmHg and temperature at 37 ~ C) was calculated automatically by the instruments on mixed venous blood, as was the Ps0"in vivo" (i.e. the Ps0 at the patient's value of pH, PCOz and temperature), using Siggaard-Andersen's algorithm. The data were compared by the one-way ANOVA test and by the t-test for paired and unpaired samples. Results: The mean resulting values (in mmHg) with the statistical differences are shown in table I. In addition, the time series analysis shows the mean Ps~st values as statistically below the PSin vivo" in the septic patients while the opposite is shown for the cardiac patients. No differences in the time analysis are demonstrated for the second group. A possible clinical significance may be drawn from these different behaviours. Objectives:Toxemia degree and humoral immunity condition have been studied in 36 patients aged from 19 to 72 with progressive course of sepsis and polyorganic insufficience. Methods: Such toxemia and humoral immunity findings as lencositlcindex of toxication (LII), level of oligopeptides of the middle molecular mass registered at the wave length of 25nM(MMI) & 280nM (MM2), distribution index (ID), immunoglobulins A,M,G, concentration of circulating immunocomplexes (CICI & CIC2) and also some clinical and biochemical findings on the 1,3,5 day after the operation serve as criteria for treatment effect. Results: It was founded that in intensive therapy and detoxication, level of LII is successively decreased from 12.6~1.4 to 2.6+.5 on the 5-th day after the operation. True decrease of the level MM2 from .704~.09 to .402+.08 un & optimal density and increase of distribution index from .96 to 1.09 are argued. Conclusions: In studlng the dynamics of the immunoglobulin's spectrum and the true increase of immunoglobulin G level from 8.4+.6g/i to i0.8+.7g/i on the 5-th day after the operation simultaneously with the decrease of CIC2 from 806.9~60 to 624.7~54.8(P .05) were founded. Some stages of the investigation true increase of lymphocytes from 9.0+2.6% to 15.0+1.8% was noted and it appeared to be a favourable prognosis finding for disease outcome. High correlation dependence between bacillus-and segmentonuclear neutrophils and immunoglobullns G & M (r=.5-.7 in P<.05) was discovered and it also showed positive dynamics of the course of the disease. A 34 year old male patient was admitted to the ICU with severe paraquat poisoning. Treatment consisted of gastic lavage and oral administration of Fullers earth. Because of very high plasma levels hemodialysis together with charcoal hemoperfusion was started within one hour after admission. This treatment was further continued by continuous veno-venous hemofiltration in order to remove the circulating paraquat and also circulating cytokines. Nevertheless patient s condition worsened necessitating artificial. ventilation and hemodynamic support. Patient died 24 hours after admission of acute multiple organ failure due to paraquat poisoning. Serum levels of paraquat were determined by colorimetric method (table) . Levels of interleukin 6 (IL 6) and 8 (IL 8), tumor necrosis factor (TNF-alpha), interleukin I receptor antagonist (IL 1 Ra) were determined both in plasma and ultrafiltrate ( Q~!ectives : Evaluate in critically ill patients the effects of tow-dose dopamine on gastric mucosal blood flow (GMBF) using Laser-Doppler Flowmetry, a continuous non invasive method of assessing microcirculation. Methods : 6 patients requiring both mechanical ventilation and pulmonary artery catheterization for multiple trauma (n=3), ARDS (n=2) and pancreatitis (n=l) were included. In each patient, the Laser-Doppler (LD) probe was inserted through a naso-gastric tube. The LD signal is proportional to the number of red blood cells moving in the measuring volume and the mean velocity of these cells. When the LD signal was satisfactory, an aspiration was created into a catheter which was fixed in parallel to the LD probe, to maintain the tip of the probe against the gastric wall at the site of measurement. Data (systemic hemodynamic parameters and GMBF) were obtained at the end of a 30 rain resting period (baseline), then 30 min after dopamine (2 mcg/kg/min) infusion, and finally 30 rain after the end of dopamine infusion (recovery GMBF 185 _+ 23 (Perfusion Units) GMBF 0 ~A% vs baseline) * p < 0.05 vs "baseline" and "recovery". Conclusions : 1) Despite a slight increase in CO (+13%), the dramatical increase in GMBF (+87%) with dopamine, strongly suggests a selective vasodilator effect of low-dose dopamine on gasaic mucosal perfusion. 2) Laser-Doppler Flowmetry appears a promising method to assess gastric microcircalation in critically ill patients. Increasing evidence suggests that the activation of iNOS is the final common pathway for vasodilation in human sepsis associated with endotoxic shock. Activation of the cellular immune system induces the excessive release of the pteridines neopterin (N) and 7,8-dihydroneopterin (NH2) by human macrophages/monocytes. Besides the well established diagnostic value of pteridines in several inflammatory diseases, it is speculated that these substances per se exhibit biochemical functions. Thus we hypothesize that pteridines can modulate iNOS gene expression in vascular smooth muscle cells (VSMC) in vilro. Cdtured rat aortic VSMC from female Wistar Kyoto rats were incubated with N (20 pM), NH2 (20 ILM), lipopolysaccharide (LPS, 5 ~g/ml), and interferone-~/(IFN-~/, 100 U/ml) for 9 h, respectively, iNOS gene expression was measured by competitive reverse transcription polymerase chain reaction. The results are summarized in the table. The present study demonstxates a neopterin induced increase in iNOS mRNA expression at the transcriptional level in VSMC. While coincuhation of cells with N + LPS resulted in an additive effect on iNOS gene expression, N + IFN-7 seem to have a more than additive effect NH2 did not alter iNOS mRNA synthesis, but it suppresses the LPS as well as the IFN-yinduced augmentation of iNOS gene expression. We speculate that this pteridine-mediated modulation of iNOS gene expression is involved in the regulation of the vascular tone in endotoxic septic shock. The relationship of sepsis and coagulation abnormalities is well known, mainly in severe sepsis and septic shock. Still farther, the extreme expression of hemostasis abnormalities (disseminated intravascular coagulation) in sepsis, has been extensively described. We studied the changes in several coagulation and fibrinolysis markers in septic patients, trying to correlate them with the evolution of the sepsis phenomenon, with an emphasis in its early stages, where therapeutic intervention might be more drastic. In 64 patients, 30 with sepsis, 22 with severe sepsis and 12 with septic shock, as well as in 14 healthy volunteers (control group) we measured : platelet (PTL), coagulation markers [Fxii, Fvii, Fviii, Fvw, Fibrinogen (Fibr) We conclude that all parts of the coagulation system are gradually changed during the evolution of sepsis phenomenon , even in the earliest stage of sepsis. The expression of an inducible nitric oxide (NO) synthase (iNOS) plays a major role in the pathophysiology of septic shock (SS). Inhibition of iNOS could therefore be of therapeutic value. However, such an inhibition has been shown to be detrimental, increasing tissue anoxia (and end-organ damage), possibly through the simultaneous blockade of constitutive NOS (cNOS). Thus, selective inhibition of iNOS might be more suitable. We evaluated the effects of L-Canavanine (CAN), a more potent inhibitor of iNOS than cNOS, in an animal model of SS. Method: In 30 anesthetized rats, catheters were placed in the femoral vein and artery. 23 rats were given an iv bolus of lipopolysaccharide (LPS, 5 mg/kg), at baseline (TO). After 1 h (T1), rats received at random an infusion of either CAN (20 mg/kg/h; CAN group, n=l 1) or an equivalent volume of 0.9% NaCI (2cc/kg/h; NaC1 group, n=12), giyen over 4 h (T1-T5). A third group (sham group, n=7) received 0.9% NaC1 in place of LPS, and then was treated like the NaC1 group. Mean blood pressure (MBP), blood lactate and nitrates (NO3) were measured each h. Glucose, creatinine and ASAT were also measured in 18 rats (n=6 in each group). The CAN 98_+9* 304+52"t 3.2+1.1"~ 3.6+_1.4"t 138+46" 48+10" *p<0.05 CAN vs NaCI ?p<0.05 vs sham CAN suppressed the hypotension, reduced the hypoglycemia and hyperlactatemia, and attenuated the biological signs of renal and hepatic dysfunction induced by endotoxemia. These effects were associated with a lesser elevation of blood NO3, confirming a partial inhibition of iNOS. Conclusion: L-Canavanine attenuates the hemodynamic and metabolic consequences of endotoxemia in the rat. These effects may be related to a partial inhibition of iNOS. They contrast with the deleterious effects described with non selective inhibitors of NOS. L-Canavanine could become a new tool for the treatment of septic shock. ROCALC1TONIN :MARKER OF SEPSIS, II~FLAMMAiiUr% t~ BOifi .Cheval*~ JF.Timsit*, M.Assicot**, B.Misset*,/.Carlet*, C.Bohuon** Saint Joseph Heap, Paris**Biochemistry Institut G Roussy, Villejuif, CE bi~)l~i~ttectives_: High serum levels of procalcitoaln (ProCT) have been shown to be ~ss-ocinted with bacterial infection. However, few data exist about the ability of ProCT to differenciate septic shock and shock from other origin in which an activation of intlmmamtory mediators has been also demonstrated. Methods: Thirteen patients with bacterial septic shock (SS), 11 patients with non septic shock (NSS), 14 patients with bacterial infection without shock (1NF) and 8 ICU patients without shock and without infection (Control) were compared for ProCT levels at Dayl, 2, 3, 7, 14 . Patients were classified blindly and independently fi'om ProCT results. Twelve patients were excluded because any classification was impossible due to mixed pathology. ProCT was measured with ebemoluminescenee (BRAHMS diagnostica-Berlin). Results: Dayl, ProCT levels are significantly different between the four groups. Dayl proCT levels are correlated with SAPS (p=0.0002), infection (3.7+_3 vs 61_+25,p=0.0007), shock (13_+10 vs 60+.-27,p=0 .002), death at Day28 (12_+7 vs 96_+44,p=0.003). When shock and infection are introduced in multifactor &NOV& only infection remains correlated with Day 1 ProCT levels (19=0.003) In patients with shock, Dayl ProCT levels are correlated with SAPS, infection and death at day28, but not with arterial lactate levels (p=0.37), white blood calls (p=0.2) or fever (p=0.1). ProCT levels remain higher i~i septic shock patients at Day 1, 2 and 3( Figure) . I C0edpsion: Procalcitonin levels in the first three days of shock are differen[" between septic and non septic shock patients. In patients with diseases known to induce acute an inflammatory process, procaldtonin seems to be a marker o~ infection. Obiectives-To evaluate the effect of endotoxic shock on the distribution of blood flow between the mucosal and the muscular layer of the intestinal wall. Methods: In 10 fasted pigs, mean aortic pressure (MAP, mm Hg), cardiac output (CO, ml/min-kg),superior mesenteric artery flow (Q SMA, ml/min.kg), and phi, where measured before (Control) and after i.v. Endotoxin (10 gg/kg). The blood flow to the mucosal and the muscular layer was measured in 3 regions (proximal jejunum (PJ), mid-small intestine (MI) and terminal ileum (TI)) by colored microspheres, using 8 adjacent samples in each region. The muscular layer was separated from the mucosa by blunt dissection, and the flow determined independently in each layer. Results: Endotoxin with fluid resuscitation induced the expected decrease in MAP (95.8_+3.0 vs 53.9-+2.4, p<0.01), and phi (7.29!-_0.02 vs 7.13_+0.01, p<0.01), with a constant CO (133_+4 vs 144_+8, p=0.28) and QSt, AA (21.5_+0.8 vs 22.8_+0.9, p=0.09). The results of regional pertusion are presented in the table. (Flow in ml/rain 9 100g of tissue; mean _+ SEM ; * p<0.001 vs Control by two-way Anova) Conclusions-These data indicate that the mucosal flow increased during septic shock. They suggest that a decrease in pHi may be due to hypoper~usion of the muscular layer or to metabolic alterations within the mucosa, despite a 50% increase in flow. acute increase in WBC count (from a mean of lO.OO01mm a to 42 O00/mm~), between the 3rd and the 7th day of therapy. There was a decline of the WBC count to an average of about 25.0001mm a after decreasing the daily dose of the medication to 200 mcg There was no increase in tile absolute number of the eosinophils during the whole course of the medication. There was a slight decrease in the C3 complement between 0.26 to 0.29 g/I. Normal values 0.5 to 0.9 g/I There was no change in C4 values. Conclusions : An early increase in WBC count was observed (3rd day) without subsequent increase in the number of immature types from bone marrow, probably due to the mobilization of WBC from the periphery and this increase was dose dependent. There was a slight decrease in C3 fraction of complement, probably due to the consumption of this fraction in the process of opsonization. No adverse effects of the medication were observed, during the treatment with the above dose. These data sugest that CM CSF may be a useful complement to tile main antimlcrobial treat,nent ~ of septic [CU patients. Objectives: As part of a large multicentric, placebo-controlled, randomized clinical trial investigating the effects of interleukin-1 receptor antagonist (II-lra) in the treatment of severe sepsis and septic shock, this substudy evaluated in dem.il the acute hemodynamic effects of II-lra in patients who were invasively monitored. Methods: In a total of 71 evaluable patients in whom vasoactive support was little altered, hemodynamic measurements were performed at baseline (twice), and I hour, 2 h, 3 h, 4 h, 8 h, and 12 h after the administration of 1 mg/kg (N=20) or 2 mg/kg (N=22) of I1-1 ra or the corresponding placebo (N =29). 58/71 patients (82 %) were treated with adrenergie agents and 66/71 (93 %) with mechanical ventilation. Data were analyzed by a Kruskal-Wallis test. Results: During the study, there was no significant difference with time or between groups in arterial pressure, cardiac filling pressures, cardiac index or left ventricular stroke work (figure). Burmester, "~ Man8 and H. Djonlagic Medical University (Internal Medicine, "Cardiology, *'Microbiology) and "**Southern City Hospital, Lfibeck, Germany Obiectives: Evaluation of the incidence of bacteremia and sepsis in patients with nontyphoidal salmonella (S.) infections, Specification of risk factors, need of ICU treatment, clinical course, and mortality in the group of the patients who developed septic complications. Methods: Data of all patients with microbiologically proven S. infections hospitalized in the Medical University of LObeck and in the Southern City Hospital of L0beck from 1992 to 1994. Results: Within the observation period S. was isolated from the stool cultures of 748 patients. In 13 patients (g m, 4 f, median age 52 yrs) S. could be detected in blood cultures (9 S. enteritidis, 4 S. typhimurium). In addition, in 10 of these patients S. was also isolated from other specimens (urine, liquor, and tissue fluids derived from abscess punctures). In all 13 patients with positive blood cultures the clinical course of S, infection was complicated: ? patients developed MOF (acute renal failure, ARDS, hemodynamic instability, DIC) and required ICU treatment for at least 4 up to 62 days, 4 of the 13 patients died. The predisposing disorders in the patients with S. bacteremia were (n=): AIDS (3), immunosuppressive drugs (2), chronic alcoholism (2), malignancies (2), none (4). Septic complications in patients with nontyphoidal S, infections are relatively rare (in this study < 2 % of all hospitalized patients with microbiologically proven salmonellosis) but severe (mortality of approx. 30 %). Patients at risk for a complicated clinical course are predominantly those with predisposing disorders but occasionally also patients without evidence for an underlying disease. age (yr) 65 + 9 58 + 14 Death (n) 22 18 Duration of Shock (h) 27 + 37 54 + 44 Noradrenaline (rag/h) 5,9 _+ 6 2 + 5 Temperature (~ 38,2 + 2 38,1 + 1 PVR (dynxsecxcm -5) 1195 + 408 572 + 418 CO (LJmin) 4,2 _+ 1,6 9,9 + 3,6 Lactate (mmol/L) 7 + 5,9 9,2 + 9 Interleukin-6 (pg/ml) 829 _+ 798 1331 + 888 Interleukin-1 (pg/ml) 9,3 _+ 9,4 9,8 + 7,3 TNF-alpha (pg/ml) 23,5 + 31,7 166 + 209 Neopterin (nmol/L) 43,2 + 43,7 218 + 193 CRP (rag/L) 131 _+ 97 233 +_ 138 Pro-CT (ng/ml) 22,6 + 50,5 77,9 + 160 There was no positive correlation between serum lactate levels, degree of shock, hypoxemia and pro-CT positivity. Pts with septic shock of bacterial origin entirely developed hyperprocalcitoninemia, whereas pts with cardiogenic shock, who expired within 24 h did not. However, in late cardiogenic shock (>24h) all pts developed fever of unknown origin and consecutive hyperprocalcitoninemia. These data suggest bacterial inflammation and/or mucosal translocation of bacterial products in pts with prolonged cardiogenic shock. The use of a loading dose of quinine (16.7 mg/kg base in 4 h) is recommended in previously untreated patients (pts) with SFM, particularly in multi-drug resistance areas. This protocol is difficult to validate, since the viability of microorganisms is not assessed routinely in parasitology laboratories. Objectives: to examine the evolution of parasite viability during the early phase of therapy of SFM. Methods: from 02/1993 to 12/94, pts with SFM (WHO 1990) treated with IV quinine for less than 6 h were included prospectively. Blood samples were collected at O, 6, 12, 18, 24, 36 and 48 h Viability was assessed by culturing parasitized red blood cells in the presence of 3H-hypoxanthine, and radioactivity was determined at 42 h by scintillation counting. Viability was expressed as the percentage of radioactivity compared to the initial sample. Plasma quinine was determined by liquid chromatography. Tile ratio plasma quinine (pmol/1)xlO00/iCso for quinine (nmo]/]) was called the parasiticida/ index. Results:5 pts were included, 42• SAPS1 18.6-+4.9. The initial parasitemia was 2t.4+7.2%. Complications of malaria were coma (4 pts), shock (3 pts), renal failure (2 pts) and acute lung injury (2 pts). All strains were sensitive to quinine (ICso 174--67 nmol/1). In 2 pts who were not given a loading dose, parasite viability increased by 63 and 157%, with concomitantly low quinine levels (22 and 19 #moW] at 12 h); 1 pt died. In 3 pts that received a loading dose (serum quinine at 12 h = 33.1--2.0 ~mol/]) a marked decrease of parasite viability (by 73+_10% at 12 h) was shown. Viability was inversely correlated with plasma quinine (r=.677, p-.O11) and parasiticidal index (r=.678, p-.O1). Conclusions: Even with fully sensitive strains, the use of a loading dose of quinine seems warranted in severe falciparum malaria in order to reach rapidly adequate plasma quinine ]evels, necessary to inhibit significantly parasite viability. L nkka, E RuokonelL J Takala. Critical Care Research Program, Department of Intensive Care, Kuopio Univ Hospital, Finland Objective: To determine the incidence of positive blood cultures, their microbial subgroups and to evaluate the outcome of ICU patients with different bacleremias. Material and methods: We analysed all positive blood cultures in 3077 consecutive admission to a university hospital ICU in 1992 -93 and the ICU and hospital survival of the bacteremia patients. During these years 73 patients had 176 positive blood cultures that were considered as clinically relevant, excluding colonizations or contanfinations. Results: Patients with positive blood cultures had an ICU survival of 65.8 % (vs. 92,7 % in all ICU patients) and six month survival of 50.7 % (vs. 85.8 % in all ICU patients). The most common bacteria were enterobacteriaceae (27,3 %), staphylococcus aureus ( 18,8 %) , coagulase negative staphylococci ( 14.2 %), pseudomonas ( 14.8 %) and slIeptococci (9.1%). Obiectives: To evaluate prognostic factors and mortality in consecutive patients (pts) with HIV Infection and septic shock. Methods: From 03-1991 to 12-1993 , records of consecutivepts with septic shock (Crit Care Med 1992, 20: 864-74) admitted to the ICU were reviewed retrospectively. Results: Among 76 pts with septic shock admitted during the study period, 28 had HIV infection-26 of whom had AIDS-(gr. I) and 46 were HiV-negative (gr. Ill. Ten gr. II pts (21%) were irnmunosuppressed because of neoplastic or immune dlsease. Mechanica] ventilation was required in 89% gr. I and 83% gr. II pts In 8 gr . I pts (29%) A multivariate analysis demonstrated that HiV infection and SAP5 I were independently predictive of death in pts with septic shock. ~onclusions: Evidence of increased mortality, number of organ failures and higher severity scores (SAPS I does not take into account immunosuppression) is demonstrated in HI V-positive pts, Infection with HIV appears to be an independent prognostic factor in pts with septic shock. The frequency of opportunistic infections (often responsible for delayed diagnosis and treatment) may contribute to the poor prognosis in this population. Obiectives: To determine interleukin (IL)-I 0 levels in plasma of patients with sepsis and septic shock. To analyze the relationship between plasma IL-10 and the proinflammatory mediators, tumor necrosis factor-aIfa (TNF) and IL-6, the underlying severity of the disease and the evolution of patients with sepsis. Methods: We studied 94 critically ill patients (63 men, 31 women; 18-86 years old) in three diferents groups. Group I: 23 patients without evidence of infection, group I1:34 patients with sepsis and 37 with septic shock (group III). We measured plasma IL-lO, TNF and IL-6 levels in the first 12 hours of diagnosis. Severity of illness was estimated with the Acute Physiology and Chronic Health Evaluation (APACHE II) scoring sytem. Results: Plasma levels of IL-10 were higher in group III (median, 51 pg/mL; range, 5-6000 pg/mL) than in group II (median, 10 pg/mL; range, 2-970 pg/mL; P <.001) and group I (median, 5 pg/mL; range, 2-133 pg/mL; P <.001). Median IL-10 concentrations did not differ among patients who survived (median 7 pg/mL; range, 2-6000 pg/mL) and those who died during the overall follow-up period (28 days) (median, 15; range, 5-5400 pg/ML); but patients who died in short-term (< 24 hours) with catecholamine-refractory hypotension showed the highest concentrations of IL-IO (median, 1200 pg/mL; range, 51-5400 pg/mL). In patients with bacteriemia (34%), levels of IL-10 were higher (median, 51 pg/mL; range, 2-6000 pg/mL) than in those with negative blood culture (median, 8,5 pg/mL; range 2-5.400 pg/mL; P< .001). There was a good correlation between plasma IL-IO concentration and levels of TNF (r= .59; P < .001) and IL-6 (r= .60; P < .001). The correlation between levels of IL-10 and the APACHE II score was significant only in the septic shock group (r= 0.48; P <.005). Conclusions: In septic shock, IL-IO and proinflammatory citokines are released in high concentrations. The significant correlation observed in patients with septic shock between IL-10 levels and APACHE Ii, short-term death and bacteriemia can possibly be explained by the massive inflammatory response in septic shock with fulminant course. Intensive Care Department -Calmette Hospital -59037 LILLE -FRANCE. In septic shock, inadequate splanchnic blood flow may play a prominent role in the pathogenesis of multiple organ failure. Measurement of gastric pHi has been propose to evaluate tissue oxygenation in splanchnic organs. Objectives: to compare gastric phi values with hepatic ICG clearance, an index of liver blood flow and function ; to determine if one of these two methods could be proposed to assess the entire splanctmic peffusion in septic shock. Methods : 6 patients (age : 65• years ; SAPS II : 46• were prospectively investigated (septic shock : Bone criteria). Following parameters were collected during 12 hours : systemic hemodynamic parameters (Swan Ganz catheter 93A434H -REF1 computer -Baxter Lab.), calculated systemic oxygen transport (DO2), oxygen consumption (VO2) by indirect calorimetry (Deltatrac Datex Lab.), gastric intramucosal PCO 2 (PCO2ss) and phi (TRIP -NGS catheter -Tonometrics Lab.) and plasma disappearance rate of ICG (PDR dye) (femoral artery fiberoptic/thermistor catheter 2024, Cold Z021 computer -Pulsian Medizintechnik, Germany). Correlations were performed using a linear regression. elevated in all days with the highest value in second and third days of treatment. Nonsurvivors had higher values of these parameters than survivors but differences did not reach statistical significance. Another trend of changes were observed in selectin P (GMP-140) concentration. In all patients concentrations measured were elevated but in survivors after not significant decrease this parameter in second day another one had simmilar values. In patients who died we noted significant decrease in third day (p < 0.05) whereafter prominent increase, significant after seventh day, in comparison to third day value and value in survivors group. ICAM-1 concentrations in all patients reached high levels and in nonsurvivors after four day of treatment significant increase in comparison to survivors we found. Conclusions: Multiple trauma complicated with sepsis induce rapid elevation of concentrations of IL-6, IL-8 and increased expressior of adhession molecules (selectin E, P, ICAM-1) Measure of ICAM-1 and selectin P concentration determine lung injury severity and prognosis as to health and life. (ClP) .Pathophysiology of CIP is unclear, but changes in regional bloodflow may be a ~ignificant factor. Nerve blood flow (NBF)is reduced in rat models of hemorrhagic shock (g),but no information is available in sepsis. We studied the comparative effect of acute endotoxemic shock {ETX)& H on perfusion of rat sciatic nerve. Methods: 20 male Sprague-Dawley rats were anesthetized with pentobarbital (ip), instrumented with a tracheostomy, carotid arterial & venous catheters and mechanically ventilated (Fi02=0.5). The left sciatic nerve was surgically exposed. Monitored variables included: a) mean arterial pressure (MAP,mmHg) ,b) NBF (ml/1O0 g/min) by laser Doppler flow meter,c) nerve internal arterial diameter (ID ~ m) by video image shearing and splitting method. After stable baseline measurements were obtained, acute hypotension was induced by randomly assigning the rats to ETX (0.25 B6, Difco) in saline at 1 mg/kg or H. Both interventions produced 50% reduction in MAP within 3 min., which recovered to baseline values spontaneously in ETX group, & by reinfusion of heparinized withdrawn blood in M. Data were analyzed by linear regression, two-way repeated measures analysis of variance followed by Bonferroni-t method. Experimental stages were:(1)baseline, (2) mid-point of MAP reduction; (3) nadir of hypotension, (4)midpoint of MAP recovery, & (5) after stable recovery of MAP. Both ETX & H induced shock result in similar reduction in NBF consistent with lack of autoregulation in peripheral nerve vessels independent of etiology. Since CIP is primarily associated with sepsis, it is not likely that acute reduction in NBF alone causes CIP. Direct & indirect neurotoxic effects of mediators of sepsis need to be evaluated. .':_.~::::o4o:oc4., Objectives : Evaluate the relationship between IL-10, a cytokine which inhibits TNF, production and protects mice from endotoxin toxicity, and the other proinflammatory cylokines, TNF~, IL 6 and ILs in severe sepsis and septic shock. Methods : Twenty-eight ICU patients (19 M, 9 F, mean age 54 + 17 y) were studied as soon as they developped a severe sepsis (n = 16) or a septic shock episode (n= 12) as defined by a conference consensus in 1992 (1). TNF~, IL 6, IL s and IL-10 plasma levels were measured by immuno-radiometrie assays from Medgenix (Fleurus, Belgium). Lc mean and range. Results : The comparisons between cytokine levels in severe sepsis versus septic shock were made using the logarithm of the value in order to normalize the distribution of data, and student test. IL-10 plasma levels were higher in patients with septic shock than in patients in severe sepsis. There was a significant correlation (p < 0.05) between IL-10 and TNF a (r= 0.6), IL-10 and IL~ (r = 0.73) and IL-10 and IL s (r = 0.65) as well as between IL-10 and Apache n score (r= 0.52). Patients who died (n = 13) had IL-10 levels higher than patients who survived but this difference was not statistically significant (114 pg/ml vs 34.5 pg/ml; p>0.05). Conclusions : During severe sepsis and sepsis shock, IL-10 seems at least to follow the same evolution (increase in plasmatic level) with the severity of sepsis as the other cytokines. Reference : (1) Crit Care Med 1992;20:864-74. Objectives: To evaluate the effects of steroids on hemodynamics and mortality in septic patients with konwn levels of cortisol concentration. Methods: Retrospectively we analyzed data ofpatients with documented septic shock who received steroids after assessment of adrenal function. In all patients hemodynamic parameters as well as the necessary vasoactive medication were assessed, before and 24 hours after corticosteroid medication. Immediately before administration of corticosteroids adrenal function was evaluated with cortisol levels before and after synthetic corticotropin (0.250 mg). Finally we studied mortality. We defined a positive respons on corticosteroids as an elevation of MAP of at least 30 mmHg and/or a decrease in the necessary vasoactive medication of at least 50% within 24 hours. Adrenal insufficiency was defined as a cortisol level after stimulation of less than 500 nmol/l. Results: 15 of 23 patients were found to respond to steroid medication, 8 did not. Mean cortisol levels before and after corticotropin were 534 • 366 and 737 • 396 nmol/l in the responder group (RG) and 583 • and 907 • 301 nmol/l in the non responder group (NRG). In the RG 9 out of 15 (60%) were found to have an adrenal insufficiency, in the NRG 3 out of 8 (37%). In the RG 2-weeks mortality was 6.7% (l out of 15), the overall mortality 33% (5 out of 15). Mortality in the NRG was 62% (5 out of 8) (p <0.01) and 75% (6 out of 8) (p <0.005) respectively. Conclusions: In patients in septic shock there is a beneficial effect of steroids in case of adrenal insufficiency, but also in a subgroup with normal adrenal f{unction. Obiectives: Intercellular adhesion is a critical step in the accumulation of leukocytes. Postischemic cardiac lymph has the capacity to stimulate ICAM-I. In the coronary microcirculation neutrophils can be trapped and in many cases obstruct capillaries, Previously we found that Troponin T (S-TnT) a marker for myocardial iechemia, was increased in septic patients. The aim of the study was to follow slCAM-1 and S-TnT levels continuously starting at the beginning of sepsis. Methods: 19 Patients were inGluded in this institutionally approved study after relatives had given their informed consent. All patients were included within 24 hrs following the beginning of sepsis. Blood was drawn every 4 hrs in the first ;~4 hrs, after 48 hrs, followed once per day for 7 days. S-TnT, ICAM-1, ELAM (ELISA's, Boehringer Mannheim Inc, R&D Systems Ltd.) arterial and venous blood gases were determined, an ECG and a complete hemedynamir measurement including cardiac output were obtained. All patients received adequate volume and catecholamine therapy (norepinephrine, dopamine, dobutamine; median (range) 0.6 (0.0-1.66), 3.12 (2.4-12), 6.29 (0.0-15.3) pg/kg/min, respectively). Statistical analysis: Wileoxon signed rank-sum test. .45 (0.06-7.6) 0.0003 13 patients had S-TnT levels >0.2pg/L. 11 of these died, whereas only 2 of 6 patients died with S-TnT values <0.2 pg/L (p=0.0296). All patients that died had elevated sJCAM-1 levels (232 Ilg/L:cut-off ) whereas in the survivor group only 50% had elevated ICAM-1 levels (p=0,043). Conclusions: Increased slCAM-1 and S-TnT levels were found during early sepsis in the majority of patients, A high sICAM-1 and S-TnT value was associated with a higher mortality. The research of the noninvasive haemodynamic monitoring accelerated recently all over the world. The aim of our study was to test whether the changes of the haemodynamk parameters measured by impedance cardiography (ICG) were corresponded to clinical changes in septic patients. Investigations were performed on 20 critically ill postoperative septic patients (their multiple organ failure score was 8-9/with ICG monitor. In 9 cases the inveStigation~ were performed in septic shock. The measured parameters were: heart rate (HR), mean arterial pressure (MAP), cardiac output (CO), peripherial resistance (SVR),preejection period (PEP), and ventricular ejection time (VET). These parameters were measured during 3-72 hours in every 30 minutes, depending on the patients Cl~tnical condition. Results: At the septic patients the HR and the CO ]~reased. In septic shock the CO was significantly higher the SVR lower than in the septic group. In the HR there was no difference between the two groups. In septic shock Noradrenalin influenced more effectively the measured parameters than Dobutamin. Conclusion: The trend of the measured ICG parameters correlated with the clinical changes of septic patient's state. The noninvasive haemodynamic monitoring by impedance cardiography helps the planning and leading the adequate intensive therapy of these critically ill septic patients. To evaluate the development of SIRS, sepsis and septic shock in hospitalized patients with fever, a prospective study was performed on 300 patients using previously defined criteria. Methods: 300 normotensive patients with fever (temperature >38.0 ~ axillary), admitted to the Department of Internal Medicine were evaluated for the existence of SIRS during the first three days of the study and sepsis at inclusion. During a follow-up period of 7 days the patients were daily evaluated for the development of sepsis or septic shock. Results: Most patients (69%) had or developed SIRS within the first three days, 16 patients (5%) did not. Sepsis was present in 25% at inclusion. In patients with SIRS, 76% did not progress to sepsis or septic shock, 24% progressed to sepsis (mean interval 2.55 • 1.97 days), and 1 patient (<1%) directly progressed from SIRS to septic shock. In patients with sepsis, 17% progressed to septic shock (mean interval 2.08 • 1.56 days). Sepsis was preceded by SIRS in 40%. Septic shock was preceded by sepsis in 92% and by SIRS in 8%. Conclusions: 94% of patients with fever in an internal medicine department develop SIRS, or sepsis. Furthermore, progression from SIRS to Sepsis or septic shock is poorly predicted by fever or SIRS. Nevertheless, all patients with septic shock were preceded bySIRS or sepsis. Taken together, this may indicate a severity hierarchy of the syndromes. However, fever, SIRS and sepsis are relatively poor indicators of development of septic shock. This supports further research on additional predictors of septic shock. B. M.Manuylov, V.B.Skobelsky (Moscow) In recent years sodium hypochlorite (SH) has been successfully used to eliminate pyo-septic complications. Moreover, the mechanism of the SH effect on the immune system has not been sufficiently studied. The aim of the present investigation was to study the mechanism of SH effect in inflammatory pulmonary diseases. 20 patients with double pneumonia were subjected to the evaluation. SH in the concentration of 600 mg/l in the volume of 400-800 m1/24 hours was administered by drop infusion into the central vein. To evaluate one of the defence systems the leukocytes activity by the chemoluminescence technique was studied. In all the patients baseline secondary immunodeficiency which was indicated by the decrease in the luminescence level was established. Even 1 hour after the SH administration the leukocytes activation exp-ressed by the enhancement of their chemoluminescence 0.5-5 times was observed. This supports the available findings that accumulation and liberation of the oxygen active forms (OL'OH, '02, H202) are accompanied by the increased phagocytosis, i,e. the signs of "the oxydation explosion" testify to the favourable SH effect on the course of inflammation processes. The use of SH permitted to decrease the percentage of lethality in double pneumonia by 15% in the intensive care unit over the year. At the same time, excessive activation of free radical oxygen may be a damaging factor. Therefore, precise individual control over the choice of concentration, dosage and the preparation administration rate is required. Prospective, double-blind, placebo-controlled, trial of ATIII substitution in sepsis R. A. Balk Objective: Pilot study to evaluate the efficacy and safety of ATIII substimtion therapy in patients with sepsis. Efficacy assessed using change in mortality or organ failure/dysfunction. Adult patients meeting a definition of sepsis and cared for in a tertiary care academic medical center in Chicago were identified and prospectively randomied to receive either ATIII (Kybernin P) or placebo in a double-blind treatment protocol. All other therapy and patient management were under the direction of the patient's attending physician. All patient's were followed for 28 days and the organ dysfunction/failure were scored using published scoring systems (Jordan et al Crit. Care Med. 1987 , Goris et al Arch. Surg. 1985 , Kuaus et al Ann. Surg. 1985 Colldusions:Wha~ we met the Shomaeker objectiv% the mortality and the pro~os[s were I~ttc*. Those criteria were obtained with file tradititmal t~ctor likr doht~mme, hut C.~VH ~,as ca1 in~aertam measure. They ac~s sMxergically in the optimizatic~l of the fell vmtrictdar work index, tad fimdameatally CAVH seox~s to have an impo.aat role in the better respiratory ev-altmtioa, leaving yet the possibility to coltrol the flui& r Althou~l eomproved it's not aec~pt~xl file importmlce h* the diminution, of the sepsis modiat~lrs llke FNT and IL-6 with h~wmotiltrafi(al, stopphlg the evolution to nmltiorganic failure mid de~easethe mortality. With ours clhlicals results, we could saythat CAVII in multiol~atlie disfut~oa septic patieats, se~r~ to be an c4Xilna] supoa or troatmeat maesure. of Anaesthesia and Intensive Therapy, Medical University of Prcs, P~csf Hungary. Objectives: Since some biological effects of bacterial endotoxin require an interaction between the LPS molecule and a serum factor(s), we hypothesized that LPS-induced NO production and cGMP accumulation in vascular smooth muscle cells (VSMC), a mechanism ~thought to underlie cardiovascular collapse associated with septic shock, is modulated by serum factor(s). Methods: Cultured VSMC from rat aorta were challenged with E. Coli LPS for 4-6 hours either in the presence or absence of fetal calf serum (FBS), and NO production was monitored by radioimmunoassay determination of cGMP content of HCI extracts. Results: In the absence of serum, 1O00 ng/ml LPS was required to increase cGMP levels, whereas the presence of 10 % FBS shifted the LPS concentration curve i00 times to the left. Similarly to FBS, human serum also potentiated LPS-induced cGMP accumulation. In contrast to LPS, serum had no effect on cGMP accumulation elicited by sodium nitroprusside, a NO releasing agent, suggesting that the sensitivity of VSMC to generate cGMP in response to exogenous NO is not modulated by serum. Heat inactivation (>80 ~ 30 min) but not removal of small molecules (<10,000 D) from the serum by dialysis, reduced the potentiation of cGMP accumulation by serum. Time course studied indicated that serum is required within the first 120 min of LPS exposure to increase cGMP levels. To investigate whether the effect of serum is specific for LPS, we treated the cells with increasing concentration of interleukin 1-~ (IL-I). 10% FBS shifted the IL-Iinduced cGMP responses five times to the left. Conclusions: Our study suggests that lower concentrations of E. Cell LPS and IL-I require a heat labile macromolecule in the serum in order to elicit NO production. This factor is present in the human serum and it may play a potentially important role during NO synthesis induction in VSMC. Objective: To evaluate the factors of acquisition and the outcome of methicillin resistant Staphylococcus aureus (MRSA) bacteremia in an intensive care unit (ICU). Methods: All patients in which bacterermia due to staphylococcus aureus developed >72 hours following admission to our ICU, during a 10 year period ( january 83 through january 94) were reviewed. 30 patients (pts) were included, mean age 68,1y (SD 13,1), SAPS 2 35,9 (SD 11,1), Mac Cabe (1 and 2) 53%, mortality directly due to sepsis 30%. 16 pts had MRSA bacteremia and 14 methicillin susceptible Staph. aureus (MSSA) . Both groups were compared using the chi square (with correction of Yates), Fisher's exact, student's t or Wilcoxon test. Results: There was no statistically significant difference between MRSSA and MSSA regarding at age (71,8+ 4,8 vs 63,9+ 1,8) , SAPS2 (33,6+6,6 vs 38,2+14,1), use of vancomycin (94% vs 71%), mechanical ventilation (94% vs 100%), number of days (d) before the drawing of the first positive blood culture (median 20 d, range 7-150 d vs median 30 d, range 7-120 d). More MRSA than MSSA pts had previous use of nonsteroidal anti-inflammatory drugs (NSAID) (25 % vs 0% p<0,001), central venous catheter infection due to Staph.aureus (62,5% vs 14% p<0,01), but previous use of antibiotics was not significantly different (37,5% vs 21%). The outcome of the bacteremic pts was not statistically different: SAPS 2 at the first day of bacteremia (33,6+_.7,2 vs 40,7+14,5), severe sepsis and septic shock (31% vs 28%), persistence of the bacteremia (43% vs 78%), mortality directly due to bacteremia (25% vs 45%). Conclusion: Previous use of NSAID, infection of venous central catheter are more frequently associated with MRSA bacteremia. Thus, similar to others studies (Hershow Infect Control Hosp Epidemio11992; 13:587-593) , these results do not indicate that MRSA is associated with increased virulence. Objectives: To closer definition of MOSF formation mechanismes in nosocomial sepsis (NS) the complex clinicobiochemical, microbiological, immunological, functional exaroination of 62 cases with NS had been done. Methods: Examination of cellular and humoral immunity, nonspecific immunologic reactivity, systemic and hepatic circulation, microbiological examination of blood,electro-and echocardiography, sonography and computer tomography of chest and abdomen organs were obligatory. Autopsy findings of 5 dead cases had been analized. Results: In 45 cases (72,6%) opportunistic pathogen microscopic flora ( Staphylococcus anreus,Staphylococcus epidermidis, Staphylococcus saprophyticus) had been found out in blood inoculations. In 36 cases (58%) side by side with destructive process in lungs the bacterial endo-and myocarditis with blood circulation failure had been determined.In 21 cases (34%) simultanious lesion of three organs (heart,lungs,liver) had been found. Morphologic examinations of 5 dead cases (8%) internal revealed involvement of them in MOSF-syndrome.Hyperplasia of adenohypophysis;sclerosis of adrenal glands cortical layer;perivascular brain oedema,paralysis of brain capillaries and plasmorrhagia, cerebral thrombosis and cerebral abscess,necrobiosis of epithelium tubules of the kidney,pletora of hepar, fatty and granular degeneration of hepatocytes had been found.Atrophy of white pulp and hyperplasia of red pulp, supress of lymphoid tissue, plethora and formation of infarctious had been found in spleen. Mentioned changes in spleen were indispensable in NS. Conclusion: In NS spleen can not secure it functions to support and appropriate detoxication potencial of organism,elimination of microbes,toxines,antoallergenes. Insolvency of immunological link of antimicrobic defence is the starting mechanism of MOSF developmentin NS. %Neviere, JL. Chagnon, B. Vallet, D. Mathieu, N Lebleu, F. Wattel ] ept of Intensive Care, Hop Calmette, Lille, France ~everal studies have described tiypoperfusion of intestine during sepsis. 4owever, it is unknow whether the mesenteric blood flow is associated with nucosal hypoperfusion. Additionally, the effects of resuscitation on the ntestinal microcirculation remain controversial. 3bjectives : to describe the effects of endotoxin in a porcine model during ~hock and resuscitation. ~ethods : Ten pigs (30 kg) were anesthetized and instrumented for "neasurement of cardiovascular variables. Gastric and gut oxygenation vere assessed by intra-mucosal pH and microvascular laser Doppler lowmetry. After baseline data collection, a 30 minute intravenous infusion )f Escherichia colt (serotype 34H4113, Sigma, St. Louis, MO) was begun ~t a rate of 150 pg/kg. An infusion of either saline at 1.7 ml/kg/min (Group ; n=5) or saline and dobutamine at a rate of 5 pg/kg/min (Group II; n=5) vas begun 30 mn after the end of the endotoxin infusion. tesults : TO Td0 1120 T180 ~ fl0w Fluid Ioadin,q alone Sfyras D, K Perreas, E Douzinas, K Spanou, M Pitaridis and C Roussos Critical Care Dpt, Evangelismos Hosp., Athens Univ, School of Medicine. Obiectives: Much controversy exists concerning the beneficial effects of CVVH on sepsis. We studied the effects of CVVH application on septic patients with reference to the following parameters: i) survival rate ii) cytokines' removal and iii) timing of CWH onset. Methods: 20 patients with sepsis (criteria according to ACCP/SCCM, 1992) underwent CVVH as soon as they developed renal failure or dysfunction (urinary output<250 ml/8h, Cr>2.5 mg/dl and BUN>60 mgd'dl ). Specimens were collected: Blood samples before CVVH and therafter both blood and ultrafiltrate (UF) samples on 24, 48 and 72 hours. Cytokines TNFa, I1-1 and II-6 were measured by the immunoassay method in all specimens (UF and plasma -P) and sieving coefficient ([UF]/[P]) and 24 h solute mass transfer of TNF and I1-6 were calculated (V24 h x [UF] ). The APACHE II score before CVVH onset, the duration of ICU stay and the timing of CWH application related to the sepsis onset in days (TA) were recorded.With respect the mortality two groups were formed, i.e. Group A (survivors) and Group B (non-survivors) . The morbidity period in days of those septic patients who died in the past year and were not subjected to CWH (Group C) was compared to that of group B. Results: Group A included 8 pts and group B 12 pts with mean+SD age (65 _+19 vs 64_+9, NS) and APACHE scores(24_+2 vs 24-+2.2, NS). The mean TA-+ SD was 3.6+2 vs 10-+6, p<0.05. The mean_+SE morbidity period of Group B vs Group C was 20_+4 vs 8_+0.8 p<0.05 . The mean values of cytokines are presented in the following figures. The sieving coefficient for TNF was 0.2 and for I1-6 was 0.25. The solute mass tranfer was 6-fold the actual plasma content at a given time. . o Conclusions: i) early application of CVVH seems to favourably affect the outcome of septic patients, ii) cytokine plasma levels do not decrease although cytokine removal is substantial, iii) it seems that CWH application in sepsis of any stage helps to buy time for further treatment. The most commonly monitored variables in shock stages iDclude : arterial pressure, heart rate, central venous pressure, pulmonary artery wedge pressure and cardiac index. With vigorous therapy it is possible to bring these values back into the normal range in both survivors and nonsurvivors. Therapeutic goal in septic shock stages is to maximize the values of cardiac index, 02 delivery (DO2) and 02 consumption (C02). Objectives: The main purpose of this article is to determine the relationship betwee~ 02 delivery an 02 consumption as a sign of hypoxia. Fifteen patitents with septic shock were treated with intention to maximize the value of CI,D02 and V02. We compared the levels of these parameters between the survivors and nonsurvivors and found no significant differences after 24 hours. High levels of DO2 and V02 may not guarantee against tissue hypoxia in early stage of septic shock. ZJar~iic, Dj Janjic, Lj. Gvozdenovic, A.Komareevic. T.Petrovic, &Marjanovic, INSTITUTE OF SURGERY, Novi Sad, Yugoslavia Objectives: Evaluation and mutual comparison of clinical signs, laboratory data and microbiological monitoring in the patients with burn sepsis. Method: Retrospective analysis of the recorded data of all burn patients treated in our Department between January 1989 and December 1994. Specially attentions were given to data considering wound infection, positive haemocultures, positive urinocultures and characteristics of septic state. Results: Out of 372 patient there were 324 (87,09~) adults and 48 (12,9(3~) children. Almost two thirds of the patients (238 -63,97~) were males. The predominantly cause (68,759~) of children's burns was scalding b~Y hot liquids and flame burns ~63,97~) in adult patients. The most frequdntly species isolated from surface swat~ were Pseudomonas aeruginosa (64"1796 in adult patients) and Staphyloccocus epidermidis (78,57% in children). In only five patients (1,349~ 1 the haenmcultures were positive -Pseudomonas aeruginosa was isolated in three and Staphyloccocus aureus in two patients. Urine infection was diagnosed in 6,72% of all patients. The treatment protocol included use of Imipenem and polyvalent Pseudomonas vaccine again~ Pseudomonas aeruginosa and Vancomycin and aminoglycosides against Staphylococcus aureus. Total mortality rate in this group of burned patients was 6,98~, but the mortality rate caused of sepsis was low (I 34%) . Conclusions: Early detection of any signs of wound infection and symptoms of septic state is a foundation for prevention and treatment of burn sepsis. The burn sepsis could be reliable detected by continuously monitoring the patient's status and by systematic microbacteriological monitoring of the burned patients. HYPERDYNAMIC VASOPLEGIC SEPTIC SHOCK P.F. Laterre, P. Goffette, J. Roeseler, J.P, Fauville, A. Poncelet, P. Lonneux, M.S. l~eynaert. Dept. of Intensive Care, St. Luc Univ. Hospital, Brussels, Belgium. Splanchnic ischemia is described as a common feature of septic shock and could determine the development of MSOF. Therapy such as noradrenaline (NA) aiming at improving blood pressure is expected to worsen splanchnic ischemia by its vasoconstrictive effect and subsequent reduction in intestinal blood flow. Ob[ective: Evaluate the effect of NA on splanchnic blood flow. Material and method : In a patient admitted for variceal bleeding, ARDS and sepsis with positive blood culture, a fiberoptie catheter was positionned in the portal vein after recanalisation of its portosystemic stent shunt. Blood pressure (BP-mmHg) , CI, SVR, DO 2 (Vigilance ~ Baxter), V02 (Indirect colorimetry), arterial, mixed venous and portal vein blood gases, pHi were determined before (TO) and during (T1) NA infusion (0, 1 to 0, 19 hcg/Kg/min.) . Changes in splanchnic flow were assessed by changes in portal oxygen saturation (SP02) and arterio-portal oxygen saturation gradient (SaO, -SPOe Laterre, ,LP. Pedgrim, Th. Dugernier, V. Delrue, Ph. Hantson, P. Mahieu, M.S. Reynaert. Dept. of Intensive Care, St. Luc Univ. Hospital, Brussels, Belgium. Aim of the study : Prospective determination of plasma levels of in patients with SS and their correlation with the type of microorganism and outcome. Material and methods : In 19 patients (pts) with SS and severe sepsis, plasma levels of TNFtI, ILl-b, IL6 and IL8 were determined every 8 hours for 3 days and on day 7 after fulfilling the criteria of SS and severe sepsis. Results : In 9 pts, sepsis was caused by a Gram (-) microorganism, in 6 pts by a Gram (+) and in 4 pts no microorganism was identified. There were 12 survivors (63%) (S) and 7 non-survivors (37%) (NS) . Cytokines profiles and levels were not different between Gram (+) and Gram (-) sepsis. ILl-b levels were seldom elevated whatever the group studied. TNFot and IL-6 were significantly higher in NS than in S ( Objective: To evaluate the effects on the nitric oxide synthase inhibitor L-N~ HCL (546C88) on myocardial performance in human septic shock. Method: Septic shock was defined as severe sepsis with either persistent hypotension (mean arterial pressure; MAP<70mmHg) or the requirement for a noradrenaline (NA) infusion >_ 0.i ]tg/kg/min with a MAP _< 90mmHg. Cardiovascular support was limited to NA _+ dobutamine (DB), 546C88 was administered for up to 8 h at a fixed dose-rate of either 1, 2.5, 5, 10 or 20 mg/kg/h iv. During 546C88 infusion, NA was to be reduced and if possible withdrawn, whilst maintaining MAP above 70 mmHg and the cardiac index (CI) as clinically appropriate. Assessments were made at baseline (t = 0); at i h from the start of treatment (t = 1); and at the end of treatment (t = 8) with 546C88. Conclusions: 546C88 can restore systemic vascular tone in patients with septic shock enabling NA therapy to be reduced and/or removed. The CI tends to fall whilst LV performance is sustained over time. 546C88 is a novel vasoacfive agent for the treatment of septic shock, which is undergoing further clinical evaluation. Laterre, F. Thys, E. Danse, J.P. Pelgrim, E. Florence, Z Roeseler, M.S. R eynaert. Dept, of Intensive Care, St. Luc Univ, Hospital, Brussels, Belgium. Therapy aiming at improving blood pressure and cardiac index in septic shock (SS) might have deleterious effects on regional blood flow. Objectives : Compare the influence of volume loading (VL), Dobutamine (Dobu) and Noradrenaline (NA) on sushepatic oxygen saturation (SHOe) and SVOe-SHO, gradient in treated SS. Material and methods : In patients with SS, CI (Thermodilution) , DOe, SVO,. SHO,, SVOe-SHO e gradient and Lactate (L) were determined before (TO) and after (T1); VL, Dobu and NA. Results: In 5 patients with treated SS, 16 tests were performed (VL n=8; Dobu n=4; NA n=4 Method: Septic shock was defined as severe sepsis with either persistent hypotension (mean arterial pressure; MAP<70 mmHg) or the requirement for a noradrenaline (NA) infusion ~> 0.1 ~g/kg/min with a MAP _< 90mmHg. Cardiovascular support was limited to NA + dobutamine (DB), 546C88 was administered for up to 8 h at a fixed dose-rate of either i, 2.5, 5, 10 or 20 mg/kg/h iv. During 546C88 infusion, NA was to be reduced and if possible withdrawn, whilst maintaining MAP above 70 mmHg and the cardiac index (CI) as clinically appropriate. Assessments were made at baseline (t = 0); at 1 h from the start of treatment (t = 1); and at the end of treatment (t -8) with 546C88. Conclusions: 546C88 is a novel vasoactive agent that can sustain MAP in patients with septic shock, enabling NA support to he reduced and/or removed. There is a tendency for the CI to fall during treatment, which may be reflex in response to the increase in systemic vascular tone. 546C88 is a promising new therapy for septic shock, which will now be evaluated in a randomised, placebo-controlled safety and efficacy study. K. Guntupalli Objective: To evaluate the acute effects of the nitric oxide synthase inhibitor L-N~ HC1 (546C88) on selected indices of organ function in patients with septic shock. Method: Septic shock was defined as severe sepsis with either persistent hypotension (mean arterial pressure; MAP < 70 mmHg) or the requirement for a noradrenaline (NA) infusion --> 0.1 [xg/kg/ min with a MAP _< 90 mmIrlg. Cardiovascular support was limited to NA + dobutamine. 546C88 was given for up to 8 h at a fixed dose-rate of either 1, 2.5, 5,10 or 20 mg/kg/h iv. During 546C88 infusion, NA was to be reduced and if possible withdrawn, whilst maintaining MAP above 70 mmHg and the cardiac index (CI) as clinically appropriate. Indices of organ function were assessed at baseline (t = 0); at the end of treatment (t = 8); and 12 h after treatment (t = 20) with 546C88. Results. -median values (* assessment made at 8 h or when 546C88 discontinued). Conclusions: There was no appareut dose-dependent adverse effect on these indices of organ function either during or after exposure to 546C88. The plmelet count tended to fall whilst creadnine appeared to increase over time in all dose cohorts. This novel and promising therapy for septic shock will now be evaluated in a randomised, placebo-controlled safety and efficacy sludy. Pharmacokinetics of 546C88 in patients with septic shock preliminary results Z. Hussein, B. Jordan, C. Fook-Sheung, K. Guntupalli Objective: To evaluate the pharmacokinetics of the nitric oxide synthase inhibitor L-N~ HC1 (546cg8) given by continuous infusion for 8 h in patients with septic shock. Method: Septic shock was defined as severe sepsis with either persistent hypotension (mean arterial pressure; MAP < 70 mmHg) or the requirement for a noradrenaline (NA) infusion --> 0.1 ~tg/kg/min with a MAP _< 90 mmHg. Cardiovascular support was limited to NA • dobutamine. 546C88 was administered for up to 8 h at a fixed dose-rate of either 1, 2.5, 5,10 or 20 mg/kg/h iv. Plasma was collected from each patient over a 24 h period and analysed for 546C88. Pharmacokinetic parameters were derived from plasma concentration-time profiles using non-compartmental pharmacokinetic analysis. Results: The (Cm~ -maximum plasma concentration; AUC -area under curve; CL -plasma clearance; V,, s -steady state volume of distribution; t'/2 -plasma elimination halflife). Conclusion: The pharmacokinetics of 546C88 in patients with septic shock are dose-independent at infusion rates up to 2.5 mg/kg/h. At higher rates, clearance of 546C88 decreases without any marked change in volume of distribution. 546C88 metabolism may be partially saturable at dose-rates above 2.5 mg/kg/h. Obiectives: Investigate the effect of the NO synthase inhibitor, L-Nt-methylarginine HC1 (546C88) on the haemodynamics and survival rate in a conscious mouse model of endotoxin shock. Methods: Female CD-1 mice (25-35 g) were instrumented under gaseous anaesthesia (isofluorane, 2%) and connected to a swivel tether system for continuous monitoring of blood pressure and drug administration. Results: After 24 h recovery, endotoxin administration (E. col• 026:B6, 6-12.5 mgkg -1 i.v.) elevated the plasma concentration of nitrite/nitrate (NOx) and caused a progressive fall in mean arterial pressure (MAP) from 101 + 5 to 59 + 4 mmHg (n=5, P<0.05) at 12 h, with a survival rate at 24 h, 48 h and 72 h of 80%, 40% and 20% respectively. 546C88 administered as a 24 h continuous infusion (3 mgkg-th -t i.v., n=5), 4 h after endotoxin, inhibited the elevation of plasma NOx and attenuated the fall in MAP from 105 + 2 to 70 + 3 mmHg (n=5) at 12 h, with an improved survival rate at 24 h, 48 h and 72 h of 100%, 100% and 60% respectively. Conclusions: This study suggests that overproduction of NO is involved in the hypotension and mortality characteristic of septic shock. Inhibition of NO synthase using 546C88 represents a novel and promising treatment for septic shock. Cultures of E.coli (19,5%) and Candida(8, 3%) were olso received from autopsy material of children;P.aeruginosa,unspored anaerobes,Proteus Sp.,S.aureus,B.pneumonia were found in the few cases. In adults the spectrum of bacterioflora was mo~ re limited speaking about the number of species and cultures. In generalized forms of bacterial pyo-septic pathology a wider specific spectrum of causative agents was revealed usua fly with associations. E.coli and K.pneumonia played the leading role in children as well as in adults. In general,K.pneumonia (23,7%cultures) and common E.Coli(18,6%)prevailed according to the date of microbiological investigations of authopsy material in pyo-septfc pathology in 1994. Objectives: .In spite of all clinical exertion sepsis is still the reason for high clinica! lethality. This study is characterizing the group of patients which survived a septi~ shock. Methods: During a period of 12 months all surgical patients on ICU were registrated prospectively, More than 270 parameters for each of them were documented'daily in a paradox file. Results (see table 1): 20 of 286 patients fulfilled the criterion of a septic shock (R. Bone, 1991 ) , 11 of them died at the 2 lth day, while the surviving group of patients stayed almost 51 days at ICU. Obiectives: To compare the effects of 6 and 10% pentastarch solutions to a human albumin solution on oxygen delivery (DO2) in septic patients. Methods: This stud}, included 41 septic patients with fever (T > 38~ tachycardia flqR > 90/rain), tachypnea (RR > 20/min) or mechanical ventilation, leukocytosis (WBC>12000/mm 3) or leukopcnla (WBC<400()/mm 3) and a clinical source of infection, who required a fluid challenge. In each patient the pulmonary arterial occlusion pressure (PAOP) was < 12 mmHg. Patients were randomized to receive 400 ml of 4% albunun (N:I4), hydroxyethyl starch (HES -Mw200/D.S. 0.5) 6% (N:14) or t0% (N=I3); 33 patients were also treated with adrenergic agents. Results Cardiac index (C1) increased significantly only in 10% lIES (Table) Hemoglobin (Hb) decreased significantly at 40 min in the same group. There was not significant change in oxygen delivery ( DO2 ). Baseline CI Alb 3.7::1.3 (l'min/m 2) HES 6% 3.9=0.9 HES 10% 3. Polyneuropathy of the critically ill (PCI ) is a well recognized complication, acquired in the course of severe illness. We undertook a prospective study, to estimate the severity, extension and time of onset of PCI in a selected group of 25 patient with established septic shock ( Bone's criteria ). All patients received inotropic circulatory support and were mechanically ventilated. None received relaxants or aminoglycosides. PCI was diagnose 1% or administration of at least 1 ICU-dependent therapy)'. 1063 consecutive admissions aged < 18 years old were included. Overall, observed and expected mortality were in good agreement (p > 0.5). Between hospitals, crude mortality showed wide variations (mean 7.1%, range 1-10%). However, in each center, observed and expected mortality were similar (mean ratio 0.99, range 0.8-1.5). In tertiary care centres, severity of illness corrected mortality in high-risk patients was less than in non-tertiary care centres; paradoxically, in low-risk patients the opposite was found. Probably the large proportion of low-risk tertiary care patients suffering from severe, incurable chronic disease, explains the higher mortality in this group. This indicates that simultaneous assessment of circumstances of dying and of long term morbidity in similar future studies is imperative. The average proportion of efficient ICU days was 72%, however large variations between units were found (range: 22-95 %). In conclusion differences in mortality rates among pediatric ICUs were explained by differences in severity of illness. High efficiency rates in combination with adequate effectiveness, found in several centres suggest that admission and discharge decisions might be improved by a better selection of high risk patients requiring ICU-dependent therapies, especially in less efficient centres. Objectives: Previously published studies showed that serum lactate levels correlated with outcome of severe ill adult, 'We hypothesized that critically ill newborns are often incurred hypopeffusion manifested by elevated lactate levels. These initial blood lactate levels should be related to NICU outcome. Design: Prospective study with Ethical Comfnittee approval. Setting: The 14-bed Neonatal Intensive Care Unit of a University Hospital Material and method: A total of 209 consecutive outbem newborns admitted to NlOd from 01,10.1991 to 31.,19.1992 were enrolled to the study. Babies who died or were discharged from the unit within 48 hours of treatment Were excluded from the study, Mean birth weight was 2040g (+/-820R), mean gestatational age was 35 weeks (+/-3.5 wks), mean age at the admission was 50 h (+/-110hi. Multiple (~_2J organ system failure occurred Jn 38.3% of babies at the admission./~tertal lactates were measure/at the admission, among 22 -26 hour and 46 -50 hour of N[C'LJ therapy. Outcome was defined as a mortality and length of NICU stay. Results" Survival rate was 68.4%, mean length of NICU stay for survivors was 17.8 days (+/-15.1 day). We found high lactate levels at the admission in 82.8% babies (~9.2% with levels above 5.0 retool/I). The mean arterial lactate concentrations for nonsurvivors were signiftcahtly higher than for survivors durin~ consecutive da~ as follows: Objectives: The purpose of our research was to analyze the frequency of bronchial asthma (B.A.) exacerbations in pregnant women and health status of infants. Methods: The research was based on the epidemiological investigation and prolonged observation of 23 pregnant women with B.A. during the gestation period. Remission of B.A. before the pregnancy in excess of 5 years was recorded in 9 patients (39.1%), 6 patients (26.1%) reported a 2-3 year remission and 8 patients (34.8%) had a remission lasting less than 12 months before they became pregnant. Results: Seven patients (30.4%) developed medium attacks in the second half of pregnancy, four patients (17.4%) experienced light attacks of B.A. Asthma attacks were most frequently caused by acute respiratory diseases and stress factors. In two cases with grave manifestation of B.A., the pregnancy ended in abortion within the first 16-18 weeks due to the frequent and heavy choking attacks. To fight B.A. attacks, five patients used 132adrenomimetics (salbutamol, becotid) in sprays, six women were administered theophyllinum and salbutamol in the form of tablets during 1-2 weeks. A significant portion of pregnant women with B.A. (78%) exhibited frequent complications during pregnancy (toxemia, late gestosis, threat of miscarriage). Our findings prove that babies born from women with B.A. of domestic and pollen origin had a low body weight (2800-2500 gr), functional immaturity and chronic antenatal and intranatal hypoxia twice as often as the infants born from healthy women without allergic background. Conclusions: Preventive treatment of women with B.A. prior to pregnancy is required to maintain a stable remission of the disease, which is a key to having healthy children delivered by mothers suffering from B.A. Introduction. Intracerebral hemorrhage (ICH) is a common event in human prematudty, affecting about 25% of newborns weighing below 1500 g who are born before 32 weeks of gestation, However, little is known about the pathogenesis of ICH with exception of the prematurity of the brain itself, (birth) trauma, and asphyxia. The postischemic production of oxygen free radicals (OFR) dudng reoxygenation as a cause of brain damage has been demonstrated in 9 animal research. Since almost all preventive antioxidant activity of plasma is associated with ceruloplasmin and transferdn we investigated the association of such iron-oxidizing resp. iron-binding proteins and ICH. We could demonstrate significantly reduced levels of both, iron-oxidizing and iron-binding proteins, in premature asphyxiated newboms pdor to development of ICH. An increase of suparoxide after hypoxia in the presence of iron ions facilitates the formation ofthe highly reactive hydroxyl radicals. Our data support the theory that ICH may be caused by OFR, which can damage any sensitive tissue including growing endothelial cells. The estimation of transferrin-saturation and measurement of ceruleplesmin levels might help to identify an infant at dsk before the onset of ICH. With the new Medos | HIA-VAD | cardiac assist system the missing tool in the armamentarium of cardiac surgeons is available in two pediatric sizes: I0-ml and 25-ml pump volume. The right sided pumps are 10 % smaller for biventricular use. Between February 1994 and May 1995 we implanted this assist system in 6 children. The indications and demographics are indicated in the following table (Left ventricular assist device-LVAD, right VAD-RVAD univentricular VAD-UVAD, post cardiotomy cardiac failure-PCF, dilated cardiomyopathy-CMR Bland White Garland syndrome-BWG, Tetralogy of Fallot-TOF, hypoplastic left heart syndrome-HLHS). Objectives: Evaluate tile effeci'of Inhaled Nitric Oxide (NO) as puhnona] T vasodilating agent ill tile posloperalivc period after correclion of congenital heart defects in 3 infant. Patient n.l: 4 kg, 7 lnonlhs, Down syndrome undenvcnl rep~fir of atrioventricular septal defect (AVSD). After surgery the puhnonary arlcry pressure (PAP) slowly rose to tile syslemic dcspilc tnaximal eonvcnlional fllerapy (fentanyl 4 mcg/Kg/h, hypocapnia of 27 mmHg and metabolic alcalinization). NO was delivered into tile inspiratory branch of!be breathing circuit at 10 ppm, and the gas aoalyser for NO and NO2 (Polylron Dmger) were situated at the espiratory branch, a rapid dccrcasc of PAP Io I/3 of systemic was obtained with a dramalic improvement. NO was continued at 5 ppm for six days and the baby was exlnbated If! days after surgery and discharged from the ICU 5 days after. Patient n.2 : 4.5 kg, 6 monlhs, onderwen! repair of AVSD. The day after surgery the systemic oxygen salnralion was 76% wilh a PAP at 75% of systemic. Two hours of c01wenlional therapy failed 1o improve Ihc patient and NO administration was slarled at 10 ppm. SO2 dramatically incrcased to 95%, but the PAP dropped only to 50% of syslemic. Nevertheless Ihe clinical conditions improved and the NO administration could be reduced at 5 ppm in the following 6 days. She was extubaled 8 days after surgery and discharged from the ICU 20 days after. Patient n.3:12 kg, 3 3'ears. underwen| hearl tral~splantalion for congenital heart disease with moderate hypoplasia of pulmonary arlcrics. At the end of cardiopulmonary bypass the transpnlnlonary al~erio-venoas gradient yeas higher than 7 mnfflg and we speculaled !hat w'ls due to a degree of puhnonary vasocostrictiont. The nsnal dose of NO was otilised, however no significant modilicalion of pulmonary pressure or systemic oxygen saluralion was noled, and after 1 h NO was discontinned. Tile palienl was carried Io the ICU with maximal inotropic support, extubated after 4 d;b's and disclmrged from the ICU after 15 days. In all patient no major adverse effect relaled to NO admilfistration ",','as holed. Conclusion: In our experience NO ms a pulmonary vasodilaling agent is effective and easily adjustable to tile palienls requiemenls, however its use remains limited ill those palienl ill whoIn tile alnonll! of fixed InlllllOJlIIFy vascular resistance is predominanl. We report the use of ECMO support in two unusual cases of severe tracheal disruption in which it had become impossible to achieve adequate ventilation. Case 1: Severe tracheal laceration due to aspiration of a share forelan bodv: A previously healthy 13 month old toddler was referred for ECMO following aspiration of a porcelain foreign body (with razor sharp edges) which had become embedded in the right mainstem bronchus with massive extrusion of air. This was removed on veno-arteda[ ECMO support, as the patient was unventilatable prior to bronchoscopy due to ongoing airieak. ECMG was continued after bronchoscopy to permit airway healing without the presence of an endotracheal tube. Unfortunately, an extensive pulmonary haemorrhage on day 4 of ECMO necessited re-exploration of the airway. This revealed a posterior tracheal tear from the cricoid to the middle of the right lower lobe. Following repair the patient was left on ECMO support together with high frequency oscillation ventilation (HFOV), the latter being used to minimise potential aideak and maximise alveoli recruitment. ECMO was weaned after 17 days (420 hours) -the patient was extubated 7 weeks later. Case 2: Tracheal wound dehiscence due to seosls -tracheal transelant on ECMO: A 4 month old infant with a c[inically significant congenital long segment tracheal stenosis and left pulmonary artery sling underwent resection of the stenosis, followed by primary reanastomosis. This was complicated, 5 days later, by severe mediastinitis and complete dehiscence of the anastomosis. An autologous pericardial patch was used to repair this, however, the tracheal wound again dehisced 4 days later making mechanical ventilation impossible. In view of ongoing sepsis and a severely disrupted trachea ECMO was the only possible form of support. Following resolution of the local sepsis (4 days) a definitive procedure in the form of a tracheal homograft (transplant) was undertaken on ECMO. The patient was managed on ECMO and HFOV for a further 3 days, the HFOV being used to optimize rapid lung inflation. Unfortunately this patient died 9 months after weaning from ECMO due to complete disintegration of the homograft, which was not deemed reparable. Conclusions: 1) ECMO can be used in the acute management of oxygenation when there is major airway disruption making mechanical ventilation impossible. 2) HFOV was a useful adjunct in aiding recruitment of lung volume on ECMO in these two patients. Backoreund: Persistent pulmonary hypertension of the newborn (PPHN) consists of a heterogenous group of diseases ranging from transient reversibTe pulmonary hypertension to fixed primary malformations of the lung (primary pulmonary dyspfasia-PPD). Inhaled Nitric Oxide (INO), a selective pulmonary vasodilator, has been proposed as a treatment for severe PPHN. Obiective and Methods: INO was administered to 23 near term neonates with severe persistent PPHN, oxygenation index > 25 and echocardiogrephic evidence of pulmonary hypertension, in order to further determine the clinical role of INO in the treatment of PPHN. The response to INO was also analysed retrospectively to examine whether this could be of diagnostic value in differentiating at an early stage patients with reversible from fixed causes of PPHN Results: Twenty one of the 23 patients studied responded to the initial trial of 1NO (20ppm x 20 minutes), as defined by a greater than 20 percent improvement in Pad2 as well as a fall in the el to < 40. These 21 patients were continued on INO therapy, with 3 patterns of response emerging: Pattern 1 babies (n=8) continued to show a sustained response to INO and were successfully weaned from it within 5 days -all survived. Pattern 2 babies (n=9) failed to sustain their response to INO over 24 hours, as definded by a rise in the el > 40. Six survived, five with ECMO. Pattern 3 babies (n=3) had a sustained dependence on INO for 3 -6 weeks. All three died and lung histology revealed severe primary pulmonary dysplasia (PPD). Patients with PPD (pattern 3) not only required INO for longer periods of time than did the sustained responders (pattern 1), but also required significantly higher doses of INO We report on the air transport of 32 paediatric intensive care patients. These transports fall into three categories: 1) Retrieval of critically ill neonates and paediatdc patients referred for either ECMO or inhaled nitric oxide (INO) (n = 12). One patient was transferred on INd. Mean transfer time 2.2 hours (SE +0.6hrs). 2) Long distance international transport using chartered aircraft (n = 11). The indications for these transfers included both urgent retrievals for cardiac surgery and semi-elective transfer of stable patients back to their referring unit following treatment in tertiary centres. Mean transfer time 4.4 hours (SE +0.4hrs) 3) Long distance international transport using commercial aircraft (n = 9). Indications for transfer were either semi-elective retrieval for tertiary treatment or the return of stable chronically ventilated patients to their referring hospitals. Mean transfer time 14 hours (SE _+1 .fhrs, longest 24 hrs). The transport team consisted of a paediatric intensive care doctor of at least registrar grade and a registered sick chidrens nurse with intensive care experience. The administrative components of the transfer (ambulances, airlines, customs) were managed in collaboration with companies specializing in air ambulance transfers. Outcome: All the patients were safely transported to their destination without mortality or morbidity. Complications durino transfer ir~lv~; 1) Patient complications -semielective endotracheal tube change and central access needed in the only patient brought to the commercial aircraft by the referring hospital (all others retrieved directly from referral hospital), seizure in patient with known encephalopathy, severe cyanotic spells in patient with Fallots Tetralogy who was retrieved for urgent surgery for this indication 2) Mechanical compfications -ventilator failure, incubator battery failure, oxygen regulator failure -all occurred with equipment sent from referral hospital, this was unfamiliar and unchecked by our transport team -it was not the decision of the transfer team to use this equipment on this single occassion. 3) Administrative complications -confiscation of incubator battery by airport security police, excessive delay by custom officials (2 hours) in the airport. The incidence of such problems were felt to be low and unpredictable. In conclusion: mechanically ventilated paediatric patients can be safely transported on both chartered and commercial airlines. These transports are best accomplished by trained intensive care medical and nursing staff with the backing of an air ambulance organization competent in arranging the necessary administrative details. It is essential to use your own equipment and to retrieve the patient _directly from the referrin(] hospital to minimise ootential complications. OUR EXPERIENCE WITH ANAESTHESIA FOR PAEDIATRIC ELECTROMYOGRAPHY _W_._Pla_ti_k_a_n_o_v, R.Eousseff, K.Pavlova, D.Marinova Dpts. of Anaesthesiology and Int. Care and Clinika] Neurophysiology, Med. University, Pleven, Bulgaria ~)_b_j#~ti_v~. To t~st a " heavv sedation " regimen of anaest-es~a for the purpose of paediatric electromyography D#s~gil~ Non-randomized,non-blinded human trial in the seting of an uriiversity hospetal. _M_a_t_eri_a_Is_a_nd_ M_e_th_od_s_.110 children,ASA I-If,median age 6 years,range 9-13 who undervent eleetrcmyography required anaesthesia. They recieved low-dose Ketamine + I~iazepam or Midazolam via musculary route( 25 children,age 0 -3 yrs,Ketamine 2,5 mg/kg, Diazepam 3-6 mg total dose ) or per os ( 85 children,Ketamine 5-7 mg/kg,Diazepam 0,3 mg/kg or Midazclam 0,4 -0,5 mg/kg ) _Resu_l_t_s. 20 -25 minutes after medication a state of heavy sedation with weak spontaneos and stimuli-provoked movements was achieved in all children, that lasted 30 -60 minutes and allowed adequate needle EMG and nerve conduction investigation. 11 children recieved additional 0,6 -1,0 vol.% Halothane during the placement of the needle. Non -invasive blood pressure , breath and heart sounds and Hb Sad2 by pulse oxymetry were monitored.None of the older children disclosed memories of pain when asked after they regained adequate verbal contact.No complicationes were observed. ANTENATAL MATERNAL STEROIDS REDUCE THE RISK OF PERIVENTRICULAR-INTRAVENTRICULAR HEMORRHAGE IN VERY PREMATURE NEONATES TREATED WITH NATURAL SURFACTANTS. I.Apostolidou, C.Papagaroufalis, G.Touloumi, M.Xanthou, N.Kalpoyannis A' and B" Neonatal ICU "AG. SOPHIA" Children" s Hosp. Athens, Greece. Dept of Hygiene and Epidemiology, Athens University, Greece. Obiectives: The aim of the study was to evaluate the association of periventricular-intraventricular hemorrhage (P-IVH) in surfactanl treated premature neonates with pre-and postnatal variables. Methods: The population of the study was 88 neonates admitted during the years 1990 to 1992, with gestational age _< 32 weeks and severe respiratory distress syndrome (RDS) (mechanical ventilation and arterialalveolar oxygen tension ratio (aJAPO2) <0.22), who received rescue therapy of at least two doses of natural surfactants (Alveofact or Curosurf) and examined with ultrasound and/or autopsy for the presence of P-IVH (Papile's classification). The examined factors in each neonate were the following: gestational age, birth weight, sex, multiple pregnancy, antenatal maternal steroids (complete and incomplete course of betamethasone), a/APO 2 before the administration of the 1st dose of surfeclant, delivery, Apgar score at 5min, type of surfactant, pneumothorax and patent ductus arteriosus. The statistical methods used were x 2 and one-way analyses of variance followed by logistic regression medels, Results: The incidence ot P-IVH was 31.8%. Three factors were found to have an independent relation to P-IVH (final logistic regression model): gestalional age, a/APO 2 before surfactant administration, and antenatal administration of maternal steroids (complete and incomplete courses). For every 2 weeks of lower gestational age the neonates had an almost doubled associated risk of P-IVH (OR: 1.92, 95% C1:1.14, 3.22). For every 0.02 on average decrease of a/APO 2 before surfactant administration the risk of P-IVH in the neonates was 1.27 times higher (95% CI: 1.02, 1.58). The neonates whose mothers received antenatally steroids had only one tenth of the risk of P-IVH of the neonates whose mothers had not (OR: 0.10, 95% CI: 0.01, 0.82). Conclusions: Our results suggest that the antenatal administration of maternal steroids, even less than 24 hours before delivery, reduce the risk of PqVH in very premature neonates treated with natural surfactants, whereas the small gestational age and the lung immaturity still remain the main risk factors tor the development of P-IVH. We analysed retrospectively the management of 103 (51 boys, 52 girls) accidental ingestions of foreign bodies in children (mean age : 2.8 years, range : 7 months-10 years). No child had ingested more than 1 foreign object. The majority of the ingested foreign bodies were : coins (n : 44), toy parts (n : 11 ), jewellery (n : 3), batteries (n : 16), "sharp" materials such as needles and pins (n : 21), "large" amounts of food (n : 8). Impaction of food occurs more frequently in children after oesophageal reconstruction in cases of oesophageal atresia. Although according to literature "Coca-Cola" is reported to be effective, this was not seen in our experience. 28/103 patients had minor transient symptoms at the moment of ingestion, such as retrosternal pain. Only 4 children experienced severe manifestations (cyanosis, dysphagia). In these children, endoscopy revealed oesophageal and gastric erosions. Children were seen at the emergency ward within a few hours after the accident ( mean : 3 hours, range 20 min. -28 hours). Chest and/or abdominal X-ray was performed as first-line investigation ( 93/103 objects were radio-opaque), and revealed an (unexpected) oeeophageal impaction in 6 children. In 87/103 the foreign body was in the stomach. Batteries, sharp objects and objects trapped in the oesophagus were removed, either by endoscopy or by magnet-extraction whenever possible. The outcome of the patients was excellent. No complications were observed. Extraction is recommended in symptomatic patients, and whenever the foreign body is trapped in the oesophagus, or if the foreign object is "sharp" or a battery. Objectives: Two strategies were used for management of malignant diphtheria in children aged from 0.5 to 13 years. Methods: Protocol N1 consisted of intravenous administration of diphtheria antitoxic serum, prednisolone (2 mg/kg bw/day), plasmapheresis and supportive care. Protocol N2 included the use of antitoxic serum against the background of high-dose dexasone (2-3 mg/kg bw/day), hemocarioperfusion and a preventive use (before the clinical manifestation of myocardial damage) of inotropic medications, inhibitors of angiotensin-converting enzyme and pentoxyphylline. Each of protocols included the monitoring of serum toxin (diphtherin) levels. Results: The group of patients treated according to the protocol N1 consisted of 17 children with malignant diphtheria, 11 of them with severe malignant diphtheria (grade 2 and 3). All patients exhibited the circulation of toxin during at least three days after the start of treatment. All 11 patients with severe grade of disease demonstrated heavy cardiovascular disturbances associated with malignant diphtheria. Of the 11 children in the group died seven. The children of the second group were treated according to the protocol N2. Out of total of 22 patients of this group. 11 patients had severe malignant diphtheria. In all children a significant reduction in serum toxin level was revealed after hemocarboperfusion. In all but one case the satisfactory control of cardiovascular function on was achieved. Of 22 children admitted to the trial 21 survived, one child with malignant diphtheria of grade 3 and congenital filbroelastosys of the left ventriculum died. The severity of neurological complications was similar in each of groups. Conclusions: The use of hemocarboperfusion, high-dose dexasone and early prevention of heart failure as a adjunct to the standart treatment has been shown to be of benefit in the management of malignant diphtheria. T. Schaible, I. Reiss, J. M611er, L. Gortner Med. University of LQbeck, Children's Hospital, Kahlhorststr. 31-35, 23538 L~beck, Germany Surfactant therapy seems a promising approach for the treatment of the biochemical and biophysical abnormalities of the pulmonary surfactant system in severe ARDS. Patients and Methods: Over a 18 months period 10 non-neonatal pediatric ARDS patients (age 1-38 months) in a "pre-ECMO"-situation (OI 40 over 4 h) were treated with bovine surfactant (Alveofact| The underlying conditions-of ARDS were pneumonia (5), sepsis (2), immunosuppression (1), near drowning (1), neurogenous ARDS (1). A total of 20-120 mg/kg b.w. was applied in several fractions. Before surfactant therapy, we first tried different ventilation (best PEEP-finding, inversed I/E-ratio, HFO-ventilation) while monitoring the pulmonary mechanics. For hemodynamic stabilisation both norepinephrine and epoprostenol were used to optimize pulmonary perfusion for max. 4 hrs. If there was no improvement of the OI by at least 10, further treatment with surfactant was initiated. In addition to surfactant all patients received a treatment with dexamethasone of 1 mg/kg in 2 doses. Patients with no benefit (OI remained unchanged or increased within the max. 2-4 hrs) were taken on ECMO. Results: Nine patients improved within 4 hours after surfactant therapy: The OI decreased from a level of 41 (mean, range 22-100) before our treatment to a level of 16 (mean, range 6-60) thereafter. In 6 patients we were able to continue the positive effects of our treatment and they could be weaned of the respirator within 3-10 days. The other 3 patients got worse despite respiratory improvement, they suffered of multiorgan failure of more than 3 organ systems. The last patient did not benefit from surfactant, he had to be put on ECMO, but died because of a complication (hemopericard)after 10 days. The autopsy of the ECMO-patient showed a pulmonary fibrosis, but the other 3 death were not due to pulmonary failure. Conclusion: A different sequential ARDS treatment integrating surfactant therapy can reduce the number of patients requiring ECMO. But ECMO as a therapeutic tool should be available in centers involved in ARDS treatment. L.Blindl, T.P.Le, H.Weinzheimer, Centre for Paediatrics, University of Bonn, Germany Selective reduction of elevated pulmonary vascular resistance by inhaled Prostacycliu (PGI) has been reported in adults with acute lung injury, neonates with persistent pulmonary hypertension and in one infant with idiopathic pulmonary hypertension. We report on the effect of aerosolized prostacyclin in two children with secondary pulmonary hypertension. Patient 1: In a boy with Down's syndrome an AVSD had been surgically corrected at 11 month of age. At 5,6 yr of age a catheter examination revealed a pulmonary vascular resistance of 70 % of systemic vascular resistance in room air and at an Fin2 of 1.0. Prostacyclin (0.5 mcg/ml) was administered with a jet nebulizer at an Fin2 of 0.21. PVR declined to 0.4 systemic vascular resistance and returned to baseline after stopping PGI-inhalation. Subsequent intravenous infusion (5 ng/kg rain) had to be stopped after 5 minutes because of systemic arterial hypotension. Patient 2: A 8 month old male infant with bronchopulmonary dysplasia developed suprasystemic right ventricular pressure inspire of therapy with oxygen and nifedipin. While he was spontaneously breathing 60 % oxygen via face mask PaO2 was 37 mmHg, arterial pH was 7.35. Systolic arterial pressure was 85 mmHg, a RV-RA gradient of 100 mmHg was measured by CW-Doppler. While FiO2 was maintained aerosolized prostacyclin was administered over 30 minutes. RV-RA gradient was 70 mmHg, systemic blood pressure 75 mmHg, PaO2 58 mmHg. Two hours later nitric oxide (20 ppm) was inhaled at an FiO2 of (3,6. RV-RA gradient declined from 100 to 65 mmHg, systemic systolic blood pressure remained stable at 105 mlnHg. Discussion: Sporadic experience shows that aerosolized prostacyclin selectively reduces elevated pulmonary vascular resistance in some patients. In patient 2 the poor response to inhaled PGI1 compared to inhaled nitric oxide may be explained by the fact that the action of PGI is not independent from endothelial function, limiting it's effect in severe vascular disease. During the last two years (1993-94), 51 infants weighing less than 1000gr. admitted to our referral unit. Thirty four of them (67%) survived, (28% of infants weighing 500-700g and 78% of infants weighing 701-1000gr survived) for the years 1986-87-88 the survival of these infants was 53% and for the years 1976-77-78, 14% (p<0.01). We analyzed the perinatal and neonatal factors influencing the outcome of these infants. The comparison among neonatal survivors (1) to neonatal deaths (2) shows: Gestational age: 27.6 w (1) to 26.4 w (2) (S). Birth weight: 923.5g (1) to 724.7 (2) (S). Apgar score: 7,90 (1) to 7.56 (2) (NS). Presentation and mode of delivery: Breech presentation is associated with higher incidence of neonatal deaths. I.V.H. (at the age of 8 weeks): No one of the survival infants had evidence of I.V.H. Respiratory problems: Intubation, at the admittance of the infants 32.3",,(1) to 95% (2) (S) Use of surfactant: 70% (1) to 95% (2). BPD observed in 62% of the babies and only one was dependent on oxygen at home. Antenatal Betamethasone was given in 20% of the mothers. In conclusion: 1) A great improvement in the survival rate observed in these infants the last 20 years in our unit. 2) Factors with positive effect are increasing gestational age and birth weight, the absence of I.V.H. and the use of surfactant. The breech presentation and the severe respiratory problems increase the incidence of death. Animal experiments demonstrated, that brain temperature determines the amount of neuronal damage caused by hypoxia and that mild hypothermia may have a protective effect. Until now there is no method described and evaluated to measure brain temperature in neonatal intensive care units. We non-invasively measured brain temperature analogues, nasopharyngeal (Tnasoph) and zero-heat-flux temperature (zhT) at the temple whereby under zero heat flux surface temperature represents deep head and thus brain temperature. The aim of our study was to investigate the practicability of the method, the relationship of the two brain temperature analogues to rectal temperature (Trect) and their dependence on insulation, thermal environment, body activity and time course. We investigated 19 healthy preterms less then 2 weeks postnatal age (gestational age 315 +_ 2.1 wks; x + SD, weight 1653 +_ 370 g) in an incubator. Tnasoph was measured by a thermistor within a feeding tube, advanced to the nasopharynx, zhT temple by a thermistor and a heat flux transducers both covered by an insulating pad, and Trect Thermal environment was characterised by operant temperature (Tair .0.4 + Twall 0.6). Body activity was video taped. Measurements were performed during the following interventions: i/ insulation increased by turning the temple with sensors onto the mattress (15rain). ii) insulation increased by a cap (30 min), iii) 30 min after its removal, iiii) increased operant temperature by 1.6 + 0.5~ (60min). Results: Seven children with EA had a gasless abdomen, the endoscopic procedure excluded (6) or diagnosticated an upper pouch fistula (1). In patients who suspected "H" fistula (2) broncoscopy has strong advocated method to make diagnosis and established cervical approach. From July 1992 14 newborns with EA and lower pouch TEF received a selective transtracheal incannulation. We were not able to proceed just in 1 case with congenital subglottie stenosis. In these patients we provided gastric drainage by radiopaque and flexible 3-4 French catheter. The knowledge of the precise anatomic position of TEF consent to adjust the tip of the endotracheal tube in order to achieve best ventilation. The presence of the catheter through the fistula helps the surgeon to identify, it quickly. No complications were correlated to the procedure and no babies had early pneumonia. Alimentary continuity was achieved in all patients (30 primary anastomosis, 2 resections of TEF, 6 oesophagocoloplasty and 1 died with gastrooesofagostomy). The late mortality 7.7% (3) was only directly related to the severity of associated malformations. Conclusion: The advantages of this technical approach are unquestionable for the anaesthesiologist and the surgeon. In our experienc e the procedure improves perioperative management of babies and appears to be safe. Relation between cytokines, prethrombotic markers and endotelial injury markers in children with septic shock Objectives: To establish the relationship between cytokines (TNF, IL-1, IL-6) prethrombotic markers (D.D., PCam) and endothelial injury markers (TM, uWF) in pediatric patients with sepsis and bacteriemia without shock, and patients with septic shock. Design and methods: Prospective study, 18 children (9 months-16 years) were admitted in our PICU in 1994 with the following diagnosis: bacteriemia (4) sepsis (4) and septic shock (10) according to Jacob's R F criteria. Measurements: IL-1, IL-6, TNF, TM, vNF, D.D. PCam and routine laboratory data on admision, 12, 24, 48 hours and on discharge. The PRISM (Pediatric Risk of Mortality Score) was also recorded. Results and conclusions: Two patients in the septic shock group died. Significant differences were found between non-shock and septic shock patients in relation to Tm, DD, PCam, IL-2, IL-6 and TNE High levels of TNF and IL-6 are closely associated with the severity of septic shock with purpura in children. Low levels of PCam on admission were associated with severe shock. who underwent open hea~nt surgery, hypervotaemia with or without oliguria was the most frequent reason to start PD (77%). In 10 patients PD lasted less then one week and there were no complications; in 3 patients it lasted 13 -29 days (one child had a peritonitis). Instillation of dialysis fluid into the peritoneal cavity was associated with a significant increase in central venous pressure. There were no significant changes in cardiac output or arterial oxygeu saturation. In all patients PD dhnJnished fluid overload or improved the metabolic status. 5 patients (38%) survived the postoperative course and all had complete reintegration of renal function. Conclusion: PD is a useful method to treat the fluid overload and acute renal failure in paediatric patients following open heart surgery with file effects of little importance on the cardiovascular fimction. Obieetives: With the marketing of computerised systems for lung function testing in newborns, there has been an increasing interest in clinical approaches. Percentile curves of pulmonary parameters permit an appropriate and clinically useful interpretation. However, the manual evaluation of the results using different curves is an impractical technique. Therefoi'e a computer programme was developed. Methods: The percentiles (5%, 10%, 50~ 90%, 95%) of the most important pulmonary parameters were determined non-parametrically in 6 weight-classes. For the calculation we have taken results of our own as well as other laboratories using a meta-analysis of reference studies. In all, individual data of 300-600 healthy newborns ageing between 1-28 days were collated. Using these percentiles, for every parameter in relation to the body-weight the cumulative distribution was calculated approximately using piecewise linear and exponential functions. As shown in the figure the results of computing are represented numerically as well as graphically and can be included in the patient report. Conelusions: Clinic~d experiences with the programme have shown that representation of all measured parameters on standardised 100% scales allows an easy interpretation at first sight and improves the detection of pathologic patterns in the parameters. ")Supported by BMFT, FP "Risikoneugeborene" PRISM (Pediatric Risk of Mortality) score is a well known, already validated scoring system that quantifies severity of illness based on 14 routinely clinical and laboratory variables measuring physiological instability. Once computed the score by summing up the weights corresponding to the most abnormal value recorded during the first 24 hours, the overall risk of mortality can be predicted by using the coefficients estimated by a logistic regression where PRISM score is the main independent variable. (Pollack MM et al, -Pediatric risk of mortality (PRISM) score. Crit. Care Med. 1988; 16:1110 -1116 . To assess the applicability and validity of PRISM in the Italian setting we launched out a prospective data collection in a sample of 33 pediatric ICUs. Measures of calibration (goodness of fit statistics) and discrimination (receiver operating characteristics and area under the ROC curve) are planned to be adopted in the cohort of patients recruited during 1 year period. As the validation study started on July 94, data collection is still on going and validation analyses will be carried out on July 95. Up to now 23 centers recruited 1116 cases. At present, characteristics of the sample recruited are the following: most of the patients were male (62%); the mean age is 3 years with 30% of patiens having less than 30 days; more than half were medical cases (59%) admitted from emergency room or from hospital floor (51%); 52% cases were admitted with an organ failure while 48% to be intensively monitored. ICU-mortality was l 1%. The paper will present final results of calibration and discrimination analyses that will be carried out in the whole sample and across subgroups known to differ in terms of clinical relevance and prognosis. If calibration and discrimination assessment will produce not satisfactoty findings, a customization of the current coefficients will be made allowing a formal comparision of previous and new parameters. JF Riera-Faneao, M Wells, J Lipman. Baragwanath Intensive Care Unit, University of the Witwatarsrand, South Africa. [Background The PRISM score is designed to assess the likelihood of death in IPaediatdc ICU patients, using only acute physiological disturbances, age and [operative status to predict mortality. There is no evaluation of chronic health status, [including malnutrition. This may significantly affect its ability to accurately predict outcome in a population where malnutdtion is common. Aim To determine the influence of nutritional insufficiency, as indicated by a low weight-for-age on outcome prediction by PRISM. Patients & Methods We analysed PRISM, weight and demographic data co ected prospectively from 1528 consecutive paediatdc ICU admissions over a 6 year pedod. A proportional weight (Pwt) was calculated as a percentage from the 50th centile of the WHO weight-for-age growth charts. The Pwt was compared for survivors and nonsurvivors, and mortality compared for Pwt categodes 0NHO Wellcome classification). Multivariate statistical techniques Were used to identity associations with non-survival and to develop a modified logistic regression equation including a measure of I nutdtional status. Receiver operating characteristic (ROC) analysis was performed including and excluding patients with low Pwt for the odginal and modified equations. Results Non-survivors had a lower weight than survivors (6.8kg and 8.3kg medians p = 0 63) a lower Pwt (85% and 90% medians p = 0.6"0. The incidence of malnutdtion , in our ICU population was 33%. The mortality of manoudshed patients was' significantly increased (p = 0.001), with a good correlation with the degree of malnutrition. The accuracy of PRISM was significantly improved when malnourished patients were excluded from the analysis (ROC value increased from 0.72 to 0.79). ! Logistic regression and discriminant analysis identified a significant association between PRISM, Pwt and outcome; age and operative status were not significantly related to mortality. The use of a modified equation including the raw PRISM score, Pwt category and age can significantly improve the discriminatory power (Az dm/elopmental sample 0.82, Az validation sample 0.81). The modified formula is: legit = -2.864 +0.134*PRISM score -0.006*age + 0.463*weight category, where the probability of mortality is exp(Iog/t)/1 + exp(IogiO. Discussion Although we can improve the prediction of mortality by a modified or recelibrated formula, this still does not compare with the reference PRISM population. The need for validation of the score itself, in the association with outcome of the acute physiological variables themselves, is thus apparent. We conclude that while the odginal PRISM formula can be improved significantly, a modification of the basic variables in this and other third wodd populations may be essential. A high incidence of malnutrition is an independent risk factor of mortality, and an important cause of the poor discriminatory performance of PRISM. In order to improve the accuracy of PRISM, nutritional status should be taken into account. Objectives: To assess the value of inhaled NO to differentiate between pulmonary vascular constriction or fixed anatomical obstruction. Methods: We assessed the response to 40 ppm inhaled NO in 12 patients(9 M, 3 F, median age 4.5 months, range 1day to 17years) with signs of increased pulmonary vascular resistance, There were 5 pre and 7 postoperative patients. Patients were divided into responders(+) or non-responders(-). A positive response was defined as a 20% reduction in pulmonary arterial pressure and pulmonary vascular resistance(PVR) or in the presence of a left to right shunt, a fall in PVR accompanied by increasing pulmonary blood flow. left atrioventricular valve atresia + 12 MUSTARD PAT: pulmonary atresia VSD: ventricular septal defect ASD: atrial septal defect PDA: patent ductus arteriosus TAPVC: total anomalous pulmonary venous connection The responders(7/12) were characterised by left to right shunts or pulmonary venous hypertension(4/7). Patient#11 was weaned from ECMO with inhaled NO. Patient#2, without congenital heart disease, underwent a lung biopsy which confirmed reversible pulmonary vascular changes. Patient#1 had a pulmonary hypertensive crisis which responded to NO. All non-responders(5/12) had evidence of anatomic obstruction to pulmonary blood flow (#7,10,12)or a low PVR(#8) on subsequent cardiac catheterisation. In patient #3, lung biopsy confirmed severe obliterative vascular disease. Conclusions: Inhaled NO appears to be an effective pulmonary vasodilator. A failed response may be evidence of either irreversible pulmonary vascular disease or a residual anatomical obstruction which may be surgically remediable in the postoperative cardiac patient. Therefore, inhalation of NO may be a useful diagnostic test to differentiate between fixed anatomical obstruction and reversible vasoconstriction. RESULTS: During these 18 years, the incidence of SDRA was 1.3% of the total of admissions. The most common etiology was meningococcic septic shock. Since 1989, there is a decrease of its incidence. (from 24% to 11%) and an increase of pneumonia and immtmodeficiencies. Mean age of our patients was 2,7 years (61% males, 39% females), Total mortality by SDRA was 59% and there is an increase up to 72% since 1989 Mean time of stay of the dead was 4,3 days and 12,4 days those who survived. Although during the late years we offer in the PICU a better attendance quality to the patients with SDRA and the mean stay is longer, both for those who die and for those who survive, mortality of patients with SDRA have increased. The incidence of SDRA secondary to the septic shock of a meningococcic etiology have decreased. On the contrary, the SDRA secondary to infections by opportunistic germs in patients with congenital inmmunodeficiencies or acquired immuodeficiencies have a tendency to increase. In our series, this change of aetiology is the responsible for the increase in mortality. Hospital Infantil Unlversitario "Virgen de1 Roclo". SEVILLA. ESPAlqA AIMS:To assess the incidence, etiology, clinical course, sequelae and mortality of the patients admitted to a Paedfiatic Intensive Care Unit with the diagnosis of severe traumatism. MATERIAL AND METHOD: 60 cases of severe traumatism in children admitted to our ICU in the period from January 1990 to June 1990 were reviewed. Age of patient ranged from 4 months to 9 years, 65% were males. In our series, 53% of cases suffered traumatism due to a traffic collision and 38% had a fall from a considerable height. Only in one case was traumatism due to violence to the child. We assessed the First Assistance received in 76% of cases: where was it performed, interval of time since the accident, and steps taken. These data were also studied in relation to the latter evolution. RESULTS: 75% of our patients suffered cranioencephalic traumadsm (CT); in 53 % it was an isolated picture and in 22 % of cases was associated to other lesions. There was participation of thoracic and/or abdominal organs in 16 % of cases. 10% of cases presented important maxillofacial involvement. Only one case presented serious cervical medullar lesion. Mortality in our series was 3.3%. In 8.3% important sequelae remained. All of these patients presented TEPAS on admission equal or lower than 4. 16% of those with traumatises had slight sequelae. 71.6% of the total evolve towards healing. A polytraumatized child is a patient that benefits considerably of it admission in a Paedriatic !CU. The rapidity in receiving First Aid and its quality are essential to avoid sequelae and to make mortality decrease. After unilateral lungtransplantation 20% of the patients develop a lung failure with decrease of perfusion and increase of pulmonary blood pressure in the transplantated lung. The improvement of perfusion is an importent task in the postoperative period. Case report: A 14 year old girl with idiopathic pulmonary fibrosis received a left sided single lung transplantation. During the early postoperative period occured a higtter demand of oxygen and an increasment of the pulmonary vascular resistence in the left lung. The pulmonary ventilation and perfusion scintigraphy indicated in comparison with the right lung a reduced perfusion of only 30% in spite of a ventilation of 70% of the transplanted lung. To improve the perfusion of the transplant we administrated per inhalation prostacyclin in a maximal dose of 20 ng/kg/min. The arterial blood pressure decreased but the perfusion continued nearly at the same level. During the following administration of 10 ppm NO in the respiratory air we achieved a significant reduction of the respiration pressure f~m 40 to 32 nun H20 and of the pulmonary arterial pressure. The perfusion in the transplanted lung increased to 70cA/Of the total pulmonary perfusion. After 3 days of administration with NO we were able to withdraw the axtifical respiration without any following complications. Conclusions: The perfusion of transplanted lungs is a major proble_r~ in the postoperative period. This case demonstrated the advantage of NO towards the inhalativ application of prostacyclin. NO showed a significant improvement of perfusion in the transplanted lung of a 14 year old girl. Results: A total of 11 children with ARDS were treated with bovine surfactant (Alveofact| 9 cases were evalable. The median age was 1.3 years (range 2 weeks to 4,9 years). In six cases ARDS was associated with pneumonia, in two cases with lung hemorrhage; in one case isolated ARDS followed hemihepatectomy. The first surfactant application was performed with a median latency of 22 clays (range 3-68 days) after first symptoms of ARDS witha median doseof 84 mg/ kg (range 42-133mg/kg). In 9 patients 28 doses of surfactant were applied. During the hour before therapy, the median PaO2 / FiO2 -ratio was 70-65. Within 30 min. after application of exogenous surfactant the PaO2 / FiO2 -ratio increased to 90 with successive decrease over a period of 8 hours to 75. Accordingly, an increase in PaO2 and oxygen saturation and (less significant) a decrease in ventilation parameters could be observed. Analysis of broncho-alveolar lavage before surfactant application in children receiving repeated doses revealed in most examined cases either clear surfactant deficiency or pathological function. 2 of 9 treated patients survived (3 of the 11, respectively). 13 of the 28 surfactant doses were applied in the 2 surviving patients.Conclusions: The application of exogenous surfactant in children with ARDS caused a significant increase in oxygenation, which declined over a period of 8-12 hours. The effect often could repeatedly reproduced, in one case after 11 applications. The increase in oxygenation often allowed the reduction of FiO2 and/or the inspiratory pressure. No side effects were observed after exogenous surfactant application.In many cases the application of surfactant wag too late after first symptoms of disease (median latency 22 days). ARDS mostly due to pneumonia seemed to respond to surfactant therapy less well or not at all. Permanent junctional reciprocating tachycardia (PJRT) is the most cOmmon incesant supraventricular tachycardia (SVT) in children. It is usually drug resistant and its onset in early life has been associated with dilated eardiomyopathy. We report our clinical experience with 3 patients detected antenatally and another diagnosed at 2 months of age. Method.Diagnosis: negative p waves were detected in leads II,III and F, P'R > RP" and there was not warm-up at tachycardia onset.Clinical records, EKG,X-Rays, Echo and Holter were reviewed. EP studies were undertaken only with therapeutic purposes. Results. In a 10 year period 13 patients under 14 y of age fullfilled diagnostic criteria; 3 were detected prenatally (25-34 weeks) and one was diagnosed at age 2 mo. The 3 fetuses had intermitent SVT during gestation. All 3 of them had PJRT in the first month of life at rates between 180 and 240 bpm. They were admitted to the ICU but did not develop signs of heart failure. They were controlled with Digoxine (D); D and Quinidine; D and Propafenone in 2 to 20 days. One was in sinus rhytm until age 4y; he then showed persistent PJRT over 70% of the day on repeated Holters and underwent successful radiofrecuency catheter ablation (RFCA).The other two patients showed initially a lowering of tachycardia rate followed by sinus rhytm for over 90% of the day (follow-up 2ran and 4 y). The 2 mo. old infant was admitted to the ICU in severe cardiac failure. Echocardiogram showed marked systolic dysfunction (shortening fraction 15%) Treatment with digoxine, amiodarone and propafenone were unsuccessful despite lowering heart rate to 185; RFCA was performed at 3 m. of age with restoration of sinus rhytm and rapid recovery of contractility. All patients were given ATP at admission with transient (15 to 35 see) recovery of sinus rhytm. ff,s162 Clinical course of PJRT is variable. ATP is useful only as a diagnostic tool. Initial treatment with Digoxine + Amiodarone or propafenone is adviced. RFCA is a very useful therapeutic modality and can also be performed in young infants Twelve patients (5%) died. These were 4 meningitis, 2 head injury, 2 sub-arachnoid bleeds, 1 status epileptieus, 1 leukaemie, 1 drowning, and 1 multiple trauma. Calculated from the a 2 admission day p edialric risk of mortality score (PRISM), the probability of death (p) ranged from 0-100%. Of the 12 deaths, I 1 were predicted by pRISM analysis except for the leukaemie patient (p I%) who died from haematological complications following chemotherapy. Two children predicted to die (p 43% & 73%) survived. The median length of Stay was 2 days (range 1-34 days). 98 patlents(50%) received ventilatn~ support and 10 patienta(4%) were transferred to specialist units (5 neurosciences, 3 liver, 1 cardiac, 1 bums). This data supports the view that many paediatric patients are being adequately treated in a DGH ICU. Meningitis and other neurological illness caused the majority of deaths and respiratory problems caused most admissions. Most deaths (9 of 12) occurred within a few hours of admission. Ectopic junctional tachycardia (EJT) is one of the most dangerous arrhythmias in the postoperative setting of congenital heart defects since it does not respond to antiarrhythmics or defibrilation. The object of this presentation is to report on two patients who presented F_JT in the early postoperative period and developed intense congestive heart failure which could be controlled after treatment with moderate topical hypothermia. Two patients, 8m and 2y, diagnosed of Atdoventficular Septal Defect and Tetralogy of Fallot developed intense heart failure in the early postoperative period. Taehyeardia rate was 215 and 225 bpm. Medical drug therapy included weaning from vasoactive drugs, IV digitalization and IV amiodarone treatment. There was not response. They were both surfaced cooled by placing plastic bags filled with cold water over the patient's chest and abdomen. Temperature was monitored to obtain a central temperature of 34~ There was a gradual decrease in heart rate in the following hours (130-150bpm) paralel to the degree of surface cooling and clinical course estabilized.Both recovered normal sinus rhytm in 48 to 72 hours. There were not significant arrhytmias after the procedure and postop, was uneventful. Conclusions. Moderate hypothermia is a very useful manuever for the treatment of drug resistant EJT. Since it lacks side effects of other antiarrthymics we beleave it should be the treatment of choice for the treatment of EJT in the postoperative patient. Present understanding of the pathogenesis of sepsis, based on the theory of systemic inflammatory reaction, has risen new interest in the more invasive methods of treatment, like plasmapheresis, leucapheresis and exchange transfusion (ET). Obiectives: Evaluate the effect of ET in the treatment of neonatal sepsis. Material and methods: From September 1 to December 31, 1994 a prospective study was carried out, where the severest cases of bacteriologically proven neonatal sepsis (n=9) were treated with ET. In total 15 newborns were treated for culture positive sepsis in the intensive care unit during this study period. Diagnosis of sepsis was based on the clinical criteria of suspected neonatal sepsis, used by MC Harris et al., laboratory data and positive blood culture. Newborns with severe congenital malformations were excluded. ET was carried out with fresh (less than 24 hours old) Adsol-conserved erythrocytes, from which buffy coat had been removed, and same donors plasma, using a slow continuous two-site technique. The mean volume of ET was 164.3 ml/kg. The effect of ET was assessed as a change in the Score for Acute Neonatal Physiology (SNAP), general treatment results were compared with a historical control group of 26 newborns, treated for culture-positive sepsis in the same ICU during the first eight months in 1994. Students Ttest and chi-square test were used in statistical analysis of the data. Results: With the use of El a significant decrease in mortality was achieved: 1 death of 15 cases during the study period, compared to 9 deaths among the 26 controls (p<0.05). No baby, receiving ET, died. The incidence of severe complications did not differ in the two groups. The SNAP-score showed quick improvement by the first post-transfusion day (p. 4. Results: 10 subjects (37%) resulted positive for BO, out of which 8 were females (80%) and 2 were males (20%). The subjects with mild BO were 7/10:1 was a doctor, 3 residents and 3 nurses. The subjects with severe BO were 3/10, out of which 1 resident and 2 nurses. Conclusion: The results obtained show that BO is a condition well represented in the staff of our PICU. The category most at dsk seem to be the nurses (5 subjects), as well as residents (4 subjects), as in literature, which shows a major incidence of the syndrome in younger subjects and having a limited partecipation of functional decision. The results obtained obliged us to start a programme of serial controls so that the subjects most exposed can have a necessary psychological support to react adequately to this condition. The term systemic inflammatory response syndrome (SIRS) was adopted by the Consensus Conference to denote a type of systemic response to severe infection or otherinsults in critically ill patients. When SIRS occurs from infection it is called sepsis. Sepsis occurs more frequently in persons with perexisting illness or severe trauma. There has been tremendous advances in prophylaxis, diagnosis, and treatment of sepsis. A comprehensive model of the disease progression from SIRS to MODS should be developed giving priority to severity of illness scoring system and other predictive methods. Some recommendations for future clinical trials include: Trials should not start with humans. Before proceeding to human trials, animal studies should indicate an acceptable risk/benefit ratio. Appropriate patient populations must be defined and treatment protocols should be standardized. Full and rapid reporting of all results should be mandatory and a central repository of published and unpublished study results could be helpful. Accrual at each center should be of sufficient size, and should include the number of patients accrued, mortality rates, and patient characteristics. Pivotal trial should be preceded by sufficient pilot or phase II studies. Correct drug dosage and usage should be delineated in pilot studies. Large, multicenter, trials should be used to enhance the unversality of trial results. Analyses should be planned a priori. Definitions for the target population should be explicit, reproducible, and include illness severity scores. Outcomes should be relevant reproducible and include both measures of benefit and harm. MODS and its reversal should be considered as an endpoint. Quality of life should also be considered as an endpoint. The estimators of overall treatment effects should be controlled for base-line prognostic factors and subgroup anaIysis should only be used for hypothesis generation and not to modify the conclusoin of the trial. Economic analysis should be included as part of clinical design. Evaluatin of source control should be a critical component of any study. Standardized clinical mediator assays should be pursued. Placebo patients in clinical trials should be studied for a better understanding of the pathogenesis and epidemiology of SIRS, Evidence based medicine should be used to evaluate the validity of clinical. Introduction: Use of inhaled nitric oxide (NO) as a modulator for optimizing ventilation-perfusion or lowering pulmonary artery pressure is becoming increasingly common. NO is a free radical but little toxicological research has been published. Clearance of nebulized 99mTc-DTPA is known to be, a sensitive indicator for early function impaimaent of the alveolocapillary barrier. We investigated whether exposure to NO increased clearance of 9~Tc-DTPA from the lung. Methods: Three groups of 5 White Sealand rabbits (BW 3.5 kg) were anesthetized, tracheotomized and paralyzed. 2 groups were ventilated for six hours at pressure regulated volume control, set to deliver 10 ml/kg with a frequency of 30/rain, I/E ratio = 1:2 and PEEP = 3 cm HzO using a modified Servo 300 Ventilator (Siemens, Solna, Sweden) with computerized NO delivery system. Gas mixture per group was either 0/25 or 20/25 [NO (ppm) / FiOz]. After six hours of ventilation in these 2 groups and immediately after anesthesia in group 3 (control), 99~Tc-DTPA was nebulized into the inspiratory line of the breathing circuit and administered as a fine aerosol. Gamma counting was measured for 20 minutes, monoexponential curves were fitted to the data and the clearance half-time (T 89 was calculated. The T~/2 mean • SD of the different groups were: T~A (Mean 4-SD) h"e,i witl~ ARF 1:14 di.Ff:erent kinds, aged .Q-ore 2 mon't.hes to [4 gears o11:1 (bodi weight .~rom 4.,5 to 45 kg), IS presen .... "ed ( i,,~u::trl:e i:ibstracLive d:Lse~se...14~ 2.ARDS'-8; :~,;,,ARF o~ ::entral genes:i s .-3,0,~ :inc lud ing men ingeenceph 1 it :is-3~ Reye ' s ~yrtdro~e-..#~,bri~:Ln pes~.re~nimatior~ disease.."5). Int:Lrl~]. pa-. "iiulle'i,~s ariel regymes o+ L;MV,L;I"t"V were cle'l'.ermllled by ba- 'I~IER was. about 4. tuber,, dopamin tiara-:. t.io; was ~.".,,'.R:~r~led. CMV,CPPV d~.!"~tion raniled -~rom f to 25 dayns.,~ <6.-:in 351, 6"t2-irl lO;and>12 daVs'-in 6'l~atierr~{s I'i"ai3s:Ltiol~ o4; patterers to IMV, SIMV modee was per.R:)rmed, ~herl PIF:' decrease.d to 16-17 mL~ar, Fi02 ~ecreased to 0,4. lind less with 5a02=901/,,. i:lesq.lts:{ In pat:i.ents e{ group :L, who were tre,~d.ed w&th 2F'F'V, teoph :i. :1. l:i.r~ (iS-24.mg/kg/day), g lucecdr t icostei~oids (2 .... :~;mg/kg/day), when R exceeded in 2,2-.];,4 times normal va I tea the e aqes/,'!:l"oln 2~J,, Ite :I.~;::.!;, 8 ~ml"lrJ), it was possible 't'(' ce 'e~ e AaD]t:..~rom !1.20 2'.' 26I', 8 to !..';51,0-106, 1 mml-lg in ~}.. :~.[~ houi,!; ~d'l(:I to ru:}l",g'd!~l:i.2e i::h,:~e,'~c['el';i.stil Obieetives : This chapter will describe what is knovca of the psychlogical responses of infant and children to hospiUiisation and attendant procedures. The factors which may modify these responses will he discussed and important considemtiorts will be outlined for optimal anaesthetic management and postoperative period of infants and children which will minimised the rise of emotional upset. Methods : In this paper the autors will discttssed the probl 9 of: 1. health children (ASA I, II) facing single uncomplicated surgical elective procedures 2. various abnormal situations including neurotic children, children facing repeted operations, chronically ill, buaaes and tsaumatically impired ones 3. unfortunate young patient facing and often expoclting fatal outcome from le "ul'ukaemia, tumors, cystic fibroses or otheq" disease. : Management of each child must vary greatly, ifi general the phases of emotional conditioning include home and preadmissiun preparation, admitiun preoperated and operative care and postoperative period. The authors would be happy if the child passes all stages without any trauma which could be prolonged in the future life. Introduction iNO is used to selectively reduce pulmonary vascular resistan(~e. We applied iNO in the postoperative intensive care of patients with pulmonary hypertension and the risk of right ventricular failure after surgical correction of a congenital cardiac defect. Methods 2-50 ppm NO were added to the ventilatory gas mixture using a specially designed equipment (Messer-Griesheim, Germany/Austria). Indications for application included pulmonary artery pressure >50% systemic pressure, critically depressed right, ventricular function or an oxygenation index >10. Assessment of N Oefficiacy consisted of on-off-on measurements according to the clinical stability of the patient including hemodynamic parameters, pulmonary gas exchange, continuous monitoring of ventitatory function and transesophageal echocardiography of the right heart. Results In 29 situations (19 patients, age 3 days-71,1 years), iNO was applied 0-628 h postoperatively. Oxygenation was improved in 13 situations from 114_+51 to 171+53 mmHg pC2; pulmonary pressure was reduced in 17 situations from 70-*22% to 34_+17% of systemic pressure. In 7 situations, no reduction of pulmonary pressure was present, but measurement of cardiac output or echocardiographic analysis indicated an improvement of right ventricular function (right ventricular stroke volume +39-*12%, cardiac output +20-*11%). In 8 situations (immediately postoperativ with suprasystemic pulmonary artery pressures [n=4], multi-organ-failure [n=4]), no response to iNO could be determined. Conclusions For a special group of patients, the selective reduction of pulmonary vascular resistance by iNO has become an important part of postoperative therapy. Using this selective afterload reduction, postoperatively depressed right ventricular function can be improved. This effect of iNO seems to be the most important one in the postoperative period. Thus, iNO appears justified to be appfleo when impaired right ventdcular function could be improved even when pulmonary artery pressure is not raised or remains unchanged. Obiectives : Premature infant are exposed to danger of apaea due to anaesthesia during their tirst months of life. It is yet unknown whether prematurity is corelated to any other kind of reslgratory disorder due to anaesthesia within the tirst year of life. Methods : We theretbre researched retrospectively for respiratory disorders in all infants under 12 months of life belonging to ASA group 1. They all had been anaesthetised in 1985.95 in our clinic for the following surgical reasons: ingvinal haemia, umbilical haemia, hydrocelae testis and phymosis. Results : In 2350 cases we tbund: lafingospasm during induction in anaesthesia (0,5%), bronchospasm during induction in anaesthesia (0,22%), impaired intubation (0,1~ postanaesthetic laringospasm (0,1%), supposed aspiration (0,04%),postanaesthetic inspiratory stridor (0,05%), postinductional Inngoedema (0,03%), death after 4 months in consequative of infection pneumonie (0,12%), none of these disorders was correlated the prematurity, 3 infants suffered of post anaesthetic apnea, 2 of them had premature medical history. Concludions : Prematurity does not enhance the risk of respiratory disorders due to anaesthesia within the first year of life, except the danger of postanaesthetic almea needs spetial cosideration. It could be demonstrated that aePGI 2 lowers pulmonary vascular resistance and indirectly improves cardiac function. This effect seemed to be selective, and was comparable to iNO in the doses we have examined. Therefore, aePGI 2 could represent a clinically useful alternate to iNC. However, further research is necessary to work up the benefits of either therapeutic strategy. Objectives: Heat and moisture exchange filtem (HME) are used as artificial noses for intubated patients to prevent tracheo-bronchial or pulmonary damage resulting from dry and cold inspired gases. Furthermore they are used for the prevention of bacterial contamination of the anesthetic apparatus by the patient's exspired air. So they are considered as a time-and money-saving device in anesthesia. Filters are mounted directly on the tracheal tube, where they collect a large fraction of the heat and moisture of the exspired air, adding this to the subsequent inspired breath. The effective performance depends on the water-and bacteria-retention capacity of the filter. This study evaluates the efficiency of four different filters under clinical conditions. Methods: Four different types of filters ( DAR Hygrobac, Gibeck Humidvent, Medisize Hygrevent and PALL BB 100 ) were investigated dudng mechanical ventilation over a pedod of 24 hours. 20 minipigs with hemorrhagic shock were intubated and Ventilated for 5 days in an animal intensive care unit (ICU). After 24 hours of mechanical ventilation the filter was randomly replaced maintaining the individual ventilatory conditions. The weight of the filter was determined before use and after removal after 24 hours. The airway pressure was monitored online to record changes during use. Tracheal secretions and both sides of the filter were microbiolologically tested to see whether bacteria of the animal's respiratory system could be found on the patient's side of the filter or if they even would have penetrated the barrier. Results and discussion: Over a pedod of 24 hours 3 of 4 types of filters showed an increase in weight of 10 + 6% and airway pressure. Bactedal celonisation 0ccured in nearly all fillers (93 of 100) on the patient's side, whereas only three of four types of filters showed identical bacterial colonisation on both sides. The only filter that did not show bacterial penetration, increase in weight or airway pressure was the PALL-HME, a condensation humidifier without hygroscopic salts for moisture retention. With respect to our data one should use a condensation humidifier if airway conditions should remain stable dudng mechanical ventilation and desinfection of the anesthetic apparatus should be avoided after each patient. Aim: To assess the clinical uses of, and experiences with, the Hayek oscillator. This is a non-invasive device capable ef delivering not only continuous negative pressure (CNP) but also external oscillatory ventilation around a negative baseline (EOV-NB) using an external cuirass. This type of ventilation avoids the need for intubation and intermittent positive pressure ventilation (IPPV) and facilitates weaning in ventilator dependent patients. Patients and Methods: 21 patients in respiratory failure, age range 3 weeks to 15 years in a total of 29 patient episodes were treated using either CNP or EOV-NB mode. Duration of treatment varied from 4 hours to 8 days. Indications for use ef the device were: 1) to facilitate weaning from IPPV 2) prevent reintubation of patients following unsuccessful extubation, and 3) avoid intubation and IPPV altogether using the Hayek oscillator as the on[y means of respiratory support. Results: There was an increase in PaO2:FiO2 ratio after CNP and EOV-NB (P <0.0001, and p=0.01 respectively, Wilcoxon Signed Rank test). Patients who were in respiratory failure with hypercapnia showed a statistically significant reduction in PaCO2 both with EOV-NB and CNP (P=0.02 and P=0.01 respectively) but the magnitude of change was individually greater in the patients who were treated with EOV-NB. All patients, however, showed a fall in respiratory rate (P<0.0001) after the application of the Cuirass in CNP mode. There was no physiological deterioration related to the application of external extrathoracic negative pressure in either CNP or EOV-NB modes. Conclusion: The improvement in PaO2:FiO2, the fall in PaCO2 and respiratory rate were indicators of an improvement in ventilation. The proposed mechanisms include improvement in FRC, recruitment of additional alveolar units, and improvement in secretion clearance resulting in reduction in the work of breathing. meek to ~ month of the lifo,The bemodyuanicfacLS were defined uitb the help of tetropolar reography method!. The excretion of !he catbocholauines fCfi] mith the urine gas detertend by Taylor LL,Laoorsy ~ Iacg/dayl. hsaltL In the hypercuagulation stage of BIC we deflorteeed the acLiuutiun of the tbrubio and plasiin syAet~ mitb the increase of the inhihitnrs, In this case we registered in full uahe dot This process coabined uitb the dayl~ excreLiou with Lho urine epinopbr ne E], nor~pinopbr no tel and dophanine IO], Lbat shod the inLensificatiou of the s~nthosis prnoe-S~eS and the release of eA in blood fron hiSSue deport The actffat on of the sVnpathadrenuI systen ]SfiSl assisted to furl the b?perd~nanical rosins of the eircuidion and increase the ,icrocirculatinn, The klinicaI Sings of the insufissieutly of the circulaLion have not defined,that has been associated the conpensatury character uf the ehan~es of ~ and heludy~enic StatuS, T~e uun~u|p-Lion ceugulupatby bus been donoustraLed in the hypocougulatien stage ~bat man xauifosted b 7 the exhaust of Lhe confulaLion nod oessel-plateL heuoStasis, The consuxptton of cnnpononts tbronbln ,plnstin, kallek~eiu-kinln s~sLots and the forniration eat in fell canoe clot uaS accoqaued bS docrea,e of fl,Nfl,O, the products of the xotabolisx of C~ and the activation of xonoaninoxydasu. The decrease of the extoll'on g and the exhaust deport CO indicahd about t!e ]ou fund/anal reserve of ~fl~. It was one of the lain reason of ~bo heiod~uanic disbroed Iheat insnfissient]~] and the uicrncireulaflion lintestinal codeme with the low effectife periferal flow] and nul[iplay organ failure,The distrued deport of SOS mitb throubocytupenin no; be one of the nechanisn the disLrood of uesseJ-plaLoL heioshasis, THe correlation boLueeo changes of boIOStoSiS C~ and circulation ore reguired aduinistration nedidnS, thai resLore the love S of C~ in the blood, prevent uulLiplay organ failure and hetorrnge in children with SepSiS, ~b~ectives: Multi-measured correlative analysis of the most number of non-invasive indices of the cardiorespiratory system function was made to determine the structure of their interrelation and the ways of their adequate and effective correction. Hethods: Spiremetry, capno~raphy, oxygenography, indirect Fick method at recurrent respiration, plethysmography, integral rheography -in all 52 indices were used. The received data were processed on a computer by a standard package of statistical BMDP programs. Results: 70 women with ~H-gestosis (I group) and 48 somatically healthy pregnant women (II group) were studied. Cluster analysis has shown that the rate of the mean correlation connection between ventilation indices was 94% in the Ist group and 90% in the IInd group; gaseous metabolism -91% and 86%, respectively; central hemodynamics was 87~ in both groups. Conclusion: Cluster interpretation allowed to suggest that an increase of the rate of the mean correlation connection between the indices was characteristic of effective adaptation as the system was multi-component and well-regulated. On the contrary, the increase of the rate of strong correlation connection Between the indices reveals the rigidity of the system and the tensity of adaptation mschaniams, i.e. the proximity to decompensation. It follows from this that in cases of EPH-gestgsis, the reliability of regulating ventilation and gaseous metabolism decreases. Seve/e hypoxemia in non intubated patients represents a major contraindicafion to fiberoptic bronehoscopy (FOB) and bronehoalveolar levage (BAL), but these procedures are often required for a correct diagnosis of the causative agent of pneumonia. Aim of this investigation was to veaify the safety and efficacy of bronehoseopic procedures during pressure support ventilation administered through facial mask (FM-PSV). Five intensive care patients, all immunoeompromised, (3 males and 2 females; mean age 41.6• were enrolled in the study. All patients presented criteria for pneumonia with PaO2/FiO2 ratio ~ 100 and were responders to FM-PSV. FOB and BAL were performed afte~ topical anesthesia with FM-PSV ( PS = 16 em H20; PEEP = 5 emH20; Trigger = -lemH20) continuously admires" tered ( 10' before FOB FiO2 = .7; during FOB, FiO2 =1 and for 90' alter FOB, FiO2 = 0.7). PaO2/FiO2 ratio as well as 02 saturation (SAT) did not show signifteative changes during the procodure (Fig.l) . No complication was observed and hemodynamic conditions were stable for all patients. CMV, Pnenmoeystiis (2), Legionella and Mycobaetermm Tuberculosis were identified from BAL allowmg a prompt and targeted therapy. We concluded that Mask PSV can represent an excellea~ technique to pexform FOB and BAL in severely hypoxemic patients without deterioration of gas exchanges and avoiding endotraoheal intubation. Intensive Care Unit, Hospital General of Albacete, Albacet~ Spain. Objective: To analyze the current incidence and epidemiology of total parenteral nutrition (TPN) among critically ill patients placed on mechanical ventilation. Design: Prospective observational study. Setting: Medical Intensive Care Unit in a Tertiary Hospital. Patients: A total of 113 consecutive L'ritically ill patients with non-coronary related disease needing mechanical ventilation admitted in our ICU during a 12 months period. Measurements: Data of sex, age, diagnosis, and outcome were recorded. Severity of illness and therapeutic effort in the first 24 hours were measured using Acute Physiology Score and Chronic Health Evaluation (APACHE II) and Therapeutic Intervention Scoring System (TIES). R~ults: 113 mechanically ventilated patients, 76 male and 37 female, were studied. Only ten patients needed TPN and their main diagnoses were: five cases of multiple organ failure secondary to pneumonia (2), ARDS (2) and septic shock (1); two eases of acute panereatitis; and one mesenteric throngmsis, one status epilepticas, and one ,prolonged cholinergic crisis b~ suicidal organophnsphate insecticide subcutaneous injection. No statistically significant differences between both TPN and non-TPN groups were found: Objectives: Evaluate the efficacy of prone position in ARDS and determine its importance in the therapeutic algorithm. Methods: 43 consecutive patients with severe ARDS (Murray-Score > 2,5; paO2/ fit 2 < 160 mmHg; 29 male, 14 female, mean age 62 years) were conventionally ventilated (PCV, PEEP 6-16 mbar, I:E=I:I, Ppeak < 30 mbar). If after 24 hours pulmonary function did not improve patients were placed in prone position. Change from prone to supine position was done every 12 hours. Beside ultimate survival, parameters investigated were AaDO2, paO2/fiO2, and venous admixture (Qs/QT). Results: During the first 12 hours in prone position 39 of 43 patients showed a significant decrease in Qs/QT (25.3% vs. 17.8%) and AaDO 2 (235 vs. 187 mmHg), and an increase in paO2/fiO2 (151 vs. 201 mmttg). Changes were most pronounced in patients with high Qs/QT, and in patients with an onset of ARDS less than 48 hours before first application of prone position. After an average of 6 position changes (2 to 16) 28 of 43 patients could be weaned from the ventilator. 22 patient could leave tile hospital. I11 the later course letality was primarily determined by additional organ failures and by the severity of the underlying disease. Negative side effects were minor, including slight cardio-vascular depression and increase in p~CO2, and never posed a limitation to continuation of prone position. Especially in patients with septic shock skin lesions in exposed areas could not always be prevented, Prone position could easily be combined with all ventilation modes and with all intensive care interventions. Also immediately after major surgery and in patients with open packing prone position was possible. Conclusions: In this investigation prone position proved to be an efficient and safe method in the treatment of severe ARDS. Patients with a pronounced ventilation/ perfusion mismatch and patients in the early stages of ARDS appear to profit most from prone position. Though the immediate effect on oxygenation is striking, still more the 40% of all patients die from multi organ failure and underlying diseases. A proposed therapeutic algorithm for ARDS is as follows: If under conservative ventilation (PCV, PEEP < 20 mbar, Ppeak < 30 mbar) pulmonary function does not improve within 12 -24 hours prone position should be applied. When after 2 -3 position changes no lasting effect can be achieved further ventilation modes (e.g. PC-IRV, APRV, NO, etc.) should be used in addition to prone position. Standard intensive care principles, such as fluid restriction and optimization of circulation, apply also to patients in prone position. Objectives: Nitric oxide reacts with superoxide to form peroxynitrite, an extremely reactive and toxic species. We quantified the presence nitrotyrosine, the stable product of the interaction ' of peroxynitrite with tyrosine residues in the lungs of pediatric patients that died with respiratory distress syndrome (RDS). Methods: Paraffin embedded lung sections, obtained at autopsy, were incubated with a polyclonal antibody raised against nitretyrosine, followed by a secondary fluorescent antibody. Alveolar structure-associated fluorescence was quantified using existing methods. Results: Tissue sections from patients who died with RDS exhibited significant specific immunostaining which was uniformly distributed across the blood-gas barrier. In contrast only background levels of fluorescence were seen in the lungs of patients who died from non-pulmonary causes. Intense staining was also seen in the lungs of rats that breathed 100% 02 for 60 h, a condition known to result in RDS-type illness; no immunostaining was observed in air-breathing rats. Conclusions: Significant levels of peroxynitrite may be formed in the lungs of patients with acute lung injury. Peroxynitrite may be contributing to the pathology of RDS by damaging key components of the alveolar epithelium including the pulmonary surfactant system. Mechanical ventilation time was prolonged 16,g • 10 days in patients with ARDSvs 1,7 _+ l,4 days in control . Mean staylCUwas lg _+ 10,g days in the ARDS group vs 4,9 • 2,7 days in control group Postoperative mortality rate was 53% in ARDS patients vs 5,7% in those without respiratory failure. 1-ARDS incidence in liver transplantation is low (11,2% in our sene) but it causes high mortality (53%) PAGE, gas ventilation of the perfluorocarbon-f'dled lung, supports gas exchange and circulation in small animals (<15kg) with lung disease. We hypothesized that large animals could be supported by PAGE without adverse effects on bemodynamics. We first elucidated the determinants of gas exchange in normal sheep, and applied them to a model of adult respkatory distress syndrome (ARDS). Methods: Using the ventilator settings determined to be optimal in our pilot study (FiO2 of 0.6, PEEP of 5 cm H20, IMV of 6 bpm, IT of 50%, and TV of 16 ml/kg), sheep weighing 58.9 ~ 8.3) kg had lung injury induced by instilling 2 ml/kg of 0.05N HC1 into the trachea. Ten minutes after injury, sheep with PaO2<100 ton" were randomized to continue gas ventilation (control, n=9) or to institute PAGE (n=9). PAGE was instituted by instilling 1.6 L of unoxygenated pefflubron into the trachea and resuming gas ventilation at the previous settings. ABG's were drawn at baseline, 10 minutes after injury, 30 minutes after injury, and then every 30 minutes for 4 hours. Objectives: Inhaled nitric oxide (NO) can improve oxygenation and decrease mean pulmonary artery pressure (PAPm) in hypoxemic patients with ARDS. In severe hypoxemic COPD patients, it is not known whether inhaled NO can exert a similar effect on hemodynamics and gas exchange. Therefore, we investigated die response of inhaled NO in hypoxemic COPD patients and the results compared with those obtained in a group of ARDS patients. Methods: Ten COPD patients (age 71_+2y;FEV~ 0.98_+0.12L) and 11 ARDS patients (age 57_+5; LIS 2.8_+0.1) mechanically ventilated were studied. Hemodynamic parameters were measured using a Swan Ganz catheter. Arterial and mixed venous blood gas determinations, SaO2, SvO2, Hb and MetHb were measured (ABL 500,Osm3). Mean intratracheal concentrations of NO and NO2 were continuously monitored using a chemiluminescence analyzer (NOX 2000) . During the study the ventilatory pattern and FiOz were kept constant. The protocol was for ARDS group: basalt, NO lOppm, basal~; COPD group: basalz, NO lO ppm, NO 20 ppm, NO 30 ppm and basal2 . After a steady state of 20 rain hemodynamic and gas exchange measurements were performed. A positive NOresponse was defined as a 20% increment in PaO 2. Results: PAPm was similar in both groups and decreased significantly after NO (ARDS, basal 33.6_+9.7 mmHg, NO 29.7 +6.7 mmHg, p <0.01) (COPD, basal 27.8_+6.3 mmHg, NO-10 24.4_+5.3 nrmHg, p<0.01). All other hemodynamic variables remained unchanged after NO. Basal oxygenation was higher in COPD group (PaOJFiO 2 189_+53 mmHg) vs ARDS group (PaOJFiO 2 100_+40 mmHg)(p<0.01). After NO-10, PaO2 increased (69_+20 mmHg to 97_+40 mmHg, p<0.01) and Qs/Qt decreased (37+11% to 31_+10%, p<0.01) only in ARDS group. In both groups, significant correlations between basal PAPm and inhaled NO-induced decrease in PAPm were found. Inhaled NO-induced increase in PaO2/FiO2 was not correlated with basal PaOflFiO2. NO responders were 8/11 (73 %) in ARDS group and 2/10 (20%) in COPD group (p<0.05). Conclusions. In hypoxemic ARDS and COPD patients, inhaled NO decreased mean pulmonary artery pressure. However, oxygenation only ameliorated in ARDS group because die number of responders to inhaled NO were higher in ARDS group and this effect seems not to be related to the basal hypoxemia. These results might be explained by the V/Q abnormalities present in COPD patients. Grant FIS 95/1390. Objectives: It has been recently reported that expired COn slope as a function of time is modulated by total respiratory system resistance (Rrs) in critically ill patients (Chest 1994; 105:219-223) . In this study, we analyze the relative contribution of disease (Dis), endotracheal tube resistance (Rtube), airway resistance (Rmin), additional resistance (~Rrs), autoPEEP (PEEPi) and dylmmic/static elastance (Ed/Es) to the CO2 elimination in different clinical conditions. Methods: We have studied 37 adult patients (8 controls, 11 acute respiratory failure, 9 severe ARDS and 9 COPD) mechalfically ventilated (Servo 300 and 900C, Siemens) without PEEP. We recorded tracheal pressure, airflow and capnograms. Signals were analogic to digital converted for posterior data analysis. Objectives: Alveolar ejection volume (VAn) Can be defined as the fraction of tidal volume (Vt) with minimal dead space (Vd) contamination. According to the classical paradigm: limvd_~ [VCO2/Vt] =FACOz, VCO2 vs Vt relationship tends asyntotically to a constant slope when approaches end-tidal volume. We have defined VAn as the volume that defines this relationship until a limit of 5% variation. Methods: Six subjects with normal respiratory mechanics were studied during anesthesia for minor surgery. Two subjects, otherwise normals but having high values of total resistance and dynamic compliance, were also studied. Capnograms were recorded in steady-state at 3 levels of Vt (0.3, 0.5 and 0.8 L) and four levels of PEEP (0, 5, 10 and 15 cmH20 Objectives: Patients with ARDS presented lung abnormalities which originate an increase in airway resistance (Rmin), in additional resistance (~Rrs) and in static elastance (Ers). Application of PEEP further increases ~Rrs. Capnographic indexes reflect lung ventilation]per fusion inhomogeneities. In these conditions, the effects of PEEP on lung mechanics could be better understood by simultaneous measurement of capnographic indexes. Methods: We studied 3 groups of subjects. N: 8 normal subjects scheduled for minor surgery; ARF: 9 critically ill patients with mild acute respiratory failure; ARDS: 8 patients with early ARDS (< 72 h). We recorded tracheal pressure, airflow and capnograms. Signals were analogic to digital converted for posterior data analysis. Respiratory system mechanics was assessed by constant end-inspiratory and end-expiratory occlusions technique. At equal tidal volmne (0.5L) a PEEP level of 0,5,10 and 15 cmH20 was applied in all patients. We calculated Ers (cmH20/L), Rmin, C~Rrs (cmH20/L/s) and autoPEEP. Capnographic indexes were alveolar ejection volume (VAE)/Vt ratio and expired CO2 slope beyond VAE (SIpCO2 In contrast to synthetic surfactant natural suffactants (Alveofact|174 are able to inhibit PMN-activation. After incubation of activated neutrophils with surfactant, L-Selectin expression is decreased. These effects depends on which preparation is used. We conclude, that natural surfactant (Aveofact| can perhaps influence early recruitment (,,rolling") of PMN in patients with respiratory failure like ARDS. WITH ARDS HOrmann Cb, Baum M, Putensen C, Knapp R, Lingnau W, Putz G . Clinic for Anesthesia and General lntensiv Care Medicine, University of lnnsbruck, AnichstraBe 35, 6020 Innsbruck Objectives: In thoracic CT scans of patients with severe ARDS atelectasis and pleural effusion can be found in the dependent lung regions. By rotating these patients from left lateral position to right lateral position a redistribution of the CT densities, a recruitment of atelectasis and therefore an improvement of gasexchange is possible within a few days (1, 2). The objective of this study was to find out the mechanism of alveolar recruitment during lateral positioning by CT scanning in left and right lateral position. Methodes: After approvel by the local institutional reviewboard we investigated 7 ventilated patients with severe ARDS (Entry criterias: Murray Score > 2,5) in the CT scann of the university hospital. After a stabilisation period of 30 minutes in supine position a thoracic CT scan slice 1 cm above diaphragm was taken. Then two different positions of the patients were studied in a randomized order: a) 60 degree of left lateral position, b) 60 degree of right lateral position. Each lateral position was held for 20 minutes. At the end of each of these periods a thoracic CT scan slice 1 cm above diaphragm was taken. Quantitative analysis of CT scan data was based on the frequency distribution of the CT numbers. To quantify the alveolar recruitment during lateral positioning by means of CT scan we defined 3 compartments within the lungs: a) normaly inflated lung, b) poorly inflated lung, c) noninflated lung ( = atelectases) (3). Results: Independant of the side of lateral positioning (L) in the non-dependent upper lung a significant increase of the normaly inflated compartment (S: 45%; L: 65%) as well as a significant decrease of the noninflated compartment (S: 34%, L: 12%) was observed in comparison to supine position (S). In the dependant lower lung the normaly inflated compartment decreased significantly (S: 45%, L: 26%) whereas the noninflated compartment increased significantly (S: 34%, L: 51%). Throughout the whole studyperiode we did not observe any significant change regarding gasexchange and hemodynamic parameters. Conclusions: In lateral position the non-dependent upper lung is decompressed. Therefore a significant recruitment of atelectases is observed in the upper lung within 20 minutes. On the other hand the dependent lung is compressed by the weight of the upper lung and the mediastinum. A great amount of the alveoli of the dependant lung collapse in this short time intervall. Therefore the net effect of recruitment of one positioning maneuver is very small. When positioning patients one should be aware, that the patient is kept in each lateral position long enough to clean up the atelectases in the non-dependant lung and short enough to compress less lung tissue in the dependant lung. Objective: To analyze effects of low-dose NO inhalation ia patients with severe aeut~ respiratory distress syndrome (ARDS) over five days. Methods: We prospectively studied 10 patients (9 men, 1 woman) with severe ARDS admitted to our ICU between May 1994 and May 1995 who required NO inhalation with a dose of 5 ppm for at least 5 days. Entry criteria for NO inJaalafioa were Murray score >i 2.5 aud Pat/FIe 2 < 125 nun Hg with PEEP >~ 8 em I~O for at least 24 hours. All patients were sedated, intubated and mechanicaUy vantil~ed with volume assist-control ventilation, and had indwelling arterial catheters (pulmonary artery, and radial or femoral artery) to measure cardiac output (by thermodilufion) and relevant intravaseular pressures, and to calculate derived parameters. NO was administered between Y piece of the ventilator and endotraeheal tube and flow was adjusted to obtain 5 ppm NO in the inhaled gas. The NO, NO 2 and NO x concentrations were continuously measured at the distal end of the endoUacheal tube by the chemiluminiscence method (NOX 4000, SEe-Seres, France). Metahemoglobinemia levels were mesured daily. NO inhalation was manteined if PaOjFiO ~ improved at least 20 % and was stopped when the change in PaO2/FiO ~ was below 20% or when the patient presented a PaOJF02 > 150 mm Hg a~er 30 minutes without NO inhalation. Every day we made an ON-OFF test to determine if NO inhalation improved PaO2/FiO ~. Statistics: Analysis of vmiance. Data: mean + standard deviation. Results: The mean age was 60.1 +_ 10.2 years and mean lung injury score was 3.3 • 0.2. Mortality was 60 % (6/10), metahemoglobinemia 1.1 • 0.2 %, and NO2 concentrations zero. PaOJF~O 2 always improved significantly al~er 5 ppm NO inhalation (see :~ Conclusions: Reintubation in salf-extubated patients strongly depends on the type of meehamcal venfilatory support: the probability of needing a reintabation ff ESE occurs during fult vontilatory support is higher than ff ESE occurs during weaning. These data suggest that some patients may remain under weaning from mechanical ventilation for unnecessarily prolonged periods of time. Objective: The aim of this study was to evaluate the acute effects on gas exehonge and hemodynamics due to positional changes from supine (SP) to prone (PP) in patients with severe acute respiratory distress syndrome (ARDS). Methods: Nine intubated, sedated, paralyzed and mechanically ventilated patients with severe ARDS were prospectively studied. All had a Murray score > 2.5, and a PaO2/F~O 2 < 100 with PEEP ~8 cm H20 for at least 24 h. All patients had indwelling arterial catheters in the pulmonary artery as well as in the radial or femoral artery in order to measure cardiac output (by thermodilution) mad relevont pressures, and to withdraw blood samples. Arterial blood gases and hemodynamie parameters were measured first in SP, and then in PP after 60 minutes of stabilization. Vontilatoly parameters remaing unchanged during all the study. Statistical analysis was done by the non parametric Wdeoxon test. Data are expressed as mean ~= SD. Results: There were 6 men and 3 women with a mean age of 54.2 years (21-71) and mortality was 55 % (5/9). Main results are shown below: Objective: To describe and compare a new method for obtaining P-V loops (P-Vcv) by using a two-way Collins valve (TWV) with thosu obtained by the supersyringe method (P-Vss). Methodology: We prospectively studied 14 patients who had an aeute lung injury and were intubated, sedated and paralyzed, and mechanieany ventilated. We performed the P-Vev loops and P-Vss loops in random order, and the static inflation pressure was limited to 35 emH20 with both methods. Pressure (P) was measured at the airway opening by means of a differential P transducer, and volume was obtained from flow (measured with a pneumotacograph) integration. The P-Vse method has already been described (H~trf A,et al.BEPR 1975; 11:709-28) . The P-Vev method consists in the following: the inlet of a TWV is connected to the ventilator's Y-piece, and both outlets are couneeted to the endotraeheal tube by means of an additional Y-piece; one of this outlets has a one-way Rudolph valve in order to allow inspiration but not expiration during the inflation maneuver. Changing the TWV tap position allows basal ventilation or progressiveinflation of the respiratory system. This maneuver is as follows: during an end-expiratory occlusion, the ventilatory settings are adjusted to deliver a 100 mL V r with a respiratory rate of 20/min and I/E ratio 1:4; at the same time the TWV tap is ehonged in order to divert flow through the one-way valve. Inflation then begins alter releasing the expiratory oonlusion. Pressure and flow signals were digitized and acquired by a computer for subsequent data analysis. We analyzed the following parameters: inflation compllonee ( Objective: To analyze the variables which eventually may differentiate ARDS patients who do and do not respond to low doses of inhaled NO. We prospectively studied 10 patients (9 men, 1 woman) with severe ARDS admitted to our ICU between May 1994 and May 1995 who were treated with NO (5 ppm). The onta'y criteria for NO inhalation were Murray score >/2.5 and PaOjFO z < 125 mm Fig and PEEP >/8 cm I~O for at least 24 hours. All patients were sedated, intubated and mechanically ventilated with volume assist-control ventilation. Tidal volume was between 6 and 10 mL&g, with constant inspiratory flow, respiratory rate was 15-25/rain, and I/E ratio between 1:2 to 1:3. All patients had indwelling arterial catheters (pulmonary artery, and radial or femoral artery) in order to measure cardiac output (by thermodiintion) and relevant intravascular pressures, and to calculate derived parameters. NO was administered between Y piece of the ventilator and ondotracheal tube, and flow was adjusted to obi~a 5 ppm NO in the inhaled gas. The NO, NO 2 and NO x concentrations were continuously measured at the distal end of the endotracheal tube by the chemilumiinscenee method (NOX 4000, SEe-Seres, France). Metahemogtobinemia levels were measured daily. We considered a response to NO inhalation when an improvement in PaOz/FO 2 above 20 % was observed after the inhalation of 5 ppm NO (Group R) . When the cha~age in PaOjFi0 z was below 20 % it was considered a lack of response (Group non-R Small airways functional abnormalities have been recognized as a common feature of lung pathology. However peripheral airways contribute relatively little (~ 10%) resistance to flow and there disturbances can not be adequately estimated by conventional measurements of respiratory mechanics. The purpose of the study was to evaluate the relationship between Raw and small airways conductance following weaning from ventilator Methods. 37 patients (age:24-62 years; 22 males) with no serious complications al~er mitral or multiple valves replacements and with more than 15 hrs on mechanical ventilation have been enrolled in this study. The modified flow interrupter technique (PTG "Gould" with Fleish head #2; differential pressure transducer PM-131-TC "Statham" w amplifier "Kistler 7251") and flow-volume recording of forced expiration (Fleish head #4) have been applied before surgery and following operation on mechanical ventilation (MY), after extubation (t:XTIJB), on 2 (2 nAY) and 3 (3 DAY) days. Airways specific conductance (SG aw) has been calculated as a mean of 7-10 consequent measurements in each patient at each stage. The SAC was estimated by max expiratory flow at 50 and 25% of VC on 3-4 F-V curves (MEF .~0, MEF 25) All the data were statistically analyzed with t-test Introduction : Noninvasive ventilation (NIV) reduces the need for endotracheal intubation, the length of stay in ICU and the mortality rate in acute exacerbation of COPD. However, some patients failed to be ventilated with NIV. .ObjectiveS...; To further delineate patients who failed to be ventilated with NIV and to obtain predicted factors of failure. Patients : A cohort of 51 patients (72 • 10 years) presenting with acute exacerbation of COPD (FEVI: 610 • 396 ml, PaCO2:62 • 17, pH: 7.33 • 0.08) and nonmvasively ventilated (pressure support through a full-face mask) between April 1990 and May 1994 Twenty-seven (53%) were successfully ventilated with NIV (discharged alive without the need for endotracheal intubation) while 24 (47%) failed, requiring endotracheal intubation. .Methods : Patients successfully ventilated and those who failed were compared according to 35 respiratory and nonrespiratory variables Univariate analysis (Wilcoxon rank-sum test and Fisher-exact test) was performed to select variables included in a multivariate analysis by stepwise logistic regression. Results : Underlying disease assessed by the Simplified Acute Physiologic Score (15 • 3 vs 11 • 3, p = 0.0003), creatinine serum concentration (122 • 45 vs 86 • 25 gM/L, p = 0.005), blood urea nitrogen (BUN : 12 • 6 vs 8 9 3 mM/L, p = 0.009), age (75 • 9 vs 69 • 10, p = 0.01) were higher and encephalopathy (71 vs 30%, p = 0.005) more frequent in patients who failed. Multivariate analysis showed that encephalopathic patients (OR (Odd Ratio) = 4, p = 0.001) older than 65 years (OR = 4, p = 0.04) and presenting with BUN >_ 10 mMYL (OR = 3, p = 0.01) failed to be ventilated with NIV. Variables related to the respiratory" status (i.e. PaCO2, PaO2, FEV1) were unable to predict tile failure of NIV. Conclusion : COPD Patients older than 65 years, presenting with acute exacerbation, encephalopathy and BUN > 10 raM/L, should be carefully monitored because of high probability of failure with NIV. Methods:From February to December 1994 we studied 30 pa_ timnts,25 males and 5 females(mean age 68+/-5);18 of The se had emphysema,lO chronic bronchitis,2 dilatative car diomyopatia,with tracheostomy and emphysema.Mean paC02 at admission in ICU was 95+/-8mmHg,while when weaningbegan, 60+/-5.Mean autopeep was 8 cmH20(4-12).All patients were ventilated in CRPV as long as four hours to calculate st8 tic and dynamic cmpliance and autopeep.Then the ventila tion was continued with PSV+Cpap(Peep 7cmH20 Objectives: Analysis of the incidence of neurogenic pulmonary edema (NPE) in a population of headtrauma patients with acute respiratory failure (ARF). NPE can occur after a central nervous system insult. Differential diagnosis: cardiogenic pulmonary edema and other forms of non eardiogenic pulmonary edema. True incidence and pathophysiohigy remain poorly defined, however the role of catecholamines seems undeniable. Early onset NPE (within 12 h after trauma) is characterised by hypoxemia, transient pulmonary hypertension and bilateral central fluffy infiltrates on chestX-ray. Characteristics of cardiogenic edema or pneumonia are absent. Late onset NPE, (beyond 12 hours after trauma), is more insidious. The clinical and radiographic picture has to clear within 24 to 48 hours. (1) Methods: All headtrauma patients admitted from January 1 to December 31, 1993 in a nearotrauma ICU setting were retrospectively analyzed for ARF with as sole criterinm a PaO2-FiO2 ratio < 250. Results: 151 Neurotrauma patients were admitted during 1993.94 Patients (63%) presented with severe head injury (GCS<8), 42 patients (27.8%) with moderate (GCS 8-12) and 15 patients (9.9%) with minor head injury (GCS 12-15). Overall mortulity was 19.2% Early (within 12 h. after trauma) and delayed onset respiratory incidents were distinguished, counting for 29 (19.2%), respectively 27 patients (17.8%), 7 patients (4.6%) had early and late respiratory complications. Early respiratory insufficiency was caused in 9 patients (25.0%) by aspiration, in 11 patients (30.1%) by lung contusion, in 1 patient (2.7%) by fat embolism and in 15 patients (41%) by NPE. In the late onset group 31 patients (91.2%) presented with pneumonia, 1 (3.0%) with fat embolism and 2 (5.8%) with NPE. The NPE group, 17 patients, presented as follows: 15 patients (88.2%) developed early NPE, and 2 (11.8%) delayed onset NPE. 9 patients (53%) died within the first days after admission, showing high mortality. GCS was less than 8 in 16 patients (94.1%), indicating severity of head injuries. Conclusions: High incidence of ARF with various etiology (41,7~ was found in this population. In about 10% of all admitted hcadtrauma patients (26,9% of ARF) NPE was causing attetial hypoxemia. Occurrence of NPE seems to be related to the severity of the brain injury and thus to outcome. These data call for extreme vigilance in respect of the insidious occurrence of NPE. were included if recovering from respiratory failure and if in the opinion of the primary physician were ready for extubation. Patients were excluded if undergoing compassionate withdrawal of support or had tracheostomies. The attending physicians were blinded to the measurements. Included patients were placed on pressure support (PS) of 0 em H20 with demand-flow continuous positive airway pressure (CPAP) 5 cm H20. After a minimum of 30 minutes on the above sehiogs: gastric intramucosai pC'O2, ABG, and a P0.1 were measured. The padents were then disconnected from the ventilator for a period of one minute and the patients" respiratory rate and minute ventilation were measured using a Wrights respirometer to calculate the frequency to tidal volume ratio (f/Vt). Patients were then extubated. Extubafion failure was defined as the inability to maintain spontaneous ventilation for 24 hours for any reason. Results: Twenty patients met criteria and were studied over one month period in October 1994. Six of the twenty patients (30%) failed weaning. The mean and standard deviation is outlined in Failure 7.01+/-0.10 6.8+/-3.9 74.3+/-43.3 87.5+/-43.6 Comparison between ROC areas shows phi and P0.1 to each show a statistically significant difference from an area of 0.5 (p 10%. No chan9es in treatment protocol (Hyperventilation, Man• etc) were carried out due to this study. Results: 30 men and 5 women were studied, aged 32• yrs. At arrival at hospital, GCS were < 5 in 20 and ) 5 in tO. The incidence of high ICP() 20 mmHg) were 7Sz at the entry. The mean Therapy Index Level required to control lOP was 4~L All patients required vasopressor therapy to maintain UPP over DS mmHg. In 20 patients a S.S F Swan-Ganz fiberoptic catheter was used to obtain a continuous recording of SJO 2. In the others 15, SJ02 were intermittently controHed.The mean time of monitoring were D.8• days. Ten patients died within this period. A total of 1.240 blood samples were analized. At arrival, SJO 2 discrepancies were found in 22 patients, B2%. At 48 hours, the incidence were lower, 18/35, 51.4%. At 4th day, were H/29, 38Z and at day 7, when the catheters were retired, II[25, 44z showed discrepancies. The CT showed new injuries in g4Z of patients with differences > 10~ in Sd02 values throughout treatment period. None of those were considered for neurosurgical treatment. No correlation was found between IOP and SJO 2 values and SJO 2 differences. Conclusions: The incidence of discrepancies between SJO 2 was higher than expected in severe head-injured patients. These situation could reflect disturbances between 02 demands. When differences are known, and those Lend to change, the CT scan, nearly always, will show new injuries. Platelet-activating factor (PAF) is an inflamatory mediator implicated in the pathogenesis of bronchial asthma and acute respiratory distress syndrome (ARDS). Its inhalation in healthy subjects produces transient bronchoconstriction and mild ventilation-perfusion mismatch, together with peripheral leukopenia as a result of intrapulmonary neutrophil (PMN) sequestration. Likewise our group has shown in healthy subjects and asthmatic patients that aaibutamol (S) inhibits both pulmonary and systemic effects of PAF, suggesting that S may inhibit PAF-induced venoconstriction in pulmonary microoirculation. The aim of the present study was to investigate if S inhalation decreases PMN by lung sequestration induced by PAF. We studied 8 healthy, non-atop• nonsmoking subjects (6M/2F, 24+4 yr), which were pre-treated with S (300,ug) or placebo, with a randomized, double-blind, crossover, design, before PAF (24,ug) inhalation. We measured the respiratory system resistance (Rrs) by forced oscillation, arterial btood gases and both total white cell and PMN count every 4 min over a 30 min. period. Simultaneously, we recorded continuously the lung dynamics of Inm-neutrophil and Tc99m-erythrocytes activity, with a gammacamara. After placebo, PAF inhalation decreased white cells (from 5410 2 1125 to 33022934x109/L), and PMN(from 29752693to 1222 _+ 767 x109/L), and increased AaPO 2(from 2.1 _+9.5 to 14.7+ 12.2 mmHg, p0.15-0.20 has been shown to occur in normal volunteers and in stable COPD patients with a specific imposed breathing pattern. Its role, however, in hypercapnic respiratory failure is less certain. We studied 10 failed weaning trials in 5 COPD patients in which breathing pattern, tension-time index (TTi) of inspimtory muscles, dynamic PEEPi, dynamic lung elastance, lung resistance, and arterial PaCO2 and pH were measured at the beginning and end of a T-piece weaning trial. In addition, the change in esophageal pressure during a Mueller maneuver (APes max) was measured. A weaning trail has been prospectively defined to have failed if one of the following criteria was met: a rise in PCO2 >20mmHg from baseline accompanied by a fall in pH<7.35; a respiratory frequency (f) >30/min; excessive accessory inspiratory muscle recruitment; and a marked increase in dyspnea. Values are expressed as mean • SE. Weaning failure was characterized by a more rapid, shallow breathing pattern, worsened mechanics, hypercapnia and respiratory acidemia despite an unchanged Tri and Pes max. We conclude that in this setting hypercapnic respiratory failure is not a consequence of inspiratory muscle fatigue. Rather the adopted breathing strategy and resultant hypercapnia may represent an adaptation to forestall the onset of muscle fatigue. Concerning the investigated ELF-par~eters, no stadstically signhqcant differences were detected between the PGI2 and the control group. Histopathologlcal changes occured in both groups and consisted in rare focal flaaaning 0f tracheal epithelium with loss of cilia and slight inflammatory cell infiltration, as well as slight swelling of alveolar typo4 pneumoeytes. Sections of generation 5, 10 and 15 from bronchial tree were free of pathological changes. Conclusion: Alter 8h inhalation of P~JI2 no signs of respiratory-lract tissue damage caused by the aerosol could be detected. The minor pathological findings in the trachea are most likely due to mechanical irritation by bronchoscopy, changes of the alveolar epithelium are known for long-term mechanical ventilation 3. Objectives: The aim of this study was to evaluate of efficiacy of ganglion stetlate blockade in patients with respiratory failure. Methods: Two groups of patients were investigated: group I (n = 15) trauma patients with acute lung injury (ALI), group If (n = 15) patients with asthmatic status. In all cases continuous mandatory ventilation (CMV) was used with Bennett 7200 AE. In both groups bilateral ganglion stellate blockade with antero-lateral approach was performed, using 0.375 % Marcain. The following parameters were analysed: PaO2, SaO2, PaCO~, PIP and C~t~t. Results: In trauma patients with AIJ after bilateral ganglion stellate blockade short -lived and slight improvement of PaO 2 and SaO2, decrease of PaCOz and PIR and increase of static compliance of respiratory system were found. In second group bilateral ganglion stellate blockade interrupted the asthmatic status and significant statistical improvement of parameters of oxygenation, ventilation and respiratory system mechanics were observed. Conclusions: We suggest that the bilateral ganglion stellate blockade is a very useful method in treatment of patients with obstructive respiratory insufficiency. The aim of the study was to analyse whether there exists serum and urine electrolyte disorder in patients(pts.) with acute respiratory insufficiency(ARI). The study included t8 pts. with ARI (PaO2:8,24@1,49 kPa. PaCO2: 5,01i-0,77kPa, pH:7 42~:0,59, HCO3: 26,3:~8,10 mmol/1, SaO2 : 90,4~-7,42%) who were hospitally treated due to pneumonia(9 pts.),emboly of the pulmonary artery(3 pts.) and severe attack of bronchial asthma (6 pts). Among tham there were 12(66,7%) males and 6(33,3%) females, average age 51,5~:16,1 years, otherwise previously healthy. Electrolyte concentracions were measured at the onset of the disease in serum and urine collected during 24 hours (sodium-Na,potassium-K, chlorine-C1, calcium-Ca,magnesium-Mgand phosphorus-P). The measured serum and urine electrolyte concentrations were compared with respective referent values (RV). By serum electrolyte analysis, the following average velues were obtained: Na:l 4o,94 The object of our investigation was a group of 21 pts with massive pneumonias, 14 males (66.6%), 7 females (33.3%),mean age 55 yrs.Thirteen (62%) of them were smokers,8(38%) nonsmokers. Only 1 pt (4.7%) had pre-existing chronic respiratory disease, and 20 (95.2%) were admitted for the first lime,with no previous respiratory anamnesis. Diagnose was based on anamnestic data of productive cough in 15 pts(71.4%),physicaly ~onchial breathing in 19 I~S (90.4%),white cell count onder 10 x 109 /L in 18 pts(85.7%). Radiographicly, bilateral massive homogeneous shadows were found in 7 pts (33.3%), onilateral in 12 pts(57.1%),pleural effusion in 2 pts (9.52%). Abnormal renal function was found in 14 pts (66.6%). Sputum culture was positive in 8 pts (38%): Slr.pneumoniae, Str.pyogenes, Pse'udomonas aerug, in 4, 2, 2 cases respectively. All patients had remarcable hypoxernia (PaO2 range from 4,75 to 8,1 kPa) without hypercalmea. All patients needed oxygenotherapy together with antibiotics and other .symptomatic therapy. Nineteen pts had anaelioration of general condition and normalization of blood gas analyses, while 2 pts with the lowest hypoxcmia died.in conclusion, massive pneumonias are frequently followed by respiratory insufficiency which is one of the markers of pneumonia severity. As existing hypoxemia complicates the course of the disease,prolonges the recovery, makes therapy more complexe and may be cause of death , frequent blood gas measurement is recomanded. We studied the effects of bosentan (BOS), an ETa and ETb receptor antagonist, to examine if endogenous ET mediates pulmonary hypertension in anesthetized and ventilated dogs with acute lung injury due to oleic acid (OA). The gradient between pulmonary artery pressure (Ppa) and occluded Ppa (Ppao), and gas exchange (evaluated by arterial blood gases and SF6 intrapulmonary shunt) were measured at controlled flow. In 8 dogs (treatment), data were collected at baseline, during long injury (obtained 90 rain after intravenous administration of OA 0.06 ml/kg), and again after BOS (10 mg/kg intravenously). In 5 dogs (pretreatment), data were obtained at baseline, after BOS and then after OA. In treated dogs, OA increased (Ppa-Ppao, mmHg, table, means + SEM, * P < 0.05 vs base) and deteriorated gas exchange. After OA, BOS did not affect pulmonary vascular tone nor gas exchange. In pretreated dogs, BOS had no effect on baseline pulmonary vascular tone but prevented the increase in (Ppa-Ppao) after OA. The deterioration in gas exchange after OA was not influenced by BOS pretreatment. Objectives: The alveolar 02 tension is measured by the application of the alveolar air equation in which the arterial PCO 2 is used or by the simplified form of this equation in which the respiratory exchange ratio is taken at the value of 0.8. The purpose of this study was to estimate the effective alveolar 02 tension (PAO2eff) during spontaneous breathing with a new bedside technique which is simple non-invasive in 14 normal subjects and 27 patients with chronic bronchitis-emphysema. We also compared these values with the ideal alveolar PO 2 (PAO2(i)), measured from the alveolar air equation in which PaCO 2 was substituted by the effective alveolar PCO 2 (PACO2eff) and with the alveolar PO 2 measured from the simplified alveolar air equation (PA02). This study is complemantary to previous work for the estimation of PACO2eff. Methods: The subjects breathed quietly through the equipment assembly (mouthpiece monitoring ring, Fleisch transducer head) connected to a pneumotachograph and a fast response 02 and CO 2 analyzer. The method is a computerised calculation of the effective alveolar PO2quite similar to that of PACO2eff, obtained from the simultaneously recorded at the mouth expiratory flow, 02 and CO 2 concentration versus time curves. Results: The results showed a mean difference (PAO2eff-PA02(i)) of -0.061 kPa in normal subjects and -0,711 in patients. The mean of the difference (PAO2eff-PAQ 2) and (PAD2(i]-PAO z) was much greater than 0.281 in all subjects. The limits of agreement for the difference (PAOzeff-PA02(i))were -0.691 to 0.568 kPa in normal subjects and -2.040 to 0.596 in patients, while those for the differences (PAO2eff-PAD 2) and (PAO2(i)-PAD 2) were very large ( > -1.5 to > 1.7) in all subjects. Conclusions: The effective alveolar PO 2 is very close to the ideal one in normal subjects, tn patients PAO2eff may excessively deviate from PA02(i) due to the observed significant difference between the alveolar/tidal volume ratio for O 2 and that for CO 2. The alveolar PO 2 measured from the simplified alveolar air equation (PAO 2) differed substantially from PAO2eff and PAD2(i) in all subjects. The essential role of glucoprotein hormone erythropoietin is to control red cell production. Hypoxemia, reduced blood 02-carrying capacity and increased affinity of hemoglobin for 02 are the primary stimuli for erythropoietin production. Both anemia and hypoxemia induce rapidly erythropoietin secretion. Kidney erythropoietin RNA levels correlate inversely with Hematocrit and directly with plasma erythropoietin level. Similarly, hypoxemia increases kidney erythropoietin RNA and plasma erythropoietin. The effect of hyperoxemia (Pa02>lO0 mmHg) on erythropoietin secretion isn't very well understood. The purpose of this study was first to evaluate the erythropoietin secretion in patients with acute respiratory failure and second to determine the effect of hyperoxemia on erythropoietin secretion in patients with and without anemia. Sixteen patients with acute or acute on chronic respiratory failure needed mechanical ventilation were included in this study. These patient were divided in two groups. The patient who developed anemia were included in group I and the patients without anemia in group I1. Erythropoietin was estimated in venous blood in three stages. The first sample was taken during hypoxemia, the second during hyperoxemia and third during normoxemia. All the patients had high erythropoietin level during the hypoxemia period (mean value 98• mU/ml). During hyperoxemia etythropoietin levels were reduced in both groups ( mean value 21.6+15.2 mU/ml in group I, 36.8• mU/ml in group II). In normoxemia stage, erythropoietin increased again in anemic patients, and decreased more in the patients of group I1. We conclude that hyperroxemia inhibit erythropoietin secretion in spite of anemia and tow arterial oxygen content. Hyperoxemia may be a factor of the insisted anemia in with oxygen treated ICU patients. The purpose of this study was to determine the relationship between clinical features of acute lung injury (ALl) and parameters like total proteins, total and individual phospholipids, the presence of PAF, and acetylhydrolase activity in BAL of mechanically ventillated patients. Acetylhydrolase catalyses the cleavage of acetyl-group from the second position of the glycerylether backbone of PAF, leading to its inactivation. Mechanically ventillated patients were divided to three groups. Group I includes patients without ALl; group II, comprisespatients with moderate degree ALl, (1.0 2.5). Broncoalveolar lavage (BAL) was obtained after infusion of normal saline at 37~ to intubated patients and cooled immediately. Cells were removed after mild centrifugation (350 x g, 30 min, 4oc). Aliquots from the supernatant were used for total protein, phospholipid and PAF analysis and determination. Acetylhydrolase activity was assessed after incubation of BAL with 3H-PAF labelled on the acetyl group. Released label was measured by liquid scintillation counter in the supernatant after trichloroacetic acid precipitation of the non-reacted substrate. Kinetic characteristics of the enzymes were also studied. Total phospholipids appear reduced in BAL of patients with ALl, while total proteins increase. These factors appear to correlate with the severity of ALl. PAF was not present in BAL samples pretreatad with equal volume of 20% acetic acid to denaturate acetylhydrolase. Detection limit for PAF under our experimental conditions: 60 pg PAF/ml BAL. Instead, acetylhydrolase activity was detected in amounts increasing with the total protein content. Background: Intubated patients without lung injury or impaired breathing control normally display an inspiratory peak flow of below 1L/s. The aim of our study was to investigate the inspiratory peak flow generated by patients with acute respiratory insufficiency (ARI). We had to take into account that both an inspiratory pressure support (IPS) and the resistance of the endotracheal tube considerably influence the flow pattern generated by the patient. Patients and methods: To investigate the non-influenced flow pattern we developed a new ventilatory mode which automatically compensates for the flow-dependent resistance of the endotracheal tube (automatic tube compensation, ATC). Furthermore, the mode maintains a constant tracheal pressure in inspiration and expiratio n . Consequently, the measured flow pattern exactly corresponds to the flow pattern generated by the patient except that the ventilator modified for this mode (EVITA, Driiger Liibeck, Germany) was not able to deliver a gas flow of more than 2L]s. We have investigated 10 patients with ARI arising from different reasons. Results: The inspiratory peak flow measured in the ATC-mode was 1.7L/s _+0.3L/s. The maximal deliverable flow of 2L/s was obtained in 3 of 10 patients. The figure shows the flow pattern under ATC and IPS in [~s] oi:) one of these patients. Conclusions: Patients with ARI display a highly increased inspiratory peak flow. Ventilators used for spontaneous breathing should therefore be able to deliver a gas flow of more than 2L/s. An overproduction of NO and reactive oxygen species (ROS) has been demonstratred in septic shock. ROS and nitric oxide (.NO) are free radicals which are known to react together leading to peroxynitrite anions that can decompose to form nitrogen dioxide (NO2) and hydroxyl radical (OH~ Thus, NO has been reported to have a dual effect on lipid peroxidation (prooxydant via the peroxinitrite or antioxidant via the chelation of ROS). In the present study we have investigated in different models the in vitro and in vivo action of NO on lipid peroxidation. Copper-induced LDL oxidation was used as an in vitro model of lipid peroxidation. LDL (100 ~g ApoB/ml) was incubated with Cu 2+ (2,5 ~tM) in presence or absence of NO donor (sodium nitroprussiate or Glutathione-NO) from 10 to 500 ~M. Oxidation of LDL was monitored continuously with conjugated diene formation (234 nm) and 4 hydroxy nonenal accumulation (HNE). Exogenous NO prevents in a dose dependent maner the progress of Copperinduced oxidation. Ischaemia-reperfusion injury (I/R), characterized by an overproduction of ROS, is used as an in vivo model. Anaesthetized rats were submitted to 1 hour renal isehaemia following by 2 hours of reperfusion. Sham operated rats (SOP) were used as control. Lipid peroxidation was evaluated by measuring the HNE accumulated in rat kidneys in presence or absence of L-arginine or D-arginine infusion. L-arginine, but not Darginine, enhances HNE accumulation in I/R but not in SOP (<0.05 nmol/g tissue in SOP versus 0.6 nmol/g tissue in I/R), showing that in this experimental conditions, NO produced from L-arginine, enhances the toxicity of ROS. This study shows that the pro-or antioxydant effects of NO are different in vivo and in vitro and could be driven by environemental conditions such as pH, relative concentration of NO and ROS, ferryl species...These conditions are impaired in circulatory shock. Methods:" The diagnostic and therapeutic approach was Standardized so that data collected over a 10-year period were comparable. A progressive deterioration of clinical conditions and/or pulmonary gas exchanges was considered as indication for MY. Variables potentially predicting the need for HV were derived from clinical and arterial gas data, extrapulmonary diseases, use of drugs, chest x-ray and ECG abnormalities. Results: RV, performed with external and/or internal ventilators, was necessary in 130 patients (22%). At the hospital admission, PaC02 was higher and pH was lower in patients requiring RV ( Pneumomediastinum, pneumothorax, ateleetasis and myocardial infarction are rarely seen in bronchial asthma. These complications occur as a result of the severe asthma.The aim of our retrospective study was to analyse the complications Seen in acute asthma attacks. During the years 1990 through 1994, 244 patients were admitted to hospital in acute asthma episode. There were 11 (4,5%) pts with complications; mean age of 27 yrs; 6 females (54%). Clinical history, ECG and chest radiogr~hs were analysed. The mean duration of bronchial asthma was 14 yrs (range from 2 months to 17 yrs), All patients were atopics. There were four ex-smokem and one smoker. The worsening of asthma symptoms begun two days before the admission (range from 1 to 7 days). On ECG all patients had tschycardia. Rightward shift of the QRS axis and ST-T changes indicative of right ventrieutur strain were found in three pts. These were the transient fmdings that improved after curing the acute asthma attack. Non-Q myocardial infarction oeeured in one patlent and resulted from the hypoxaemia of asthma. Hyperinfl~ion was the usual finding on the chest radiograpk Pneumomediastinum and subcutaneous emphysema were apparent in five pts and required no additional treatment Unilateral pneumothoraccs were present in two pts and needed eontimous intrapleural drainage; one of these patienst died in eardiorespiratory insufficiency. Ateleetasis of right upper lobe was present in one patient. It oceured due to inspissated secretions and needed no additional treatment All these patients, except one who died, improved on lreaanent with oxygcr~ steroids, beta-two agonists, theophylline and antibiotics. In conclusion, complications occur in acute asthma episodes as a result of the severe asthma Mediastir,*l emphysema and atelectasis are not serious complications. Pneumothorax and myocardial infarction are very serious life-treatening complications and always have to I:m considered in taati~ts with sev~ asthma. Acute bronchial asthmatic episodes represent one of the most common respiratory mnergendes, Its maximmum expression "status asthmatiens" is one entity of low incidence, still it is a risk to the physical integrity of the patient. During 1993 a total of 52 patients with diagnosis of status asthmabcas were hospitalized. Out of these palients six had a near-fatsl asthma and they were subjected to a complex examination. Near-fatal asthma was defined as either respiratory arrest or acute asttuua with PaCO2 greater than 6,7 kPa and/or an altered state of consciousness. Mean age was 56,2-d:16,2 yrs, four male and two female sex. At presentation two patients suffered from coma, others were confused. They exh'bited severe dystmoes, diffieul~ speaking, used accessory muscles of respiration, increased whee~tg while two cases had silent chest on auscultation. Cyanosis indicated a very severe asthma attack in all six patients. Mean respiratory rate was 28~4/min and puts rate 118.d: 12 bts/imn. Arterial blood gases revealed a PaO2 of 6,95~1,33 kPa, PaCO2 of 7,87• kPa and pH of 7,274-+-0,132. ARea-careful evaluation they received conventional therapy (immediately continuous oxygen, impelled nebulization with high doses of betatwo agonists and ipmtropium bromide, intmvanous st~oids and theophylline). In two eases signs and symptoms of deteriorating airflow and respiratory muscle fatigue determined the need for mechanical ventilation. Out of six near-fatal attacks aggressive lrealanent was suscessfull in four patients and fatal in two eases. One patient admittcxl in coma died in severe hypoxae~a upon one hour and one mechanicaly ventilated died from cardiac arrhythmia. Life-threatening attacks in asthmatics in our group developed gradual worsening despite neatment which r symptoms in most other patients. One patient had "brittle asthma", other long-standing acute episodes Ireated with systemic steroids. Conclusions: Idantitiechon of fatality prone subjects may lead to fttrther Muetion of seveze episodes. Respiratory affest and coma upon admission, severe dyspnoca with silent chest on ausouhation, oyanusis and use of accessory muscles of respiration constitute the basic cfinieal picture. Hypoxasmia must be immediately eon'ected.The patients and physicians should be able to assess the severity of asthma, a major factor in near-fatal and fatal asthma attacks. Objectives :Our purpose was to asses if the evolution of patients with a adult respiratory distress syndrome (ARDS) ,shows any relation to the pulmonary or systemic origin of the disease and whether or not there were differences in the frequency of the syndrome in both groups. Methods : Randomized prospective study in multidisciplinary ICU. One hundred and sixteen patients with a high risk developing ARDS were distributed into two groups. One was named systemic origin group(SO) and the other pulmonary origth group (PO).AI1 patients only showed one cause (pulmonary or systemic) with potential risk of ARDS.The patient's hemodynamic and respiratory status was evaluated every 6 hours the first day and every 12 hours the second and third day. At the end of 72 hours the patients were diagnosed as ARDS or non-ARDS. Measurements and Main Results : Of the total 116 patients, 57 were finally included in the SO group and 59 in the PO group.Patients in SO group and PO group had comparable ages (p<.01).PEEP in both groups was comparable (=.06) at the mmnent of admission to the study. There were no statistically significant differences for Cardiac Index and Systemic Vascular Resistances. The Pulmonary Vascular Resistances (PVR) showed significant differences at 48 h.(p<.05) and 72 h. (p<.03).The oxygen comsumption (VO) in patients of the SO group showed statistically significant differences at 48 h. (p<.05) with respect to initial values.Fifteen cases of ARDS (26.3%) in the SO group and twenty five cases (42.3%) in the PO group were identified. The time of onset of ARDS was 35_+ 14 hours in the SO group and 11 + 4 b hours in the PO group.The final outcome was very similar th both groups : mortality of 36% in the SO group versus 37% in the PC group. Conclusions : The pathogenesis of ARDS depends on whether the lesion is originated at or outside the lung. The PO group showed a sborter thne of onset of ARDS, a faster and more severe increase of pulmonary shunt and a higher percentage of patients developing ARDS compared with patients of the SO group.The SO group showed a higher and faster increase in puhnonary resitances tbat PO group and a decrease th oxygen comsumption earlier and more severe than in the PO group. These data thus seem to show that there could be two mechanisms involved in the genesis of ARDS depending on the cause. The fact that the ARDS genesis is shorter in the cases of pulmonary etiology with faster impairment of pulmonary shunt, and a slower increase in pulmonary resistances in this pulmonary group, would indicate that the underlying mechanisms responsible for the hypoxemia are different to those which thitiate the increase in pulmonary resistances. Finally, the exclusive inapairinent of oxygen consumption, which appears earlier than the onset of ARDS in the systemic origth group, could show the generalized character of the process in this group. Perfusion of prostacyclin (PGI2) to treat pulmonary hypertension in adult respiratory distress syndrome (ARDS) worse pulmonary gas exchange due to a marked impairement of ventilation/perfusion mismatch. Recently has been shown that if prostacyclin is given by aerosol instead of intravenous the net effect is an improvement of arterial oxigenation due to a redistribution of blood flow to well ventilated areas. Objectives: To asses the effects of inhaled proatacyclin on pulmonary haemodynamics and gas exchange in patients with severe ARDS. Methods : Two patients with severe ARDS (Murray Score >3) recived inhaled PGI 2 at 15-20 ng.kg.min "1 using an ultrasonic nebulizer. Haemodynamic measurements, arterial and mixed venous blood gas analysis were performed before and after 30 rain of PGI inhalation. Results: Short-terro P~I 2 inhalation improved pulmonary g-~ e-'~hange in both patients. Arterial oxygen partial pressure (PaO2) increased from 101 to 166 mmHg in patient 1 and from 87 to 108 in patient 2, the ratio PaO to the fraction of inspired oxygen increased from 1262 to 207 (patient 1) and from 124 to 154 (patient 2). Venous admixture decreased from 36% to 29% and from 34% to 27% in patient 1 and 2 respectively. Mean pulmonary artery pressure decreased slightly from 25 to 23 mmHg in patient 1 and from 41 to 37 mmHg in patient 2. No effects on systemic haemodynamics were observed in any patient. Conclusions: PGI 2 inhalation improves gas exchange and produces selective pulmonary vaaodilation, thus can be an alternative therapy for the treatment of pulmonary hypertension and hypexemia in patients with severe respiratory falllure. Methods: We treated 67 ARDS-patients (age 41 yr (16-75) mean, range) during 1991-94. The lowest PaO2/FiO2-ratio was 74 (29-140), the worst Murray score 3.0 (2.3-4.0), ICU-stay 41 (1-121) days and hospital mortality 40%. The costs of intensive care were calculated according to intensivity of patient care as assessed by TISS-scoring (Therapeutic Intervention Scoring System). The more intensive the care, the higher are the costs. Costs per year of life saved (=life-year" in US $) were compaired by other medical treatments (1-4). It is assumed that the mean expected length of remaining life in ARDS-survivors after intensive care is 25 years. Treatment life-year ($) ' Bone marrow transplantation (acute leukemia) 65 000 Lowering cholesterol using Iovastatin 51 000 Treating hypertension using nifedipine 32 900 Heart transplantation 28 000 Intensive care of ARDS-patients 3300 Conclusions: Intensive care of patients with severe ARDS is highly more cost-effective as compared with many other routinely used medical treatment strategies, The usually good recovery and the reasonable quality of life in survivors justifies investments to care of these patients (5). There is a close correlation between these two methods of measuring EVLW. However there is an underestimation of 38.5 % in this kind of pulmonary edema ( oleie acid induced ) with the double dilution method. Although the size of the sample is small, in normal lungs there appear not to be this underestimation. The effect of PEEP on EVLW has been studied with contradictory results, probably as a consequence oft differences in methods of measuring EVLW, variations in the type and severity of lung injury, and different timings of PEEP application. Objective= 1) to analyse the effect of different levels of PEEP (0, 10 and 20omH20) on EVLW during HPE; 2) to establish whether increases in intrathoracic pressure due to high PEEP levels can obstruct lymphatic drainage. Material and Methodet HPE was provoked in 3 groups of dogs by inflating a Foley catheter in left auricular to a pressure of 24-26 r~uHg. PEEP levels of 0, i0 or 20 m~Hg were applied. Resultst Objective: To assess the effect on extravascular lung water (EVLW) of the application of PEEP and the reduction of Vt in an oleic acid pulmonary edema model in pigs, using three ventila~ary strategies. Material and Methods: Twelve adolescent pigs (weighing over 30 kg) were randomly divided in three gmups immediately alter infusing via a central vein 0.1 ml/kg of oleic acid to produce a permeability pulmonary edema. The ventilatory parameters for each group were as follows: GROUP I (n=4) : Vt: 10-15 ml/kg; ZEEP. GROUP 2:(n=4) : Vt: 10-15 ml/kg; PEEP: 10 cm H20. GROUP 3:(n=4) : Vt: 5-10 ml/kg; PEEP: 10 emil20. (Resulting in permissive hypereapnla) After a four-hour period of ventilation the animals were killed and the lungs excised to calculate gravimetrically the extravascular lung water using a standardized procedure ( hemoglobin content method ). Ill EVLW (ml/kg) GROUP Obiective: In the postoperative period, maintenance of adeguate arterial oxygen tension is a major problem in morbidly obese patients probably because of a large reduction in functional residual capacity (FRC). The aim of this study was to evaluate the effects of PEEP on respiratory mechamcs and gas exchange in this kind of patients. Methods: In nine postoperative mechanically ventilated morbidly obese patients (BMI>40 kg/m 2) we partitioned the total respiratory system mechanics into its lung (1) and chest wall (w) components using the airway occlusion technique associated with the esophageal balloon, during constant flow inflation (JAP 1989; 67: 2556) . At three different levels of PEEP (0, 5, 10 cmH20 ) we measured: compliance (Cst), airway (Rim) and "additional" (DR) resistance, FRC and gas exchange. Obiectives. To describe the use of prone position in our ICU we analyzed the clinical records of all patients admitted in 1993-94, selecting adult patients with ARF defined as: intubation and PaO2/FiO2<250 mmHg plus an FIO2>0.5 or PEEP>5 cm I120. Results. 146 patients met the ARF criteria: 40 of them (27.4%) underwent prone positioning (P+). Prone position use began in the early phase of ARF (3.5• days from the beginning, range 1-32, median 2).25 out of 40 P+ pts were treated with controlled ventilation (CPPV or PCV), while 14 were on assisted ventilation (SIMV+PS) and 1 on spontaneous breathing (CPAP). Only 2 pts were awake when turned prone, while 11 pts required adjuncts of sedation to tolerate the change of position. The duration of prone positioning was variable (average lenght 4.7• h, range 0.5-12 h). Only minor side effects were observed (eyelids and facial edema, chest and facial pressure bruises). We consider responders (R+) those patients presenting at least 12.5 mmHg increase in PaO2/FiO2:35/40 patients (87.5 %.) were responders when first pruned. The PaO2/FiO 2 changes induced by prone position are reported in the figure. PaO2/FiO 2 increased when patients were pruned (*p<0.001) and remained higher than baseline values when returning supine(*p<0.001). PaCO 2 remained unchanged. Prone positioning was used at least twice in 21/40 ( Conclusions. This retrospective analysis confirms that prone positioning improves oxTgenation in the majorib' of ARF patients. Altough we have no available criteria to discriminate in advance R+ from R-pts, we now routinely consider the use of prone position in the treatment of severe ARF. Palo A, Otivei M*, Galbusera C, Veronesi R, Sala Gallini G, Zanierato M, Iotti G, Braschi A.Servizio Anest. e Rianim. I, *Laboratorio Biotecnologie e Tecnologie Biomediche IRCCS S. Matteo, Pavia, Italy Inhaled NO can improve arterial oxygenation and reduce pulmonary hypertension in ARDS patients; little information is, however, available about the dose-response curves. Methods Seven ARDS patients (LIS 2.7+.5) submitted to mechanical ventilation randomly received 8 inhaled NO doses in increasing or decreasing sequence: 0.5, 1, 5, 10, 20, 50 and 100 ppm. Reference measurements were obtained before and after the entire period of NO inhalation. Hemodynamic parameters and blood gases were measured after 25 min in each condition. CMV was administered under sedation and paralysis, with constant ventilation, PEEP (lOL-_2 cmH20) and Fit2 (.56+.14). The changes in Vt and Fit2 due to the NO (1000 ppm in N2) injection in the ventilator external circuit were compensated for. Results .34 The dose of 0.5 ppm, ineffective on PAPm, significantly improved oxygenation. The increase of Pat2 and the decrease of Q'va/Q' and PAPm were nearly maximal at 5-10 ppm. No deterioration of arterial oxygenation was observed at NO doses as high as 100 ppm. CO2 exchange was not influenced by NO inhalation. Systemic hemodynamic variables did not change throughout the study. These results suggest that a concentration around 10 ppm is adequate for obtaining maximum effects on hypoxemia and pulmonary hypertension in patients with ARDS. Low-dose inhaled nitric oxide (NO) induces redistribution of pulmonary perfusion in patients with severe ARDS and causes improvement of oxygenation [1] . However, addition of exogenous lowdose NO in the inspiratory gas mixture might be only a replacement of missing atmospheric NO (2-130 ppb) in hospital central-supplied medical air. [2] We have realised nitric oxide measurements in ten healthy volunteers, (4 smokers and 6 non-smokers) breathing with a mouthpiece and occluded nostrils through a ventilator circuit, with separation of inhaled and exhaled gases by a valve. NO concentration was measured with a double-chamber chemiluminometer (Environnement SA, France) and with charcoal/silicate purified compressed air. There was no nitric oxide detectable in the inspirat0ry limb of the ventilator. Unfiltered central supply medical air contained :20 -50 ppb of NO and 10 -30 ppb of NO2, whereas central supplied oxygen was NO/NO 2 free. Samples were taken after equilibration periods of 5 minutes, with increasing Fit2 levels of 0.21, 0.50 and 1.0 for subsequent 5 minutes periods; paired values were recorded every 30 s. The mean NO value was 4.57 ppb (SD 2.51) and n o significant differences were found for different Fit2 levels both in smokers and non-smokers. These data suggest that the NO concentration of pulmonary origin in the exhaled air of' healthy volunteers is probably lower than that reported by other Authors [2] and that, previously reported, differences between smokers and non-smokers are not always striking [3] . We suggest the use of activated charcoal/silicate filters for clinical trials in order to achieve standard conditions. [ Objective: To compare efficacy and safety of two doses of salbutamol. Methods: Sixteen adults who had severe acute a~hma were randomly assigned to receive either 10rag (n=9) or 5rag (n=7) of nebulized sulbutamol. Both groups were similar with respect to age, duration of a~hma, duration of attack before arrival at the hospital and severity of a~hma according to baseline measurements (table) . Evaluation was performed 30, 60, and 120 rain after the start of nebulization. Results: Compared with 10mg regimen, 5mg regimen resulted in the same improvement in peak-flow and Fischl index (figure). The changes in heart rate, respiratory rate and pace2 did not differ significantly between both groups. The incidence of side effects, which included tremor, palpitations, cardiac arrythmlas and other symptoms, was not sJ~ificanfly different in the two populations. Conclusion:The results of this study suggest that nebulization of 10ng of salbutamol is not more effective than 5rag in the initial treatment of acute severe asthma in adult patients. The prognostic factors of neutropenic patients admitted to the ICU remain poorly known. The aim of this study was to determine the respective weight of underlying malignancy and organ system failures on the outcome of these patients. Patients and methods: The charts of 107 neutropenic patients (WBC < 1000/mm 3 and/or PMN < 500/ram3), admitted to the ICU between 1986 and 1990, were retrospectively reviewed. The characteristics of the neoplastic disease (h~emopathy or solid tumor, tumoral evolution, duration of cancer disease and of neutropenia), the Mac Cabe's score, the organ system (respiratory, hemodynamic, renal, neurologic, hepatic) failures and the severity scores (SAPS, SAPS II ,OSF) were registred within the 1 st day in the ICU. When discharged from the ICU, the patients were classified as alive or dead. Results: Fifty-seven patients (53.3%) had a h~ematologic malignancy, and 50 (46.7%) a solid tumor. Fifty-nine of the 107 patients died (55.1%); the mortality rate did not differ between both groups (61.4 and 48% respectively, p = 0.16). With univariate analysis, none of the tumoral features is linked to the prognosis; only the respiratory (p < 10 -4) and cardiovascular (p < 10 -3) failures, and the number of organ system failures (p < 10 -4) are associated to the risk of death. The SAPS (p < 10 -3) and SAPS II scores (p < 10 -4) were higher in patients who died. With multivariate analysis (logistic regression), only the respiratory failure is correlated to the risk of death (p = 10-4); neither the features of the underlying malignancy (p > 0.8), nor the duration of neutropenia before admission in ICU (p = 0.83), nor the severity scores fIGS II: p = 0.068) are linked to the outcome. Conclusions: The tumoral characteristics do not modify the prognosis after admission to the ICU. They should not influence the decision to admit or refuse a cancer patient in the ICU. Respiratory failure at ICU admission has the predominent weight on the risk of death in the ICU. Patients with respiratory acidosis due to asthma occasionally require levels of mechanical ventilation that place them at risk for barotrauma. A few case reports have described the use of an extra-corporeal membrane oxygenator(ECMO) circuit as an alternative means of CO 9 removal. Generally, this has been used for short periods of time (<24h) without serious complications and with low blood flows through the extra-corporeal circuit. We report a case of refractory asthma who could not tolerate even small-volume breaths from a mechanical ventilator due to severe bilateral airleak. ECMO therapy was initiated at the referring hospital prior to helicoptor transport. High blood flows were used (70% of the patient's cardiac output), sufficient to achieve both CO 2 removal and oxygenation. Satisfactory gas-exchanged was accomplished (pCO2=50-60 mmHg) with nearly total lung rest for a prolonged period (60h). However, the long ECMO duration was associated with two severe complica-ti0ns:1) bilateral hemothoraces due to anticoagu!ation in the extra-corporeal circuit, and 2) prolonged weakness as a result of neuromuscular blockade for six days. The patient was discharged from the hospital in good condition. We present the respiratory and hemodynamic features of this case aw well as the potential complications of ECMO therapy in asthma. Objectives: Parameters derived from tidal expiratory flow ~e) and volume (VT) can be used to detect airflow obstruction in COPD patients who might be unable to perform forced spirometry (e.g., ICU). However, indices such as AVe/V T and At/re are highly variable (Thorax, 1981: 36; 135) . Methods: We investigated whether the standardized for V m effective time (teff~) of a tidal breath, which is derived by asimple mathematical procedure (teff,= J'Vdt/VT2), is a more reproducible and sensitive detector of airways obstruction, We studied nine normal subjects (5 male, 31 -+5yr) and 13 COPD patients (4 male, 61 -+10yr) in the seated position, with a noseclip on. They breathed quietly, through a pneumotashograph to measure flow (V). Volume was obtained by numerical integration of thellow signal. Each subject had an initial 10-15 min trial run, in order to become accustomed to the apparatus and procedure. When regular breathing had been achieved, all breaths over a5 min time interval were recorded. The mean value of six consecutive breaths (ERS criteria) for each subject was used for analysis under the condition that within session variation of tidal volume (VT) was <10%. Lung function tests were: in normals (mean-SD), FEVl%pred = 100• FEVl/FVC%=81-+3% , and in COPD patients, FEV~%pred=53__.20 and FEVI/FVC%=51 --.12%. Results: Values are shown as mean-..+-SD in the following A su~ve~ os literature sources p~oves that t~aditlona], i.e. medicinal medication and physiothe~apeutic methods os t~eatment often p~ove to be insufficientl~ effective both currently and in the ~emote future. The goal of this study was to investigate the efficacy os t~eatment of b~onchial astI~ma patients by means os speleo-and artificial sp~ay therapy. Speleotherapy t~eatment was conducted in the conditions os mic~oclimate os salt mine in Solotvino hospital. A~tis sp~ay the-~apy was conducted by means os a self-made device. Ou~ method is based on the p~inci-~ le os using the majo~ facto~ of speleo-he~apy -highly dispe~sed sp~ay 0s sodium chloride. The obtained ~esults ~e~e analyzed in five g~adations. At the end os the speleothe~apy improvement and considerable improvement was observed in 75,0~ os patients; inconsiderable improvement -in 15,9~ os patients. Having evaluated the e~s os t~eatment using a~tis sp~ay therapy the indices a~e 68,7H and 20,5~ ~espectively. Remote ~esults of t~eatment a~e an important index os t~eatment, the ~esult os ~hich ~e~e studied by means 0s a ~uestionnaive-method. Patients ~ho had been t~eated by speleothe~apy mo~e f~eguently ~e-po~ted a ~elapse in disease 3ust afte~ the course o~ t~eatment (29,3H). Ho~eve~, in a ]ate~ phase the ~emission ~ould last ]on-~e~ (s 6 months in 84,5~ os patients, till one yea~ in 69~9~). In 12,5~ os patients who passed the co~se os a~tificial sp~ay therapy a ~elapse was ~egiste~ed immediately as the co~se os t~eatment. Then thei~ condition stabilized ~hile in 73,5~ os patients a period os ~emission lasted s ha]s a yea~. 42,9~ of patients dida't ~epo~t a ~elapse of the disease du~in~ one yea~. Evangelismos Hospital, Critical Care Department, Athens, Greece Method#: 19 mechanically ventilated patients (5 COPD, 6 ARDS, 8 other pulmonary diseases) were studied in two phases: 1) During the acute phase of respiratory failure; 2) During recovery 5-73 days later. We measured MIP and monitored the pattern of breathing while the patients were breathing spontaneously through the respirator (pressure support mode with 3-8 cmH20) until either the point they were unable to sustain spontaneous breathing (SB) any longer (phase 1) or for two hours when they could sustain SB indefinitely (phase 2). Subsequently the patients were sedated, paralyzed and mechanically ventilated. Then we simulated the pattern of SB at the end of the SB trial by manipulating the variables of the ventilator and assessed respiratory mechanics b y the end-inspiratory and end-expiratory occlusion technique. 1. During recovery, a combination of reduced inspiratory load and increased venfilatory capability makes a patient previously unable to sustain SB to breathe spontaneously. 2. Inspiratory load is reduced during recovery, mainly because both intrinsic PEEP and breathing frequency are diminished. Obiectives: Although elevated concentrations of a few cytokines have been shown to be present in the bronchoalveolar lavage (BAL) fluid (BALF) of patients with the adult (acute) respiratory distress syndrome (ARDS), the pethogenesis of ARDS is largely unknown. Leukemia inhibitory factor (LIF), a growth factor recently recognised as a polyfunctional cytokine integrated in cytokine networks was measured in unconcentrated BALF of patients from different patient groups. Methods: LIF was measured in BALF by means of a specific and sensitive ELISA (detection limit 10 pg/ml)in BALF (lavage of 3 x 50 ml in the right middle lobe). Results: LIF was not detected in the BALF of 13 healthy control patients and in only one (34 pg/ml) out of 26 patients at risk for ARDS (after cadiopulmonary bypass surgery) who underwent BAL 4 h after the end of the extracorporeal circulation. High and detectable levels were found in the unconcentrated BALF of 10 out of 12 patients with full-blown ARDS (196 + 80, mean + SEM, range 10-985 pg/ml). There was a good correlation between the level of LIF in the BALF and a number of markers of inflammation: neutrophils/ml (R:0.70, P= 0.01), albumin ( R:0.75, P=0.008) and protein level (R:0.74, P=0.006). Conclusions:The biological role of LIF in these BALFs is not readily explained by its currently known actions and it is unkwon whether LIF contributes to or is a response to local tissue damage. Our results indicate that this cytokine with Lots of Interesting _Functions is a pert of the inflammatory cytokine cascade in ARDS. Background and Obiective : We recently demonstrated that cisapride -a new prokinetic drug -enhanced enteral feeding in a heter0genoas group of ventilated ICU patients by significantly accelerating their gastric clearance (Crit Care Meal, 1995 ; 23 : 481-485) . It remains unknown, however, whether certain subgroups of patients might benefit more from adding cisapfide to their enteral nutrition regimen than others. Patients with chronic obstructive pulmonary disease (COPD) might represent such a subgroup since their illness and its specific treatment put them at risk for gastric emptying disorders. Design and setting : Prospective, consecutive sample study in an adult medical intensive care unit in a university hospital. Patients : 10 mechanically ventilated and hemodynamically stable COPD patients. Interventions : Gastric emptying was evaluated by bedside scintigraphy and expressed as the time at which 50% of a Tcg~-labelled test meal was eliminated from the stomach (T 1/2). Baseline data (dO) were recorded after enteral nutrition reached 1500 to 2000 ml daily. Scintigraphic measurements were repeated 4 days after cisapride (10 ml orally, q.i.d) had been added to this regimen (d4). Patients were considered cisapride responders when gastric clearance improved by more than 50% from baseline. Results : Normal values for the test meal and for scintigraphic acquisitions obtained in the supine position were found to be 31 + 15 min. in healthy volunteers (Crit Care Med, 1995 ; 23 : 481-485) . Five patients responded to cisapride (T 1/2 : 81 + 31 rain vs. 26 + 10 min at dO and d4, respectively) and five did not (T 1/2 : 36 + 18 min vs. 33 _+ 11 rain at dO and d4, respectively). In contrast with non-responders, all five responders had clinically significant maldigestion at baseline (excessive (> 150 ml) gastric residues, vomiting (> 3 times/day and abdominal distension) which disappeared in 4 of them after the administration of cisapride. Conclusion : COPD patients who tolerate enteral nutrition well have basal gastric emptying times which are comparable with those of healthy volunteers and are not influenced by cisapride. However, cisapride treatment provides both scintigraphic and clinical improvement in those COPD patients who exhibit clinically obvious gastric emptying disorders. Cernv V., Dostal P., Zivny P., Zabka L. Dept. of Anesth. and Critical Care, Charles University, Faculty Hospital, I-Irade~ Kralove 500 36, Czech republic Objective: The aim of the study was to evaluate the effect of early entera nutrition started within 24 hours of injury on the incidence of multiple orgar failure (MOF) in trauma patients requiring vantilatory support. Methods: After institutional approval 25 patients were enrolled in the study Enteral feeding was begun within 24 hours of injury in 14 trauma patients (EN group) admitted to ICU. Nasuenteric tube was placed as soon as possible after admission into the distal duodenum under endoscopy. Additional parenteral nutrition was used to meet patients energy and protein requirements. The control group (PN) consisted of 11 patients fed during this period paretuerally. Severity score Apache II, Trauma score, cumulative balance of nitrogen (g), incidence of MOF (three and more organs) and length of ventilatury support (days) were calculated. Values are expressed as mean + SD. Results: Tab Introduction : Parenteral nutrition (PN) is an important aspect in the optimal treatment of patients on gastroenterology or intensive care. The aim of this bi-center study in 38 patients has been to assess tolerence and efficacy of a new protein-lipid mixture for PN from a simple preparation. Patients and m~hods : Patients were selected in two hospitals (Tenon and Saint-Lazare, Paris) and were divided into two groups : group A (Gastroenterology~ 1 l short bowel syndrome) and group B (Intensive Care, 27 surgical patients). All patients likely to require Pig for a period of 10 days (group A) or 7 days (group B) were studied. The PN regimens administered were the following : combination with 50 g of MCT/LCT fat emulsion end 9,6 g of nitrogen, in 1 liter end glucose requirements were met by imfizsion of l liter of glucose 20-30 % via a "Y " connection. Lipid thus provided 3040 % of the non introgen calories. Total daily calorie intake was 1540 to ] 940 kced. This study monitored, before and at the end of infusions, the sennn Albumin (Alb), preAIburtun (preAlb), triglycendes (TG), cholesterol (CS), and the serum ammotransferases (SGOT and SGPT) end alkaline phosphatase (ALP) activities. Statistical significances were calculated using the Wilcoxon-tost. Introduction: Many 1CU patients present a catabolic illness in response to inflammation and infection, characterized by a rapid loss in skeletal-muscle mass despite optimal nutritional support. Growth hormone (GH) is responsible for a rise of lipolysis, enhancing the energetic balance, and of protein synthesis. Recombinant human GH (rhGH) is nowaday available for clinical use, but its cost is very high. Therefore, rhGH should only be prescribed to ICU patients when its efficacy can reasonably be anticipated (ie. when the patients are catabolic or stressed, but in order to avoid overprescription for unstressed patients and for those who are overly catabolic). Hence, we, as others, recently demonstrated that rhGH had no favorable effect in highly stressed ICU patients. Objective: To detect on a clinical basis, low (LS), mild (MS) and severe stress (SS) states in ICU patients and validate this clinical judgement by objective metabolic mesurements, in order to select early those ICU patients potentially able to benefit from rhGH therapy. Methods: 36 consecutive ICU patients were prospectively stratified as LS, MS and SS by two experienced ICU senior consultants (temperature; agitation; heart rate; arterial blood pressure; presence of an infection; respiratory rate; exogenous catecholamines). Anabolic (insulin, IGF-1, GH) and catabolic (cortisol, ghicagon) hormones, and nitrogen balance were determined for each patient within 8 hours after admission in the ICU. Metabolic and clinical data were then compared. The clinical stress states determined by ICU physicians correlate with an objective metabolic assessment. Therefore, the patients who will more likely benefit from adjuvant rhGH therapy can be detected simply and early. A prospective study on rhGH therapy in MS ICU patients is in progress. Berger MM MD 1, Chiolero R MD 1, Pannatier A PhD 2, Berger L 2, Cayeux C 1, Voirol P 2, Hurni M MD 3. 1 Surgical ICU, 2 Pharmacy, and 3 Cardiac Surgery, CHU Vaudois, CH-IOtl Lausanne, Switzerland Objective. Nutrition of the compromised cardiac surgical patient is challenging. Numerous factors influence the gastrointestinal (GI) absorption function, among which gut perfusion, which depends largely on the systemic hemodynamic status. Patients in hemodynamic failure are prone to organ failure, and may benefit from an early jejunal feeding. The study was designed to assess the absorption function after cardiac surgery in patients with adequate and altered hemodynamic status, using paracetamol as tracer of GI absorption. Methods. After cardiac surgery, 24 patients, aged 63_+8 years (mean_+SD) were assigned to 2 groups (anaesthesia: fentanyl 20 gg/kg + midazolam): Group 1 (n=10): reference group, with normal hemodynamic status, easy recovery. Group 2 ('n=14): patients in low output syndrome, cardiac index < 2.5 I/m2 on day 1 (D1) after surgery, requiring prolonged intensive care, mechanical ventilation + nutritional support. Paracetamol 1 g, was given intragastrically on D1 + D3: plasma levels measured (H.P.L.C), at administration (TO), T30-60-90-120-180-240 and 480 rain. Hemodynamic status assessed with pulmonary artery catheter. 5 healthy subjects served as controls. Results. Compared to healthy controls, absorption was strongly reduced on D1 in all patients (no difference between groups). On D3, peak paracetamol level was significantly lower in Group 2 (low cardiac output): in Group 2 the area under the curve on D1 and D3 were similar. There was a large inter-patient variability, reflecting the hemodynamic status. Conclusion. GI absorption was decreased on D1 in all patients, and reverted to normal between D2 and D3 in case of normal cardiac function, but not in case of low output syndrome. The decrease on D1 can be attributed to fentanyl, known to slow down the GI transit. In patients with cardiac failure, correction of altered absorption was correlated with the hemodynamic status, suggesting that GI absorption is dependent on adequate splanchnic perfusion. The aim of the work was to define specific significance and evaluate efficiency of enteral component of infusion therapy in the intensive care of gastroenterotogic patients of surgical profile with pyo-septic complecations. There were used the methods of radial diagnostics and polyelectrography; the laboratory control on oxygen-transporting function, volumetric and hemodynamic state, changes in metabolic, hormonal and immunologic status was conducted. From January, [992 till November, 1994 there was carried out the randomized study of 155 patients with general purulent peritonitis; among them 70 persons constituted the control group and 85 -the main one. In the main g~oup the intestinal lavage, enterosorption, enteral introduction of nutrient solutions with gradual turn to enteral nutrition by equalized mixture "Ovolaet" were started from the first hours after operation. The data obtained allowed to define the specifity of the program of artificial medical nutrition in the group of examined patients, based on necessity of individual selection of media for enteral introduction depending on the stages of intestinal insufficiency syndrome. It was shown that inclusion of enteral component into the program of infusion therapy during early periods stabilized circulation in the regime of moderate hyperdynamia, considerably decreases the deficiency of circulating blood volume, normalizes the values of oxygen transport, consumption an}d extraction, provides the optimal level of mycardial adaptive possibilities without tension of its compensatory functions and pulmonary circulation overload. Due to combined application of parenteral and enteral nutrition the metabolic processes are shifted towards anabolism. This is supported by decrease to normal values in the contents of blood aggresive hormones (ACTH,hydrocortisone) and increase in somatotrophic hormone. The complete parenteral-andenteral nutrition influences positively on restoration of cellular and tumoral immunity, activates the factors of organism nonspecific protection and recovery from immunodepression, prevents the development of immunodeficiency. IMPACT TM VS CONTROL. S Atkinson, N Maynard, R Grover, E Sieffert, R Mason, M Smithies, D Bihari Departments of Surgery and Intensive Care, Guy's hospital, London, U.K Objectives: Comparison of the effect of an immunonutrient enteral feed versus a control on the outcome of a mixed intensive care unit (ICU) population. Methods: Admissions to this multidisciplinary adu)t ICU thought likely to stay more than three days and with tube access to the GI tract ~r randomised to receive either Impact TM, a feed with supplemental arginine, dietary nucleotides and omega-3 fatty acids, or an isocaloric and isonitrogenous control feed. Study end points included mortality and ICU stay. Approval was obtained from the hospital ethics committee. Rosults: 390 patients were entered into the trial. The two groups were well matched for age, sex, and admission APACHE II with an overall mean admission risk of death of 30.8 (std. dev. -+ 22.9). On an intention to treat basis, there was a no significant difference in ICU mortality, ICU stay or standardised mortality ratio (S.M.R.) between the two groups (see table) . Similarly, there were no differences after stratification for patients receiving 5 or more litres of feed. Conclusion: There is no evidence of an effect of Impact@, an enteral immunonutrient feed, on pre-determined end-points (ICU mortality, ICU stay or standardised mortality ratio) in a mixed intensive care unit population over that of an isocaloric, isonitrogenous control feed. ObJeeflves: Evaluate changes of blood laatate levels according to patient medical status after CVVHD initJ,~ion using dialysate solution containing lactate. Method: Review of medioal records of 20 consecutive patients ~eated by CVVHD (dialysate solution Hmnosol LG2, Hospal,UK, lactate concentration 40 retool/l). Date obtained 1 hr before and 4 -6 hrs at~er CVVHD initiation were analysed. Results: All data are presented as mean + SEM. In one patient, pre end post filter lactate levds were measured during standard CVVHD setting (blood flow 100ml/mlu, dialysate solution flow I 1/hr), and approximate daily lactate flux into the patient was calculated to be as high as 920 mmol/d. Lactate leveh measured after CVVHD initiation increased significenfly compared to baseline levels (3.80+0.67 axtd 2.88+0.68,respectively; p<0.01,paired T-test). When patiente with increased basal lactete (~-7) were compared to paliente with normal basal values (n=13), no difference in laotete increase was fmmd (p=0.22, MANOVA). Patiente with severe liver dysfunction (2 points in MOP scomlg, n=9) had higher basal laotate levels than patiente with normal or slightly abnormal liver teste (0 or 1 point in MOF scoring, n=ll), rite values being 4.04 + 1.17 and 1.84 + 0.23, respectively (p<0.05, Student T-test). Increase in blood lactate did not differ between these two groups after CVVHD was stetted (p=0.86, MANOVA). In 11 pafiente with invasive hemedynamio mo~, no oorrelation batween changes in lactate levels and eitlm" changes in oxygen ddivery (t2=O.Ol; p--O.81) or oxygen consumption (reversed Fie, k) (r2-q).O7;p--0.66) were found after CVVHD initiation. Conclusion: Blood lactate increases on CVVHD with dialysate soh~on rich in lactate. This increase is predominantly caused by influx of lactate into the blood via the filter end does not seem to depend on the liver fimotion and/or oxygen metabolism changes. Objectives: The study was designed in order to determine the effect on plasmatic proteins, of two types of aminoacids solutions of parenteral nutrition (PN) adapted to stress, having different concentration of branched chain aminoacids (BCAA), when applying to politraumatized critical patients. Methods: A prospective study was performed using a randomized double blind design of 20 polytraumafized patients, split in two groups of ten patients each, with mean ages of 35 _+ 17 an 45 -+ 20 years. Due to their condition, all patients required P.N. for at least 9 days. Both groups were subjected to isocalorie and isonitrogenous solutions (45 cI/kg/ day and 0.24 g of Nitrogen/Ks/day), varying only in the concentration of BCAA; solution A having a 10 % concentration and solution B 23 %. Blood samples determinations during days 0, 3, 6, 9 after the beginning of treatment with P.N. were total proteins., Albumin, Trandferrine, Protein binding retinol; Prealbumine and Fibronectine. The anova test (One and two way) was used to compare the values between the two groups. Results: The administration of solution A, showed statistically significant increases in the determinations of the values of Protein binding retino] (p < 0.05) and prealbumin (p < 0.05). No significant increases were observed in the values of Total Protein, Albumin, Transferrine and Fibronectin. Solution B produced statistically significant increases only in the values of total proteins (p < 0.05). The remaining proteins did not changed from their control values during the whole period of PN administration. Comparing both groups, no statistically significant differences were observed related to the type of diet. Nevertheless, differences were found in total proteins, albumin, protein binding retinoI, fibronectin (p<0.05) and prealbumin (p < 0.005) in relation to the time course of PN therapy. Only the albumin values showed significant differences (p < 0.01) when considering the interaction of both the type of diet and the time course of PN. Conclusions: 1. Solutions of PN adapted to stress, can maintain the control values of slow turnover proteins and improve the values of rapid turnover proteins. 2. No significant differences on plasma proteins were found between the two solutions having 10 % or 23 % concentration of branched chain aminoaeids. &Determination of rapid turnover proteins does not seems useful for discriminating different solutions of BCAA during PN. Obiectives; The hormonal changes in the post-traumatic situation often leads to an elevated blood glucose and a negative nitrogen balance. To reduce the elevated glucose production by aminoacids the apprication of xylitol may be an alternative energy source. In a double-blind randomized study we investigated the effects of a xylitol/glucose solution (group A: aminoacids 50 g/I; glucose/xylito180 g/40 g/l) on metabolism and particularly on pancreatic and liver enzymes compared to a glucose based nutrition solution regimen (group B: aminoacids 50 g/I; glucose 120 g/I). Methods: The clinical trial was carried out after the approval by the local ethical committee on 31 patients with severe brain injury. There was no difference in Body Mass Index BMI (group A: 25.9 +/-2.7 kg/m 2 and group B: 25.1 +/-2.4 kg/m=), age, and sex. Daily individual energy expenditure was measured by indirect calorimetry (Deltetrac "~). Nutrition was started 24 -48 hours after trauma or surgery with carbohydrates and aminoacids. Fat was added 24 h after nutrition had started. To analyze the effects on pancreatic and liver enzymes we investigated the following parameters for 4 days: blood gtucose, serum lipase, serum amylase, ASAT, ALAT, ~GT, AP, and serum cholinesterase (CHE). Results: Due to the daily indirect calorimetric measurements energy requirements were satisfied. There was no difference in blood glucose concentration and cumulative nitrogen balance between the two groups. Neither were there any significant changes in ASAT, ALAT, AP, and CHE for 4 days in both groups. Serum tipase steadily rose to 202 lUll in group A and 320.2 lUll in group B, respectively. Conclusions: There was no measurable influence of either nutrition solution on liver enzymes. The xylitol/glucose nutrition regimen does not have any advantage over the glucose based nutrition solution concerning blood glucose level or nitrogen balance. The elevation of serum lipase to a 2-fold level in either group needs further investigation on trauma patients. The effects of fat emulsions in lung function, particularly in lungdamaged patients, have been attributed to alterations in pulmonary vascular tone caused by eicosanoid production modificatione. As the eicosanoid production may depend on the fatty acid profiles of the intravenous fat emulsion, haemodynamic, pulmonary gas exchange and plasma levels of prostanoids were investigated in Acute Respiratory Distress Syndrome (ARDS) patients, during different intravenous lipid emulsions (providing different prostanoid precursors). We studied in a randomized double-blind design 3 groups (n=7 each) with ARDS. Group I (LCT) received a fat emulsion with long chain triglycerids (LCT-20%), Group II (MCT) an emulsion containing a mixture of medium and long chain triglycerids (MCT/LCT 50/50-20%) and Group IiI placebo (control), during 12 h (2 mg/Kg/min each). We measured before, at the end of 12 h infusion, and 12 h after the end of the infusion: lipaemia, arterial and venous blood gases, pulmonary and systemic haemodynamics, and plasmatic levels (arterial and in mixed venous sample) of eicosanoids (TXB=, 6-keto PGF~,, and LTB4). At the end of the fat emulsion, groups (I and il) to 1,10• to 0,51 • mmol/I), The PaOz/FiO z remained unchanged in the three groups; no changes in intrapulmonary shunt (Qs/Qt) were shown; neither in the mean pulmonary artery pressure. In contrast, only in the LCT group: cardiac output and oxygen consumption increased significantly (12.5% and 19%) (p<0.05). Eicosanoids were increased at baseline compared to reference values (p<0,05). A decrease (p1000 IU/1. Etiologies were: traumatic and ischaemic 24, infectious 4, toxic 4, excess activity 1. Factors studied were: simplified acute physiologic score (SAPS: 10.3+1.1), Organ Systemic Failure (OSF: 1.02_-!-0.2), diagnosis delay (D: 33+_5h), clinical parameters (sepsis, dehydration), blood chemistry data (CPK, BUN, creatinine, potassium, phosphorus, calcium, proteins, hematocrit) and urinary pH. Severity of RH was estimated by Ward Score determined according to phosphorus, albumin, potassium, CPK, dehydration and sepsis. Urea appearance rate (UAR) and creatinine index (CI*) were determined over a 24 hours period. ARF was observed in 25 pts. In non-ARF and ARF groups respectively, SAPS (5.5_+0.5 vs 11.8+1.3), deshydratation (0 vs 11), sepsis (0 vs 12), phosphorus (1.03+0.16 vs 2.21-+0.21), calcium (2.1+0.07 vs 1.8_+0.07), Ward score (4_+0.65 vs 11.8+0.8) were significantly different. However, no significance was observed in UAR (310-+89 vs 210-+35) and CI (26_+7 vs 34_+3). 16 patients required hemodialysis (HD) (85:2 sessions) and 9 remained dialysis free. Only OSF (1.1_+0.1 vs 1.9-+0.23), Ward score (9.2_-/-0.95 vs 13.25_+0.92) and CI (29+_3 vs 41-+3) appeared significantly higher in pts requiring HD. 5 pts died from associated disease. All patients suffering from ARF recovered a normal renal function. We confwmed that an elevated Ward score (over 7) is a good predictive index of ARF. In addition we found that CI is a severity factor for ARF requiring HD. Thus, patients suffering for RH with elevated Ward Score and CI, have a fair chance of dialysis and should be treated more intensively. * CI (expressed in mg/kg) = (CAR + Feces Creatinine) / Weight. Where CAR: creatinine appearance rate; Feces cr~t..= mean plasmatic creatinine x 0.043. Tr~er K., Cetin T.E., Tugtekin I., Georgieff M., Ensinger H. Universit~tsklinik fLir An~sthesiologie, 89070 UIm, Germany Introduction: Endogenous as well as exogenous adrenergic agonists have a profound effect on carbohydrate metabolism in human critical illness. In this study the effects of noradrenaline (NOR) and dobutamine (DOB) on carbohydrate metabolism during a 4 hr infusion were investigated. Methods: After approval by the local ethic committee 14 healthy volunteers were studied. Hepatic glucose production (HGP [mg/kg/min]), using 6,6-D2glucose as stable isotope tracer, as well as plasma concentrations of glucose (GLC [mmol/I]) and lactate (LAC [mmol/I]) were measured prior and during infusion of NOR (0.14 pg/kg/min) and DOB (6 pg/kg/min). Blood samples were drawn before and during the agonist infusion. Results: No major changes in insulin and gtucagon plasma concentrations could be found during the study period. ::i:::: :iiiii~ 8~ i ::i: ~:: : :: i:ii. Mean-+SD are shown. # p<0.01, ANOVA for repeated measurments. Conclusions: The effect of NOR on HGP and GLC were smaller as compared to adrenaline (I) with a similar time course. In contrast to the effects of adrenaline and NOR, DOB had a different effect on carbohydrate metabolism: a decrease in HCP and GLC, which is uncommon for a /3-adrenoceptor agonist. Since HGP is an energy consuming process that might deteriorate hepatic oxygen balance in critical illness, the differential effects of adrenergic agonists may be of importance and need further clarification. The nutritional insufficiency often accompanies post-operative hypercaloric states, inanition, serious infections and weakening chronic illnesses. That is why the early nutritional support, sufficient and appropriate for each individual base, is a fundamental component of intensive care unit as an indispensable factor for recovery. Per this reason, our Unit, developed a software for the implementation and nutritional control of t~e assisted patients. This software is incorporated is an expert system called ~I~Su, designed and developed by the computational division of our Unit. This system arrives to inferred diagnoses such as : respiratory, hepatic, renal(with and without dialysis) dysfunctions, pancreatitis, ARDS, decrease of consciousness, diabetes. According to these data Objectives: To compare the effect of short term enteral feeding versus parenteral nutrition, when a isonitrogenous and isocaloric feeding solution is administered by either mute. Methods: In a prospective controlled clinical trial 30 patients were studied; all exhibited moderate degree of malnutrition, normal liver and kidneys, and a functi6ning gastrointestinal tract. The patients were randomized to receive a free amino acid and small peptide diet (15 patients) or an isonitrogenous isocaloric parenteral support (TPN) (15 patients) (total energy: 2880 Kcal, nitrogen: 14.5 g, carbohydrates: 380 g, fat: 112 g, N/non protein calories: 1/175) at least for 10 days. Results: There were no significant changes in anthropometric parameters within either group. Nitrogen equilibrium was aqhieved by day 3 in the TPN group and by day 5 in the enteral group (66.6% of the enterally fed patients and 80% of the TPN patients maintained in positive balance the day 10 of the study). There were no significant changes in serum albumin within either group. Serum level of transferrin reached a significant increase in both groups (p=0.003). Thyroxine-binding prealbnmin rose significantly in both groups as well (p=0.019 and 0.004 respectively). Statistically significant rises in lymphocyte counts (p=0.003 and 0.001 respectively), in levels of C 3 (p=0.009 and 0.01)1 respectively), IgA (p=0.002), IgG (p=0.004 and 0.003 respectively) and IgM (p=0.004) occurred in either treatment group. There was a high incidence of negative skin tests at the start of the study in the enteral group (73.3%) and the TPN group (60%). By the end of the study the incidence of negative responsiveness was 40.0% and 26.6% respectively. Despite maintenance of similar glucose levels in both groups, TPN led to significantly higher serum insulin levels. The serum insulin increased almost linearly over the study period and eventually prevented fat mobilization and lipolysis, so that free fatty acid levels had fallen significantly. A significant elevation of the liver enzymes over the study period occurred in 73.3% of the TPN group, but not in the enterany fed patients. Conclusions: The present findings provide no evidence that enteral diets containing free amino acids and small peptides, as their nitrogen sources, are in any way inferior to isonitrogenous isoealoric regimes parenterally given. Aim: The aim of this study is to describe and explore the expectations of the functions of the critical care nurse to enable the formulation of guidelines for the scope of practice for the critical care nurse with a South African context, Methods: Phase I was to determine the expectations of the critical care nurse, the nursing service managers and the doctors with regard to the functions of the critical care nurse. A focus group interview was held with a group of experts in the field of critical care. The results were used to compile a questionnaire. This questionnaire was sent to the critical care nurses, the nursing service managers and the doctors in South Africa for completion. From these results the functions of the critical care nurse were determined. Phase II was to formulate guidelines for the scope of practice for the critical care nurse within a South African context. Through usage of the date (Phase I) the scope of practice was formulated. Guidelines were formulated for the practise, education and research regarding the limitations of the professional-ethical authoration and the implementation of the scope of practice for the critical care nurse. Objectives : High output gastric aspirates arc occasionally observed during fasting in critically ill paticnts, preventing any attempt of feeding via the enteral route. Although these patients are often said to suffer from "gastroparesia", the motor correlates of this condition arc lurgcly unknown. In this stud?', wc recorded the gastrointestinal motility of critically ill patients with abundant (>250 mL/24 hours) fasting gastric aspirates. Methods : Antral (4 sites separated each other from 1.5 cm), duodenal (1 site) and jejunal (1 site) contractions were recorded simultaneously by ~eans of a multihimen tube assembly positioned trader fluoroscopic control (perfused catheter technique). Tracings from prolonged recordings were obtained on a multichannel recorder (7758A recorder, Hewlett-Packard) then anal)7,ed visually, with a special attention for the following abnormalities which are characteristic of intcstinal pseudoobstmctiou: l) absence or aberrant propagation of the migrating motor complex (MMC), 2) presence of bursts (> 2min) of nonpropagated phasic pressure and 3) presence of sustained (>30 min) uncnardinate pressure activity. 11 patients with a volume of gastric aspirates of 731 • 506 (SD) [median 5001 mL/24 hrs were investigated for 538 4-271 [median 4551 minutes. Results : Only one patient had no detectable motor abnormality. MMCs were either absent (n=4) or migrated abnormally (retrograde propagation : n=4; retrograde and stationnary : n=2) in 10 pts. Bursts of nonpropagated phasic pressure activity were present in the duodenum in 9 pts and sustained uncoordinate pressure activity was found in 2 pts. Additional abnormalities included episodes of prominent pyloric activity. (n=l) and sustained antral pressure activity (n=2}. Conclusion : Critically ill patients with large volume of gastric aspirates have manometric evidence of intestinal pseudoobstruction. Prokinetic therapy in these patients should thus focus not only on enhancing gastric motility, but also on restoring a normal propagative contractile activity in the intestine. This prospective, open-label, randomized placebo-controlled study included 20 patients with hypokalemia in whom rapid potassium replacement (20 mEq KCI in 1 h) was performed: 14 patients received Mg sulfate (6 g in 3 hours) and 6 patients received a corresponding saline infusion. Measurements were made at time 0, +1, +3 and +6 hours RESULTS: K levels increased more in Mg treated patients than in the patients who received saline infusion at time 1 and 3 h (p < 0.05 -Students-Newman-Keuls). (Table 1 ). Introduction. Dual lumen uaso-gastrojcjunal tubes are a major ads'ance in nutritional therapy of mechanically ventilated critically ill patients since the3" authorizc jejunal feeding with concurrent gastric decompression, there,, reducing the risk for aspiration. Unfortunately, placcmem of these tubes in the jejunum regularly dictates to resort to endoscopy in order to facilitate pyloric intubation. Recently, the remarkable gastrokinetic properties of the well known macrolide antibiotic er}lhromycin have been demonstrated in gastroparetic critically ill patients 1. Aim. In the presem stu~,, we evaluated the feasibility of placing dual lumen naso-gastrojcjunal feeding tubes at the bedside without endoscopy, using eDthromycin to help Iranspy'loric migration of the tube under fluoroscopic control. Methnd Each patient admitted in our ICU during a 2 months period and requiring artificial ventilation and enteral nutrition for a period of at least 3 days was included in the study.. After inserting the tube (Stayput| Sandoz, USA) in the gastric anmnn, e.rythromycin (200 rag) was aduunistored intravenously, to help fluoroscopic positioning of the tube into the jejunum. The total duration of the procedure (from nasal intabatiun to jejunal placement), as well as the duration of ftuoroscopy were recorded in each patient. Results. 15 patients (male/female : 13/2: Mean age : 56.9 + 22.2 years; Mean Apacbell score : 23.t • 7.0) wore enrolled into the study.The procedure was performed within the 2 dab,s following institution of mechanical ventilation. Jejunal access was obtained in all 15 patients without resort to enduscopy in 10,81 • 7.31 min.(total duration of the procedure). Mean duration of fluoroscopy was 3.54 + 2.97 rain. Conclusion. We conclude that placement of dual lmnen naso-gastrojejunal tubes can be obtained in mechanically ventilated critically ill patients without resort to endoscopy., provided that e rythromycin is used as gastrokinetic agent to help pyloric intubation. The following AD and DIS parameters were considered in all patients: -mid arm circumference, triceps skinfold thickness, serum transferrin, albumine and lymphoeites and urinary creatinine/height index. Patients whose results were bellow 80% of normal values in 3 or more of the 6 above criteria were considered undernourished (UND).Statistical analysis was performed using %2 analysis.Statistical significance was established at p median lenght of stay 18 days; 2 UND at AD and UND at DiS = > median lengbt of stay 22days; Nutritional status and age at admission: 1-Age > = 60 years : NOU (13) , UND (53) 2-Age < 60 years: NOU (26), UND (37) Nutritional status and age at discharge: 3-Age > = 60 years : NOU (12) , UND (54) 4-Age < 60 years: NOU (36), UND (27) We observed a p 5 days) were randomized and allocated to the SDD group (n=32) or the control group (n=33). In their general intensive care theraw, there were no differences between the groups. The SDD regimen consisted of the four times daily administration of 50 rag polymi~ 80 mg tobramycin and 500 mg amphotericin B in the nesc, mnoth and stomach. Systemic prophylactic ~dmini~/rution of antibiotics was not part of the SDD regimen. Smears were taken from the nose and the rectum twice wceldy and from the pharynx and trachea once wceldy, and tested for MRSA. Further samples were taken as clinically reqnircR Results: 625 smears were examined in the SDD group. MRSA strains were detected in 66 samples (10.5%) from 7 patients, and in 5 patients they were detected for a period of up to 4 weeks. The positive smears were districted as follows: tracheal 11/117 (9.4%), nasal 28/199 (14.0%), pharyngeal 15/111 (13.5%) and rectal 121198 (6.1%). Severe MRSA-induced infections were observed in 2 patients (infection rate 28.6% of the colonized SDD patients). 560 smears were examined in the control group. IvlRSA swains were r in 15 samples (2.6%) from 6 patients, but only repeatedly over a period of up to 10 days in 3 patients. The po~tive snmars were distributed as follows: traclmal 1/114 (0.8%), nasal 8/174 (4.6%), pharyngeal 4/98 (4.0%) and rectal 2/174 (1.1%). There were no MRSA infections in the control group. Conclusion: The data collected support the view that the use of SDD promotes a selection and persistence of MRSA strains. Longer-term colonization with MRSA and sovere systemic inf~ons were only found in the SDD group. Although the clinical and epidemiological impact of resistance develol~ng when SDD is applied ~maine unclear, this question should be given close scrutiny. Tazobactam/Piperacillin (TAZ/P1P) is a new broad spectrum antibiotic, in which the acylaminopenicillin Piperaeillin is protected by the betatactamase inhibitor Tazobactam from hydrolization by bacterial enzymes. TAZ/PIP has shown to possess a high antibacterial activity against almost all clinically relevant bacteria and is a registered drug in Germany. Obiectives: Purpose of this investigation was to evaluate, whether FAZ/PIP 4.5g is suited for efficient antibacterial monotherapy of severe infections and what influence dosage frequency reveals on clinical efficacy. Methods: 2151 hospitalized patients have been documented in this multicenter trial during a 2 year period. As this investigation should reflect the usual clinical treatment, the only criteria for enrolment were the typical signs of infection as e.g. temperature > 38~ leucocytosis or an isolated pathogen. Exclusion criteria did not exist and the patients were treated in accordance to the severeness of infection, underlying diseases, risk factors etc. with TAZ/PIP 4.5g t.i.d, or b.i.d. Results: Patients suffered in most cases from infections of the lower respiratory tract (n=926), followed by intraabdominal (n=765) and skin and soft tissue infections (n=460). 61% of the 926 LRTIs wvre nosocomial acquired and in 75% the treatment was conducted as monotherapy. In 53% the LRTI was treated with TAZ/PIP b.i.d, and in 45% t.i.d. Pseudomonas spp. (n=138) and Staph..aureus (n=134) were the most isolated pathogens pretrcatment. The clinical response rates (cured/improved) after treatment with TAZ/PIP 4.5g b.i.d, and t.i.d, were 89% and 81% respectively. Results for intraabdominal-and skin and soft tissue infections will be presented. Conclusions: In hospitalized patients with severe infections successful treatment with TAZ/PIP in monotherapy is possible. In this population a reduction of the dosage frequency to 4.5g b.i.d, revealed equivalent clinical response rates. Objectives. Retrospective evaluation of 9 cases of severe generalized tetanus (SGT), treated in our ICU the last 7 years. We review 9 cases of SGT (6M, 3F), Mean age 66.7 years. In 5 eases the entry site of C.Tetanus was a skin laceration, in 1 case it proved to be the external genitalia, while in the rest no portal of entry could be determined. In the first 6 cases incubation period was short (3-11days) and so was the period of onset (1-6 days). All patients needed mechanical ventilation (range 15-58 days), initally through an orotracheal tube,and later through a tracheostomy, performed 6• days after admission. Clinical manifestations of SGT included muscle rigidity and i generalized spasms, persisting for up to 6 weeks in the most severe cases. Significant autonomic nervous system dysfunction was present in 3 cases occurring 5-12 days after the admission and following the time course of generalized spasm. Besides general supportive measures, specific treatment included passive +active immunization, Penicillin G, Magnesium Sulphate and sedation in a variety of regimens. Neuromuscular blockade was required in 5 cases. Nosocomial infections occurred in 7 eases, with sepsis and MOF in one. Average stay in the ICU was 18-62 days. One patient died with severe septic complications and one was discharged with severe 9 disability due to anoxaemie ancephalopathy, after a cardiac arrest on admission. ~ DISINFECTANT In suspension test, without presence of organic load, 12 disinfectants showed efficacy on LM. In the carrier test, in the presence of organic load, 6 out of 14 examined disinfectants did not exposed efficacy on LM. The results of examinations clearly showed that evaluation of disinfectant's efficacy partly depend on the used test method. Antun BaSi6, Intensive care unit, KB Firule Split Spin~ideva 1! Jugoslavia Bacteremia and sepsis are frequent complications encouuntered in severe ICU patients.Microorganism identification with hemoculture presents the basis for adequate and successful antibiotic treatment.In many patients damage and vulnerability of the peripheral veins presents an obstacle for obtaining the blood culture from the central venous (CV) catheter sample could be also used. Material and methods Blood cultures were perfomed in lo4 patients on blood samples simultaneously obtained from the peripheral vein and CV catheter three times in a 24-hour period.Criteria for the suspected bacteremia were body temperature above 38 C and leucocytosis above Ioooo leucocytes/dL. The site for venipuncture and the CV catheter stopcock port were cleansed with povidon iodine.After the initial 5 mL of blood were discarded,lo mL were used for the blood culture.Standard laboratory technique for blood cultures was used. Results and discussion In 76 (73%) patients hemocultures was negative at both sites,whereas in the remaining 28 (27%) they were positive.For twentyone (26~176 of the positive patients the same results were obtained at both sites (peripheral vein and CV catheter),whereas in 7 (6.7%) patients the blood culture were positive only for the CV catheter samples.The CV catheters were in place for less than 4 days in 81 patients and for more than 7 days in 23 patients.From 7 patients with positive blood culture from the CV catheter,one patient had the catheter for three days,whereas the other 6 had the catheter from 6-1o days. We neither found significant differences in hemodynamic dates : Objectives: 1, To count and evaluate bacteria isolated from endotracheal (ET) suctiori samples (with and without saline). 2. To establish the exogenous source(s) of pathogens isolated from carer's hands and the equipment involved in sampling in order to reduce the incidence of contamination and infection. Method~: This prospective study included 20 consecutive ventilated patients (15 male and 5 female, 56_ + 16 yr; APACHE II score 19-+7) over a period of 3 months. ET aspirated samples with and without saline were taken daily from Day 0 of intubation until pathogen~ were presented in counts of _> 105 per ml. At the same time, samples from both carer's hands were taken before and after ET suction and a swab from the ventilator tube. Results: The overall length of intubation varied between 3 to 65 days. Bacterial transfer between staff and patients was noted in 80% of patients until Day 5 of intubation. There was no significant correlation between severity score and appearance of colonization. The incidence of pneumonia in studied patients was 45% with an overall mortality rate of 30%. Acinetobacter anitratas (No 15), Staphylococcus aureus (No.15), Klebsiella pna~moniae (No.9) and Pscudomonas aeruginosa (No.4) isolates predominated in all our specimens. We noticed increased resistance to most antibiotics with the exception of imipenem for Gram (-) bacteria and vancornycin for Gram (+) bacteria. Conclusions: I. Tracheobronchial colonization appears directly in the maiority of intubated patients. 2. There is a close relationship between the microflora of personnel, patients and equipment. 3. Bacteria transfer was noted both to and from patients. 4. Strict hand disinfection policy remains an important measure for the proper care of mechanically ventilated patients to reduce respiratory infections. Nnseeomial pneumonia is the most common nnsocomiai infection in the ICU-settiag, reported in up to 20% of patients admitted to the ICU following surgery. It is associated with significant mortality that ranges from 30~ to 70%. Enteric gram-negative bacilli have been implicated in 75% to 85% of ventilntor-associated pneumonias and Pseudomonas aeruginosa accounts for 27% to 40% of these pneumonias. Importantly, epidemics of/3-1actamnse-pruducing Enterobacter spp or Klebsiella spp that are resistant to extended spectrum cephalosporins or penicillins, pose serious obstacles to effective antibiotic choices. Carbapenems provide in ~tro activity against a wide range of Enterobacteriaceaeand other gramnegative aerobic bacteria, except Steaotrophomonns maltophilia. In vitro meropcnem is more active against Pseudomonas spp than imipanem (especially P. aeruginosa and P. cepacia), Imipenem and meropenem are effective against more than 95% of strains responsible for nnsocomial infections. All major pathogens associated with LRTI are usually covered by the carbapenems, exceptions are pathogens involved in so-called atypical pneuomouia like mycoplasma, chlamydia and legionella. Carbapenems are highly stable in the presence of most chromsomal and plasmid-mediated Blactumases and usually offer a postantibiotie effect lasting for three hours against most of the Enterubacteriaceae. Reeent studies comparing imipenem/cilastatin with other ~-lactams and fluoroquinolones in severe LRTI in ICU patients resulted in favourable clinical cure rates and good tolerance, but development of resistance in P. aeruginosa and 5;. aureus during treatment were of some concern. Meropenem offers the advantage of greater stability against enzymatic degradation, so no concomitant administration of an enzyme inhibitor is necessary, and meropenem appears to be associated with a lower risk of seizures, particularly when used at high doses. Results from studies with meropenem in LRTI, especially in critically ill patients with acute exacerbations of chronic bronchitis, demonstrated excellent cure rates and better gastrointestinal tolerance of this new carbapenem. Both earbapenems are effective candidates for use as empiric monotherapy in nosucominl infections of critically ill patients. Qbl~ctives A favourable effect of IV immunoglobulins in septic surgical patients has been reported, but not sufficiently validated. We conducted this study on trauma patients to: i) investigate the effect of IVIG on septic complications and il) quantify this effect by means of serum bactericidaI activity (SBA) assessment and iii) to explore the effect of temperature increase (from 37 to 40 ~ C) on the SBA Methods: Twenty trauma patierIts matched on admission for age, sex, inju~ severity score and Glasgow coma scale, were allocated to receive either WIG (IVIG group; I0 patients) or equal volumes of human albumin 20% (Control group; 10 patients). WIG (Sandoglobulin) was administered in a total dose of 1 g/kg divided in a four time regimen on days 1, 2, 3 and 6 post-admission. Three blood collections were performe& before the first dose (day 0) and 24 hours after the third and the fourth dose (days 4 and 7 respectively). Complement, lgG fractions, the SBA at 37 ~ and at 40 o C and clinical parameters were recorded. Results-Similar lgG and IgG] serum levels were found in groups IVIG and control on day 0 (743+_130 vs 898• NS and 394+103 vs 472+101, NS), whereas they were significantly higher (p<0 05) in the 1V1G group on days 4 (1700_+_274 vs 799+197, p<0 05) and 7 (1740_+227 vs 864+I64, p<0.05). The various complement-fractions increased in both groups without inter-group differences The mean (• SBAs (37 ~ C) at 30 rain in IVIG group vs control group were: -53_+32 vs -56• NS for day 0, 9_+46 vs -54_+46 p<005 for day 4 and 7_+34 vs -54+47 p<005 for day 7. The mean (+SD) SBAs (40 ~ C) at 30 rain presented a significant improvement over those of 37 ~ C but for the control group remained negative a~d were respectively as following: -~8• vs -26+33, NS for day 0, 22+_39 vs -29_+35, p<0.05 for day 4 and 24_+31 vs -27_+36, p<0.05 for day 7. The increase of temperature induced a 3-fold improvement of SBA in IV1G group and 2-fold ofcontrol-~oup Positive blood cultures, and the product of the infectious episodes number multiplied by days of occurence, were significantly lower (P<0 05) in the IVIG group than in the control (2 vs 7, and 440 vs 1900, respectively). Conclusions: Our study shows a significantly favourable effect of IVIG administration on septic complications and on SBA of trauma patients. The increase of temperature results in a significant improvement of SBA of patients that received IVIG, which theoretically means a farther prevention of infection in the febrile state. Pharmaceutical Microbiology, University of Bonn, Meckanheimer AUne 168, D-53115 Bonn, Germany Infectious diseases in intensive care patients are common in comparison to patients on other wards and out-patients. The main difference is that intensive care patients are much more sensitive even to less virulent bacteria. Thus, the spectrum of infecting organisms is different. Strains often regarded as pathogens with low virulence cause serious infections in these patients. Strains such as Serratia, however, have intrinsic resistance to most commonly used agents such as 3rd generation eephalosporins. Furthermore, the common pathogens like staphylococci, Psoudomonas aeruginosu, enterocneei and gram-negative bacteria, enterobacteriaeceae as well as the non-fermenters are less sensitive if isolated from intensive care patients. It is difficult to generalize on intensive care units as different patient groups are in different ICUs aud there are great changes from one hospital to another and from one country to another. If we take S. aurens strains from one study from 1990 the'overall resistance in intensive care units towards oftoxacin was 22 %, whereas in other hospital wards the percentage of resistance was 5.3 %, in out-patients, however, only 2.$ %. The same trend was true for Entercnecus faecnlis, coagulase-negntive staphylococci, and other bacteria as well as other drugs. One most striking difference was found with Klebsialla pneumoniae and gantamycin resistance, which was $ times higher in intensive care units as compared with outpatients, whereas in the same species no difference was to be seen with the resistance towards carbapenems. However, differences between countries seem to be even more striking, as example gantamycin resistance and Staph. anrens is given. The extreme difference is more than 60 fold. Thus, it is evident that there is a general trend towards higher resistance in intensive care units, but no generalizatiouis possible. Therefore, surveillance studies in intensive care units are needed and the antibiotic policy has to be adapted to the specific needs of the unit. In the ICU setting the most potent antimicrobial agents are required to address problem organisms including those resistant to penicillins, cephalosporins and aminoglycosides. Carbapanems would appear to present a useful option in this setting. Objectives of this study was the evaluation of systemic candid•177 in postoperative cardiac surgery patients (pts) with prolonged ICU stay. Methods: Out of 2617 postoperative adults pts of mean age 61.1+8.9 years old, with a mean ICU stay of 1.6_+0.6 days, following an open heart surgery from July 1993 to April 1995, 54 pts (2%) remained in ICU for more than 10 days because of severe perioperative complications. Patients were included in the protocol if they had clinical signs of infection or sepsis, and fungi isolated in blood culture or in culture from at least three different sites. The patients who developed systemic candidiasis received IV fluconazole (800 mg/day) (10 patients) or amphotericin-B for at least four weeks, and then they were closely monitored. Results: Out of 54 postoperative pts with prolonged JCU stay, 11 pts (20.3%) developed systemic candid•177 usually after the 20th postoperative day. They were 8 males and 3 females of mean age 64+_7.4 years old. This group of pts had prolonged bypass and aortic cross-clamp time compared to control group (119 min vs 84, and 64 vs 49 min). All these pts received inotropes per• (mean value=2.3). During their ICU stay, 9 pts developed sepsis of bacterial origin, while the other two severe infection, and received antibiotic regimens for prolonged period. The patients were submitted to mechanical ventilation for a median period of 50 days. The median ICU and hospital stay was 58 and 60 days respectively. All pts have been improved and finally negative cultures were obtained. Conclusions: 1. A significant percentage of patients who remained in the postoperative iCU for more than 10 days developed systemic candidiasis. 2. All patients who developed systemic candidiasis had received antibiotics because of sepsis or severe infection, for prolonged period. 3. Fluconazole seems to be a very good alternative to amphotericin-B. 4. Fluconazole is a safe antifungal agent with few side effects. Botulism is the most severe and an odd food poisoning. Although it is more commonly related to preserved meat derivatives, preserved fish and vegetables are also responsible for a number of cases. Obiectives: To evaluate four familiar outbreaks of botulism . Methods: We study the patients that were admitted in our hospital because of botulism from May 1982 to February 1995. Results: The thirteen pacients involved had a previous history of home preserved beans ingestion. After a 24-hours incubation period, gastrointestinal symptoms (abdominal pain, vomits, constipation) appeared and lead them to hospital consultation in the 4th to 7th day after ingestion. Two patients died (acute respiratory failure before admission), seven were admitted in ICU, two in ward and two of them were discharged from emergency room. Clinical symptoms and the previous history of the ingestion established the diagnosis, that was EMG confirmed. In all cases, symptoms were consistent with B-toxin botulism. B-toxin was isolated in serum and food proceeding from the third outbreak, and the serum was negative in the other ones. Neurological symptoms were predominant: midriasis (100%), dry mouth (100 %), dysfagia (100 %), asthenia (55 %), palpebral ptosis (55 %), accomodation paralisis (66%) and urinary retention (55%). Muscle weakness lead to acute respiratory failure in three patients (one of them required mechanical ventilation). Four patiens developed infections (respiratory, urinary and phlebitis). Both died patients and one another presented severe hypertension. All admitted patients were treated with polivalent anti-toxin. The two patients who underwent a more severe muscle weakness received also guanidine hydrochloride, with no answer in one case and provoquing a cholinergic crisis in the other one. ICU length of stay was 10 days. At hospital discharge, patients continued symptomatic, mainly with dry mouth, disfagia and impaired vision. Conclusions: Although botulism is a serious illness, the pronostic seems favorable if treatment and support measures are avaible. Usually neurological symptoms we predominant and at discharge some of them could still persist. The Arrow "Hands-Off" (AHO) thermodilution catheter (TC) is completely shielded during balloon testing, preparation, and the insertion procedure. In order to assess the value of the AHO thermodilution catheter in the prevention of systemic infections associated with pulmonary artery catheterization (SIAPA), we conducted a randomized prospective study over an 18-month period. Methods : The patients (pts) were randomly assigned to two groups : Group I for a standard TC customarily used in the department, versus group 2 for the AHO thermodilution catheter. The diagnosis of SIAPA was determined on the basis of a positive culture of TC and bacteremia with the same organism, with out any other nearby focus, in association with regression or disappearance of the clinical signs of infection after removal of the thermodilution catheter. Results ( Objectives: The mortality rate (MR) of TB requiring mechanical ventilation (MV) is high (70-100%). The aim of the study was to evaluate MR, associated factors, and prognostic significance of MV and hemodynamic disorders from TB in ICU in 35 patients with TB. Methods: Clinical parameters on admission, and complications in ICU were related by univariate analysis to ICU, hospital, and 6 month outcome. 18 patients required MV; 10 were immunocompromised (IC) including 8 HIV. TB was pleuropulmonary in 24, disseminated in 9 and meningeal in 2. Results: MR was 31% in ICU, 34% in hospital and 47% at 6 month. 12/16 (75%) <0.001 Mortality was associated with a high SAPS score, initial shock, MV and nosocomial septicemia. The MR dramatically increased when ARDS occurred during illness, despite the lack of correlation between MR and initial PO2/FiO2 ratio or initial Murray score. The site of infection did not influence the MR. Surprisingly, the mean therapy delay was shorter for non survivors. MR was not related to IC status, nor HIVstatus, but was only related to previous steroid therapy. Conclusion: MR of TB requiring ICU is high (47% at 6 month). Need for MV increased mortality (72% vs 18%). General severity and respiratory dysfunction seem to be major prognostic factors in ICU rather than TB per se or than therapy delay. In spite of the improvement in the prognosis of pneumococcal meningitis (PM) with Third Generation Cephalosporins (TGC), this infection still presents a great mortality which could be increased with the appearance of antibiotic resistant Streptococcus Pneumoniae. OBJECTIVES: To asses intensive care mortality and morbidity of PM and to define patients (pts) at risk of complicated evolution. PATIENTS AND METHODS: A retrospective evaluation of PM cases (all diagnosed by CSF culture) admitted in our ICU from January 1985 tit March 1995. In all pts we analized: demographic data, underlying disease, APACHE II score, clinical symtomps, treatment, complications and outcome. Statistical analysis was done using BMDP sofware package. RESULTS:A total 0f42 pts were studied, 26 males; mean age 55,8 _+ 16 (16-81); APACHE II score 16,6 + 7,9; Glasgow Coma Scale (GCS) at admission 12,5 _+ 1,7; 17 (40%) pts suffer from cronic pathology; 5 (12%) pts diabetes mellitus (DM), 4 (9,5 %) pts had had a previous cranial traumatism. In 22 cases the source of infection was otic and also in 22 (52 %) episodes of PM there were bacteriemia. In 21 out of 26 (80%) pts that CT was performed no radiologic abnormalities were shown, 3 of them presented cerebral oedema and 1 pts a cerebral abscess. Twenty-eight percent presented seixures, 14% hemiparesia, 46,3% respiratory failure, 17,5% shock, I5% renal failure, 5,1% multiple organ failure (MOF). As for treatment refers 5,5% pts recieved only Penicillin, 69,4% pts only TCG, 11,1% pts TCG followed by Penicillin and 8,3% pts TCG+Vancomycin. Seventy-five percelat of pts recieved Corticosteroids and 25,6% vasoaetive drugs. The mean ICU stay was 7,5 5:6 days (1-28). Twelve (28,5 %) pts died, two of them presented PM relapse (resistant Streptococcus pneumoniae) and another two pts developed neurological sequelae. Factors associated statistically with bad prognosis were DM, the use of vasoactive drugs, shock, MOF, the APACHE II score at admission, the GCS at the 48 and 72 hours from admission in the ICU but not the GCS at admission. Didn't resulted statistiealy signifcative age, previous eronie pathology, seizures, baeteriemia, renal failure and coagulation disorders. CONCLUSIONS: Mortality was high and associated to APACHE II score at admission, to GCS at 48 and 72 hours after admission, shock, vasoaetive drugs and MOF. Objectives:The aim of the study was to analyse some of significant immunologycaI changes in surgical patients,requiring intensive health care,and to determinate the possibility for evaluation,dynamical examination and importance of immunologycal problems for treatment. Methodes:The study concerns a number of 30 patients with expanded surgical intervention or serious postoperative complications.The results has been carried out with fiowcytometryc analyses of lymphocytic suhpopulations and routins methods for investigation of humeral immunity.The"panel" for evaluation of 2(} immunologycal parameters has been offered:T-calls total/CD3+/;T-helper/CD4+/;T-supressor/CD8+/ Th/Ts ratio;B-cells/CD19+/;naturaI kilier/NK/cells;skin test for cellular immune function;phagocytic and oxidative activity;serum levels of immunogiobulins-G ,A,M;protease inhibitors;C-Reactive Protein.All patients have been studied during suffering and after surgical procedures dynamicaly. Results:There have been estimated significant changes in immunologycal parameters especially:decrease of T-cells: CD3+mean=37.62%/14.3%-47.9%/and CD4+mean=22.11%/9% -28.8%/;inverted Th/Ts ratio ,mean=O.72/0.37-0,90/;reduced or negative skin teste;reduced phagocytic and oxidative activity before septic complications. Conclusions:Dynamical examination of immunologycal parameters shows,that the prolonged T-total,T-helper lymphocytopenia with functional deficience of ceils-mediated immunity correlates with the stage of clinical condition of the patients and has prognostic importance.It's clear,that immunologycal monitoring gives a possibility for immunecorrection. Patients (pts) with the Human tmunodeficiency Virus (HIV) infection have a decreased immune response and are particularly susceptible to infectious endocarditis (IE). The aim of our study was to analyze the prevalence of IE, its clinical and therapeutic implications in a HIV population 9 We prospectively studied 245 pts, 9.4% (23/245-Group IE+) with IE during the clinical course of this disease. We analyzed the following parameters: age, gender, race, type of HIV, CDC classification, number of T4 and T8 type cell population and its ratio, therapeutic with AZT, type and number of opportunist infections (INF, Mycobacteriosis (MB), Neoplasm's (NEe) 9 The echocardiographic parameters were LV internal diastolic and systolic diameters, LV percentage of fractional shortening, interventricular and posterior wall thickness, the degree of valvular regurgitations and the presence of pericardial effusion. El was located at the MV in 2.7%, TV in 6.0%, AV in 2% and PV in 0.9~ and was multiple in 2.0%. HIV El+ pts had larger LV diameters and more frequent significant valvular regurgitations (39% TR, PE 33%, Mortality 32%). These two groups differed significantly in the following clinical parameters: The typical symptoms were watery diarrhea, high fever, tachycardia,luekocytopenia and oligouria within 7th postoperative days. The patients with MRSA enterocolitis had positive MRSA culture from the many materials except feces.MESA strains frequently had coagulase type 2,enterotoxin A and toxic shock syndrome toxin-1 .Eight of 1 6 patients had postoperative organ failure.Most of the MRSA strains in Japan were similar in coagulase type to our hospital and our department.All of MESA strains were susceptible to vancomycin and arbekacin,tbough most of them showed resistant to many other antibiotics.We have employed guidelines for therapies such as oral or enteral administration of vancomycin and correction of the hemodynamics for dehydration and circulatory failure due to diarrhea from 1992.Futhermore we have placed colonized or infected patients in private room,worn gown and mask,and carefully washed our hands from 1992. These countermeasures for prevention of nosocomial infections after 1992 significantly reduced the incidence of MRSA enterocolitis. Conclusions:Earlier diagnosis and treatment, and distric prophylactic measureres against MRSA infections are very important. -- CDO IVDA Leptespiresls affects all the organs with widespread hemorrhage that Is more prominent in skin, mucosa, skeletat muscles, liver and kidneys. Lung involvement is usually mild and less common. SUlI, it is very uncommon acute respiratory failure to be the pr6sontirlg symptom. A case with leptosplrosl..,s which was presenting with acute respiratory failure is described. A 36 year-old man admitted to ICU becauso of fever, myaigla, aevere c~, hemopty~s. His blood gases showed: PaO2:46mmHg with FIO2:1.0, PCO2:27 mmHg, pH:7.4, HCO3:20mEcl Chest x-ray film demonstrated diffuse bilateral alveolar pattern occupying beth lung4/4). Trarmamlnase, bllllrubln, ~ and ESR were elevated, WBC was 8.7001mm8, Platelet: 40.0001ram3, Hematesrlt:30%, Hemoglobin: .Sgrldl=. There was no clinical or ecttlographlc evidence of left heart failure.Patient fulfilled the criteria for diagnosis ARDS He was found to have an ~lutinatlon tlter for leptoq~lral antigens(indirect he~lutlnatlon atomy, IliA} very high (1/800, negative of patients admitted with PNM in our ICU during the same period (1990-94): Group A, 19 patients HIV+, and Group B, 152 patients HIV-. APACHE II was identical in the 2 groups (p=ns). Group A required more often mechanical ventilation (p=0,O07), had a higher P(A-a)O2 (p=0,004) and metabolic acidosis was more frequent (p=0,001). Regarding laboratorial parameters Group A had a lower no. of linfocytes (p=0,02), a higher LDH (p=0,04) and a more marked hypoalbuminemia (p=O,03). Mortality was higer in Group A (52,6%) than in Group B (29,6%), (p=0,04). Analysing the A group patients, we found no significant differences between alive and deceased patients, with exception for albuminemia, which was lower in the deceased patients (p=0,02). In conclusion, the HIV+ patient's PNM have a more agres sive behavior when compared with community acquired HIV-patient's PNM. The prognosis was not influenced by the APACHE II. Perhaps other parameters such as P(A-a)O2, metabolic acidosis, linfocytes, LDH and albumin shoud be more evaluated as possible predictive indices. Some prognostic factors, usually accepted as predictive in the analysis of HIV+ patients do not seem to be worth in the late stages of AIDS, mainly when they reqquire intensive care. Intensive Care Unit, Onassis Cardiac Surgery Center, Athens, Greece. Objectives of this study was the comparison of two different antibiotic regimens as prophylaxis in cardiac surgery patients. Methods: In a prospective randomised comparative study, two different forms of antibiotic regimens were investigated : a single dose of cefuroxime (Zinacef, 3 gr) (group A) given during the induction of anaesthesia, versus a four days combination of amoxicUline (Amoxil, 2 gr tid) plus netilmicin (Netromycin, 150 mg bid) (group B). A total of 926 patients (pts) (767 males and 159 females, of mean age 60.6+8.7 years old) were included in the study over a period of one year; 424 in group A and 502 in the group B. Patients were checked for the occurrence of infection during the first postoperative month. Results: The total rate of infection in cardiac surgery pts was 5.8%; 5.4% in group A and 6.1% in group B (p=NS). 34 pts (4.7%) developed infection following CABG, 17 pts (7.9%) following valve replacement and 6 pts (17.6%) after other cardiac surgery. They were 43 males (5.6%) and 11 females (6.9%). Endocarditis has occurred 0.4 % in group A and 0.2 % in group B. Severe wound infection was recorded in 0.4% in group A and in 0.8% in group B. One case of sepsis (0.2%) in group A and in group B (0.2%). Respiratory infection occurred in 11 pts of group A (2.6%) and in 11 pts of group B (2.2%). Two cases of urinary tract infection was in group A and one in group B. Catheterrelated infection was occurred in 5 (1.1%) in group A and 6 (1.1%) pts in group B. 3 pts (0.6%) had fever of unclear aetiology in group B. Conclusions: There was no statistically significant difference regarding the rate of infection in both groups. A single dose administration of cefuroxime is accordingly just as effective as a four days regimen of amoxicilline plus netiimicin. Legionella pneumophila is a common bacteria of the environment, and it is an agent responsible for severe community acquired pneumonia (CAP). We analyzed the 8 patients with LpP admitted in our ICU during the last 8 years (1986) (1987) (1988) (1989) (1990) (1991) (1992) (1993) (1994) . They represented 4.6% of CAP. Seven patients were males and 1 female, with mean age 46.7+12.1 years. Tiss was 24.1+10.9 and APACHE II 21.0+5.2. All, but 1 patient, were under mechanical yen tilation (MV) during a mean period of 11.7• (min-l, max-44) days. Two pneumonias occurred beyond the season, while 4 patients had an epidemiological history. Only 1 patient had no risk factor. In all the others tobacco smoking and alcohol abuse was quite frequent. Diagnosis was based on serologic test and culture or direct fluorescent antibody staining of bronchial secretions. Seven patients had a multisystemic disease with hepatic dysfunction in 5, renal failure in 4 (due to rhabdomy~ lysis in 3). One patient had a prosthetic valve endocarditis and another developped ARDS. Nosocomial septicaemie occurred in 3 patients. Mortality rate was 50%. Deceased patients had initially higher APACHE II, (A-a) 02, and lower natriemia. Comparing LpP with the other CAP (n=84), both submitted to MV, mortality rate was similar (57,1% versus 54.7%). In conclusion LpP can occur all over the year. There was a high incidence of severe complications and outcome was similar to the other CAP when requiring MV. Prospective specimen brash (PSB) with culture > 10 CFU 10 3 CFU/ml. Broncho-alv~lat lavage (BAL) ~= 104 C'FU/rnl or positive blood culture. 10 were excluded for rapture of treatment ; 63 were analysed (shift with oral antibiotic8 : 3 ; prohibited antibiotics associations : 5 ; resistant germ : 2). Clinical data : age 60,6 • 18,7 ; SAPS 12 • 2,86 ; MAC CABE I : 76,2% -II : 22,2 % -III : 1,6. 63,5% of the patients were intubated and under mechanical ventilation. The pneumoaiae were : primitive in 35 (55,6%), COPD 9 (14,3%), aspiration pneumonia 19 (30,2%). 75 germs were isolated (PSB 67, BAL 1, blood culture 7) : S. pneumoniac 28 (37,3%), H. influeazae 14 (18,7%), Sttep~:occns 10 (13,3%), SAar6ns 10 (13,3%), Enterobaetdrindr 5 (6,7%), Mosexella catarrhalis 2 (2,7%), othem 6. 71/75 (94,7%) were sensitive to freatment. The ltentment was 100 mg/kg/d of AMPIClLLIN and 50 mg/kg/d of SULBACTAM in continuous IV adminisU'ation during at least 10 days. Clinical eff~ienev : success 46 (73%), Failures 17 (27%) with superinfeetion 7, worsening or relapse 3, dead 5, side effects 2. There was no difference between etiologies : primiti~;e~ 74,3%, COPD 77,8%, aspiration pneamoniae 68,4%. The bacteriological effieieacy was evaluated only for 41 patients with eradication 30 (73,2%), eradication but super~ection 6 (14,6%) : with Pseadomoaas a&ogiuosa 2, Eater~ac~ 3 ; beeteriological failure 5 (12,2%). In conclusion, the aasor AMPICILLIN -SULBACTAM is effective for the I~eatment of severe acquired community pneumonise. Objectives : To assess the efficacy of chlorhexidine (CL) gel or suspension applied in the nose and in the OP for the prevention of the tmcheobronchial colonization. Methods : Thirty-seven patients expected to be intubated for > 48h were randomized to received topical application oga CL suspension (2%) qShrs, a CL gel (1%) q6hrs or a placebo. In addition all Vpts received a nasal and a OP spray (2%) of either CL or placebo administrated according to the same schedule. Semi-quantitative cultures of the anterior nares, the oropharynx (OP) and the trachea were obtained on admission and once a day until extubation (just before the next application). The results were assessed according to the following criteria: success = no acquisition of GNB in the trachea ; failure = acquisition of GNB in the trachea. Acquisition was defined by a follow-up culture positive for a GNB not present in the trachea on admission. Results : SUCCESS failure nosocomialpneumonia overall morality CLsusp. placebo CLgel placebo n=8 n=10 n=9 n=10 5/8 6/10 7/9* 3/10 3/8 4/10 2/9* 7/10 1/8 2/10 3/9 4/10 0/8 1/10 2/9 2/10 i *p = 0,03byFisher'sExacttest Conclusions : these results suggest that topical CL gel administered q6hrs may prevent tracheal colonization by GNB. F. Daumal*, M. Daumal**, C. Plot**, V. Vurmmen ~ E.Colpurt**, B. Manonry** * Hygiene Hospitali&e, ** Service de R6enmmtion, * Service des Admissiens-Urgeuces Centre Hospitalier G-6ndral -02 321 SAINT-QUENTIN -FRANCE Obiectives: Evaluate the nosocemial risk due to peripheral venous inserted short catheters, and the quality of care. Patients-Methods: The intensive tare unit (I.C.U.) is a 9 beds unit. The prospective study includes all the patients comn~ in from 01/01/1993 to 30/03/1995. The recruitemont uses an evaluation schedule of local clinical signs. The nurses aimed to create this evaluation data which includes the place of entry site, the duration of catheterization and the cause ot withdrawal. Only patients staying longer than 2 days in the I.C.U. are accounted for. The diagnosis of uosoenmial infection is assured by the physician taking care of the patient and by the hospital epidemiologist on the next signs: evident pus at the catheter entry site, positive culture of the strain, with or without the same pathogen in the blood sla'uam,the patient having no other distant source of infection. Analyses were performed on EPI/NFO. Results: The occurrence of 3 nosoeomJal inthrtions: I abcess and 2 bacteremia during the first part of the study lent the medical staff to modify the protocol of insertion end survey of the device. So we analysed 2 different periods: Period 1( 1/01/93 to 31/10/93 ) and Period 2 (01/11/93 to 30/03/95 ) for all 1 .e peripheral catheters inserted in the I.C.U. Period 1 44,9 % 46,2 % en infection due to peripheral venous device is a daily threat. The severity of some clinical situations requiring admission in ICU proves it. The motivation of nurses for rigid adherence to established protocol, the daily survey of the entry site, the withdrawal of the peripheral catheter every 72 hours aimed to reduce significantly the local signs of inflammation end infection of peripheral catheters inserted inside the I.C.U. Objectives: To investigate the use of a new metabolic monitoring device for different IPS levels by comparing oxygen consumption (VO2) to measurements of the mechanical work of breathing (WEB) and P0.1. Methods: The study was approved by the institutiotml ethics committee. Eight patients were investigated during weaning after prolonged mechanical ventilation (6-75 days) for various diagnoses when the clinical physician judged the patient to be ready fur weainag. IPS was setto 15, 10, 5, 0 mbar far 20 rain periods each. All patients had a PEEP between 5-8 mbar.. Respiratory frequency (f), tidal volume (Tv), minute ventilation (Ve) were read from the ventilator display (7200ae, Puritan Bennett, Carlsbad, USA). Flow and airway pressure were measured at the endotracheal tube site. Esophageal pressure was measured using an esophageal balloon catheter (Fa. Ruesch, FRG). WeB was determined as the area subtended by the pleural-pressure-vohime curve. P0.1 was determined by using standard occlusion technique and graphical analysis of the airway pressure tracing. VO2 and VCO2 were measured using the PB 7250 metabolic monitor (Puritan Bennett, Carlsbad, USA) connected to the PB 7200ae ventilator. All data are given as mean• deviation for each IPS level. Comparison between the different IPS levels was performed using ANOVA for repeated measurements. Significance was considered at p<0.05, compared to IPS 0 mbar. Results: The values for breathing pattern, WeB, P0.1, VO2 and VCO2 are given in the table for the different IPS levels; significance is indicated by ~. Objectives: Fluidized beds are often used in the management of critically ill mechanically ventilated patients. Critically ill patients are increasingly colonized with resistent pathogens [ie: P. aeruginosa, methicillinresistent S. aureus (MRSA), extended spectrum I~-Iactamase producing enterobacteriaceae ] that can ultimately cause nosocomial infection. Methods: We prospectively monitored bacterial colonization of mechanically ventilated patients and of the fluidized bed (Clinitron) inwhich they were treated. Multiple samples for quantitative bacterial cultures were taken from oropharynx, trachea, feces and bedsores. Samples of ceramic beads from the bed were also taken both during and after patient stay (after bed operation in the absence of patient). Re,~ults: 13 episodes in 12 consecutive patients (mean age: 57.6 years) were analyzed. All had bedsores and/or urinary catheters and fecal incontinence, 7 patients had nosocomial pneumonia, 6 had urinary tract infection [2 with extended spectrum IMactamase producing K/ebsie//a pneumoniae (KI~LSE)], one had positive blood cultures with MRSA, and one patient had a KI~LSE found in high concentrations (103 -10 s CFU/ml) in 2 occasions in feces. Patients were heavily colonized: the , samples from ceramic beads showed no growth or became sterile without any sterilisation procedure (even in one case of presence of Kf~LSE) during the patient stay. Conclusions: Fluidized beds do not put patients at high risk of acquiring nosocomiN pathogens, and cross-contamination between patients seems unlikely, even when multiple resistent organisms were initially present. The recommandation from some manufacturers to undergo extensive sterilization of fluidized beds after use does not seem warranted, at least with the bed used in this study. Ant. Koutsoukou, A, Tahmitzi, P. Kithreotis, M. Koutonlidou, K. Stavrakaki, Kainis E, G. Vlahogiorgos and E. Eliopoulos ICU-Centre for Respiratory Failure -Chest Diseases Hospital of Athens. The cost-effectiveness issue is becoming vital in modern medicine and may lead to moral dilemmas since sometimes certain groups of patients may not have access to highly specialised modalifies. Objective: Our study compared the mean daily cost for antimicrobial medication in COPD patients treated in ICU versus all other patients in the context of relevant epidemiological, prognostic and outcome data. Methods: Age, sex APACHE II score, length of ICU stay (LOS) and in -ICU fatality were retrieved from the files of all ICU admissions over 1994. Mean daily cost for antimicrobial therapy per patient (DCAT) was estimated. These variables were statistically compared between COPD and non-COPD patients. Significance was assumed at p<0.05 Results: 140 of the total 178 admissions were fully evaluable. 38 of them (27%) were COPD patients. Data (m---SD) results for statistical test are given in Table I . COPD patients were significantly older spent more time in the ICU and presented with significantly higher APACHE II scores. Outcome and DCAT were comparable in the two groups. Objectives: The use of heat and moisture exchangers (HMEs) during long term mechanical ventilation (MV) is increasing. In ICU patients, they are routinely changed every day, according to the recommendations of the manufacturers, but the clinical basis for such a daily practice is lacking. We therefore prospectively assessed whether changing HMEs (DAR Hygrobac, SpA, Mirandola, Italy) every 48h only would affect their clinical and bacteriological efficiency. Methods: Two consecutive groups of patients requiring MV for >48h were compared: Group1= HME replaced every day, n= 61 episodes of MV in 61 patients; Group 2 = HME changed every 48h, n=68 episodes in 64 patients. Tubings were not changed in the same patient during the whole length of ventilatory support. Diagnosis of nosocomial pneumonia (NP) was based on a positive quantitative culture (~103 CFU/ml) of a protected specimen brush in patients with clinical signs of pneumonia. Quantitative cultures of pharynx, trachea and y-cannector were performed every 48h. Results: The groups were similar in terms of age, indication for and overall duration of MV (10+_8.6 vs 10+_9days, p=0.9), and severity of illness (SAPS: 16---4.9 vs 16.4+_5.5, p=0.6). The maximal values for peak airway pressure were identical in both groups (33.4-+7.8 vs 33.7• cmH20, p=0.9). Obstruction of the tracheal tube was observed in only one instance in a Group 1 patient who had tracheal bleeding. Circuit colonization was very rare, and of low grade in both groups. The level of patient colonization and the type of organisms were identical in both groups. More importantly, the incidence of NP was the same (6/61 vs 8/68, p=0.7), as was duration of MV before the occurence of pneumonia (9• vs 10.5+_4.7, p=0.6) and overall mortality rate (17161 vs 17168, p=0.7). Conclusions: the clinical efficiency of this HME does not seem altered after 2 days of use. Indeed, replacing this HME every 48h only neither affect circuit and patient bacterial colonization nor the incidence of NP. Therefore, substantial savings could be obtained changing HMEs every other day only. Obiectives: To evaluate the usefulness of different paraclinical investigations for the diagnosis and prognosis of acute viral encephalitis in ICU patients. Methods: We reviewed 13 patients (pts) admitted to our ICU from july 1989 to december 1993 with the diagnosis of acute viral encephalitis. All were in coma and were initially treated as presumed herpes simplex virus (HSV) encephalitis. The causative agents were: HSV (2 cases), Herpes zoster varicellae (1), measle (1), rabies (1), unidentified (8). Eleven pts survived and three presented neurologic sequelae. Twelve pts were investigated by MRI, and eleven also by SPECT and multi-modality EPs. including brainstem auditory EPs (BAEPs). These investigations were obtained as soon as possible following admission and were repeated during ICU stay when possible. The clinical outcome was noted. Results: Six pts (6/12) had an abnormal MRI. Among them, 2 pts made a complete recovery, in comparison with 5/6 pts with a normal MRI. In one HSV infected patient, MRI remained normal despite clinical deterioration and bad outcome. When repeated, MRI became abnormal in 3 cases (with poor outcome in one) and was improved in one. SPECT was found abnormal in 10/11 pts (among them, 4 pts had thus a normal MR/). The correlation regarding the topography of brain lesions was poor between MRI and SPECT. The findings of SPECT could not be correlated with a poor outcome. The BAEPs confmned in 56% of the pts the clinical diagnosis of brainstem involvement. Changes in visual and somatosensory EPs were mild in all the pts and were not helpful for the prognosis. EPs were otherwise interesting for the follow-up of the coma in these sedated and ventilated pts. Conclusions: The value of MRI and EPs for the diagnosis of acute viral encephalitis is of limited interest. SPECT seems to show early modifications, even in pts with a normal MRI, but this test is poorly specific and does not correlate with MRI changes when present. Concerning the prognosis, larger studies should probably confmn that a normal MRI could usually result in a good outcome. This serie illustrates also that HSV encephalitis could be demonstrated only in a small number of cases and that the prognosis of non HSV encephalitis is not easily assessed. Objectives: To study the influence of Gram (-) bacterial lung infections on liver function i~ MV ICU pts. Pts and methods: We studied 102 pts, 68 # (66,7%), 34 (33,3%). Hean age:48,3• years (16-82). Mean stay in ICU:13,3• days (8-75). They were divided in 2 groups: A(44 pts) who did not suffer from pneumonia and B (58 pts) who developed a Gram(-) bacterial pneumonia. Both groups were consisted of pts with same age, sex and disease distribution and same systemic failures. We measured SGOT, SGPT, total bilirubin(TB), direct bilirubin (DB), alk.phosphatase (Al.ph.), V-GT and albumin (Alb.) 3 times: on days O, 4 and 7 of the pneumonia for group B and respectively for g~oup A. Conclusions: 1) In elderly intubated pts of an ICU, KP is isolated more frequently than in ICU pts<65 years (p 0,5 IJg/ml. Results: Gentamicin was administered by the ET and IV routes in 18 and 7 separate sessions respectively. A total of 107 samples were assayed, 69 in bronchial secretions (bs) and 38 in serum. The ET route resulted in higher GM levels in the bronchial secretions compared to the IV route (3,26 + 2,86 vs 2,1 _+ 2,1 pg/ml respectively, p = NS ). Adequate bronchial GM levels were achieved in 100% of patients after ET administration, compared to 66% after IV aaministretion. The blood levels of GM were significahtly lower after the ET vs the IV route (1,56 + 1,95 vs 5,56 • 1,96 pg/ml respectively, p _< 0.01). The ET administration resulted in toxic bronchia~ GM levels in 47% of the specimens. 66% of these samples were from patients with renal failure, however toxic blood levels were reached in only 12% of these. Gentamicin seems to be a safe and adequate alternative route of treatment for the LRTI. However, in patients with renal failure the ET administration of the aminoglycosides should also be modified and continuously monitored. In order to evaluate the pathogenic role of anaerobes in nosocomial pneumonia (NP), we investigated the systemic humoral response in patients who developed a NP with anaerobic bacteria, especially Prevotella species. Methods: Blood samples from 4 groups of patients were tested. Group I: 13 patients with a NP in which Prevotella spp. was isolated from protected specimen brush (PSB), Group Ih a control group of 30 patients with a NP without anaerobic bacteria, Group Ill: a control group of 27 patients with dental stumps but without pulmonary infection, Group IV: a control group of 30 healthy voluntary people with Prevotella spp. isolated from the dental plaque. An ELISA was used to evaluate the total antibodies level against a mixture of four Prevotella strains and a Western-blot method was done to identify the antigenic proteins. Results: Data are expressed as means .+ SD. The antibody levels in patients of group I (63• was statistically higher (p=O.O05) than in the 3 control groups (respectively: 29+25, 32_+25, 31_+23). Using Western-Blot method, the intensity of the response was roughly superposable to levels obtained by ELISA and the profiles were different according to the Prevotella species. The occurence of a NP with anaerobic bacteria (Prevotella species) isolated from PSB leads to an antibody response which seems specific of the Prevotella species isolated. Fever is common in the Intensive Care Unit, but is not always related to an infection. We sought to define the epidemiology of febrile patients in a general medical/surgical ICU. Methods: We prospectively analysed the source of fever (T >38.2 ~ C) in all adult patients admitted for >-48 hours in the ICU during a two month period. These patients were studied for 14 consecutive days. and werc classified in 3 groups according to the evidence of infection (Center for Disease Control criteria) after complete evaluation: Documented infection: CDC criteria + isolation of pathogen (D); Possible infectron: CDC criteria without isolation of pathogen (P); Unlikely infection: patients who did nol meet the CDC criteria (U). Results: Of a total of 208 patients studied, 74 dec'eloped fever (35 6%). including (after complete evaluation) 39 D, 15 P and 20 U palients. Both the highest temperature in tile first day of fever and the maximal temperature were higher in D than in U (38.7•176 versus 38.5•176 and 39.2-~0.9~ versus 38.64-0.4, respectively p= 0.05 and p= 0.003). Most common sources of infection in D were the lungs in 25 patients (64%) and urina .ry tract in 4 (10%). 14 of these patients had positive blood cultures (36%). The overall mortality was 27% (23% in D, 40% in P and 25% in U. differences NS). Antibiotics were given in 100% of D, 73% of P and 15% of U (3 patients). In P there was a non significant lower mortality." in patients who received antibiotics (3/11 (27%) versus 3/4 (75%) patients, respectively). Conclusions: In febrile ICU patients both the highest first day" temperaturc and maximal temperature are significantly higher in infected than in non infected patients, but the differences are too small to be useful clinicall). Mortality rate is not significantly influenced either by the presence of an infection or by the administration of antibiotics, Obiective: Retrospective study to determine the influence of candida infection on ICU outcome. Methods: 126 patieet with a stay of more than 7 days in inteaasive care were screened for candida infection. 70 patients were treated with antifungal therapy due to either an increased antigen titre of -> 1:8 or clinical evidence of candida colonization. Serological Candida-antigens (RAMCO, Pastorex) and antibody titres (hemagglutination, lgG-, IgM-ELISA) were examined routinely. Seroconversion was defined as a threefold increase of antibody titre or a titre of 1:640 or higher. Results: The median length of stay was 37 (ranging from 8 to 132) days, the mean APACHE II score on admission was 18 (+_ 5.8 SD) points. Of 126 patients 31 patients died (24.6%). In the group treated with antifungnls (71 patients) 19 patients died (26.7 %). Although of the 126 patients only 51 (40.4 %) developed a candida infection as defined above the mortality in the group that showed signs of infection was significantly higher (37.2% vs. 14.6%, p < 0.05 [Chi-square-test]). In 34 patients an antigen concentration-> 1:16 was measured. Seroconversion was found in 41 patients. The most common fungus was Candida albicans (66.4 %). Furtberm0re, Candida glabrata was found in 21.1%. Most of the patients were treated with 2 x 200 mg fluconazole (66 patients). In 38 patients therapy was changed to amphotericin B/flucytosine. In 5 patients therapy was started with amphotericine B and flucytosine. In 40 patients a threefold decrease of candida antigen titre was found. 27 patients showed a decrease of candida antibody titre. Conclusions: Meticulous screening for eandida infection seems to be necessary since the number of patients with fatal outcome is significantly higher in the group with signs of fungal infections and thus requires immediate antifungal treatment. Objective: Early diagnosis of patients with ventilator-associated pneumonia (VAP), and subsequent identification of causative microorganism, and selection of the appropriate therapy are critical important points that affect morbidity and mortality. The results of the quantitative bacterial cultures are not available for at least 24 hours, while a two hours period, since the specimen are obtained is enough to know the gram stain results. The aim of this study is to determine the usefulness of gram stain in specimens obtained by bronchoaiveelar lavage (BAL), through the bronchoscope. Material and Methods: We studied 47 patients (36 males and 11 females, age 49 + 23) with suspected ventilator-associated pneumonia. The BAL gram stain was considered positive when the specimen after a centrifugation at 1500 rpm for 15 min revealed: i) more than 20 leukocytes per optic field, ii) squamous epithelial cell less than 1 percent and iii) one or more microorganisms per optic field on 1000 magnification. All patients had been receiving antibiotics, with no change during the last 3 days, prior to bronchoscopy. Results: 8 patients had VAP and 39 patients did not. In 5 cases the BAL specimens (quantitative bacterial cultures) established the diagnosis of VAP In the remaining three patients the VAP diagnosis was established by other procedures (blood or pleural fluid culture, clinical outcome, autopsy). Apache fl score in patients with VAP was 15,7 -+ 5,5, while in patients without VAP was 17,9 + 6,4. There was a significantly higher incidence of VAP in patients who had i) coma (GCS <8) and II) been receiving neuromuscular blockade (p<0.05) . The sensitivity of the gram stain for VAP diagnosis was 75%, the specificity 89,5%, the positive predictive value 60%, and the negative predictive value 94,6%. Conclusion: Our data indicate that the gram stain of BAL specimens is useful for the early diagnosis of VAP and the subsequent administration of the appropriate treatment. The role of anaerobes in mechanically ventilated patients with pneumonia (MVP) have been poorly investigated Aim of the study : Analyse the prevalence of anaerobic isolation in MVP. Methods : Between october 1992 and february 1995 all suspected MVP were investigated using protected specimen brush (PSB) technique. Brushes were rapidly transported in Shaedler broth to laboratory. A special care was tooken for anaerobic isolation. Results : Among the 153 PSB performed for suspected MVP (132 nosocomial and 21 community-acquired pneumonia), 81 yielded at least one micro-organism (positive PSB : 53%). 63 of positive PSB demonstrated only aerobic bacteria and 18 (23%) yielded with anaerobes. In 14 out 18 patients, anaerobes were associated with aerobic bacteria. Anaerobes were mostly isolated in nosocomial pneumonia (17/76 positive PSB). 27 strains of anaerobes were isolated. Prevotella species represent 19 out these 27 strains (70%) The most frequent anaerobic species were Prevotella oralis (6) P. intermedia (5) and P. buccae (4). Comments:Using adequate methods, anaerobic bacteria are frequently isolated in MVP. It could be off importance to take in account anaerobes in the choice of empirical antibiotic therapy in MVP. Objectives: The majority of patients with multiple trauma are considered immunocompromised. The aim of this study was to identify risk factors of pneumonia in mechanically ventilated patients with multiple trauma or after surgery. Methods: In this prospective study we studied 64 multi-trauma patients (mean age 58 + 19 years, Apache II 16.5 + 6), admitted to a general intensive care unit (ICU). All patients were intubated and mechanically ventilated. We were considered that a patient had Ventilator associated pneumonia (VAP) when the specimens of Bronchoalveolar Lavage (BAL) or Protected Specimen Brush (PSI?,), ebb'ned through the bronchoscope, had one or more microorganisms in concentrations greater than 105 and 103 cfu/ml respectively. All patients had been receiving antibiotics, with no change during the last 3 days, prior to bronchoscopy. Results: 14 patients had VAP, and 50 patients didn't. In the bivariate analysis, the Glasgow Coma scale (GCS)<8 (x2=4.15, P<0.05), the administration of neuromuscular blockade (x2=7.9, p<0.05), the duration of mechanical ventilation to be greater than 5 days (x2=5.5, p<0.05), the flail chest (x2=4.1, p<0.05), the parenteral nutrition (x2=5.6, p<0.05), the ARDS (x2=3.9, p<0.05), the Abbreviated Injury Scale (AIS) of more than 4 for thorax (:,:2=5.9, p<0.05), the pneumothorax (x2=5.1, p<0.05) were statistically significant related to development of VAP. In multivariate regression analysis, using the stepwise technique, three of the seventeen studied factors showed to have an indepantent association with the development of VAP:The administration of neuromuscular blockade (F: 4.8, p<0.001), flail chest (F:3.0, p=0.003), and GCS (< 8) (F:2.1, p= 0.039). Conclusions: In patients admitted to ICU for multiple trauma or major surgery, the administration of neuromuscular blockade, the flail chest, and the GCS (<8), in the population under study, were the indepedent risk factors for VAP. MOF is a sereous complication of differem states: infection, sterile inflamation, extensive fissure injure, intoxication, ets. There is close correlation between extension of MOF and death, developement of nasocomial infection. immunologic disfunction. In order to prgnose probability of risk of MOF development among the patients with sepsis and septic shock, we achived an eqation, allowing to recive a coeficient, closely connected with this probabiliti. We have used retrospective analisis of 160 cases of sepsis. Diagnosis of sepsis was based according to Bone's criterions of sepsis. MOF was assessed as disfunction of 2 or more systems according to Bone's classification of MOF. Having used correlation analisis we have estimated factors which have had high correlation coeficient with the probability of development of MOF. There were: APACHE-II score points, evidenceof septic shock, endocrinopathy. With the help of multyple regression analisis we acheved next equation: y= 0,2042 + 0,0243x~ +0,2931x 2 +0,1504x3 , Were x I-APACHE-II score points, x2-evidence of septic shock, x3-endocrinopathy. The explanatory power of this quation was evidenced by ROC of 0.88, SE (V 0 -0.033 Introduction: The presence of liver dysfunction in the process of multiple organ failure is associated with an adverse outcome, particularly when it becomes progressive to liver failure. Disturbances of liver function may occur early and their detection may be of significant importance for the further development of organ failure. Routinely used liver function tests appear to be inconsistent indicators of hepatic damage. In this study, we used p_lasma disappearance rate (PDR) of indocyanin-green dye (ICG) as an early estimate of liver function. Methods: We serially evaluated PDR and routine liver function tests (serum bilirubin, SGOT, SGPT), as well as acute phase and non-acute phase proteins (CRP, transferrin) in 26 patients during the first week after trauma or the onset of sepsis. Patients: Group 1: (n = 11) multiple trauma ISS > 30, Group 2: (n = 15): abdominal sepsis, acute necrotizing pancreatitis (ANP) grade III. Patients were selected on the basis of clin4cal estimates that these patients would require continued ICU observation. PDR was determined by means of a fiberoptic catheter and a computerized system (COLD Z-021, Pulsion), which permits repeated bedside measurements. The initial values of PDR, serum bilirubin and transaminases were not significantly different in trauma, sepsis and ANP. In trauma patients PDR improved during the first week. In patients with sepsis and ANP PDR remained low and worsened with time. The decrease in PDR preceeded an increase in biochemical liver function tests in these patients. 1+4.4 110&-_11 7(4-9) Discussion: Routinely available blood tests of liver function are usually altered several days after injury. However, they are generally non-specific indicators and they are influenced by extrahepatic factors. PDR seems to be useful to evaluate impaired liver function early after the onset of sepsis and trauma. Objectives: To study frequency of organ system failure (OSF) and it's influence on outcome in granulocytopenic patients with hematological malignancies and septic shock(SS). Materials and Method: Retrospective review of medical records of 52 granulocytopenie(WBC<0,5xl09) patients with hematological malignancies and SS, who were admitted to the Intensive Care Unit (ICU). Frequency of OSF before and after SS was analysed. The patisnts were categorised on survival and non-survival. Results: Signs of OSF were observed in 28.8% of patients before SS and in all patients after SS. Only 3 patients presented with hypotension refractory to inotropic therapy. Nevertheless there was a significant increase of frequency of acute respiratory failure (ARF), acute renal failure (ARenF) and liver injury (LI) after SS occurred(showed on the figure). Only 5 Frequency of organ failure before and after Objectives: Statusmetria allows to define the effective level of oxygen status and accordance to it means of carbon dioxide and elec-trolyte8 in critical care. The conception of syndrome int~ive care (SIC) is exhausted itself and invariable outcomes of SIC of MUltiergan System Failure (MOSF) confirms that. Therefore, an alternative to SIC should be advanced. Methods: Efficlenoy of treatment has been asscsaed in 257 patients with MOSF using value of metabolic rate and ability of an organism to cover it by oxygen and substrate supply. Oxygen pulse (OP) and Index of Efficacy of Oxygen Transport (IETO2) was monitored. ~Lt~.Lntenaive care is considered to be Homeostasis-Securing Therapy (HST) if energostructure deficit is eliminated and necessary for recovery regeneration rate is .restored. OP in patients with MOSF was 0.8 mt-m "2, and le,~ and IE'I~ w~ 2.4 units in SIC. We managed to maintain OP of 1.0-1.7 ml.m "2 and IETO2 of 1.9-2.6 units in HST. 41 patients from 135 with MOSF survived in SIC and 96 patients from 122 survived in HST. Efficiency of HST appeared to be two times as much as efficiency of SIC. Cr of homeostasia-se-'uring therapy is advancing. The conception provides restoration of regeneration rate due to effective then in SIC elimination of en=gostructure deficit. The conception may be a basis of new technology for treatment of MOSF. Helen F Goode PhD, Nigel R Webster PhD. Anaesthesia & Intensive Care, University of Aberdeen, AB9 2ZD, UK. Objectives: Xanthine dehydmgenase is converted under conditions of ischemia, reperfusion and endothelial damage to xanthine oxidase, with superoxide anion as a co-product of its catalytic activity. Multiorgan dysfunction syndrome is associated with splanchnic vasoconstriction resulting in significant and prolonged gut ischaemia. Aggressive volume resuscitation with prompt restoration of blood flow results in reperfusion of the tissue and is likely to cause xanthine oxidase-mediated release of oxygen-derived radicals. This study investigates xanthine oxidase activation and oxygen-derived free radical-mediated damage in such patients. Methods: Fourteen consecutive patients on ITU who met established criteria for septic shock and secondary organ dysfunction were studied. Serum xanthine oxidase activity was measured using oxidation of a chromagen in a dual enzyme system and plasma malondialdehyde was measured using a specific spectrephctometdc assay. APACHE II scores, blood pressure, SVR, cardiac output and 28 day survival were also recorded. Biochemical data were compared with results from 20 healthy subjects. Results: Xanthine oxidase activity was 6.30 + 1.59 units/I in patients (mean :t: SEM) and 0.74 + 0.12 units/I in controls (p 4 failing organsysterns was 80% the only exception being the subgroup of trauma patients where mortality under these circumstances was 5O% Conclusions: Mortality in surgical ICU patients receiving RRT for ARF is high. No significant difference in mortality is found between RAAA and EVS. Mortality increases with the number of failing organ systems. The subgroup trauma patients shows a lower mortality compared to the group as a whole, even with > 4 failing organ systems. To look for the most accurate scoring system to measure the severity of the complications occuring in the early phase ( first 30 day) of kidney transplantation and to asses their prognostic value. Methods: In our retrospective study we applied the APACHE lI and the Goris scoring system for the kidney recipients who developed multiple organ failure (MOF) as a consequence of their pulmonary and. cardiovascular complications following kidney transplantation. We evaluated the recipients The distribution of the women and men ( 60% ~ 40 % ) was the same as in the kidney recipients. Applying the APACHE II system most of the patients had their score between 10 and 19, and the function of 2,3 or 4 organs were affected at the time of the onset of MOF. The APACHE II system gave adequeate information about the disturbance of the function of other organs beside the kidney failure even at the time of the transplantation. The scores and the number of the affected organs correlated with the condition of the patients in the Goris scoring system but not as sensitively as in the APACHE II scoring system. Conclusions: Both the Goris and the APACHE II scoring system can be applied to measure the severity of the multiple organ failure occuring during the early phase of kidney transplantation. However the APACHE II system is more suitable to follow not only the stateof the patients at the time of the admission but also the changes occuring in their condition during the complication. V.V.Erofeev, V.V.Ivleva Scientific Research Institute for General Reanimatulogy Russian AMSci, Moscow, Russia Objectives: The analysis of SSC and results of their treatment in patients following critical states showed the necessity of developing a combined antibacterial therapy. Methods: According to the protocol 97 patients (18-60 years old) with combined trauma and massive hemorrhagy following vast aml traumatic operations were examined. Microflora's composition and resistence to up-to-date antibiotics was studied using the anaIyser IEMS Reader by "Labsisteme"(Finland). General clinical, bacteriological, immunological indices, as weil as the duration of the treatment and recovering rate served as criteria of the combined antibacterial therapy effectiveness. Results: It was proved expedient to administer antibiotics in staphylococcus infection in the following combinations: Riphampizin with fluoroquinolones; I-II degeneration, cephalosporins with aminoglycosides; cephalosporins with fluoroquinolones. In case of singling out the exciters of the Euterobacteriaceae family, including the Pseudomonas aereginosa, -fluoroquinolones combined with modern amynoglycosides; fluuroquinolones with ureidopenicillines; ureidopenicillines with amynoglycosides; amynoglycosides with the II-III generation cephalosporins; cephalosporins with fluoroquinolones. In severe SSC caused by combined infection (including anaerobes) clindamicin with modern amynoglycosides was prescribed. Conclusion: The combined antibacterial therapy allows: 1) to increase the effect on microbic agents and the efficacy of treatment in combined infections; 2) to lessen the possibility of the exciters'resistence to antibiotics; 3) to prevent the development of superinfection: 4) to decrease the doses of medicine and its toxic effect. Objectives: Two methods of blood volume measurement in a group of critically ill patients were compared to investigate the practical possibilities of a new easy to use method based on carbon monoxide (CO) uptake. Methods: All patients had multi-organ failure and haemodynamic monitoring with a Swan-Ganz catheter. Mean APACHE II score was 19 (10-25). When indicated, 9 patients had blood volume measurements simultaneously based on the techniques of, i) dilution of 5~Cr labelled red cells, and ii) inhalation of carbon monoxide gas with measurement of the rise of carboxyhaemoglobin produced. The CO was administered via a newly designed, ventilator driven, fully closed circle system ensuring CO retention and CO2 removal with automatic addition of oxygen to m}ttch patient uptake. A portable computer performed all necessary calculations. Results: Volumes obtained by CO uptake were compared with the "gold standard" radiolabelling method. Mean blood volume determined by the CO method was 6310ml (4710-7959ml) compared with 4690ml(3755-5778ml) with SlCr labelled red cells (r=0.9). Regression analysis produced an intercept at 769ml. The slope of the regression line was 0.62 (0.33-0.9, 95 % confidence limits). Discussion: The CO method produces volumes in excess of the radiolabelling method. There appears to be a systematic error, and one possible explanation is CO binding to substances other than haemoglobin. Conclusion: The CO method is easier to use than radiolabelling and of the lower cost, since COHb measurement only is required. Aceuraey is sufficient for clinical use and our preliminary findings suggest this system will meet the requirements. Objectives: This study was conducted to determine the role of nitric oxide (NO) in the pathophysiologic alterations and multiple organ damage, and the possible effects of " " " (L-N -monomethyl-L-arglnlne NMMA) on hemodynamics and mortality in rats caused by a prolonged hypovolemic insult. Methods: A prolonged hemorrhagic shock (30-35 mmHg for 180 rain) was induced in anesthetized rats followed by adequate resuscitation. L-NMMA was administered intravenously at doses of 2.0 mg/kg or 20.0 mg/kg at the end of resuscitation. Results: Infusion of 2.0 mg/kg L-NMMA diminished the fall in mean arterial pressure, significantly increased the cardiac index (CI) and stroke volume (SV), together with remarkable protection from multiple organ damage compared to the controls. The 48 h survival rate was significantly improved from 26.7% in the control group to 68.8% in the treatment group (p<0.05). In contrast, the high dose of 20.0 mg/kg L-NMMA resulted in a strong blood pressure response but a marked reduction in CI and SV concomitant with an increased total peripheral resistance index within the observation period, and caused severe damage to various organs at 2 h after treatment. In addition, marked elevation in both endotoxin and TNF levels were observed in animals subjected to shock insult. Conclusions: These results suggest that NO induced by hemorrhagic shock in rats is an important mediator for pathophysiologic alterations associating with cardiovascular abnormalities, multiple organ dysfunction, and even lethality. Thus, regulation of NO generation and use of NO inhibitors might provide new aspects in the treatment of hemorrhage related disorders, and the use of L-NMMA would be either deleterious or salutary in a dose dependent manner. (Hebert, Chest-1993) . The purpose of this study was to assess the risk factors for hepatic dysfunction in MOSF. Methods: 733 patients have been hospitalized in our ICU from January 1992 to May 1. 994, 198 (27%) with MOSF. Among MOSF pati~ts, 57 (29%) have had hepatic dysfunction defined according to Hebert (bilirubin ~ 60 ttmoP1, Chest 1993). Thirty six of these 57 patients acquired hepatic dysfunction after admission in the ICU. These 36 patients were compared with 36 MOSF patients without hepatic dysfunction selected blindly. Chrorfic diseases, severity scores, eanse of admission, clinico-biologieal and hemodyunrrfic parameters, use of vesopressors, use of hepaiotoxic drugs, use of nutritional support and mortality were compared for hepatic failare and non hepatic failure groups.Twenty nine patients had postmortem hepatic histologic examination, Results: Univaciate analysis: only parameters with p _< 0.05 are pre~nted. Including these paramet~'rs in a multivariate analysis, anly c~hosis and vascular surgery remain independent risk factors for hepatic dysfunction. In particular, PaO2/FiO2, arterial lactate, DO2 were not different between the two groups, Some de~'ee of histological abnormalities was found in all liver samples, despite a normal bilirubin level in 15 % of the cases Conclusions: In our patients, conU'ary to previous studies, hypoxic and hemody~anfic parameters were not independent risk factors for hepatic dysfantion. This might be due to the inadequacy of the usual biologic definition of hepatic dysfunction as well as to the poor sensitivity of general hamodynamic parameters. Critical states of various origin are complicated with the mldtiorgan farm (MOI~ oceuzr~ce. Due to their and functional features the lungs become the primmy damage target in various critical.states. ARD8 that occurs in such states is associated with pulmonary edema development because of capillary permeability increase mediated by humeral and cenular responses to 0amag/~5 factors exposure. R nmst be emphasized that mediators and effecto~rs of this respo~e affect not only puknonary capillaries, but other organs capiU~es as welLenhancing their permeability. Orsans edema is a conmm~ finding at the autopsy of patients died from MOF.Clinical and radiolosial findings allow to have a diagnosis of pulmonmy edema before ~mi!ar lesions in other organs occm. Additionally, there are some techniques that permit quantitative assessment of pulmonary edema flv.id (EVLW) volume. In conclusion, we suggest that EVLW changes in .dyn~rmcs in patients with MOF are considered as a critical state severity measure which reflects indirectly the edema in other organs. Objectives: We compared three different dialysis membranes to find out whether or not there were differences between their clearance characteristics on substances such as inuline, creatinine, urea, and phosphate to be eliminated in acute renal failure (ARF). Moreover, if a loss of clearance did occur we were interested in whether this was due to heparinization and a high production of the thrombine-anti-thrombine-complex (TAT). Methods: We carried out a randomized controlled study on 13 consecutive critically ill patients presenting with ARF, most of them in association with multi-organ failure, to be treated by continuous pump-driven arterio-venous renal replacement therapy on continuous low-dose heparinization. Three different types of high-flux filter membranes (F 60 TM [Fresenius] , CT 190 TM [Baxter] , and Filtra116 TM [Hospal]) were assessed. Each filter was changed intentionally after a 24 hours" use. Together the data of 54 filters were evaluated, each at three different times (immediately after its onset [0 hi, after 8 h, and after 24 h). The clearances of creatinine, urea, phosphate, and inuline were measured. Results: There were some significant differences in clearance characteristics of inuline, creatinine, urea and phosphate between the filters (p<0,05) showing the F 60 TM membrane excelling Filtra116mand CT 190 TM the more. The loss of inuline clearance (3 mI/min/m 2) after 24 h, however, was insignificant for all 3 filter types. A continuous low-dose heparinization scheme was applied without any relevant prolongation of the aPTT. Even lower losses were noted for the clearances of creatinine, urea, and phosphate. We found the TAT-producfion increased after 8 h (p<0,05), but it did not rise any further. Conclusions: As we could demonstrate in our study the clearance data of different types of filter membranes applied during continuous renal replacement therapy do show significant differences. On the other side, no relevant loss of clearance occurs during a 24 hours" period indicating a high efficiency over time. To consider commercial aspects as well it shows that inexpensive conventional filter membranes can successfully be applied even for a longer renal replacement period, if needed. A retrospective study was performed on 100 patients with acute renal failure (ARF). We analysed survival in continuous (CD) and intermittent dialysis (HI)). Mean age of the patients was 60 years (y), 57 patients (57 %0) were <65 y, 43 patients (43%) were >= 65 y. The incidence of dialysed ARF in our mixed Intensive Care Departement is 3%/admission/y. Statistics: Fischer's Exact Test, Mann-Whitney-U test. Efioloev: the contribution sepsis, cardiac failure and aminnglycosidcs was respectively 70%, 44 % and 35 %. Treatment: CAVH (CD) or CVVH (CD) was used in 40 patients (40%), hemedialysis (HD) was used in 60 patients (60 %). Data: Mean Apache 2 scores were the same for CD and HD (27 for both groups), Patients treated with continuous dialysis techniques had significantly (p=65 y (26 vs 30; p<0.05). Patients<65 y had significantly (I}<0.05) more coagulation disorders (53 % vs 17 %) and elevated bilirabin (81% vs 52 %). There was no significant difference in vasopressur need and ventihatio~ between age groups. Outcome:. HI) had a better SR compared to CD (43 % vs 15~ p<0.05). Patiants>=65 y had a comparable SR vs patients<65 y (3") */e vs 28 %; p----a.s.). Tha global survival rate (SR) was 32 % (32 patients). Conclusions : Diaiysed ARF has a well known lowsurvival rate (32 %): Hc~raedialysed patients had a better survival rate than patients treated with continuous dialysis. This can be explained by the fact that the latter were in a worse condition considering organ failure (more vantilatian, elevated bflirubin and need for vasepressurs), Apache 2 score couldn't illustrate that. Patient~65 y with ARF have the same survival rate as patients<65 y: Although Patients >=-65 y have a higher Apache 2 score they have less organ faille. The Avacbe 2 score is not a good oredictor of survival in P with organ failure. Departments of Surgery and Intensive Care, Guy's hospital, London, U.g-Obiectives: A randomised controlled trial of a management protocol utilising the regular measurement of gastric intramucosal pH (pHim) to control the administration of dopexamine. Methods: Patients admitted to a multidisciplinary teaching hospital intensive care unit (ICU) undergoing insertion of a pulmonary artery catheter were managed according to a resuscitation protocol. Randomisation was to either the protocol alone or to insertion of a nasogastric tonometer and subsequent management guided by pHim. pHim < 7.32 initiated volume and inotrope resuscitation and, if unsuccessful in elevating pHim, dopexamine was commenced. Approval was obtained from the hospital ethics committee. Results: 94 patients were considered for analysis and the two groups were well matched for age and sex. Overall, there was a high hospital mortality of 64.9%. There was no difference in ICU or hospital mortality between the two groups (see table) . Objectives: To compare Cardiac Output (CO) measurements between continuous termodilution (CCO) by thermal wire on pulmonary artery catheter (CCO/SvO2 Vigilance. Baxter Critical Care), and CO measurement using a trans-esophageal Doppler (DCO) ultrasound system (ODM II, Abbott Laboratories), in the immediate postoperative period of cardiac surgery. Methods: 15 patients undergoing myocardial revascularization were monitored with CCO by a Swan-Ganz catheter and an intra-esophageal DCO probe, after induction of anesthesia. Exclusion criteria were: Aortic valve disfunction, previous valvular surgery esophageal disease, absense of sinus cardiac rhythm, and need of ventricular or intraaortic assistance. Hemodynamic parameters, CO by both CCO and DCO, SvO2. SaO2, diuresis, pHa, and hemoglobin were repeatedly registered during the first 6 hours after surgery, as the patients were kept under sedation and mechanical ventilation. Results were compared using the method described by Bland and Altman. Results: 176 measurements of CO were obtained, ranging 2. Objectives: A decreased tissue oxygen delivery is responsible for a higher morbi-mortality rate among surgical patients; this diminished oxygen delivery/consumption rate (DOJVO2) may origin the lactic acidosis observed in the gastrointestinal tract, reported in patients undergoing hypothermic cardiopulmonary extra corporeal surgery, and can be registered by tonometry as result of the gastric mucose pH. The purpose of this study is to evaluate the reliability of the intramucosal pH (pHi) measurement by a nasogastric catheter as indicator of the DO2/VO > its co> relation to other parameters of DO2/VO 2 disturbance, and with postoperative complications and clinical course. Methods: 20 patients (16 male, 4 female) undergoing cardiac surgical procedures were included (16 myocardiai revascularizations, 3 valvular substitutions, 1 constrictive pericarditis). Mean age was 63 + 12 years, mean weight 70 _+ 10 kg. A nasogastric probe (TRIE Tonometrics) was placed after anesthesia induction; pHi values were registered in the postoperative period (90', 120', 240", 360' and 18 h after surgery end). The corresponding hemodynamic parameters, venous oxygen saturation (SvO2), diuresis and arterial pH (pHa) were also recorded. Results: pHi values ranged 7.20 to 7.65 (mean 7.40 (0.8); the mean values of clinical evolution were: extubation time, 20 _+ 12 hr.; discharge from postoperative care unit, 88 4-50 hr.; and hospital total postoperative time, 10 _+2.2 days. Complications registered were: 2 perioperative acute myocardial infarctions, 2 cases of respiratory insufficiency, 1 occlusion of coronary bypass, an 1 ease of hyperamilasemia. All patients with severe complications needing specific treatment showed either a low pHi value, or a considerable descent in comparison with the initial register. Statistic correlation between low pHi and presence of complications was found; the low significance (p > 0.05) degree may be due to the low population size. Conclusions: phi measurement in cardiac surgery patients is a non invasive, uncomplicated method for prediction of DOz/VO 2 disturbances, thus reflecting risk of increased major complications, and may precede changes in other usual indicators (SvO2, pHa, Cardiac Output, ...). Work-in-progress with a greater population size may offer more significant results. References: (1) Gutidrrez G: Lancet 1992; 339:195-202. (2) Landow I: Acta Anaesthesiol Scand 1994; 38: 629-639. The haemoglobin-level (hb) is besides the arterial oxygen saturation and the cardiac index one of the relevant parameters of oxygen supply to the tissue. In contrast to otherwise healthy patients, there is no agreement on tile so-called transfusion-trigger in critically ill patients. In I?ont of this background the question arises, whether and to what extent blood transfusion in critically ill patients improves oxygen supply Io tile tissue. This study was performed in 34 critically ill/septic patients in the postoperative period alier an inlcclive/scptie revision operation of the hip or knee joint. On cardiac/seplic reasons monitoring consisted beside other measures of a pulmonary arlery catheter and of an indwelling arterial line li~r measurering/calculating standard haem~dynamic as well as systentic oxygen parameters. The indication for blood transfusion was given by hb together with the cliuical slatus of thc patienl (ASA-scorc and multiple organ dysfunction (MOI))). Statistical analysis w~ks performed by MANN-WHITNEY-U-test. by FISHER's exact-test and by WII.COXON-test: statistical significance was set with p<0.05. According tu the pretransfusion value of hb and of lactate (lac) palicnts ;,,'ere divided into groups as follows: A: hb<8 and B: >Sg/dl: I: 1ac<2.8 and II: >2.SmM. In either group blood transfusion results in zt significant increase in hb (A: 7.5_+0.4 to 9.4+0.8 g/dl; B: 9.(~0.8 tt, 10.5+0.09 g/dl; I: 8.0-+1.0 to 10.2-+0.6 jdl; I1:8.4-+0.9 to 10.1+0.7 g/dl). Wlailc, however, haemodynamic parameters do not difl)r significantly from each other before and alter blood transfusion, oxygen delivery (DO, -ml/min x m-') increases significantly hi either group studied (A: 467-+86 to 581-+158; B: 521+125 to 577+137; 1:512 -+1 13 to 599-+141; I1:516-+123 to 632-+214), In contrast oxygen consumption (VO~ -ml/min x m e) does not change significantly in either group (A: i68-+148 to 158-+38; B: 180-+162 to 175-+52; I: 171-+38 tu 179-+35; 11: 199-+60 to 219+_71); oxygen exlraction ratio decreases. This study in critically ill/septic patients demonstrates, that in this group of patients studied blood transfusion at a base-line-value of >7.5-+0.4 g/dl expectedly rises DO~, however, it does not improve VO=; even not in septic patients with elevated lac-values. Paclitaxel in a new anticancer agent, extract from the bark of the yew tree (Taxus brevifolia), employed against breast and ovarian cancers resistant to chemotherapy. It promotes the polymerization of tubuline, and disrupts the normal microtubule dynamics. Hematologic toxicity, hypersensitivity reactions (bronchospasm, urticaria and hypotension), and peripheral neuropathy are the main reported toxic effects. Cardiac side effects are rare: atrioventricular blocks of higher degree are reported in 0.1% of patients; congestive cardiotoxicity was discussed only in one trial in patients treated with paclitaxel and doxorubicin. We describe the history of a 48-years-old worn an with a breast cancer, diagnosed in 1989, initial staging T3NIM0, treated with mastectomy, axillary lymphadenectomy, andchemotherapy with a cumulative dose of anthracyclines of 678 mg/m2 until August 1994. The patient complained of dyspnea and severe hypotension immediately after an intravenous infusion of 100 mg paclitaxel, given over 1 hour for the treatment of bilateral, malignant pleural effusion. At echocardiography die left ventricular ejection fraction was reduced to 20%. She died 20 days later because of a severe cardiac low output with hepatic and renal failure; an impressive hepatic cytolysis was observed. The post mortem examination confirmed the dilatation of the cardiac cavities, especially of the right ventricle, bilateral pleural fluid, and ascites. The histology was suggestive for a cardiomyopathy secondary to anthracyclines. The electron microscopy revealed a deposition of an unusual pathological pigment in the myocytes; subsarcolemmal deposition or membranous were absent. We hypothesize that paclitaxel was the cause of a major hypersensitivity reaction with shock and severe hepatic cytolysis, worsening the myocardial damage induced by anthracyclines. The possibility that a low doge of paclitaxel could directly increase anthracyclines cardiotoxicity -as decribed in the medical literature -will be discussed. Objectives: Activated endothelial cells release soluble intercellular adhesion molecule-1 (sICAM-1), vascular cell adhesion molecule-1 (sVCAM-1), and E-selectin (sELAM-1). sICAM-1, sVCAM-1, sELAM-1, and inflammatory cytokines were determined. Methods: sICAM-1, sVCAM-1, and sELAM-1 were determined by ELISA. TNF-a, IL-6, and IL-8 were also measured by ELISA. Endotoxin was measured by an endotoxin-specific Endospecy test after pretreatment of new PeA method. Results: The sICAM-1 and s VCAM-I levels were significantly higher in the septic multiple organ failure (MOF) and sepsis groups than in the non-septic MOF group. The sELAM-1 level was slightly higher in the septic MOF group than in the sepsis withut MOF group and non-septic MOF group. The increases of soluble adhesion molecules were not in agreement with changes of plasma endotoxin level. Levels of soluble adhesion molecules were correlated with the levels of plasma TNF-a and IL-8, but the level of IL-6. Discussion and Conclusion: The slCAM-1 and sVCAM-1 levels in septic patients closely reflected the severity of the pathophysiological conditon. It was possible that the release of sluble adhesion molecules were not stimulated by plasma endotoxin, but endotoxin in the local infectious region. TNF-c~ and IL-8 also were suggested to be involved in the release of these soluble adhesion molecules. Obiectives: Cardiopulmonary bypass (CPB) surgery is associated with a systemic inflammatory response attributable to the release of various inflammatory mediators and the activation of complement or coagulofibrinolytic system. In addition, adhesion molecules, such as ICAM-1, ELAM-1, and VCAM-1, appear to be of central importance in the inflammatory process following CPB surgery. We previously reported the effects of a synthetic protease inhibitor, FUT-175, reduced release of inflammatory cytokines (TNF, IL-lg, IL-6), activation of complement (C3a, C4a) or coagulofibrinolytic system (TAT, PIC, FPA) and protected platelet function (GPIb, GPIIb/llla) following CPB surgery. Methods: In this study, we analyzed FUT-175 on soluble adhesion molecules following CPB surgery. 20 patients undergoing CPB surgery were divided into two groups, Group A consisted of 10 patients who received 1Omg of FUT-175 in priming solution, followed by a continuous infusion at 2mg/kg/hr during CPB in addition to initial heparin dose of 3mg/kg. Group B, a control group, included 10 patients who were injected with heparin only. The plasma slCAM-1, sELAM-1, and sVCAM-1 concentration was measured by ELISA. Results: Every soluble adhesion molecules decreased during CPB in both groups, and rose after CPB. sELAM-1 and slCAM-1 reached their peaks on 3 hours after CPB and on POD 1 respectively in both groups, but they remained lower in group A (sELAM-I: 32.1+11.5 vs. 38.6• ng/ml, p<0.05, slCAM-I: 327• vs. 483• ng/ml, p<0.05), sVCAM-1, in both groups, remained lower than preoperative levels, but did much lower in group A. Conclusions: FUT-175 reduced adhesion molecules and suggested to be the effect on postoperative organ dysfunction. In the last few :,'ears the conditions of treatment in continuous hemofiltration/hemodiafiltration were discussed controversially. A significant removal of TNF-alpha and IL-I could be demonstrated in CVVHD. The aim of our study was to investigate the elimination of TNF-alpha, 1L-2, IL-6, IL-8, s-CD-14 and IFN-gamma in CVVH by measurement in plasma and hemofiltrate of 10 critically ill patients with an acute renal failure. The patients of our study were treated with a continuous veno-venous-hemofiltration (polysulfone-filter, blood flow: 100-130 ml/h, filtration rate 500 ml/h). The samples, hemofiltrate and plasma, were taken one hour after the start of treatment. The patients suffered from septic shock (4), the so called hepatorenal s~aldrome (3) and a severe pancreatitis (3). The cytokine concentrations were measured with ELISA-method. In contrast to elevated concentrations in plasma for TNF-alpha (5 cases), SCD 14 (9 cases), IL-2 (l case) and IL-6 (4 cases), hemofiltrates contained no activities. Only IL-8 was removed in significant amounts with even higher levels in hemofiltrate than in plasma. This phenomenon was described so far for TNF-alpha and IL-1 and may be due to the absence of metabolic properties (possibily enz~natic) in hemofiltrate. It can be shown, that TNFalpha, IL-2, IL-6 could not be eliminated in CVVH with a filtration rate to 500 ml/h. In contrast to findings of other investigators with a higher filtration rate (> 1000 ml/h), we found no significant concentrations of TNF-alpha and IL 6 in hemofiltrate. We conclude, that for a significant removal of important cytokines higher filtration rates (>1000 ml/h) are necessary. Objectives: Multiple organ dysfunction syndrome including liver and renal impairment is a fatal complication in patients with the diagnosis of sever sepsis. This study focused to the effects of removing toxic substances from inflamnatory tissue by hemodiafiltration. ~4ethods: Eleven patients were admitted to the ICU in emergency center and met the criteria of systemic inflammatory response syndrome in association with infection. All patients developed liver and renal dysfunction and were treated by hemodiafiltration with high flux membranes (FB-U:Nipro). The hemodiafiltration were performed 19 times using nafamostat mesilate as an anticoagulant in 5 hours with 12 L of substitution fluid (HF-B:Fuso). The serdm levels of endotoxin, cytokines, endothelin-i (ET-]), human neutrophil elastase ~1-proteinase inhibitor complex (HNE-PI), fibronectin (FN), lactate, and amino acids were measured before and after the hemodiafiltration. The hemodiafiltration would be effective to renal dysfunction by reducing endothelin and beneficial to tissue metabolism represented in Fisher's ratio, but might be harmful to respiratory function by activating neutropila in patients of severe sepsSs. Background : Intermittent HD may be poorly tolerated in the early phase of ARF in hemodynamically unstable patients (pts). This technic may fail to achieve steady state urea low levels in hypercatabolic pts. Method : Nt = 25 consecutive pts treated with HD; N 2 = 25 consecutive pts treated with CVVHF. Hemodynamic unstability is defined by arterial hypotension and requirement of inotropie support despite adequate filling. Rate of change in urea (U), ereatinin (Cr), K + , pH were computed from a linear regression .analysis of data vs time in each treatment group during the 5 first days of application of the two technics (ANOVA). Dally worst values were recorded. Results : HD-group : Apach% score = 22 _+ 7; mean number of organ system failure (OSF) = 1.5 -+ 1; mean blood pressure (MBP) = 75 • 22 mmHg (first day of application of HD). CVVHF-group : Apachen score : 25 + 6; OSF = 3 -+ 1; MBP = 57 + 20 mmHg (first day of application of CWHF Discussion : During the 5 first days of application of HD/CVVHF, U and Cr decreased much more rapidly in the CWHF-group. K* and pH were maintained within normal range in the two groups. Initial MBP which was much lower in the CWHF-group significantly improved during the application of CVVHF while MBP remained unchanged in the HD-group. Conclusion : Despite higher severity of disease in CVVHF group (Apachen score, OSF, lower initial MBP), we obtained a better performanco with CVVHF regarding the decrease of U and Cr and the improvement of MBP. In relation to the different and continuous renal replacement techniques, the continuous venovenous one is the alternative method to continuous arteriovenous for critical patients with acute renal failure (ARF). We present you our experience with CVVH in patients with MOF. In our Intensive Care Unit (ICU) 20 patients with MOF were treated with CVVH in the period betWeen January in 1992 to March in 1995. The mean (• age of our patient population was 52,1• years, being 65% male and 35% female The whole patient population was with MOF iust at the moment the technique was accomplished; 75% was in mechanical ventilation, 90% needed vasopressor support and 65% required both of them (mechanical ventilation and vasopressor support) Apache II score mean of the patient population was 17,84~:6,97 (range 5-28) and ati of them were with ARF oligoanudc. Technique: CVVH was accomplished using a single-d~al Iumen catheter, ptaced in either a temoral or subclavian vein by the stand ard Seld{nger technique. Pol{sultone hemofitiers were also used, and the extracerporeal circuit used standard arterial-venous blcod tubing. Blood flow and hence oltrafiltration pressure, within the circuit was generated by a roller blood pump. The modulus has a roller pump, a pressure transducer connected in an arterious and venous line, such as an air-transducer which is adapted to a drip-chamber in the return way. The replacement used was a peritoneal dialysis solution. Medicine 1, St. George's Hospital Medical School, London. England. Hepatic sinusoidal endothelium shows a major inflammatory response in porcine sepsis that can be attenuated by the administration of dopexamine hydrochloride. Dopexamine is a beta 2 and dopaminergic receptor agonist. The specific beta 2 adrenoceptor antagonist ICI 118551 has been shown to reduce the protective effects of dopexamine. We investigated the effect of this antagonist on hepatic ultrastructure in porcine sepsis. Six pigs (25-30Kg) divided into 2 groups were anaesthetised and intubated. Cardiac output and portal blood flow were measured using standard techniques. The 2 groups were; placebo, (peritonitis induced); blocker, (peritonitis induced and 200 pg/kg ICI 118551 bolus infused then given hourly). Caecal content was aspirated and peritonitis induced. Colloid was infused to maintain PAWP at 10-12mm Hg for eight hours the animals culled, hepatic tissue removed and prepared for electron microscopy. In the placebo group hepatic endothelium was swollen and the sinusoids occluded by WBC. But in the ICI 118551 blocker group, much of the sinusoidal endothelium was absent and there where large extra sinusoidal spaces among the hepatocytes. An assessment of the two groups showed worse hepatic architecture in the blocker group. The b2 antagonist blocked any protective effect of endogenous beta 2 adrenoceptor agonist (adrenaline) on hepatic endothelium in porcine sepsis. George's Hospital Medical School, London. England. Dopexamine hydr0chloride, a beta 2 and dopaminergic receptor agonist reduces hepatic damage in porcine sepsis. We tested dopexamine's effect on cerebral oedema. The beta 2 adrenoceptor antagonist ICI 118551 was infused to block any protective effect of dopexamine. Nine anaesthetised pigs (25-30Kg) were randomised into 3 groups; placebo, (peritonitis induced); dopexamine, (peritonitis induced and 5 ~tg/kgdar of dopexamine infused); blocker, (as in dopexamine group but in addition 200 pg/kg ICI 118527 bolus given then infused at that rate hourly). Caecal peritoneum was induced and colloid infused to maintain PAWP at 10-12mmHg for eight hours when the animals were culled, cerebral tissue removed, prepared for electron microscopy and digitisation. Digitisation of the area of oedema surrounding the blood vessel and expressed as a percentage of the micrograph. 30.5_+4.1, Dopexamine 13.5+2.9", Blocker 31.5+3.7. Data expressed as mean + SD. Significance p<0.05. * dopexamine compared to placebo and blocker. In the dopexamine group the area of tissue oedema was significantly lower than either the placebo or blocker groups. There were no significant differences between the placebo or blocker groups. The 62 antagonist completely blocked the protective effect of the drug on cerebral oedema in porcine sepsis. Beta 2 adrenoceptor stimulation is protective of cerebral oedema in porcine sepsis. Objectives: The hemodynamie~ of hepatic circulation during multiple organ failure (MOF) have not been suffleienly studied. We investigated liver hemodynamics in two subgroups of patients with MOF, those with either liver or lungs as the main organ of involvement. Methods: Three groups of patients were created: i) MOF-hepatic involvement (MOF-HI) (7 patients) with bilirubin >3.5 mg/dL and lung injury score <1.8, it) MOF-ARDS (9 patients) with respective values <2.0 and >2, iii) 5 patients with head injury with respective values <2 and <1, served as group control. All patients were in haemodynamieally stable state with an oxygen delivery index >300 ml/min/m2 prior to measurements. Two Swan-Ganz catheters 'were inserted, one in the hepatic veins and one in pulmonary artery and the following measurements were determined: the hepatic vein free pressure (HVFP), the hepatic vein wedge pressure (HVWP), CVP, PAOP and CO. The gradient of HVWP-HVFP represents liver perfusion pressures. By injecting contrast media at dose of ImL/lOkg with the balloon inflated to achieve sinusoidai image, the hepatic blood flow (HBF) was concluded by the time in seconds of media removal after balloon deflation. Results: The CO, CWP and CVP were comparable to all three groups. Namely, for MOF-HI, MOF-ARDS and control groups the mean (+SD) value of CO was 7.2_+0.8 vs 6.9_+0.3 (NS) and 6.3_+0.6 respectively, of the PAOP was 8.7+_1.6 vs 10+:3 (NS) and 8.2+3.1 respectively and of the CVP was 12.+.2.3 vs 11.6+2.3 (NS) and 5.8 respectively. In contrast the two MOF groups were different after the cut-offinclusion criteria ie the mean (+SD) value for bilirubin was 6.8+9.5 vs 1.20+0.7 (13<0.05) and 0.8_+0.2 respectively and lung injury score was 1. Objectives: Oxygen delivery (DO2) and oxygen consumption (VO2) are increasingly monitored parameters in the ICU. There still remain controversies about an oxygen supply dependency in critical illness particularly with respect to VO 2 determination by either indirect calorimetry (VO2m) or tick calculation (VO2c). The purpose of this study was to investigate the changes in VO2m and VO2c following DO 2 increase. Methods: The relatives of 24 critically ill patients (mean age 63 years, mean APACHE II 24, mean MOF-score 9) gave their written informed consent to participate in this institutionally approved, prospective study. DO 2 was increased by fluid loading (hydroxyethylstarch 10%: mean volmne 750 ml, mean duration of infusion 80 min) and catecholamine support (dobutamine: mean dose 14,3 ~g/kg/min). Changes in VO2m and V02c were recorded sinmltaneously before, during and following interventions. Calorimetry was obtained with the Metabolic Monitor 7250 integrated in the Ventilator 7200 (Puritan Bennett, Carlsbad, CA Adaptive endocrine response of organism to septic shock consisting in activation of the production of adrenal hormons, renin -angiotensin -aldosterone system (RAAS) and other hormonal systems has an influence over microvascular changes in these states and for development of multiple organ failure (MOF). In 25 patients with peritonitis of different origins (18 nonsurvivors and 7 survivors) were followed the changes in cortisol level and RAAS by radioimmunological methods and many variables for evaluation of respiratory, renal, hepatic function, coagulation etc. as a signs of MOF. It was observed significant increase of the level of cortisol (1099 +_ 4,83 nmol/ I), aldosterone (0,895 • 0,687 nmol/I). By factorial statistical analysis we found significantly high correlations between hormonal changes and respiratory function (for example r=-0,539, p < 0,02 between cortisol and PaO2; r = 0,817, p < 0,001 between cortisol and D (a-v) 02; olso renin -CaO 2 r=-0,824, p < 0,001, renin D ~,vl O2 r = 0,626, p < 0,001). Such significant correlations was found and for RAAS with respiratory, renal function, byproducts of arachidonic acid thromboxan B 2 and P6Fla, soluble fibrine degradation products etc. These correlations between the degree of endocrine changes and multiple organ failure in patients with septic shock produced by peritonitis suggest that their effects upon peripheral vascular resistance and constriction of the splanchnic, splenic, renal and other organ vasculatures are not always with physiologic expediency and there are perhaps the possibilities of therapeutic influence. Intredu~on : Dopexamlne has previously been shown to control hyperkalaemia ia patients with acdto renal failure (ARF), however effects on the subsequent course of ART are undomunente~ Ob_iectlv~ : To evaluate clinical progress in patients with acute renal failure (ARF) in an Intensive Care Unit (ICU) with regard to biochemical control, need for -and time to -dialysis, and outcome in patients receiving dopexamine. M~ods : 14 consecutive patients meeting standard criteria for diagnosis of ARF were included in the study. Full cardiovas~dar, biechemical and intervention/outcome details were recorded. Dopex.~min~ was infilsed at a dose of 2 pg/kg/min in conjunction with a regimen of inotropir support and blood volume optimization. Resn]~ : Following the intzoduetion of dopc',~mine Ilrinr vohlmes increased slightly over the next 24 hrs fzom 516 + 140 ml/24 hrs to 817 + 229 ml/24 hrs (NS). Data expres,uxl as Mean + SEM. Three patients (21%) became polyuric with urine output >100 ml/hr within 3 days and did not need dialysis. In the remaining patients the time to dialysis (to correct acid-base deficits or volume overload) was 5.09 + 0.84 days. Serum potassium levels were well controlled. Day 5 or immediate pre-dialysis levels were 4.48 + 0.19 mmol/L compared with pre-lreatment 4.67 + 0.2 mmol/L Overall mortality in this series was 4/14 (28%). Duration of acute dialysis in survivors with renal recovery was 16.8 +_ 1.82 days. 2 patients (14%) progressed into chronic renal failure and needed continuing renal replacement therapy. No adverse cardiovascular altects were seen at this low dopoxami~ dose although its competitive inhibition to adrenergic reuptake mechanisms meant that doses of pressor agents could often be reduced. : Dopcx:~minr nsed in conjunction with inotropic support and blood volume oplimiTntion, can safely postpone, or even avoid, the necessity for acute haemodialysis in ICU patients. No evidence of tachyphylaxis to the effect on serum potassium levels was seen over the duration of the study. Hen'era M., Suarez G., Dagn D., Varela A., Ramos J., Garoia JM, Aragdm C, Jurado L, Medina A. ICU. Hospital Regional. Malaga. Spain. Objective: To evaluate the haemodinamic tolerance to the veno-venous continuous hemefiltration (VVCHF) system in patients with Systemic Inflammatory Response Sindrome (SIRS), and the possible beneficial effect of this technique on the haemodinamics in these patients. Material: 13 patient admitted to the ICU, with diagnosis of SIRS and monitored with a pulmonary artery catheter at the beginning of WCHF. We performed a complete haemodinamic study to all these patients (Cardiac output, vascular resistanoss, pH and CO2 in arterial and mixed venous blood samples, saturation of pulmonary mixed venous blood, DO2 and VO2 calculations and temperature) and determined the respiratory mechanics (compliance and PaO2 /Fie2 relatinship) before starting the procedure, after 10 minutes operating with the ultraflltrate branch closed (without filtered fluid production), afler 30 and 60 minutes of zero fluid balance bemofiltration and after 120 minutes of filtration with negative balanos adjusted to the patients conditions. For the statistical analisis we have performed the ANOVA test over the mentioned variables. Results: We have not detected statisticaly significant differences of the analyzed variables before the beginning after operating the pun'@ for 10 minutes without filtered fluid production and after 60 minutes of zero fluid balance HF. Only temperature shows a meaningful decrease in time. Objectives: Among many organs, playing the important role in pathogenesis of multiple organ failure, the particular place is taken by the intestine. ~ethods: The study was carried out in 5 dogs !~n"~h PI was modelled by severe operative trauma (OT). The DCM was estimated by the indices values of work time (WT), contraction frequency (CF), mean amplitude of contractions (~AC) and motility index (MI) measured by method of tensography. "SL", created on the basis of sorbit and sodium lactate (1800 mOsm/L), was injected in the dose of 10.O ml/ kg into v. cephalica antebrachii after 24 hrs of OT. The results of the present study are the evidence of "SL" stimulative action on DCM and are experimental ground for "SL" using in complex therapy of PI in clinic. WITH SPLANCHNIC VENOUS BLOOD PC02 P.F. Laterre P. Goffette, J.P. Fauville, A. Poncelet, P. Loneux, M.S. Reynaert. Intensive Care Unit, St. Luc Univ. Hospital, Brussels, Belgium. Determination of gastric intramucosal pH (pHi) by gastric tonometry using the Henderson-Hasselback equation is expected to allow the detection of splanchnic ischemia in critically ill patients. Because of bicarbonate concentration and acidbase balance influences on the calculation of pHi, it has been proposed to use arterio-gastric PCO,_ gradient [P(Gast-a)CO,] to assess splanchnic perfusion. HTpothesis : PCOz in the gastric mucosa is in equilibrium with intraluminal CO z and with CO, in the blood leaving the stomach (mesenteric and portal blood). Objective: Mesure PCO; and pH in portal vein blood and compare its value with PCO 2 and pHi obtained simultaneously by gastric tonometry. Material and method : In a patient (55 y.), a fiberoptic catheter (Baxter R) was positionned in the portal vein after transhepatic stent shunt repermeabilisation. Hemodynamic parameters, DO, (Vigilance n Baxter), gastric CO 2 and phi (tonometrics Baxter) and portal blood gas were determined at regular intervals. Results : 19 sets of data were obtained and are expressed in mean + SD. Gastric PCO z was 46,5+18 compared to 40,4+3.5 mmHg for portal PCO 2. pHi was 7.32+._0,17 VS 7.34+._O,04 for portal pH. No correlation was found for these 2 parameters. P (Gast-a) C02 was 6.4+15 mm Hg VS 2+1.6 mm Hg for P (portal-a) COz (no correlation). There was a good correlation between DO e and P (portal-a) CO z (r = 0,61) [Figure] but no correlation with P (Gast-a) C02. Obiectives: Desaturation is a common finding during haemodialysis (HD). Pulmonary oedema might be one cause for impaired gas exchange (1). The aim of this study was to quantitate the amount of extravascular lung water (EVLW) and gasexchange in chronic renal failure patients during and after a regular hemodialysis session. Methods: 10 chronic renal failure patients without symptoms or diagnosis of cardiac or respiratory disease were studied at the start (I), at the end (II) and two hours after (III) a regular bicarbonate hemodialysis session. The double-indicator 2 dilution method, with indocyanine green and the stable isotope H20 as tracers, was used to measure EVLW (2). Arterial bloodgases and endtidal CO2 were registered. EVLW data was compared to a group of 18 renal healthy patients (0). DCP n EVLW, mL -PaO2, mmHg H~O +, nmol/L Control group 0 18 243 -2--61 l 10 332 _+ 91"* 13 -+ 3 38 _+4 CRFgroup II 10 269 -+ 84~ 11 +-2ns 34-+2"(" III 10 283 +-78t 11 _+3ns 34-+2t ** p < 0.01 dcp I from dcp 0, t p < 0.01 dcp lI or I1 from dcp i, :~ p < 0.001 dcp II from dcp I The EVLW at the start of dialysis was larger in the CRF group than in the control group. The EVLW decreased significantly to a level not different from the control group in response to the reduction in weight after HD. PaO~ was normal at the start of HD and showed a nun-signficant reduction after HD. PaCO2 (5.3+0.6 kPa) and EtCO2 (5.2+0.8 kPa) were unchanged while H3 O+ decreased and bicarbonate increased significantly. Conclusions: The elevated level of EVLW at the start of HD did not impair gasexchange. The decrease in EVLW did not inhibit the decrease in PaO2. The reduction in H30 + followed by a fall in alveolar vantilation is the most plausible cause for the decrease in PaO2 in bicarbonate dialysis. 1. Prezant Lung 1990; 168: 1-14.2. Wallin J Appl Physio11994; 76: 1868-75. A. Dona~ D. Battis& L Col~ R Danieli, D. Achill~ L Viglienz;~ C. Giov-anaini, P. Piaropao~ Oblectives: To verify if intraoperative modifications of mtramucosal gastric ph (phi) below the normal lowest value 7.32, can be predictive for important complications, as perforation, sepsis, MOF or death. Methocls: We have considered 20 patients who andenvent major abdominal surgery. All patients received the same drugs in pre-anaesthasia, the same type of anaesthesia (balanced anaesthesia) and the same treatment with H2-bloekers. After the induction of anaesthesia a gastric tonometer was positioned and a catheter was positioned in the radial artery. During the operation, every 30 minutes, the following parameters were measured at the same time: phi, arterial ph (pha), blood lactate, mean arterial pressure. In follow up we considered death and complications happened during the hospital stay, in relation to intraoperative phi falls below 7.32. Results: Among the 20 patients, 9 had a drop of phi below 7.32 during surgery. In three of them this fall was a single episode and happened within the first hour after the begiluting of the operation. After that phi rose to nomml values until the end of the operation These patients had a normal post-operative period, without complications, The other 6 patients had a fall of phi during the demolitive manoeuvres. Two paticots of them died. The first had a lowest phi=7.25 and the second 6.97. The first one ~zs operated on for hepatic istiecitoma, suffered a complete del'dseenco of the surgical wound on the 20th day after operation and died on the 25th day, the second one was operated on for a hepatic carcinoma had an intraoperative haemorrhage and died ~vo hours after the end of the operation. The other 4 patients with a fall of phi had a lowest phi=7.24. 7.18. 7.26. 7.28 respectively.The first patient,operated onfor sigmoid carcinoma, underwent on a second operation for a transmural necrosis of the colic segment on the 25th day; the second one, operated for carcinoma of the right colon, had a cardiac ischelnia on the 5th pest-operative day and a dehiscence of the surgical wound on the 8th day: the third one, operated on for a sigmoid carcinoma, had melena in 41h post~ operative da b, and finally the fonrth patient, operated on for carcinoma of the tight colon, suffered a fistula of the surgical enteral anastomosis.All these 4 patients were discharged alive from the hospital. The other 11 patients, who had not reductions of phi ditring the operation, had a normal pest-operative period, without complications. Conclusion: Phi was able to predict the arising of some complications, probably due to intraoperative ischemic events. We can say that gastric tenometry, for its low invasivi.ty, can be included among the intraoperative monitoring in patients that tmdenvent on major abdominal surgery. (ttD),t"ea~rrerJ.~ of 3 hours duraticn. All l:atients nm.'-~mS_(~Lly va~2ated in eantrol wcde ard_ la':'ad a a,~m--ganz catheter, with optic fibers for contirums mmsuremmt of SvO mic studies were performed, c~e before the hegir~ of HD, c~e 15 rain after the ~, ~ne at the middle, ~ne 15 rain before lhe erd ard one 30 rain after the erd of HD. Paired t test ~as used far slatistical eval~ti~n. Results: Daring I~D there was a significant'reductton (p AS 10.6%> NI 9.1% > ED 6.8%; p = 0.01. In-hospital mortality: 365/1853 patients (19.7%) --OTH 30.9% > NI 20.4% > AS 19.6% > ED 16.2 %; p = 0,05. Mean survival time in days after discharge: AS 620 < NI 682 < OTH 708 < ED 777; p = 0.66. Conclusions: Despite an excess in-unit mortality of secondary referrals from other hospitals the Iongtime course of this special patient group is not different to others. Solsuam, J, Marrugat*, G, MirS, J, Nolla, A, Vazqu~z-Sanchez, L Alvamz, ~ioio S Xndioina I~siw. Ir~itate l(~Icipal da l~sti~iSn l~di~*, ~ospits dal Objective: To study the influence of modifiable variables (complications derived from therapeutic activities) on the prognosis of ~atients admitted to the ICU indapeMently on thn severity of illnsss. Patients aM Methods: Between January 1990 asd ]lay 1993 data from 1,425 patients over 14 years of aqe who retained in the ICU for mare than 24 hours ~ere pr~pectively regiatered. A cohort st~ly with follo~-~ nf patients durin~ ~eir stey in the hospital was deni~.eL In all patients, reasons for a~issien, principal diagnosis sad severity of illn~s moasared by the SAPS scare vare recorded. Fastens affecting patients' outcome that my be proventsd or modified included technical :omplisafioss, heapital-acqnired infections and in~pro~riate therapeutic decisions. A logistic regression model was used to assess the relative risk (l~} for in-heapital mortality adjusted for each variable. Results: IC~ mortality ~s 17.2% and in-hospitul mortality 22.7%. Patients who died showed a higher SPAS score then survivors (15,3 ~ I0,i). After adjusting hy severity of illness, co~;licetices that statistically increased the risk of in-hospital death were septic shock secoMery to hoapitul-acqdired infection (1~ 7.18; 95% el, 1.9 to 27.1), pmo~othor~x related to mocAsnical ventilation (@ 6.28; 95% Cl, 1.7 to 22.3) and delay in the insertion of a fln~-quidod catheter (II~ 5.49; 95% IC, i.i to 26.9). Col~lusien: Registration of complicaticas derived from therapeutic activities is a valuable tool far quality central in the ICU. G, ~i~5, J.L Mle~ma, J, ~amqat*, J..~lla, A, vazquez-SalteMz, F, Alvamz , Servioia de Nndicina l~siu. I~stitutu ~icipal de ln~sti~acidn ~4i:a*, Hospital dsl Objective: To dstsr~ine the incidence of self-extebatien and its effect on ~ortality. Patients and ]~etheds: Betveen January 1990 and April 1994, all I~tiente in whom selfextubatien w~s registered were inclnded in a prospective study. Patients were divided into @nee who needed r~intabatinn within 24 hoers and those who did not. In all patients, dsmoqraphie and ciinical data were recorded as well as ICII mortality, in-hoapital mrtality and severity of illness according to SAPS score. 0eta were analyzed usi~ the cbJ-square test for cathgorical Verinbls, the analysis of varianc~ (ANVA) for aontinuc~ ~ria~les and a leqi3tic regression anal~is to estimate the relative risk (IIII) for mortality as result of celt-nxtt~ation after adjusting for severity of illness. Results: A total of 815 intnMtsd patients amre stndied. Self-extu~atien occurred in 54 (6.6%) patients and 25.6% required reintuhot~pn. When a co,arise was made between patients who did not required reint@atinn and patien~.s who did, statistically significant differences in eqe (52.1 v_s 60.4 years, P = 0.~02), ~verity of illness (11.4 ~ 13.1 SPAS score, P = 0.02), dia~isstia category (4S.6% v_s 66.7% of patients with res~iratury conditiono, P =0,002} and mean length of stay (10,9 ~ 20,7 days~ P = 0.05) were fo~M, A~ter ad~sti~ for severity, patients with self-ext@atinn who did not reqnired reintalatien showed a 0.4 IIR for mortality (95% CI, 0.i to 0.9) as co~arod with patients in when self-ext@ation did mot occur. Conclnsien: Self-~extaMtice that does not require reint@ation is associated with a isamr in-hospital natality probably dt~ to a prolonged period of weaming. Patients' admissions to ICes am often delayed doe to the shortage of beds available. @ile amaltieq ICU admission, these patients are treated in observation nits of @e emergency services which bare ,either tile structure nor the trained ~reomenl that are available in leb~. Objective: To daterdno the effect on the patient's proqusis of a delay in tile admission to the ICU when criteria for ICIJ admission are fulfilled. ~terials and Methods: Between Jme aM l?ece~ber 1993 all patients who fulfilled criteria to be alMittod to the IC0 who for wAste~r reason retained in tile observation unit for more than 24 hours were included in a prospective stedy. In all patients, des~raphic end clinical dabs amre recorded as well as severity of illness aencrdi~j to SAPS score. A cesucontrol dasi~ was eend with a total ss~ln of 1,425 patients who suffered no delay is admission to ICII over a period of 3 years. Data wen analyzed using the chl.-squ~re test (to aeons the association hetwenn in-patienty mortality end categorical vari~lns) and a maltipln logistic reqression model to sstimta odds ratio fOR) for in-hospital mortality as result of delay in ICY admission as compared with early ad~issi| after adjusting for severity of illness end use of assisted mchenical ventilation. ~9&ults: A total of 50 patients remained in the observation nit for more than 24 hours with a del W in Igd admission of 55.8 _+ 25.1 hoers. Assisted mechanical ventilation was requited in 22% of patients and only monitericatien in 46%. Itsse patients were cspared with 112 Ntients from the tet~l sample ratchod by age, SP~ score and rennoss of admission. In-hospital mortality for cases warn 16% as compared with 17.5% for controls (P = S). After adjamtilg fen SPAS, age and mobamioal ventihtien, no statistically significant differences between both ~renpa were foaM, altho~b there was a tendency towards a higher mortality amen@ patients with delay in ICU admission (OR = 0.779; 95% CI, 0,2 to 2,5). Conclnnien: ~se findings suggest that prognosis of critically-ill patients is no worse as a result of admission to the loll being deln~d for 24 borers. All data appropriate for the calculation of the APACHE Ii score (APS) together wi'th other specific cardiac details relevant to these .patients were collected daily, verified and enter~ into a computer database. Results: 150 patients were studied. Six patients died and five of thee underwent cardiac surgery. The mean APS was 9 for survivors and t6 for non-survivors (P < 0.001). The mortality ratio was 1.1 and the major markers of mortality were APACHE ![ score, presence of chronic ill health, mean duration of ventiiation, mean length of ICU stay and need for emergen~ surgery. Sixteen percent (233) of ICU bed days were occupied by 4% of patients (non-sarvivors) which resulted in cancellation of 60 cardiac sot#cat sessions in 6 momhs. Conclusions: This study concludes that APACHE 1t could be used as an audit tool in a cardiac surgical ICU and demonstrates the severe compromis~don of cardiac surgical throughput by a few non-survivors, ORGAN To determine the number of organ failure free days (OFFD) in a cohort of survivors and non-survivors with sepsis syndrome followed over a 30 day period. 2) To determine sample size requirements for clinical trials utilizing a increase in the number of organ failure free days as the primary outcome as opposed to mortality. Methods: Beginning December 1990 through to April 1992, patients who met inclusion criteria of the "Cardiopulmonary effects of Ibuprofen in Sepsis Syndrome" and who did not have HIV/AIDS. brain death or moribund state were prospectively identified. Presence or absence of failure of 7 organ systems (pulmonary, CVS, renal, hepatic, GI, hematologic, & CNS) was recorded daily until death or until 30 days. A score of one was assigned to each organ system free of organ failure in patients still alive, ie, maximum daily OFF score=7, maximum 30 day OFF scorn=210, Sample size estimations were performed for variable detectable differences in OFF scores (Delta). Alpha was set at 0.05 (Two-sided), with n/group = 2[(Z a +Z b ) o Conclusions: A clinically relevant increase in OFF days may be detected with as small a sample size as 30 to 50 patients per group. This represents a significantly smaller sample size than needed to detect a change in mortality from 40% to 30% (25% relative risk reduction) where the n/group=356. Scoring patients in this manner prevents a lethal inte~entien from providing an improved organ failure score. In addition, an intervention that prolongs survival must also provide greater organ failure free days in order to be counted by this scoring method. Survival as an outcome provides no information about the quality of that survival. OFF days provides a measurement of burden of illness. Interventions which lessens this burden may be just as valuable as those that decrease mortality by providing a measure of the quality of survival and by decreasing costs of care. They may also prove to be an accurate surrogate marker of mortality. The advantage of this approach is that the event rote is much higher and sample size requirements are subsequently smaller. This would mean that clinical trials can be completed faster and at lower cost. Outcomes such as mortality could then be assessed at a later date utilizing recta-analysis. We suggest that the use of OFF days is a valid outcome measure that may be utilized in clihieal trials of sepsis syndrome. The ICU is perceived by many as being a stressful environment for both patients and staff. Stress has been defined in three ways: a stimulus producing a particular response; the physiological and psychological response to a stimulus; an interaction butwom an individual and their environment. Stress is currently thought to be a dynamic system of stimulus and. response which takes into account the individual's perception of the stimulus and their ability to respond effectively. Stress may, therefore, be positive and allow personal development but an individual unable to respond effectively to a stimulus will experience negative effects or strain. Critical illness is an intense stimulus to which the body needs to respond effectively. Physiological responses are vital and most of intensive care involves supporting these. Alternatively, blocking them, for instance with atom(date, increases mortality. Psyehological responses are also vital but often poorly appreciated because of communication problems. Many of the problems patients experience in an ICU are evidence of psychological strain. This can be exhibited in various ways, for instance, anxiety, depression, passivity and confusion. Dealing with critically ill patients is perceived as stressful. We recently studied occupational stress in our ICU. Most aspects of intensive care were not generally perceived as stressful indicating a self-selectien of ICU staff. The most stressful aspects of ICU work for nursing staff were the structure of the organization and career opportunities. Medical and nursing staff had different stressors and different coping strategies. Support for occupational stress, therefore, should focus on the individual and concentrate on information and communication. atmosphere, and especially at intensive care units, we face up to daily decision making. In most cases these are taken on the basis of personal opinion and the processing of a very limited amount of information. Rising need to optimize the results of medical attendance becomes necessary to set structured system of d@cision making in which ethical basis have a sp@dial significance in view of next considerations: -We live into a pluralist society in which the importance of values is different. -Most persons consider health as the first value only in the event of illness. -Medical resources available are limited, whereas medical, attendance demand from population increases in a way many people consider it unlimited. In consequence, it becomes necessary to set up priorities in patients treatment. Ehtical basis that rule decision making are essentially these ones: I. Beneficence: To provide the patient that is being treated the highest profit. 2. Non maleficence: It is our first duty to avoid hurting or damaging the patient."Primum non nocere" 3. Autonomy: In every particular medical attendance, the patient has ability to decide by himself. 4. Justice: As equity: To provide the same treatment for those who have the same pathology, ignoring another factors such as age, sex or race. Severe application of these principles can cause difficulty, which resolution requires a systematization of decision making. (1-84) . The lenght of stay between survivors and non survivors didn "t show statistical significance (p =0.51 ). The mean AIII score when considering all admissions was 59,9 (8-153) . The initial score between survivors and non survivors showed ststistical difference (48.6 vs 92.5) respectively (p < 0.0001). Univariate logistic regresion analysis demostrated a 90% increment in death probability for every 10 points augmentation in the AIII score with a sensitlbity of 94.9% and specificity of 62.7%, The ROC curve showed that the best cut off point for death prediction was 75 points with a sensitivity of 75.6% and specificity of 79.9%. If a patient is classified as high risk (> 75) the Bayesian analysis showed a 52.8 probability of death and for one class(fed as low risk (<75) a death probability < 10%. Conclusions: The first day AIII score in this population showed to be a good discriminator between survivors and non survivors, and the risk of death augments as the AIII does. In this population an AIII score > 75 points is asociated with a greater risk of death. Using the AIII score in conjuntion with the clinical judgement will help clinicians reducing uncertainty in the every day decision making and better predict outcome, The results from this study should been taken with caution because the data were obtained from a small sample. Objective: the quality of life has been considered a "uniquely personal perception" resulting from a mixture of health related factors and social circumstances [T. M. Gill, JAMA 1994, 272: 619] . The aim of this study was to evaluate two measures of PQOL in Intensive Care Unit (ICU) admitted patients. Patients and methods: during ICU stay and six-months after hospital discharge, 160 co-operative ICU admitted patients were directly interviewed about their PQOL. We administered ftrstly the Uniscale (PQOLU) [Sage et al Crit. Care Med. 1986, 14: 777-782] and then a 5 step verbal scale (PQOLV): best, good, fair, poor, worst. Of the 160 studied patients, at the first interview, 116 were able to use both scales, but 44 (27.5%) understood only the verbal one. At the second interview, 8 patients were not able to answer, 113 used both scales and 39 only PQOLV. Statistical analysis was performed using Wilcoxon signed ranks, Spearman rank correlation, Student's t and chi square tests. Results: Of all cardiac surgery pts, 42 pts (1.6%) died in ICU. They were 33 males (1.5%) and 9 females (1.6%). Their mean age was 66 (+7) years and mean EF was 0.38 (+0.1). Nineteen pts (45%) had low (<0.35) preoperative EF. Mortality was 0.9% in the Coronary Artery Bypass Grafting (CABG) group (N=2014) and 2.8% in the Valve Replacement (VR) group (N=359). In the CABG +VR group, mortality was 8.4% (N=95), and 3.3% in the remaining pts (N=149). Cardiogenic shock was the sole cause of death in 24 pts (57%), septic shock in 6 pts, whereas sepsis in combination with ARDS in 4 pts, sepsis and stroke in two pts. In addition, 6 pts died from cerebrovascular accidents, one from ARDS and one from pulmonary embolism. The pts who died in the ICU had a significantly longer bypass and aortic cross clamp time and received more blood transfusions (p<0.001) than a matched control group that survived to ICU discharge. The duration of mechanical ventilation and length of ICU stay were greater in the pts who died in the ICU than in the control group. Conclusions: 1. Although cardiogenic shock is the main cause of death (57%)in cardiac surgery pts, sepsis and cerebrovascular accident are relatively frequent causes. 2. Patients who died in the ICU had longer bypass and aortic cross clamp time and received more transfusions, compared with the control group. 3. Although renal or hepatic failure contributed to death in some pts, they were not the primary cause of death in any patient. Objectives: Evaluate the acute and follow-up outcome of 27 patients (pts) treated with primary PTCA (without prior thrombolysis) in acute myocardial infarction (AMI) after 12 and up to 24 hours after onset of typical thoracic pain ("late" primary-PTCA). Methods and patients characteristics: From 12/89 to 4/95 364 consecutive pts with AMI were treated by primary PTCA in the Wuppertal Heart Center 9 27 pts (7,4%) were admitted to our hospital > 12 hours and < 24 hours after symptom onset with ongoing chest pain and typical ECG-changes.Mean age was 62 years (49-78). 23 pts were male, four female. 37% had an anterior wall myocardial infarction, 63% suffered an inferior/postero-lateral wall myocardial infarction.Two pts were in cardiogenic shock at admission. Singlevessel-disease was documented in 70.4%, multi-vessel-disease in 29.6%. Average time of onset of pain to recanalisation was 929 min (720-1440). Angiography revealed TIMI-flow 0 in 85.2% of the pts, TIMI-flow I in 11.1%, TIMI-flow II in 3.7%. Average follow-up (FU) period was 12 months (4-28 months). TIMI III LV-EF ~ 30-day major late re-late flow p.i.* aeute/FU mortality bleeds infarction mortality 92.6% 58%/63% 7.4% 7.4% 3.7% 0% Early mortality occured in the two pts, who were in cardiogenic shock at admission 9 No pt required emergency coronary artery bypass grafting.Restenosis >50% was seen in 37% of the pts. Conclusions: "Late" primary PTCA achieves a favourable high recanalisation rate of about 90% (TIMI Ill-flow) in our study group. Additionally, there seems to be a trend for LV-EF improvement in follow-up. Early high mortality is influenced by the patients admitted in cardiogenic shock. There might be a trend for increased major bleeding complications. Objective: To assess the validity of SAPS II (New Simplified Acute Physiology Score), comparing it with the previous version, (SAPS), in a sample of patients recruited by GiViTI, a network of 128 ICU's representative of the Italian ICU system 9 Methods: Measures of calibration (goodness-of-fit statistics) and discrimination (receiver operating characteristics curve and area under the curve) were adopted in the whole sample and across subgroups differing in relevant prognostic characteristics. Of the 3004 patients recruited during one month period, a total of 1813 patients were included in this study. For the purpose of the comparison of the two scores, patients with less than 18 years, or having cardiac surgery or staying in the ICU less than 24 hours were excluded. Vital Status at ICU discharge in the whole sample and at hospital discharge in half cases wher adopted as outcome measure. Re$01~: SAPS II fits the data equally well compared to the older version (goodness-of-fit P=0.29 and 0 9 in the new and old versions, respectively) but its performance is somewhat better in terms of capability to distinguish patients who live from patients who die (areas under the curve 0.81 and 0.73, respectively). Furthermore, SAPS II is better in terms of uniformity of fit across relevant subgroups, although substantial over prediction of mortality was observed in trauma patients and in patients admitted without organ failure to be intensively monitored. SAPS II performed very wet] also in the subsample where hospital mortality was the dependent variable.Satisfactory measures of calibration (goodness-of-fit P--0.47) and discrimination (receiver operating characteristics area=0.80) were observed. C0nr SAPS II, a multipurpose scoring system developed in an international study, retains its validity in this independent sample of patients recruited in a large network of Italian ICUs. Although it has shown a good performance when adopted to predict ICU and hospital mortality in the entire sample, further investigations are warranted. The observed over prediction of mortality in a few subgroups indeed call for a through assessment of the impact of confounders and biases on model performance when SAPS II is adopted in samples that do not reflect the "average" ICU patient. Objectives: 1) Assess the effectiveness in a group of Intensive Care Units by means of a quality performance index (QPI); 2) Assess the efficiency by means of a resource use index (RUI); 3) Evaluate the performance of individual ICUs with respect to both indices (clinical and economical) while controlling for severity of illness. 1270 critical from 17 UCIs in Catalonia patients alearic Islands have been included in the study. Inhospital mortality and weighted hospital lenght-of-stay (LOS) have been considered the outcome variables. Severity of illness has been measured with the MPM II at admission. In each ICU, expected mortality has been obtained adding the probabilities of dying for its patients. Expected LOS has been estimated adjusting a second order polynomial to the severity of illness. PerfOrmance indices have been obtained by dividing the observed by the expected outcomes. Re~ult~: The overall QPI was 1.15 and it ranged from 0.58 to 2.05 in the 17 ICUs. The overall RUI was 1 and it ranged l~ont 0.61 to 1.51. There was not a trade-offpattern between clinical performance and resource use. Objectives: Teaching hospitals often provide [CU care across a variety of specialized services. Overall, this approach appears to result in the best risk adjusted survival rates, but at the highest cost (Critical Care Medicine 1993;21:1432-42): Recently, there has been increasing focus on markers of overall hospital performance. However, in large teaching institutions, such markers may fail to detect intra-institntional variation at a large tertiary care medical center. Methods: First intensive care unit (ICU) day, Acute Physiology and Chronic Health Evaluation III (APACHE III) and active Therapeutic Intervention Scoring System (TISS) data were collected on 1621 random admissions to 8 specialty ICUs with 90 beds (range 8-14) between February I and December 3 l, 1994. Post-operative solid organ transplant recipients were excluded. Units included 2 general medical, 2 general surgical, and trauma, neurosurgery, cardio-thoracic surgery, and coronary care units. Data were analyzed for risk adjusted outcomes: ICU and hospital mortality and length ef stay (LOS); risk of requiring active 1CU treatment; and ICU readmissinn using APACHE III risk prediction models. Results: The study ICUs cared for a diverse group of patients. Mean APACHE III scores ranged from 36.9-55.5; predicted risk of hospital death ranged from 8.5-21.1%. Standardized mortality ratios ranged from 0.40 to 1.24 with 4 ICUs performing significantly better and 1 performing worse than predicted (p<0,05). LOS ratios and ICU readmission rates ranged from 0.95 to 1.09 (NS) and 2.1 to 13.2% respectively. Patients predicted at low risk of requiring active ICU treatment ranged from 6,6 to 45.8% Conclusions: There was wide variation in the mean level of patient severity between ICUs. After controlling for this severity, outcomes also varied widely. No clear pattern of overall institutional performance was evident. These data suggest that efforts to assess performance, improve quality, and maximize efficiency must be focused within individual units. Programmatic evaluation of outcome allows for focused review of the processes of care contributing to good outcome (best practices) and where to focus ongoing quality improvement and cost reduction activities. Background and method : We compared ICU mortality in different age groups presenting with the same severity of disease. We assessed severity of illness by the physiological day 1 -Apache~ (physio-Aa) score (thus excluding the age related points). For each of the following physio-A n score intervals (0-5; 6-10; 11-15; 16-20; > 20) , we compared tCU mortality within 6 age intervals (< 40; 41-50; 51-60; 61-70; 71-80; > 80 years -10, 11-15, 16-20) . In these groups mortality may be twice higher in the > 60 years patients than in the _< 60 years. Mortality does not vary with age in low (physio A n = 0-5) and high (physio A n = > 20) risk groups. In the low risk group, mortality is low in all the age intervals because of the begninity of illness. In the high risk group, extreme severity of disease probably blunts the impact of age and leads to high mortality rates in all age intervals. Introduction: To access the actual social/clinical outcome of the patients who undenvent intensive care therapy OCT) is rather difficult, quality of lilr is not easih.' defined and ohserver subjectivity is a prime factor in the evaluation. Mortality ratio after discharge must be established and its causes understood. Obieetives: The propose of this stud)-is to look into the mortality ratio that occurred on a series of patients that undorwent ICT at our Unit from of the ~iew point of severity of the original illness and the diagnostic groups. Material and Methods: During the period of one )-ear (1994), 216 patients were treated at the Unit, 45 of them died, and 16 ~ere not matched in our series because os incumpletc records. Thirteen patients died in hospital after their reference to other departments, Twelve patients were lost after discharge. Thus. at the end. only 142 patients were evaluated on the FU. The, were classified into the follov4ng three groups: acute medical, elective surge D 9 and acute and emergency postoperative. The patients were seen at 3, 6 and 12 months after discharge. The, were evaluated in accordance to their abili~, to being self supported in their daily life and capecity to fully return and hold to their pre~4ous jobs. Apache 11 scores were evaluated for each of the three groups and correlated to the ICU dead, Hospital dead, and mortality after hospital discharge, SPSS package was used for statistical analysis. Remlts/Conclasions: Data shows that 19/142 patients died after discharge from the hospital, of ~itch nine died in the first three months. Seventy-eight per cent of the patients were fully self supported in their daily life and 20% showed some kind of handicap. Fosty-nine per cent of the patients wore on retirement either due to age or some form of chronic disease, when admilled to our Unit. Thirty-two peg cent had not been able to return to work, because the" were incapacitated on discharge. Only 7% had return to their fully jobs but the period of the stu~, is not enough for all of them to be fully physically recovered. Preliminmy statistical analysis shows us significant differences among groups. The aim of the present study is to compare the prognostic performance of five general severity indices ou coronary patienta and to find out if a proper ntatistical hundling of these indices could provide better results in these patients. Methods: SAPS II, MPM II (MPM II0 i MPMP II24), APACH II end GAPRIK were evaluated o~ 456 patients with acute myocardial infurction admitted to 17 intensive care units from Catulunye. Calibration and Discrimination were calculated for each index. Calibration was calculated by th Bosmer-Lemeshow test. Discrimination was evaluated by the area under the relative operating characteristic (ROC)curve. If a model did not show a good performance it was customized using multiple logistic regression. Finally, tworeduced models were developed, one fro~ the MPM series (MPM II24cor) and one from the group APACHE-SAPS (SAPSIIcor).Their performances were again evaluated. Results: Discrimination was high enough for all models. Neverthelees, oelibration of APACHE II, SAPS II and MPM was not satisfactory. Thus,MPM II24, SAPS II and GAPRIK were customized for coronary patients using the logits of both models, and obtaining good calibrations. MPM II24, and APACHE-SAPS were adapted and reduced to 5 (MPM II24cor) end to 4 variables (SAPSIIcor), respectively . Both models showed better oalibrutions end discriminations than the original models. Conolusion| Models developed for multidisciplinary patients show a good discrimination when applied on aoronar I patients, but some needed customization in order to improve calibration. The number of variables of the principal model can be reduced (even to 5 or 4 variables) without loosing prognostic accuracy. Objective: to compare the ability of two methods to predict outcome for intensive care patients. Methods: We included 343 consecutive intensive therapy unit (ITU) admissions with an ITU stay>24 hrs in a 18 month prospective study (exclusion criteria: burn injury and age <16 yrs). Data were coUectsd applying the criteria described by the developers [1, 2] . The definition of coma (MPM24II) was modified and the best assessment within 24 In's, rather than the admission score, was used. Statistical analysis included classification tables and receiver operaUng characteristics (ROC) curves to assess discriminative power, and Lemeshaw-Hosmer statistics and calibration curves to test accuracy of prediction. Results~ Average abe was 58 yrs (ranse:16-92) with a male:female ratio of 1.6:1. The actual hospital mortality was 26.8%, mean predicted death rates were 22.8% (MPMz4II) and 15.2% (AP HI). Non-survivors had siguitlcanfly higher predicted risks than survivors applying both methods (p<0.000l, t-test). The total correct classification rates (TCCR) for APACHE III were bett~r for all decision criteria applied (TCCR, decision criterion 50%: APACHE ]/I 77.1%, MPM24II 71.4%). The area under the ROC curve was 0.75 (AP III) and 0.66 (MPM24II) confirming the better discrimination of APACHE Ill. Accuracy of risk prediction was similar for both models (AP nl ~2-59, MPM24B ;(2-52, Lemeslmw-Hosmer). Showing some fluctuation, calibration curves lay close to the ideal line for predicted risks -<50% with increasing deviation for higher risk groups (s. figure) . APACHE III underestimated the risks of hospital death for almost all risk groups (curve above diagonal), whereas considerable overestimation for predicted risks >40%0 ceenred with MPM~II. Objective: to assess the goodness-of-fit of the APACHE III model for British ITU patients. Methods: We prospectively studied a cohort of 715 adult patients consecutively admitted to a medical-surgical ITU over a period of 18 months. Patients with burn injury, age < 16 yrs and ITU stay < 4 hrs were excluded. Using a eomputerlsed database, we routinely recorded 24 hrs APACHE Ill scores. Predicted risks of hospital death were computed by Critical Audit Ltd, London. Accuracy of risk prediefion was assessed by Hosmer-Lemeshaw chi square (;(2) statistics and calibration curves [1 ]. Discrimination was tested employing classification tables and receiver operating characteristics curves (ROC). Restths: The mean age of the 453 male and 262 female patients was 59 yrs (range: 16-92 yrs). Of these patients, 64% were medical admissions, 17% were admired after emergency and 21% after elective surgery. The observed hospital mortality was 25.4%, the overall mean predicted death rate was 16.8%. Mean predicted risks were siguifieanfiy greater for nonsurvivors (38.0 %o) than for survivors (9.6%, p<0.00l, t-test). APACHE III showed good calibration (Z2-~59, Lemeshaw-Hosmer). However, the calibration curve lay above the diagonal for almost all risk groups reflecting the tendency to underestimate actual mortality (s. figure) . The best total correct classification rate (TCCR) was 89.3% (decision criterion: 50%). The area under the ROC curve was 87.6% confirming the good discriminative ability of the model. Objectives: The aim of this study is to point out the discrepancies between needs and actual treatment of less severely ili patients admitted in Italian Intensive Cam Units (ICUs) requiring only intensive monitoring, and verify the substantial likelihood of data comparing those collected from a national short term study with a regional long ternl use. ~: Less severely ill patients ("observed patients") were only monitored; they did not require intubation, even if for a short period (less than 24 houm) or major cardioeiranlatory supports, and were neurologically normal. Epidemiologieal national data were obtained from GiViTI group (Gruppo italiano Valutazione interventi in Terapia intensiva); this cohort study, collected 5092 patients, in two months in summer in 1992 all over Italy. Regional data were echieved in a three years entlection (1990-I992) in Lombardia' ICUs from ARCHIDIA group (Arehivio Diagnostieo), including 10065 patients. Mortality, severity score, diagnostic category and some typical intensive procedures were analysed and compared in both studies. Patients' disgunstie categories were defined as surgical, medical and trauma, according to the main diagnosis and the presence/absence of surgical procedures. Rr162 Observed patients account for 23.2% and 22% of all ICU's patients respectively in national and regional data. Very tow mortality rate was found in national data (2.3%) and extremely low mortality in regional data (0.6%). In both studies mortality, S.A.P.S. and length of stay were much lowor in "observed patients" than in general ICU's population (mortality: 25.7% and 22.3%; 8.A.P.S. score: 10.6 and 13; Iength of stay: %7 and 9). Homologous distribution of patients in the two studies was noted for what concern their diagnostic category, aside from a slight prevalence of tranmatised patients in the GiViTI study. In the two groups the surgical patients were respectively 47% vs. 57%, medical patients were 34% vs. 31% and traumatised were 19% vs. 13%. 92% of "observed patients" in national study and 93% in the regional did not received any intensive procedure. Only a minority of these patients availed haemodynamie eonu'ol with Swan-Ganz or renal haemofiltration. Conclusions: These results underline that about one fourth patients admitted in Italian ICUs benefit an oversized slructure i, relation to the real needs of their pathology. In hot more than 90% did non received any advanced treatment and mortality and S.A.P.S. score were substantially lower respect to general population. The results obtained from these two studies are similar, suggesting an uniform distribution of the case mix in Italy, even if a different recruitment period and a different gengraphieal distribution were used. Some discrepancies in the two studies were found in the diagnostic categories moreover regarding the tranmatised patients (19% vs. 13%); this can be explained from the seasonal (summer) characteristic of the national study. Mutuality, yet very low, is different in the two groups, but these data do not allow any definite explanation. Finally these epidemiologieal survey suggest need of further studies settling more strict criteria of admission in ICU. This study aims to evaluate patients OUtcome, quality of care and effectivity of therapy in our Intensive care unit. The main goal was to indentify factors that the most influence that outcome. During 1994. the authors collected data of patients outcome and predictor variables. Overall mortality rate was 39,3%. The most common causes of death were infection. The diagnosis of Sistemic Inflammatory Response Syndrome (SIRS) and Multiple Organ Dysfunction Syndrome (MUDS) significantly correlate with death (90%). Average length of stay was 6.6 days ~. 55% patients died in the first ten hosiptal days and only 18% after 30 days. Age was directly correlated with death 50% of dead were older then sixty years. An analysis of physiological variables showed that serum levels of gl~cose (55%) and natrium (71%) were in optimal physiological values. Serum proteins (72%) and Haemoglobin (50%) levels were inversely related to death. Multivariate showed that alveolo-arterio difference in 02 content was the most informative of all mortality predictors (Mean value 22,4 mmHg in 90% patients iO>mrnHg). Factor that most influence the patients OUtcome was infection (sepsis) and MUDS. Use of predictive indicators of outcome in critically ill patients may help to assess treatment regimens and to compare patient groups. Acute physiology and Chronic Health Evaluation (APACHE If) score (Crit. Care Had. 1985; 13: 818-29) and the sepsis score of Elebute and Stoner (Br. H Surg. 1983; 70: 29-31) have been used, Objectives: To compare sepsis score and APACHE II score in predicting outcome of critically ill patients. Methods: Overall survival during the past 8 years for patients in our ICU was calculated = 62% (prior probability). The outcome of 230 patients who were admitted to our ICU for > 72 hours was observed. APACHE II score on admission, patient predicted risk of death (APACHE II risk) and the sepsis score on the first day of antibiotic course were prospectively recorded. Discriminant function analysis of the scores in relation to outcome was performed. Results: APACHE II and sepsis scores in the survivors were significantly lower than in those who died (21.6 i 7.2 v~s 25.6 • 6.5 and 10.9 • 5 v's 15.2 • 5.9 respectively P < 0.001). Correct prediction of outcome by each score is shown in Discussion and Conclusions: Although both scores have been previously evaluated in predicting outcome of ICU patients, studies of the sepsis score were conducted in small numbers of patients or involved additional measurements not routinely available. This study demonstrates that the sepsis score alone or in combination with APACHE II score is more effective than APACHE II score in predicting outcome. Objective To test the hypothesis that resuscitation titrated against gastric intramucosal pH (pHi) improves survival in critically ill patients as suggested by Gutierrez et al~. Method Emergency admissions to the Intensive Care Unit were randomized into control and intervention groups. In the control group pHi was measured at 0, 12 and 24 h while in the intervention group pHi measurements were made 4 hourly for 24 h. Both groups were managed according to the same guidelines to achieve the following targets: mean arterial pressure >70 mmHg, systolic arterial pressure >90 mmHg, urine output >0.5/ml/kg, haemoglobin >8 g/dl, blood glucose < 12 mmol/1, arterial oxygen saturation >94% and correction of uncompensated respiratory acidosis. If the pHi was < 7.35 after achieving these targets, or after maximal therapy to achieve the targets, patients in the intervention group were given fluid to ensure an adequate cardiac preload and then dobutamine at 5 then 10 mcg/kg/h, titrated against pHi. This additional therapy was continued until 24 h after entry into the study. In each year patients were subdivided in two series with random selection, so that the 1st series contained abeat 2/3 and the 2nd 1/3 of the patients. The 1st series of all the years constituted the devdoping data set and the 2nd series the validation data set. With data of the 1st series (642 patients), we created the predictive model, using stepwise logistic regression (BMDP, USA). Each patient has been evaluated in die 1st, 5th, 10th and 15th day, calculating for each lime the Apache II score (for a total of 1444 records), independent variables were, besides time and Apache II of the time ( Michaloudia G,, Melissaki A., Alexias G., Gogafi C., Kolotoura A., Krimpeni G., Pamouktaoglou F, Filias N. Objectives: To determine the medical staff's attitude towards various ethical issues Methods : Between January 1994 and February 1995,185 anonymous questionnaires were sent to 20 Intensive Care Units, all over Greece. Results : 107 questionnaires (57,7%) were replied and returned back. Of them 58,9% were answered by male and 41,1% by female. The doctors replied in the following rate :81,2% aged up to 34,80% aged between 3544 and 94,4% aged over 45. 36 questions were answered and were divided into 3 main topics, as following: 1. Admission Criteria: Limited bed availability was the main cause for refusing admission in 54,5% of ICU's. 54,5% evaluated each case's viability and only 10,3% used some prognostic score system. 21,5% of ICU's accepted all cases and a significant percentage (64%) gave in to pressure coming from their colleagues (72,7% female and 58,7% male). 2. Informing the Patient/Relatives: Only 6,5% was willing to tell the whole truth, while 39,2% had given selective information.. In the case of iatrogenic incident, 58,9% withheld it, because either they feared legal implications (34,6%), or lost of trust (46,7%). Doctors are asking consent from the patient and/or his family, in order to include him/her in research protocols, in a rate of 82,3%, while only 55,1% found informed consent necessary for the proposed treatment procedure. 3. Withdrawal of Therapy/DNR orders/Organ Donation: 80,4% were willing to withdraw complex treatment in patients with short life expectancy, except of administi'ating intravenous fluids, feeding and analgesics. In 34,6% such a decis~n was unanimous, while the percentage of those carrying it out was 69,1% (72,2% female, 63,9% male). In case of brain stem death 87,8% (67,3% female, 85,7% male) withdrew any life support. 67,3% would like therapy withdrawal to be legally established, while only 12,1% would perform euthanasia, if there was substantial legal cover. For these cases, relatives' consent was considered to be necessary from a percentage of only 11,2%. 83,2% considered organ donation to be a necessary proposal, while 10,3% refused to ask the patients' relatives for an organ donation, either because they didn't have the psychological strength for it (3,7%), or because they doubted the procedures' objectivity (4,7%). Note: In Greece, ICU beds are less than 1% from the total number of hospital beds available. Only a percentage of 35-50% of these admissions comes from the same hospital, with a potentially direct evaluation. Usually an ICU doctor has to be informed through the telephone. Finally, employment conditions in Greece are such that any changes of the medical and nursing staffare limited. Conclusions: The mathematical model we found has been validated also in the second series and the discrimination capability increases with time. Using this model we can evaluate the probability of survive at every, time. Its application at different times permits a better evaluation of haemodinamically instable patient trend. Introduction: The feasibility to assess pulmonary capillary pressure (pcap) offers the opportunity to determine the longitudinal distribution of pulmonary vascular resistance (PVR). The purpose of this study was to measure Pcap and to calculate PVR to determine whether relevant shifts in the distribution of PVR could be expected after routine cardiac surgery. Methods: The study population consisted of 25 consecutively admitted patients after cardiac surgery. Surgical procedures included coronary artery bypass graft (CABG) (n=14) and mitral valve replacement (MVR) (n=t 1). pcap was estimated by analysis of the pressure decay tracing after pulmonary artery occlusion. After estimation of pcap precapillary (Ra) and postcapillary resistance (Rv) was calculated. A complete set of hemodynamic variables was obtained at 1 hour and at 6 hours after operation. Results: There were no significant hemodynamic changes during the first 6 hours after surgery. The MVR group maintained pulmonary hypertension and higher levels of Pcap. Ra/Rv, reflecting the longitudinal distribution of resistances, remained unchanged. However, Rv predominated Ra during the postoperative period in both groups. Objectives: Evaluation of the influence of long-term continuous I.V. administration of the ACE-inhibitor enalaprilat on regulators of circulatory homeostasis. Methods: t9 trauma and 26 sepsis patients randomly received either 0.25 mg/h (group I, n=15) or 0.5 mg/h (group 2, n=15) of enalaprilat I.V. or saline solution (control, n=15) as placebo for 5 days. Plasma levels of endothelin-1 (ET), atrial natriuretic peptide (ANP), renin, vasopressin, angiotensin-II, and catecholaMines were measured before injection of enalaprilat (='baseline' values) and during the next 5 days. Results: Except for ET, plasma levels of all vasoactive substances exceeded normal range at baseline. Angiotensin-II significantly decreased during enalaprilat infusion (0.25mg/h: from 53.1• to 22.1• pg/ml; 0.50mg/h: 62.1• to 17.9• whereas it remained significantly elevated in the untreated control patients. Vasopressin increased only in the control group (p<0.01) and decreased after 0.50mg/h of enalaprilat. ET remained almostunchanged in group 2, whereas ET increased significantly in the control patients (from 4.9• to 10.t• on the 5th day). Catecholamine plasma levels (epinephrine, norepinephrine) markedly increased in the control group (p<0.01), but they did not change significantly throughout the study period in both enalaprilat groups. Conclusions: Continuous I.V. administration of the angiotensin-converting enzyme inhibitor enalaprilat beneficially influenced systemic and local vasoactive regulators of the circulation, which are normally increased in the critically ill. Thus patients at risk of (micro-) circulatory abnormalities may profit from enalaprilat infusion. Objectives: To determine the time taken for hemodynamic and gas exchange variables to a reach stady-state after a change from supine to Trendelenburg position (TrP). Methods: We prospectively studied 8 adult patients with severe sepsis or septic shock requiring hemodynamic monitoring. Usual cardiorespiratory parameters were measured at baseline, 15 min after the patient was placed in a TrP and again 15 min after the return to a supine position. A fiberoptic pulmonary artery catheter (SvO~ Oximetrix, Abbott) allowing continuous SvO 2 monitoring wa~used. During the protocol we also continuously measured SaO~ by pulse oximetry and VCO~ and VO 2 by monitoring partial concentration of O2and CO 2 ir~ inspiratory and expiratory gases (DELTATRAC metabolic monitor, DATEX). Therefore, we were able to monitor cardiac output variations by dividing VO~ with arteriovenous difference according to the Fick equation (CO-Fick). Results: No significant difference in hemodynamic status was observed 15 min after the patients were placed in TrP. Despite the fact that no significant change was observed in CO and VO~ estimated by thermodilution, CO-Fick had a tendency to dedrease continuously in TrP and then to return to its initial value when patients regained supine position. Respiratory gas analysis showed a small but persistent continuous increase in VCO 2 without a similar trend in VO 2 values. Conclusions: We conclude that no significant hemodynamic effect was detected in our patients after 15 min in TrP. Evaluation of VO 2 from respiratory gases analysis after a change in body's position should be interpreted with caution, since the patient may not yet have reached a stady-state after 15 rain. Since VO 2 did not change, VCO~ increase was probably due to position related changes in-pulmonary gas exchange and not to a change in patient's metabolic status. Objectives: To determine whether changes in SvO 2 and/or other hemodynamic parameters during weaning trials could be used to predict successful weaning. Methods: We prospectively studied 10 adult patients with a history or clinical evidence of cardiovascular dysfunction, who were unable to tolerate spontaneous breathing (SB) for 3 hours. For all these patients right heart catheterisation was considered necessary in order to detect hemodynamic alterations during weaning. A fiberoptic pulmonary artery catheter (SvO 20ximetrix, Abbott) allowing continuous SvO 2 monitoring was 5sod. Hemodynamic status was evaluated ~t baseline and after one hour of spontaneous breathing through a T-piece. Patients were assigned to one of two groups depending on whether they tolerated SB for 3 hours. Data were analysed by analysis of variance and unpaired Student's t-test We also used multiple linear regression analysis to determine which hemodynamic variables were correlated with the magnitude of SvO~ change and multiple discriminant analysis to determine if aSy of the above variables were associated with toleration of SB for 3 hours and/or successful weaning (S-W). (J Physiol 1995; 78." 696-701) . We tested the hypothesis that the ventilatory stimulation by dead space (VD) loading and 3% CO2 inhalation is accompanied by a proportionate cardiovascular change. Methods: Six healthy subjects, mean age, 25 year, performed three incremental exercise tests in a randomized order: 1) inspiring air without VD (air control, AC); 2) inspiring air with VD of 920 ml (AVD); 3) inspiring 3% CO2; 21% oxygen, balance nitrogen. The ventilatory responses were examined at matched heart rate (HR) equivalent to 90% peak HR. Results: Ventilation (VI) was significantly greater (p<0.0001) during the AVD and CO2 tests than during the AC test at the same work rates. End-tidal CO2 (PETCO2) and estimated arterial CO2 (PaCO2) were significantly greater (p<0.01) at 150 W and 200 W. Oxygen saturation was significantly lower (p<0.05) during the AVD test than during the AC and 3% CO2 exerdse. At matched HRequivalent to 90% peak HR, VI was significantly greater (p<0.01) during the AVD and 3% CO2 tests than during the AC exerdse (123 l, 121 l, and 91/). Conclusion: We conclude that the increase in XrI and PETCO2 due to VD loading and 3% CO2 inhalation is not associated with an acceleration in HR. Sup.ported by MRC (Canada). Objeetlve: The production of large amounts of oxygen radicals from the onset of ~en may be responsible, st least in part, for peroxidative damage to myocardial tissue. The aim of this study was to evaluate the time dependence of plasma TBARS in patients with AM] receiving Thrombolytie Therapy (TT). Patients and M~hods: Filiy eight patients admitted in ICU (46 men and 12 women; mean age 50.6 4-16.02 years) rec~ving systemic TT for possible AM] were ~died. All patients received recorabinant haman Tissue-type plasminogen Activator (r-tPA). The mean time fi'om the onset of symptoms and the be~nning of TT was 3.01 4-2.13 hours. Peripheral veao~s blood samples were obtained fi'om each patient before and serially after TT (0, 3, 6 and 9 hours). TBARS levels woe determined by using a spectrophotometrie technique. Rq~r fusion was identified by the timing of ereatine phosphate kkmse (CPK) peak (<15 hours). Table I list the variation of plasma eoneenlrations of TBARS (mean 4-sd) in groups (A,B, and C) as a function of time from the beginning of Tr. Co,arisen oftbe 0 time cuncentzatiens reveal a difference p50 ml/min). Serum samples were obtained A) before operation, B) after removal of the aortic crossclamp, C) at admission to the ICU, D) 4 hours after operation, E) 22 hours after operation. Results: TAS was significantly decreased after removal of the aortic crosselamp ( B, C and D lower than A), followed by a subsequent significant increase of LIP ( C and D higher than B). The levels of TAS and LIP returned to baseline 22 hours after operation. Methods: 10 patients with preoperative LVEF<40% undergoing coronary artery bypass grafting were studied. After surgery, a 3F femoral artery catheter was inserted and connoted to a fiberoptic monitoring system (COLD Z-02t; Pulsion Medizintechnik, Germany); this allows, with a double-indicator dilution technique, the calculation of cardiac index (CI,L/min/m2), intrathoracic bood volume (ITBV,ml/m2), pulmonary blood volume (PBV,ml/m2) and extravascular lung water (EVLW,ml/Kg). With a 7F pulmonary artery catheter, wedge (W,nunHg) and central venous pressure (CVP,mmHg) were measured, while 02 extraction ratio (O2ExR,%) and oxygen delivery (DO2,mL/min/M2) was calculed. Peak inspiratory pressure (Pawp,cmH20) and mean airway pressure (Mawp,cmH20) were measured with a Varflex flow transducer (Bicore,Sensormedics,US). The patients were studied after 60 minutes (TO) of volume controlled standard ratio ventilation (VC), and after 60 minutes (TI) of stabilisation period of PCIRV (67% inspiratory time, 0 % pause). Vt,Ve and total PEEP were held constant in every mode of ventilation. 1+_0.4" *'p < 0,05 versus TO Conclusions: These data show that PCIRV 2:1 is a safe ventilatory support also in cardiac patients with impaired ventricular function, and monitoring of ITBV is more reliable to measure and optimise circulatory volume status, than W and CVP. C.Ledeki-,G.Rldisis,S.KarotzaI,C.Micheilidis,M.Agioutantb, G.Beltapaulos. Objeolivee:To evaluate the influence of LVSWl on the well known correlation of Sr and SvO2. Paw eight patients (12 melee end 16 females) were included in this study regerdlen of the ICU ~h"niseion couse. All paints were ,'~theta~ with e fiboroptir pulmonary artery catheter connected with an Oxymetfir (R)~ SO2/CO Abbot computer.For any pulmonary artery catheter insertion, two pain= of Sr and SvO2 were obtained, one dudng inserlion and one during taking the catheter out. For any pair obtained, we eleo collected the deta concemig with the pedient's hemodynamir and oxygenation end we calculated the LVSWI. were significantly (p 40% ; n=10 and < 40%; n=4) did not alter these results. Back~ound: In man, vascular endothelium-bound ACE is expressed in concentrations greater than 50x that in serum and is believed to be the site of synthesis of circulating angioteusin IL It is unclear whether ACE inlubitors interact similarly with ACE in different vascular beds. Coronary vessels possess all the components of the renin-angiotensin system, including ACE which may be involved in normalcardiac homeostasis, as well as in the pathogenesis of various cardiomyopathies. Obiecfive: To develop a method for assaying the interaction of ACE inkibitors with coronary endothelium-bunnd ACE in man, Methods: ACE a~Aty was meas~ed in five patients undergoing CABG surgery, from the transeuronary hydrolysis of the synthetic ACE substrate 3H-BPAP. Trace mnou~ of ~FI-BPAP (4gCi) were injec~d as a bolus in the root of the aorta and simultaneously blood was withdrawn from a coronary sinus catheter into a syringe containing protease inhibitors which prevented the convession of umeaet~ aI-I-BPAP by blood ACE. The sample was later centrifuged to separate cells from plasma and the radioactivities due to formed product (~rl-BPhe) and total sH were astimated in a [B-counter. Two additional such determinations of ACE activity were perform~ the second in the presence of 1.5pg/kg E (coinjected with ~-I-BPAP) and the third ten minutes after E. Results: All subjects were hemodynamically stable throughout the course of the there were no noticeable hemodynamic effects of E. Control transcorunary metabolism of~-BPAP averaged g0-a:3%, in agreement with previously reported data. In the presence Of E, % metabolism of ~-BPAP was reduced to 21• reflecting a 85• inhibition of normal ACE activity. Ten minutes after E, ~rI-BFAP metabolism had partially recovered to 52:L:10%, representing a 50-a:15% inhibition of control ACE activity. From this data, the dissociation constant of E for coronary ACE in vivo was estimated as 6.8x10 "4 sec "l. Conclusions: We have demonstrated the feasibility of repeated, reproducible measures of coronary endothelium-bound ACE activity and of its inhibition by E. This procedure is safe and can be used to study the role of ACE in normal cardiac function and in card pathologies. Objectives. Primary pulmonary hypertension (PPH) is a progressive fatal disease of unlmown origin, with median life expectancy of less than three years after diagnosis. The responsiveness of pulmonary hypertension to a variety of vasodilator agents led to the speculation that, concomitant with vascular renmdelling processes, persistent vasoconstriction is an important feature of the disease. Long term use of Ca-channel blockers and intravenous PGIz may improve mortality in certain populations of PPH patients, but both of these treatments lack selectivity for tire lung vasculature. The aim of this study was to test the efficacy of aerosolised prostacyclin and its stable analogue, [loprost for selective pulmonary vasodilatation in PPH. Methods: In three patients with PPH, we compared aerosolisation of prostaglandin Iz (PGI2) and Iloprost to a battery of vasodilatory agents (Diltiazem, Nifedipin, inhaled nitric oxide, intravenous PGIz). Results: Nebulisation of PGI2 and Iloprost tumed out to be most favourable for achieving effective and selective pulmonary vasodilatation. Pulmonary vascular resistance decreased from 1664 + 75 to 1054 -+ 93dyn*s*cm 4 (p<0.001) and pulmonary artery pressure from 63.3 + 3.1 to 528 + 3.4 mmHg (p < 0.05), cardiac output increased from 2.66 + 0.11 to 3.57 _+ 0.16 I/rain (p < 0.001), mixed venous oxygen saturation from 49.6 _+ 2.2 to 63.3 + 2.8 % (p < 0.001) and arterial oxygen saturation from 87.9 + 2.6 to 93.6 _+ 2.2 % (mean _+ SEM of 7 trials in 3 patients). 5-month Iloprost nebulisation in one patient (100 gg/day in six aerosol doses) demonstrated sustained efficacy of the vasodilator r~men. Conclusion: Aerosolation of PGI2 or its stable analogue may offer as new strategy for selective pulmonary vasodilatation in PPH. Endothelial adhesion molecules may play an important role in the pathogenesis of myocardial cell damage, and may contribute to the progression of heart failure. We measured the plasma soluble intercellular adhesion molecule-1 (sICAM-1), vascular cell adhesion molecule-1 (sVCAM-1), and E-selecfin (sELAM-1) levels in 27 patients with acute myocardial infarction admitted within 6 hours after onset. Peripheral venous plasma-samples were collected at the time of admission, 12, 24, 36, 48, and 72 hours after onset. Plasma soluble adhesion molecule concentrations were determined by ELISA. Patients were divided into 3 groups as follows: group 1; Killip's class (K) 1 and 2 without thrombolytie therapy, group 2; K 1 and 2 with thrombolytic therapy and group 3; K 3 and 4. Both plasma sICAM-1 and sVCAM-1 concentrations in group 2 and 3 were elevated rapidly and significantly and maintained at a high level during the first 3 days. Plasma sELAM-1 level did not change in any of the groups. These results suggest that the adhesion molecules ICAM-1 and VCAM-1 may play a role in the pathogenesis of myocardial reperfusion injury and may indicate its severity in myocardial infarction. Objectives: Nitric oxide (NO) is known to exert cytotoxic and negative inotropic effects on cardiomyocytes. NO synthase activity has been reported to be increased in infarcted area in animal model of myocardial infarction. These findings suggest that NO may be an important regulator for myocardial damage and cardiac function after myocardial infarction. We measured plasma NO27NO3-(NOx) levels and estimated serial changes in acute phase of myocardial infarction. Methods: Subjects were 15 patients admitted within 3 hours after onset. Venous blood samples were collected at 3-hour intervals on the first day, 6-bour intervals on the 2nd day and 12-hour intervals on the 3rd day and 4 th days after onset. Plasma NOx concentrations were determined by Griess method. Results: The time course of the plasma NOx levels (mea~+SEM) displayed a tendency to gradually increase and to make a biphasic pattern with two peaks about 24 hours and 2-3 days after onset (basal level; 32.8_+4.9, first peak; 42.0!-_7.0, second peak;47.0+7.6 raM/l). Plasma NOx concentration was not influenced by the thrombolytic therapy, and NOx values at the time of 60 hours after onset were significantly correlated with maximal plasma creatine kinase level (r=0.83, p<0.01). The levels of plasma NOx in the early stage of myocardial infarction (from admission to the 4th day after onset) did not correlate significantly with the hemodynamic parameters (left ventricular ejection fraction, pulmonary capillary wedge pressure). Conclusion: The early and late increase in NO production after myocardial infarction may be implicated in the deterioration of myocardial contractility and induction of myocardial damage in the early phase of myocardial infarction. range 36-85) fullfilling the high risk criteria of Shoemaker (colectomy 13, gastrectomy 10, pancreaticoduodenectomy 4, others 6). Patients were admitted to the ICU preoperatively. Arterial and pulmonary artery catheters were inserted and hemodynamics and oxygen transport were measured at admission and after stabilization to predetermined physiological end points. Patients were considered stable when CI >2.5 l/min/m 2, PCWP >10 mmHg, Hb >100 g/l, Sat2 >.94. Objectives: Evaluate the acute effects of 0,08 mg ipratropium bromide and 0,2 mg fenoterol (IBF) inhaled dose on pulmonary function in 17 nonsmocers (Nb:m) and 14 smocers (S) with sever (New York Heart Association class II-III), stabile congestive heart failure(CHF) and 12 healthy subjects. Methods: Pulmonary function tests were performed < 3h postprandial. The tests consisted el arterial blood gas aspiration followed by routine spirometry and pletismography, and single-breath gas analysis. After performance of these maneuvers, the patients was administred 4 puffs-ipratropium bromide (0,08rag) and fenoterol (0,2 rag). For 0,5 h, spirometry was repeated. Results: In resting, pulmonary abnormalities observer in the S group were more severe then abnormalities observere in the NSm group. After treatment with IBF the improvement in pulmonary function was even more marked in patients who had smoked. The mean changes by forced expiratory volume in 1 second(EEVT) was 8,1% (p<0,00t) improvement and 3,9% (p<0,Ob), forced expiratory flow betwen 25% and 75% of the forced vital capacity (FEF25.75) was 30,8% (p<0,001) and 23,6% (p<0,001) and maxamal voluntary ventilation (MW) was 21,7% (p<0,05) and 16,2% (p.70; p<.01) as well as regional analysis of sequential 3-DE cut planes. Conclusion: In our group of patients with the diagnosis of ischemic dilated cardiomyopathy, this new 3-DE method could be applied. Our results show that this method allows a better assessment of the LV morphology and spatial geometry, with the calculation of global and regional indices with critical clinical and prognostic value in this particular cardiovascular pathology. Simultaneous left atrial (LA) and left ventricle (LV) inflow analysis assessed by pulsed Doppler TEE illustrate the loading conditions and reflect the hemodynamics of the left heart. We performed a prospective TEE pulsed Doppler study with recordings of the transmitral LV filling and pulmonary venous (PV) flow drainage in a group of patients with Dilated Cardiomyopathy (DCM). A group of 23 DCM patients, mean age 57_+11 yrs, 74% male were studied. This population was divided according to TEE severe LV dysfunction (Group SLVD+ 62% pts; Group SLVD-38% pts) In each pt we measured the peak velocities (Vel/m/sec) and time velocity integrals (VTI/m) of the transmitral early (E) and late (A) filing waves, the Vel and VTI of the PV systolic (S), diastolic (D) and atrial contraction (C) reversal flows. 3-DE TEE evaluation of the LVED, LVES, LVSt Volumes and LVEF were obtained. We calculated other parameters, such as E/A, S/D and A/C ratios and the sum of C+A Vel, that refelect LA systolic function and LV compliance. 19+5 305-_8 0.02 Simultaneous and quantitative analytical approach of the pulmonary venous and transmitral flows and ventricular volumes improve the non invasive assessment and understanding of left ventricular diastolic function and cardiac performance in dilated cardiomyopathy patients. Objectives : To assess the hemodynamic effects of fluid loading (FL) in acute circulatory failure (ACF) due to acute massive pulmonary embolism. Methods : Hemodynamic measurements (fast-response thermistor pulmonary artery catheter) were performed at baseline (Baseline) and after a rapid fluid loading with 250 (FL250) and 500 (FL 500) ml of Dextl'an 40 (Rhemacrodex| in 12 patients free of previous cardiopulmonary disease (66 • 3 yrs) with ACF (CI < 2.5 l/rain/m2) due to angiographicalty proven MPE (Miller score > 21) . Results : are expressed as mean _+ SEM and compared by ANOVA. A significant negative correlation (r = 0.83) was observed between baseline RVEDV[ and the effects of FL on CI. Such correlation was not observed between baseline RAP and the FL induced increased in CI. Conclusion : FUsibmificantly increases CI in ACF due to MPE. However, the simultaneous decrease of arterial 02 content due to hemodilution, limits the benefits expected from improved CI on peripheral oxygenation. Obiective: To examine the hemodynamic effects of external positive endexpiratory pressure (PEEP) on right ventricular (RV) function in acute respiratory failure (ARF) patients. Methods: Incremental levels of PEEP (0-5-10-15 cmH20) were applied and RV hemodynamics were studied by a Swan-Ganz catheter with a fast response thermistor for right ventrieular ejection fraction (RVEF) measurement in 20 mechanically ventilated ARF patients (LIS = 2.6 ~-0.45 SD). According to the response to PEEP 15, two groups of patients were defined: group A (9 pts.) with unchanged or increased RV end diastolic volume index (RVEDVI) and group B (H pts) with decreased RVEDVI. Results: In the whole sample cardiac index (CI) and stroke index (SJ) decreased at all levels of PEEP, while RVEDVI , RV end systolic volume index (RVESVI) and RVEF remained anchange d. At ZEEP the hemodynamic parameters of the two groups did not differ. In group A, CI decreased at PEEP5, RVEF decreased at PEEP10 (~0.8%)~ RVESVI increased only at PEEP15 (+21.5%) and RVEDV[ reded unchanged. In group B, CI and RVEDVI started to decrease at PEEP5, 'RVESVI decreased only at PEEP15 (-21.4%), anf RVEF was unchanged. Individual behaviors of the hemodynamic parameters at the 4 levels& PEEP were studied. RVEDVI and CI were significantly correlated in 10 out of:l 1 patients in group B, and in no patient of group A. On the contrary, mPAP and RVESVI were significantly correlated in 5 out of 9 patients in group A, and in no patient of group B. The slope of the relationship between RVEDVI and RV stroke work index (RVSWI) expresses RV myocardial performance. This relationship was significant (no change in RV contractitity)in 8 patients of group B and in 2 patients of group A. In some patients of group A, increments of PEEP shifted the RVSWI/RVEDVI ratio rightward inthe plot (RV function decrease). Conclusions: In ARF patients PEEP causes more often a preload decrease with unclmnged RV conctraetility. On the contrary, the finding of increased RV volumes during the application of PEEP is related to a decrease in RV myocardial performance. Thus, these data suggest that application of PEEP might be considered as a stress test to assess RV function. RIGHT INTRODUCTION: After heart transplant (HT), the right ventricle can be subject to an acute pressure overload, especially in cases where there is a preexisting severe pulmonary hypertension. This provokes right ventricular failure and, occasionally, circulatory collapse in Intensive Care Unit. Desire the advances that have been made in systems for preserving the donor heart and in post-surgical management, we have failed in our attempts to totally avoid this problem. The right ventricular function, although it usually remains within tolerable limits in these patients during the post surgery period, represents a factor which limits the results achievable in clinical transplant programmes. OBJECTIVES: To determine the Maximum Tolerance of the Right Ventricle (MxTRV) when faced with acute pressure overload. To study the function of both ventricles of the healthy heart (donor) when faced with different degrees of pulmonary hypertension. To detect possible interactions between the ventricles in the absence of the pericardium to approximate the experimental model to the clinical model of HT. MATERIALS AND METHODS: The pulmonary artery is progressively constrained in an experimental model until biventricniar failure is detected. This experiment is performed in two diffferent situations: with and without pericardial integrity. RESULTS: When pericardial integrity is maintained the MxTRV faced with a pressure overload is 73.2 + 8.56 nun Hg. When this pressure is exceeded there is a circulatory collapse with a sharp fall in the Cardiac Output and in the Aortic Pressure. However, when pericardectomy is performed (model similar to HT), only 52 • 6.71 nun Hg is tolerated (p < 0.01). CONCLUSIONS: With the pericardium open, as in Heart Transplant, the maximum pressure that the right ventricle can support is significantly less than with the pericardium closed. The pericardium has a positive effect in protecting the systolic ventricular interaction. It is, therefore, advisable to close the pericardium after Heart Transplant. JB Prrez-Bernal, A Ordrfiez, A. HeroAndez, JM Borrego, Map Camacho, C Cruz, MaC S~nchez, J Monterrubio, C Garcia, E. Gonz~lez. Hospital Uulversitario " Virgen del Rocio ". Sevilla. ESPAIqA. INTRODUCTION: Nowadays cardiomyoplasty isused incases of cardiac insufficiency as an alternative to cardiac transplant. After surgery the patients show a noteable improvement with the aid of this "biological circulatory assistance". Some researchers suspect that the improvement could also be due to the formation of new blood vessels from the muscle that wraps the heart, nourishing the ischemic myocardium. OBJECTIVES: Our Cardiovascular Research Group has proposed as an objective, the detection of any possible myocardial neovascularization through the muscle used for cardiomyoplasty. In the case that there are new blood vessels to the diseased myocardium through the wide dorsal muscle in which it is wrapped and which aids it mechanically, it would be possible to confirm the worldng hypothesis that cardiomyoplasty not only improves the cardiocirculatory funcfinn mechanically but also by facilitating a better blood flow to the ischemic myocardium. MATERIALS AND METHODS: the cardiomyoplasty technique is described using an experimental model of myocardial ischemia. The vascular cast is achieved by injecting methacrylate simulataneously into both the coronary tree and the wide dorsal muscle, in five experiments the connections between the coronary vascular system and the vascular structure of the wide dorsal muscle are demonstrated, CONCLUSIONS: We have demonstrated that Cardiomyoplasty, as well as improving ventricular function, favours the revascularization of the myocardium. Cardiomyoplasty could be indicated for cases of ischemic cardiopathy in patients in whom it is not possible to perform direct revacularization using conventional methods. A The therapeutic cardiological manouevres necessary in cases of ischeima reperfusion have increased considerably: fibrinolysis, transluminal angioplasty, coronary revascnlarization surgery and cardiac transplant. The appearance of a specific pathology ht acute reperfusion has been related to Free Oxygen Radicals (FOR) generated by oxidative damage. OBJECTIVES: To evaluate the appearance of FOR during a conti-olled process of ischemia-reperfusion in an experimental biological model and compare it with that in clinical cases. MATERIALS AND METHODS: Transitory cardiac ischemia was performed in five rabbits by reversible surgical ligation of the Descending Anterior Coronary Artery. After 15 minutes coronary reperfusion was performed. Blood samples were taken in the basal situation, at the end of ischemia and at 5, 10 and 15 minutes after the start of reperfusion. Malondialdehyde (MDA) was measured to evaluate the degree of lipid peroxidation (oxidative damage to the membrane). In ten patients undergoing conventional cardiac surgery the production of FOR was measured after aortic clamping. RESULTS: We observed that after 5 minutes of reperfusion there was a highly significant increase (p < 0.001) in the MDA values (mean =2.00 /zmols/L). These returned to basal levels after 10 and 15 minutes of reperfusion. CONCLUSIONS: An "explosion" of oxygen free radicals was detected very quicldy, just a few minutes after post-ischemia reperfusion. Thus, if antioxidant agents are to be used to reduce the toxic effects of the FOR, these will ordy have a therapeutic effect if they are administered in the early phases of reperfusion. INTRODUCTION: Aortic connterpulsation is a ventricular assistance widely used in Intensive Care Units in patients with cardiogenic shock as a provisional ventricular assistance. Paraaortic or external aortic counterpnlsation is been investigated as a definitive veutricular assistance in those cases of terminal congestive heart failure and when heart transplantation is counterindicated. AIMS: To assess the haemodynamic effects of an aortomyoplasty in a biological model of congestive heart failure. MATERIAL AND METHOD: As specimens, we used 10 "Large White" pigs. Mean weight was 22 Kg. After the administration of conventional anaesthesia, dissection of the ladssimns dorsi muscle was performed on the samples at the Laboratory of Experimental Surgery of our Hospital. Then we performed a thoracotomy at the level of the fourth intercostal space to reach the thoracic aorta. The aorta is dissecated 7 centimetres from the exit of the subclavia and it is wrapped by the dissecated muscle. A cardiomyostimulator is provided in order to allow the synchronization between the diastole and the muscle contraction. The model of heart failure was provoked using Verapamil plus Propanolol i.v.. RESULTS: A significant increase of the aortic diastolic pressures and a significant decrease of the left ventricle telediastolic pressures were observed. This improvement in the parameters (DPTI/TTI) implies an increase of the coronary perfusion in a model of heart failure. CONCLUSIONS: Using the external aortic counterpulsation, the aortomyoplasty improves the coronary perfnsion and the heart efficiency in patients with heart failure in whom no conventional therapeutic action is possible. The permanent character of the paraaortic counterpulsation is it main advantage. The appearance of specific pathologies as a resuk of myocardial reperfasion has been related to the oxidative damage secondary to the release of oxygen derived free radicals (OFR). During the myocardial ischemia induced during heart surgery with extraeorporeal circulation, severalsubproducts of the oxygen are produced that shall cause toxic effects after the reperfusion which could be counteracted by the physiological antioxidant systems and/or provided by the medication. AIMS: To asses the OFR during heart surgery. To check whether an antioxidant treatment administered in the preoperative period make decrease the levels of OFR before and after the myocardial reperfusion and to verify whether its administration have any beneficial effect on the intra and extraoperative management. MATERIAL AND METHOD: The study comprehends 20 patients studied as two groups of 10 individuals each (A and B). All patients underwent conventional heart surgery of valvniar substitmion or myocardial revaseularization. Group A patients were administered 400 rag/8 hours of vitamin E (Tocopherol acetate) 72 hours prior to the intervention as antioxidant treatment. Group B patient were not administered vitamin E. We assessed the quantity of malondialdehido (MDA) to assess the degree of lipidic peroxidation or oxidative damage of the membrane during the myocardial ischemia and 15 nm after the reperfusion. CONCLUSION: Patients who underwent heart surgery and were treated with tecopherol acetate in the preoperative period presented levels of RLO significantly lower than those who were not administered the drug, both during the intraoperative period and after myocardial reperfusion. We detected in these patients a need for antiarrhythmicals and pharmacoIogical support with catecholaminas, although not significant, both in the introaperative period and the immediate postoperative period. Recommendations for the treatment of pulmonary embolism (PE) in the presence of right atrial thrombus (AT) are conflicting. Because of a significantly higher mortality rate due to fulminam or recurrent PE, there is a necessity to treat patients (pts) with mobile type A thrombi compared to pts with adherent Type B thrombi. Therapeutic strategies include anticoagulation, thrombolysis (T) or surgical thrombembolectomy. Combination thrombolysis (COT), predominantly used for the treatment of acute myocardial infarction proved to prevent reocclusion of the infarct related artery at a comparable rate of hemorrhagia. Benefit has been related to the alteration of hemostatic proteins by non-fibrinspecific thrombolytic s. Administration of COT in PE has been performed sporadically. In the present case, a 55-year old male with no history of prior cardiovascular disease developed acute dyspnea which was related to PE in the presence of deep vein thrombosis of the left femoral vein. Therapeutic anticoagulation was installed for a couple of days until there were several bouts of deterioration. Biplane transesophageal echocardiography (TEE) was performed and revealed a large, wormlike, hypermobile thrombus within the right atrium. Computer tomography (CT) of the chest detected a saddle embolus in the bifurcation of the pulmonary tmnk almost occluding the entire left pulmonary artery (PA) and parts of the right PAT consisted of 100 mg frontloaded rt-PA and the subsequent continuous administration of urokinase in a dosis of 200.000 U/hr for 24 hrs followed by therapeutic anticoagulation. Symptoms, blood gases and ECG improved steadily during infusion, no adverse effects, i.e. minor or major hemorragia were registered. Follow-up CT promptly after termination of T showed almost complete resolution of the saddle embelus, whereas TEE showed complete dissolution of the AT. ' Finally, the patient was switched to oral anticoagulants and had an uneventful clinical course until he was discharged. Conclusion: In the present case, COT was effective for the treatment of a complicated PE without any adverse effect. INTRODUCTION: Nowadays we can assist hearts with problems of insufficiency by techniques other than transplant. Many researchers believe that the best way of assisting insufficient heart muscle is with another muscle from the patient. This technique of ventficular assistance is known as CARDIOMYOPLASTY. We describe the surgical technique of cardiomyoplasty using a biological model. The transformed skeletal muscle is transferred to the thoracic cavity where it wraps the heart and assists it. The choice and preparation of this muscle is currently under investigation. Our group has focussed on the development of protocols for electrical stimulation to transform a skeletal muscle into a muscle which resists fatigue and which is functionally similar to the myocardium. We detect the optimum time at which this muscle has been transformed, by studying the transmembrane action potentials using intracellular electrodes. When the action potential of the trained muscle behaves like cardiac muscle we consider it ready for cardiomyoplasty. CONCLUSIONS: Cardiomyoplasty is an alternative surgical technique to cardiac transplant, which has a great future in the treatment of patients with advanced cardiac insufficiency. We describe methodology which, by intracellular techniques, allows selection of the optimum moment of transformation of a skeletal muscle trained to perform,like cardiac muscle, without suffering fatigue. Purulent pericarditis is a rare disease. Its treatment associate systemic antibiotics and drainage of the pericardium. We report a ease of purulent constrictive pericarditis in which intraperieardial fibrinolysis was use. A 38 years old patient admitted in our ICU for a constrictive pericarditis as a complication of a purulent pericarditis diagnosed seventeen days before. He had also an aehalasia and the o'esogastric endoscopy had found an oesophageal neoplasm. A fistula was not seen, indeed pericardial of flora was the same that oropharyngeal. Hemodynamie and echographic study had confirmed a constrictive pericarditis. Because of the poor state of the patient an intraperieardial fibrinolysis was prescribed (250.000 UI of streptokinase on days 23, 25, 27, 29). Fluid drainage was improved and cardiac output was also improved (day 23 : 2.53 1.min "I, day 27 : 3.58 l.min'l). No change ofhemostasis was noted. A pericardeetomy and an oesophagectomy were performed after 37 days of evolution. Eighteen months latter the patient was still alive. Intraperieardial fibrinolysis seems an interesting therapeutic way if rapidly prescribed in the purulent pericarditis course. The decrease in the systolic pressure following a mechanical breath, termed dDown (delta down), has been shown to be a sensitive indicator of preload (1,2) . However, the clinical use of this method necessitates the introduction of a short apnea. We have therefore developed a Respiratory Systolic Variation Test (RSVT) which obviates the need for apnea. The test is based on the delivery of 4 successive breaths of increasing magnitude (5, 10, 15, and 20 ml/kg). A line of best fit is drawn between the 4 minimal systolic values (one after each breath) and the downslope calculated as the decrease in blond pressure for each increase in airway pressure ( mmHg / cmH20). In 14 mechanically ventilated patients the RSVT was performed during controlled mechanical ventilation under sedation. The test was repeated after the administration of 7 ml/kg of plasma expander. The initial mean downslope of the RSVT was -.40 + .40 mmHg/cmH20. Following volume loading the downslope decreased to -.23 + .44 (ns). At the same time, cardiac output (CO) increased by .96 + 1.2 L/min (p<.02), end-diastolic area (determined by TEE) increased from 18.5 + 6.9 to 20.3 + 7.1 cm2 (ns), and PAOP increased from 12 + 8 to 17 + 9 mmHg ( p < .001). The preinfusion downslope value of the RSVT correlated significantly with the increase in the CO (r = .81) and the EDa (r = .70). Methods: An Expert System has been constructed running on a multimedia computer with the two objectives in mind, viz training of inexperienced staff, and protocol guidance with treatment regimes for all staff. The System is based on experience gained from two previous systems, the one for dealing with acid-base and electrolyte problems in ICU patients; the second for stabilisation of patients with Heart Rate and Blood Pressure abnormalities. The Training Section takes the form of a stage-by-stage account of the insertion of the PAC and displays of correct waveforms, coupled with indications of possible incorrect placements, and guidance when failing to achieve the perfect positioning. The Treatment Protocol Section extends an existing protocol for correcting abnormalities in Heart-Rate and Blood-Pressure, and now takes account of all the indices as measured by the PAC. The system will suggest treatment to correct such things as abnormal Wedge Pressures concomitant with parameter values throughout the rest of the cardiovascular system. The type of patient eg post-operative cardiothoracic or I. C. U. trauma, will be taken into account when recognising abnormal parameter values and when prescribing treatment. Results: A working system which will be improved by the finetuning being carried out. The results and lessons learnt will be presented at the Conference. Method: Septic shock was defined as severe sepsis with either persistent hypotension (mean arterial pressure; MAP < 70 mmHg) or the requirement for a noradrenaline (NA) infusion ~ 0.1 9g/kg/ rain with a MAP --< 90 mmHg. Cardiovascular support was limited to NA + dobutamine (DB). 546C88 was given for up to 8 h at a fixed dose-rate of either 1, 2.5, 5, 10 or 20 mg/kg/h iv. During 546C88 infusion, NA was to be reduced and if possible withdrawn, whilst maintaining MAP above 70 mmHg and the cardiac index (CI) as clinically appropriate. Assessments were made at baseline (t = 0); at I h from the start of treatment (t -1); and at the end of treatment (t -8) with 546C88. Conclusions: 546C88 does not appear to increase MPAP or worsen pulmonary gas exchange in patients with septic shock, when given by infusion for up to 8 h. 546C88 is a novel vasoactive agent for the treatment of septic shock which will now he evaluated in a randomised, placebo-controlled safety and efficacy study. Objectives : To compare cardiac output (Q) data obtained for thermal indicators in pulmonary artery (QTPA) and aorta (QTAo) and for the stable isotope 2HzO in aorta (Q2v~2o) with indocyanine green (ICG) in aorta (QIcG) as reference. Methods : An indicator solution of ice cold H20 (9.4 mL), 2H20 (0.6 mL) and ICG (10 mg) was injected as bolus via the injection port of a Swan-Ganz catheter. Qlco and Qzmo was measured using a dual optical system (Penn Lab Instruments, Philadephia, PA, USA). QTPA and QTAO was measured using a In contrast to the recoveries of thermal indicator in PA and 2H20 in aorta the :~covery of thermal indicator in aorta was significantly increased in group II (n= 18 boluses) over group I (n=18 boluses) (1.3<-0.3 vs. 1.0+-0.1, p=0.04). Conclusions: The "overrecovery" of thermal indicator in aorta is in agreement with " BiScks deconvolution study (i) and results in erroneous values for Q. The most pausible explanation is the distortion of the thermal curve caused by the slow response time of the thermal detection instrument as shown by Ganz (2) Objectives: To compare data obtained with the double indicator dilution method using indocyanine green (ICG) and the stable isotope 2H20 for the estimation of extravascular lung water (EVLW2Hzo) to gravimetriu lungwater data (EVLWg~). Methods: An indicator solution oflCG (10 rag) and 2H20 (0.6 mL) was injected as bolus via the injection port of a Swan-Ganz catheter. Dilution curves for ICG and ZH20 was registered in aorta with a dual optical system (Penn Lab Instruments, Philadephia, PA, USA). Cardiac output and mean tranist time was measured for both tracers (QIco, tlco, Q2n2o, t202o) (1). Data analysis: EVLWg~av was reference for EVLWzHzo calculated as Q2HZO times the difference in mean transit time between t2nzo and rico (Atm2n). As reference for Atzn2o EVLWg~,v was divided by Q~cG to obtain Atg~,. A reference distribution volume for 2H20 was calculated as the sum of central blood volume and EVLWg=v. 54 boluses were administrated in a group (I) of 6 anaesthetized pulmonary healthy sheep while Q was altered. Another 18 boluses were administrated in a group (II) of 6 anaesthetized sheep with stable oleic acid induced pulmonary oedema. EVLWg~v measurement was performed postmortem. 2 Results: For 72 boluses H20 parameters were not significantly different from their respective reference parameter: At2vao 5.3+_3.5 s vs. Atg~, 5.4+3.2 s, EVLWzH2O 332-+195 mL vs. EVLWg~,~ 338+169 mL. In group I the ratio between 2HzO parameters and respective reference parameters (n=54) were independent of Qlco from 1.4 to 7.1 L/min. Obiectives: To assess the thermo dye method using indocyanine green (ICG) and thermal indicator for the estimation of lung water (EVLWT). Methods: Ice cold indicator solution of ICG (10 mg) in water (10 mL The aim of the study was to assess left and right ventricular function in the early postoperative period after orthotopic heart transplantation to elaborate therapeutic approaches of heart function abnormalities correction. Mathefial and methods. Haemodynamic monitoring data of twenty one patients ( 19 men, 2 women ) age from 18 to 56 were studied. Cardiac output, pulmonary artery, right atrium and pulmonary wedged pressure were measured with Swan-Gans catheter. Central haemodynamic indices were calculated with the help of computer-based monitoring system. Relations of ventricular stroke work index to it's end-diastolic pressure were used for ventficular function assessment. Results. In most cases right ventricular disfunction was the main problem. Isolated fight ventficular failure with high pulmonary vascular resistance (PVR) was observed in 24% (5pts), without high PVR-in 43% Opts) and with left ventricular failure-in 33% (7pts). One of the most important reasons for fight ventricular failure was the time of heart ischemia more than 90min, which is of great importance in the ease of distance harvesting. The most effective treatment for cardiac failure was combination of dobutamine with i8oprotherenol, atrial pacing and vasodilatators in case of right ventfieular disfunction. All cases with isolated right ventricular failure were treated sucsessfully. Biventricular heart failure was a sighn of bad prognosis and the reason of death in 2 cases. Conclusion. Right ventfieular disfunetion is the main problem during transplanted heart adaptation in the early postoperative period. Optimal therapeutic management of cardiac disfunction includes infusion of dobutamine in combination with isoprotherenol, atrial pacing and vasodilatators. Cardiology-department of Clinical Centre-Kragujevac Institution for occupational Health "Zastava"-Kragujevac, SR Yugoslavia The aim of the investigate is analisis five years survives patients with A.I.M.In dependence of locality and risk-factors. We ana~sed-39~-pat~e~ts (273 males and 118 woman), average 59,8 years. For statistic evaluation we used Life-table slstem in oder to estimate prognostic determinants. Patients with respkatory muscle paralysis may benefit from respiratory assistance by abdomino-diaphragmatie pneumatic belt. We used a non invasive technique, M-mode sonography, to assess the effect of this device on diaphragmatic excursion. We measured the amplitude of right diaphragm motion in seven patients with Duehenne muscular dysl~ophy in supine position with various thoracic posture (0 ~ 45 ~ 75~ without and during pneumatic belt respiratory assistance. Without respiratory assistance, the thoracic posture had no significant consequence on the amplitude of diapttragm motion, either in quiet or deep breathing. The pneumatic belt increased the diaphragm motion amplitude from 7.1 +__ 3.6 mm to 17.71 +_ 5.5 ram (p = 0.009) at 45 ~ tilt angle, and from 8.4 + 3.8 mm to 19.3 + 5.8 mm (p = 0.009) at 75" tilt angle. The tidal volume increased from 211 + 78 to 373 + 99 rut a145* tilt angle, and from 229 + 78 to 447 + 143 ml at 75* tilt angle (p = 0.009). Two patients could not bear the horizontal position (0' tilt). In the five other patients, the pneumatic belt increased but not significantly the amplitude of diaphragm motion (9.2 + 4.9 mm to 15.5 + 7.3 ram). After an overnight respiratory assistance, PaO2 increased from 66.4 +_. 8.7 to 73 + 10.1 mmHg (17 = 0.015), SaO2 increased from 91.1 + 2.5 % to 91.9 +_. 3 % (p = 0.015), and PaCO2 decreased from 52 + 6.4 to 46.4 +_. 4 mmHg (p = 0.015) According to the ventilatory pattern result, M-mode sonography allows to measure non invasively the improvement of diaphragm kinetics obtained by pneumatic belt respiratory assistance, and may be helpful for its adjustment. Objective: To study the effect of flow triggering (flow sensitivity 1 and 5 L/min) vs pressure triggering (-lcmH20) on inspiratory effort during pressure support ventilation (PSV) and assited/controlled mode (A/C) in 8 stable COPD patients non-invasively ventilated with a full face mask. Methods: The patients were studied during randomized 15 min. runs using a Bird 8400 ST ventilator at zero PEEP (ZEEP). Trigger values for pressure (-lcmH20) and flow (1 L/rain) were the lowest allowed by this ventilator. The transdiaphragmatic pressure time product per breath (PTPdi), dynamic intrinsic PEEP (PEEPi,dyn), maximal airway pressure drop during inspiration (APaw) andl ventilatory variables (Ti,Te,TTot,RR,Vt and minute ventilation) were measured. Results: No major problems due to airleaks or to auto-triggeriffg phenomena were observed in the patients, so that all of them were able to perform all the protocol runs. Minute ventilation and respiratory pattern were not different using the two triggering systems. The PTPdi was significantly higher during both PSV (10.6+6.8 cmH:0 x sec) and A/C (10.1+2.5) with pressure triggering, as respect to PSV (8.9+7.5, p<0.02) and A/C (5.8+4.4, p<0.001) with flow triggering (1 L!m). No differences were observed between 1 and 5 L/min flow triggers. APaw was also significantly larger during pressure triggering; PEEPi,dyn was reduced during flow triggering being 5.5+1.5 cmH20 (PSV flow trigger) vs 8.1+1.5 (PSV pressure trigger) and 4.4+_1.3 (A/C flow trigger) vs'f~5+l (AtC pressure trigger). Conclusions: In stable COPD patients non-invasively ventilated, flow triggering reduces the respiratory effort during both PSV and AiC mode as compared to pressure triggering. This may be partly due to a decrease in PEEPi,dyn using a flow-by system. Objective. Cardiac output is higher during alternating ventilation (AV) (i.e. differential ventilation of the lungs with a phase shift of half a ventilatory cycle) than during synchronous ventilation (SV) of both lungs 1 . We verified the hypothesis that the higher cardiac output depended on a lower central venous pressure and intrathoracic pressure, due to a lower mean lung volume, which we attributed to part of the expansion of the inflated lung at the expense of the expiring, opposite lung 2. We studied this interaction between the lungs during one-sided inflation, which we called cross-talk. Method. In 6 anaesthetized and paralyzed piglets we applied short periods (30 s) of one-sided ventilation (10 breaths per rain, bpm), while the other lung was open to the ambient air. The air flow into the non-ventilated lung during expiration of the ventilated lung was integrated to volume. We studied 1-to-r and r-to-I cross-talk at ventilatory rates of 10, 15 and 20 bpm. The amount of cross-talk was the volume displacement in the non-ventilated lung. Results. During 10 bpm the r-to-I crosstalk was 23 _+ 4.7 % (mean +__ sd) of the tidal volume to the right lung and the 1-to-r crosstalk 31 _ 6.3 % of the left tidal volume. Both values increased at 20 bpm to 30 _ 4.1% (p < 0.05) and 39 _ 7.7 % (p < 0.01) respectively. The values at 15 bpm were in between., Conclusion. We concluded that the lower mean lung volume and lower thoracic expansion during AV compared to SV depends on partial expansion of the inflated lung into the non-inflated lung, resulting in a lower mean intrathoracic pressure as the main reason for the higher cardiac output during AV. Obiective: Natural surfactant given for RDS in premature infants leads to a rapid improvement in oxygenation, but lung compliance did not improve in most studies. However, acute effects on lung mechanics during and immediately after surfactant administration have not been studied before. Methods: A total of 13 administrations of bovine surfactant in recommended doses was given via a small catheter into the distal endotracheal tube either as a bolus (n = 8) or as a slow infusion (n = 5) in 10 infants with established RDS. Static compliance (C), resistance (R) and time constant (TC = CxR) of the lung were measured every 3 minutes with a lung function cart (Sensormedics 2600) without interrupting ventilation. 3 infants receiving synthetic surfactant were studied as controls. Results: After surfactant as a bolus or during infusion C first decreased but then increased, whereas R increased immediately with great fluctuations but did not return to baseline. This pattern was more pronounced in infusion than in bolus administration. Change of C and R varied greatly in the individual case, maximum C was > 400 %, maximum R > 700 % of baseline value. Retreatment was followed by an increase in R in all 3 patients, but C increased only in the one who was responder. Patients receiving synthetic surfactant had no change of C or R and were non-responders. OB~I31CTIVES= Acute lung injury (ALI} sometimes induces severe hypoxernla which may be refractory to conventional modes of mechanical ventilation (MV). The elm of this study was to observe some cardio-pulmonary effects of an alternative method of ventilatory management of severe ALI. Five patients with severe ALI (MURRAy scores >3) requiring MV were studied. Protocol inclusion was considered when a control-mode of MV (with a PzO~=l.0 and a PEEP level <15 cmE=O} was not able to get either a P.OJF=O= ratio >55 or a S.O= >85%. Patients were sedated, paralyzed, and a ventilator (Serve 900C) was used for pressuz'e-control ventilation (PCV). FIO= was maintained at 1.0 and PEEP removed. Continuous gas flow (250• mL/Kg] was humidified and jet delivered through a tube (7 ram ID, 18 mL capacity, 0.09 mL/cm H=O compllancel ended in a nozzle (0.8 mm IS) attached to the endotracheal tube connector. A thermodilution flcw-dlrected catheter was inserted in pulmonary artery. Following variables were recorded 15 minutes before and after protocol started: tidal volume (VT), minute ventilation (vz), Intratracheal pressures (P~w), wedge pulmonary artery pressure (WP), central venous pressure (CVP), mean arterial pressure (MAP), cardiac index (CI), arterial and mixed venous oxyhemoglobin saturation (SaO=, SvOa) , oxygen delivery (DO~) , oxygen consumption (vo2) , intrapulmonary shunting (Q./Qt) , and Oxygen extraction ratio (ERO). This observation suggests that HFPV could allow to ventilate at lower Fin2 and improve blood oxygenation during the acute phase after inhalation injury reducing toxicity risk related to high Fin2. Further studies are necessary to confima these results and evaluate the possible implications on mortality alter smoke inhalation and for other ICU pts. Objectives: To design a system for volume controlled high frequency ventilation (HFV) and to estimate the dependence of the tidal volume (VT) on frequency (f) in normocapnic ventilation in rats at frequencies 2 -25 Hz. Methods: A new system for volume controlled HFV was devised consisting of the generator of the constant flow during inspirium and the constant pressure during expirium. The ventilator allows ventilation at frequencies 2 -25 Hz with the relative inspiratory time (Ti) 0.2 -0.8. The airway pressure was measured at the proximal port of tracheostomic cannula , at the same site inspiratory and expiratory flow was measured using modified Lilly-type of pressure-differential flow sensor. Non-linearity of flow sensor was compensated on line by derived equation based on calibration at static and dynamic conditions. Flow and pressure data were evaluated on line using original software. Value of the positive end expiratory pressure (PEEP) was serve-regulated by analogous feed-back. In animal experiments white Wistar rats (400-430 g) narcotized with ketamine/xylazine with cannulated carotid and femoral arteries were kept at the rectal temperature 37~ The arterial pressure was monitored. After traeheotomy the metal cannula (2mm [.d.) was inserted, animals were curarized and ventilated at the following condition: PEEP = 0.1 kPa, Ti = 0.5. The dead space of ventilator including canula was 0.45 mL. The initial frequency was 2Hz and 10 rain after each change of the ventitatory regimen the blood gases analysis was performed. The frequency was changed according to the following schedule : 2 Hz-->4 Hz-->8 Hz-->4 Hz-->16 Hz-->4 Hz--~25 Hz-->4 Hz. VT for each frequency was regulated to maintain normocapnie ventilation with arterial pCO2 = 40 + 2 mm Hg. The arterial pO2 was always above 70 mm Hg. Results: For normocapnie ventilation in rats the following tidal volumes VT [ mL/kg] were found : VT1 = 5.91 --+ 0.30 mL/kg for ft = 2 Hz, VT 2 = 4.31 +0.12 mUkg for fz =4 HZ, VT3 =3.30 +_0.27 mL/kg forf3 = 8 Hz, Vm4=2.61+0.08mL/kg forf4=16Hz andVmt= 2.18 + 0.13 mUkg for fs = 25 Hz (presented as mean values _+ s.d., n = 6 ). The regression analysis using the mean values resulted in the equation for normocapnic VT in rats in our experiments : VTN = 37. 5 * f-e.30 . Conclusions: The described system allowing ventilation in a wide frequency range 2 -25 Hz with accurate measurements of airway pressures and VT might be useful for optimisation of artificial ventilation in new-barns with different lung pathologies. Supported by grants IGA MZ CR nr 1448-3 and GACR nr 305. S122 Intensive Care Unit. University. Hospital of South Manchester, UK. Methods: Measurements were conducted on 6 ventilated patients (Puritan Bennett 7200ac with metabolic monitor PB 7250 set to measure end tidal CO2). All measurements were repeated with the patient stabilised at 5cm. 10cm and 15cm PEEP. Inclusion criteria were: 1) haemedynamic stab(l( .ty for 1 hr; 2) pulmonaD" anon" flotation catheter in situ: 3) volume control ventilation with plateau of 0.5s: 4) FiO2~ > 11.6 to maintain PaO~. > 10 kPa with 5em PEEP: 5) Qs/Ot > 20%; 6) PaO2/FiO2 ratio <150. Measured variab!es included: r162 minute volume: plateau ainvay pressure: applied and intrinsic PEEP: fractional end tidal CO2; arterial and mixed venous blood gases and hacmod).Ttamic variables. Results: Statistical analysis was performed using repeated measures ANOVA. Significant decreases in cardiac index (Ch p<0.01), compliance (p 7 cm. One case resulted in an endobronchial intubation. The mean height of all patients were 167 cm (153-187) for males and 155 cm (140-170) for females. Of the patients with ETT tip < 3 cm from carina, the mean height was 163 cm and 151 cm respectively. ~2onclusion : Adopting the above quoted reference marks did not result in ideal positioning of the ETT in a significant proportion of cases (32.4%). We postulate that [s because our Asian population is generally shorter than those in previous studies. Objectives: To measure the changes of pulmonary mechanics before and after tracheostomy in patients with prolonged mechanical ventilation and to determine factors that predict the outcome of liberation from mechanical ventilation. Design: Prospective. Setting: Respiratory intensive care unit (RICU) in a tertiary hospital. Patients: Twenty patients with chronic lung disease requiring long-term mechanical ventilation. Tracheostomy is indicated for further care. Intervention: Tracheostomy. Measurements and Results: Pulmonary mechanics including respiratory rate (RR), tidal volume (VT), peak inspiratory pressure (PIP), intrinsic positive end ex~ piratory pressure (PEEPi), lung compliance (CLD), mean airway resistance (RAWM), work of breathing (WOB), pressure time product (PTP) by Bicore CP-100 pulmonary monitor were recorded 24 hours before and after tracheotomy. Ventilator setting parameters remained the same during surgical intervention and were also recorded for comparison. Generally, the mechanics including PIR WOB, RAW~x and PTP showed improvment after tracheostomy. But only PIP was significantly reduced (pre 33.4 _+ 11.8 to post 28.6 _+ 9.2, p < 0.05). Changes of WOBp showed significant correlation with pre-operation RR, minute volume (MV), WOBp, and PEEP(. Changes of RAW M were also significantly correlated with pre-operation PEEP, VT, and RAW M. The patients were divided into two groups according to their outcome after two week follow-up. Group 1 included eight patients who were completely weaned from ventilator; group 2 included twelve patients who still remained ventilator-dependent or were mortality. There was no difference in age, duration of mechanical ventilation, pro, post or changes of several lung mechanics between the groups of patients. Pre-Tracheostomy PEEP i and CLD showed significant difference between these two groups (1.1 _+ 1.6 vs 2.7 + 1.4 in PEEPi; 47.3 _+ 36.9 vs 28.8 _+ 16.5 in CLD, p < 0.05). Pre-tracheostomy ventilator setting in mode of assist/control also showed significant higher percentage in group 2 (3%5 % in group 1 vs 66.6 % in group 2). Conclusion: In prolonged mechanical ventilation patients with chronic lung disease, tracheostomy will significantly improve PIP and slightly reduce WOBp, RAW M and PTR Patients who used pressure support mode before tracheostomy had better underlying lung conditions (lower lung compliance and auto-PEEP) will have better chance to wean from mechanical ventilation. Forty-eight infants with congenital diaphragmatic hernia presenting within the first 6 hours of life, who underwent surgical rapair,were analysed prospectively in order to produce a reliable inde x of severity of disease that would reliably predict eventual outcome. There were 25 survivors and 23 deaths in this series (mortality 48%).Using arterialPCO 2 values measured 2 hours after surgical repairand correlating them with an index of mechanical ventilation,we have been able to clearly define two groups of diaphragmatic hernia based on their response to hyperventilation. The first group, with CO 2 retention and severe preductal shunting,was unresponsive to hyperventilation with high rates and pressures the mortality was 90%. The second group responded well to hyperventilation and demonstrated reversable ductal shunting only. Survival in this group was 97%. Arterial CO 2 accurately reflects the degree of lung development in this disease and separates those patients with severe pulmonary hypoplasia where the outcome is invariably fatal, from those with a well developed contralateral lung where there is excellent potential for survival. Respiratory Failure Unit, Dpt Medicine, Univ. Thessaloniki, Thessaloniki, GREECE The variability of arterial blood gases (Po2, Pc02) and the pH (ABG) was examined in 20 stable ICU patients, few hours before a successful weaning from the ventilator. All patients were lightly sedated and the ventgatory conti~ons were pressure support (PS) for 9 and PS plus intermitted mantatory ventilation in Ii. [n each patient, 6 speciments of ABG were measured at 10 min intervals during a 1-11 study period. At the same time with ABG the arterial blood pressure (BP), the heart rate (cf), the tidal volume (TV) and the respiratory rate (n r were measured. For all the patients, the mean coefficient of variation (C) was 3.45 percent for Po2, 3.53 percent for Pco2 and 2.27 percent for HCO3. The average SD for pH was 0.008, the corresponding C for systolic BP, diastolic BP, cf, TV, rf were 4.35, 5.21, 3.68, 2.51, 4.18 percent. We conclude that the spontaneous variability of arterial blood gases in ICU patients is not substantial ~hen they have stable the heamodynamic and the ventilatory parameters. Deptx?fA'aaesthesioiogy and Reanimation, rhe Sechenov Medical Academy, Moscow, Russia Objective: ~he prevention and treatment of hypoxia in the critical patiems. Methods: I~fusions of perphtoran -a blood substitute with gas-transporting fimclion based on perphtorhydrocarbon -in 496 patients with acute hypovolemia, microcirculatory distnrbance~ tissue gas exchange and metabolism; pulmonary iavage in 104; Iongterm extrapulmonary oxigenation with tleoroearboa oxygenator in combination whb ~trafiltra!ion, hemosorption and hemodialysis -in 73 patients. Results: Pe~htoran increases blood volume, CO,SV, decreases SVR, improves capillary blood flow, increases the blood oxygen capacity, tissue oxygen tension, 102 del, VO2 by improving the rheologic properties of blood and plasma, normalizes 02 ext., prevents and eliminates fat embolisation and ARDS. decreases the need for blood transfusions and infusions of plasma expanders by 1.3-1.4 limes. Alveolar venti!ation-perfusion ratio remains unchanged with its increased effective utilization. There was no surfactant destruction during lavage. Extrapulmonary oxygenation of small volumes of venous blood eliminates venous destruction and then arterial hypoxia and increases pulmonary oxygenation. The use of lluorocarbon cxygenators during hemosorption and hcmodialysis provides the atraumatic and iongterm oxygenation of arterial blood and increases elimination of CO2 which prevents the development of hypoxic complications. Conclusions: Perphtoran and fluorocarb~n oxygenators are effective in the correction of hypoxia in the criticat patients. Objeqtives: To determine if there are differences in oxygen consumption (VO2) during weaning from mechanical ventilation (during total ventilatory support and spontaneous ventilation with CPAP), and to compare different predictive parameters of weaning in predicting success of weaning. Methods; Prospective study in 20 critically ill patients treated with mechanical ventilation for at least 48h, who fulfilled at least 3 of 4 standard weaning criteria (Vt>5ml/kg; respiratory frecuency (f) <35; PIMax > 20 cm H20; PaO2/FiO2 > 150). Baseline measurements: t, Vt, P0.1, PIMax, f/Vt, P0.1*(f/Vt), P0.1/PIMax. Study protocol: Measurement of VO2, VCO 2 (MedGraphics), Vt, f, VE, and arterial blood gases during total ventilatory support (CMV), and after 30 and 120 minutes of spontaneous ventilation with CPAP 5 cm H20. The weaning trial was stopped, failure to wean diagnosed, and MV resumed it a patient presented significant tachypnea, tachycardia, bradycardia, cardiac rythm disturbances, hypertension, hypotension, hypoxemia or hypercapnia. Results: Four patients did not complete the weaning trial, 16 were extubatad, and 2 of them had to be reintubated before 48h, being considered also weaning failures. During CMV, VO2/kg was 4.07 + 0.2 ml/Kg/min, and 5.09 _+ 0.4 mlO-2/Kg/min after 30' on CPAP 5 cm H20 (p < 0,01 ). 14 of 15 patients (93%) with 4 standard criteria were extubated, while only 2 of 5 (40%) with 3 criteria (p<0,01). Next Objectives: Compare the extent and distribution of lung injury in dogs preinjured with oleic acid (OA) and ventilated with high TPP and adequate PEEP in the prone and supine position. Methods: Lung injury was induced with OA (0.06-0.09 ml/kg) in anesthetized, paralyzed, and intubated dogs (n=10) during volume controlled ventilation: rate=12/min, PEEP=5 cmH20, TI/TTOT=0.3, FIO2=0.6, VT=15 ml/kg. Animals were rotated during the OA infusion and the following 90 minute stabilization period to assure uniform injury. In the supine position, PEEP was set 1-2 cmH20 above the lower inflection point (as determined by the pressure-volume curve), and VT was set to obtain a TPP of 35 cmH20: Animals were ventilated in either the prone (n=5) or supine (n=5) position for four hours. Pulmonary artery occlusion pressure was maintained constant (4-6 mmHg) with saline infusion. At the end of the protocol the lungs were removed and divided by template into dependent (D) and nondependent (ND) sections for wet weight/dry weight (V~N/DW) and grading of Nstologic lung injury (HLI; scale 0-3). Oseillatron | is a pneumatic device that generates high frequency, oscillation by means of a reciprocating system in the form of a membrane. It generates sinusoidaI wave form at 2(10 to 10(10 cycles/rain. The system does not deliver gas but must be adapted to the proximal respiratory, circuit of a conventional ventilator, resulting in CI-IFO. It was developed to enhance intrapnlmona~ diffusion during mechanical ventilation and to mobilise endebronchial secretions. Methods. We measured arterial blood gases and haemedynamics during a first period of conventional ventilation (CPPV) followed by. two 30 rain periods of CHFO (sequences : 6(10 and 901) c/rain : group l, n = 1 l: 900 and 600 c/rain : group 2, n = 8). Measurements were made at the end of each period. Cardiac output was measured using thermedilution method: FlU2 and PEEP were kept unchanged throughout the study. Intrinsic PEEP was also evaluated by, means of an occlusive valve. Results. Pa02 is not significantly modified during CHFO at 600 or 900 c/rain. PaCO2 is slightly decreased at 600 c/rain (p = 0.(16). However, intrinsic PEEP remains unchanged. There is no sequential effect (Gr. l vs Gr. 2). There is no more effect of CHFO for patieets who are at a FlU2 higher than 0.50 (n = 9). No changes in haemodynurmcs are observed except a slight increase in central venous pressure (CVP) during CI-IFO (p < 0.Ol). Obiectives: To examine the effects of inspiratory muscles unloading on neuromuscular output at controlled levels of chemical stimuli. Methods: The ventilatory response to CO 2 was examined in ten normal subjects using rebreathing method. Ventilation ~) and respiratory muscle pressure output (Pmus) at the same end-tidal partial pressure of CO 2 (PETCO~) were compared with and without combined flow and volumeproportional pressure assist in two protocols (A and B). Protocol A (n = 10): Two levels of assist were studied; flow assist (FA) of 2 cmH20/I/sec and volume assist (VA) of 2 cmH20/I (Assist 1), and FA of 2 cmH20/I/sec and VA of 4 cmH20/I (Assist 2). All conditions were applied randomly. V~, tidal volume (VT) and breathing frequency (F) were measured breath by breath and plotted as a function of PETCO~. Protocol B: In 5 subjects, in addition to above measurements, esophageal (Pes) and gastric (Pg) pressures were measured and the time courses of transdiaphragmatic pressure (Pdi) and Pmus were calculated. One level of assist (Assist 2) was studied in this protocol. Results: In both protocols inspiratory muscle unloading did not change the F response to C%. Compared to control, with assist V T response was displaced upwards; at PETCO2 of 55 mmHg V T was increased significantly by 0.4+0.1 I and 0.7+0.2 I in protocol A with assist 1 end 2, respectively, and by 0.5_+0.1 I in protocol B with assist 2 (P<0.05). ~/~ responses showed similar changes as VTresponses. In both protocols the slope of V~ response (s did not change significantly with unloading. At low PETCO~ (50 mmHg), Pdi and Pmus waveforms did not differ with and without assist. With unloading, at high PETCO2 (59 mmHg), Pdi and Pmus at the end of neural inspiration decreased by 18.8-+8.3% and 11.7+15.7%, respectively, from control values. Neither change was significant (P>0.05). By theoretical analysis we estimated the expected changes in VT and ~/~ when the levels of assist used in both protocols were applied in the absence of : any change in neural output response to CO z. The predicted response was similar to that observed, indicating that the small difference in Pdi and Pmus between control and unloading runs was due to intrinsic properties of respiratory muscles end respiratory system. Conclusions: These results suggest that when chemical stimulus is controlled, respiratory motor output is not downregulated with unloading. The determinants of the response of the respiratory output to inspiratory flow rates (V~) were examined in awake normal subjects. Subjects were connected to a volume-cycle ventilator in the assist/control mode and V~ was increased in steps from 30 to 90 I/min and then back to 30 I/min. V~ pattern was square, and all breaths were subject-triggered. In six subjects the effects of breathing route (nasal or mouth) and temperature and volume of inspired gas (Protocol A) and in 8 subjects the effects of airway anesthesia (upper and lower airways, Protocol B) on the response of respiratory output to varying V~ were studied. In Protocol B, in order to calculate muscle pressure during inspiration (Pmus), respiratory system mechanics were measured using the interrupter method at end-inspiration. Independent of conditions studied breathing frequency increased . significantly and end-tidal concentration of C% decreased as V~ increased. The response was graded and reversible and not affected by breathing route, temperature and volume of inspired gas and airway anesthesia. With and without airway anesthesia (Protocol 8) neural inspiratory and expiratory time and neural duty cycle, estimated from Pmus waveform, decreased significantly as V~ increased. At all conditions studied the rate of change in airway pressure prior to triggering the ventilator tended to increase as V~ increased. The changes in timing and drive were nearly complete within the first two breaths after transition with no evidence of adaptation during a given ~/~ period. We conclude that V~ exerts an excitatory effect on respiratory output which is independent of breathing route, temperature and volume of inspirate and airway anesthesia. The response most likely is neu~'al in origin, mediated through receptors not accessible to anesthesia such as those located in chest wall or below the airway mucosa. It has been shown, in mechanically ventilated awake normal humans, that increasing inspiratory flow rate (~/~) exerts an excitatory effect on respiratory output. It is not known if this effect persists during sleep. To test this seven normal adults were studied during wakefulness and NREM sleep. Subjects were connected through a nose-mask to a volume-cycled ventilator in the assist/control mode and ~/t was increased in steps (3-4 breaths each) from 30 to 70 I/min and then back to 30 I/min. V~ pattern was square, and all breaths were subject-triggered. Forty-one trials during NREM sleep and 10 during wakefulness were analyzed. Both during sleep and wakefulness minute ventilation increased and total breath duration (TToT) decreased significantly in a graded and reversible manner as ~' increased. These changes were complete in the first breath after V{ transition. The response was significantly less during sleep than during wakefulness (P<0.05); at 30 I/min TTOT, expressed as % of that at 70 I/rain, was 110.2+_1.3% during sleep and 127.8+_3.9% during wakefulness. During wakefulness, at 30 I/min, the rate of change in airway pressure prior to triggering the ventilator, an index of respiratory drive, was 60% of that at 70 I/min (P<0.05). The corresponding value during sleep, was 86% (P>0.05). In four sleeping subjects the increase in V~ was sustained for 1.5-2 min. There was no evidence for adaptation of the response; Tro T, averaged over the last three breaths, did not differ from that obtained when Vj was sustained for only 3-4 breaths. We conclude that 1) Vt exerts an excitatory effect on respiratory output, mediated by a reflex neural mechanism and 2) the gain of this reflex is attenuated by sleep. Chest radiographs is a common complementary technique for patients in critical care units, with a low cost and easily available. However, it has certain well-known limits in diagnosis, the most important derived from the low quality of some pictures. In this paper we make a general review of some new technical approaches developed for improving the quality of the images, and so incrensing the diagnostic value of conventional radiology. We begin deaEng with the correct positioning of the patient, trough the filtering techniques, the synchronization of radiology and ventilation, and we make reference to the new computerized systems for digital image processing. Conclusions: The portable radiographic system is a device that probably with maintain for many years in critical care units as a basic non-invasive diagnostic tool. But we need an increase in the efficiency of it, applying means as simple as a correct positioning of the patient, or the use of fitlers or synchronizers. Thus we should improve the general standards of portable radiography. "Are circular circuits safe? Quantifying undelivered tidal volume in pediatrics patients". Objectives: To evaluate the overall influence of internal compliance of circular circuits on delivered tidad volume (VT). Methods: We studied prospectively 14 ASA I pediatrics patients (2 to 10 yr. old) scheduled for elective general surgery. Mechanical ventilation was supplied by an Ohmeda Excel 210 (circular circuit). The internal compliance of the circuit (Cc)-anesthesia machine plus external circuit-was determined by the supersyringe method: Corrugated DAR tubes of 10 mm. ID and 1.5 m. long (children < 30 kg), and a corrugated DAR set of 15 mm. ID and 1.5 m. long (children > 30 kg) were respectively used for Ccl an Cc2 values of 9.3 and 9.5 ml/cm H20. A VTof 10 mlg/kg and respiratory frequency was adjusted for an end-tidal CO2 (ETPCO2) between 30 35 mmHg. Tidal volumes (measured by spirometry) and airway pressure (Paw) data were recorded every ten minutes. Volumes and thorax-lung compliances were calculated as follows: (VT delivered = VTadjusted-VoL compressible, being Vol. compressible = Co x Ppeak (aw). Apparent compliance (Ca) = VT adjusted/Pplateau(aw), and true compliance (Ct) = =VT delivered/Pplatean(aw)). Comparative statistics were separately designed between calculated compliance data and tidal volumes on a paired sample ~test basis. Results: Calculated values for volumes and thorax-lung compliances were: Conclusions: Due to the elevated internal compliance of the circular circuit there is a remarkable dilference between adjusted and delivered VT: mean undelivered VT was 38.8 % and reached as high as 64.1%. TEere is also a significative error in calculating true thorax-lung compliance: its overestimation can be as high as 69.6 %. Circular circuits are considered safe and cost-saving for anesthetical practice. Nevertheless we conclude that anesthetists should bearin mind VT losses when using circular circuits, due to compressible volume. Tracheal stenosis is one of the most serious complications of patients submitted to prolonged endotracheal intubation, in which the decrease in inner diameter of upper airway makes it very difficult to achieve a correct ventilation. Objectives: Compare the results of applying high frequency jet ventilation (HFJV) to some of these patients with conventional controlled ventilation (CMV). Methods: We used a prototype of high frequency jet ventilator (Santiago-2) developed in our University, and we developed a tracheal tube in wich we modified the distal tip (conic tip). We applied this system to two patients which were initially ventilated in the operating room with usuai controlled mecanical ventilation (CMV) following the standards of our department, and then intubated with the special endotracheal tube and ventilated with HFJV. Results: We could verify a proper ventilation of both patients with CMV and HFJV. During HFJV, the airway pressures were lower than those recorded during CMV. A lower airway pressure prevents lesions due to high pressures. Conclusions: HFJV is a good method of ventilation for patients with significative stenosis of the trachea, not only during surgical procedures, but also during ventilation for long periods in critically 111 patients. The ventilatory setting is pressure support mode. The pressure level and Fit2 were kept constant during H/D. Arterial blood gas, WBC count, and mean BP was checked according to the schedule: 0'(immediately before H/D), 15', 30', 60', 120', 180', 240'. Respiratory drive (represented by PoA), tidal volume(Ti) and minute ventilation(VE) were continuously recorded by pulmonary mechanics monitor (Bicore CP-100). The mean value of the breaths 5 minutes before blood sampling were used to represent the ventilatory status of that period. ANOVA test is used for comparison between groups. For PoA, Hierarchical Cluster method is applied to divide the cases into two groups of similar change. Conclusions: our data suggest that PL is very useful, non invasive and low-expensive emergenc E support for ARF, expecially in the elderly with severe chronic pulmonary disease and relative controindications to EtI. PL seems to be an effective alternative when it is not immediatly possible to perform Etl. The multiple inert gas elimination technique (MIGET) can be used to assess the effects of any given mode of mechanical ventilation on the pulmonary and systemic factors determining arterial PO2 and PCO> However, a potential problem in mechanically ventilated patients is that the 10 L mixing box (MB-10L) placed in series in the expiratory side of the circuit of the ventilator to sample mixed expired gas may provoke substantial discrepancies between the tidal votume set in the ventilator and the effective tidal volume delivered to the patient, due to the increase in the compression volume (Vc) of the circuit. The effects of the MB-10L on the V c were compared with those produced by a new 1 L mixing box (MB-1 L) specifically designed to produce adequate gas mixing and to prevent loss of the two most soluble gases (ether and acetone) used in the MIGET. At any given peak cycling pressure (P~ak, cm H~O), the V c (ml) provoked by the MB-10L was substantially higher (Vc= 7.4*Ppeak) than that provoked by the new MB-1L (Vc= 1.4*Ppeak). At a Ppeak = 50 cm H20 ~ the V c were 377 ml (MB-10L) and 67 m{ (MB-1L), respectively (p< 0.001). In a group of 6 subjects (4M/2F, 57_+6 years), for each of six the gases used in the MIGET, the regression line between the mixed expired partial pressures simultaneously obtained from MB-1 L and MB-10L fell on the identity line. It is concluded that the new MB-1L allows adequate assessment of the effect of different modalities of mechanical ventilatory support on pulmonary gas exchange, with less potential for gas compression and thus hypoventilation. Objectives Evaluate the influence of different pressure support ventilation (PSV) levels on cardiovascular and respiratory funcion in ICU polytrauma patients. MetBed&We studied 15 polytrauma ICU patients , who were in weaning process , after long term mechanical ventilation for acute respiratory failure . Mean age 52 (37-71) yrs . They all were connected to servo ventilators Siemens 900C , and all were in stable condition , without sedation , inotropes or diuretics. The hemodynamic studies were done with continuous SVO2, Swan Ganz catheter (Oximetrix, Abbott). They all were in spontanuous mode (spent) with 5 cm H20 CPAP for at least one hour. We turned them to PSV with 0 inspiratory assistance (PSV 0 cm H20) and after 60 rain we applied PSV 10 cm H20, and after 60 min PSV 20 cm H20 . Hemodynamlo and respiratory measurements were done before and after the application of insiratory assistance. The results were statistically analyzed with ANOVA. Resets . Respiratory variables . No significant changes in minute volume (VE). Tidal volume (VT) and mean airway pressure (mPaw) increased statistically significant (p< 0.001 ) . Respiratory rate (RR) decreased significantly (p<0.01) . Blood gase showed no difference . Cardiovascular variables. Cardiac output (CO) decreased NS , heart rate (HR) had no change , central venous pressure (CVP) , mean pulmonary artery pressure (mPAP) , pulmonary capillary wedge pressure (PCWP) , increased NS , oxygen delivery (DO2) decreased NS, oxygen consumption (VO2) decreased NS. Conclusions. PSV is a very useful respiratory mode helping patients to be weaned from long term mechanical ventilation . It has beneficial effects on respiratory function and oxygen consumption without affecting seriously the hemodynamic parameters, possibly due to a decrease of the work of breathing. A. Michalopoulos, A. Anthi, K. Rellos, J. Kriaras, S. Geroulanos Intensive Care Unit, Onassis Cardiac Center, Athens. Objectives of this study was to examine the effect of different levels of PEEP on postoperative SvO2 and PvO2 values in a group of patients, following open heart surgery. Methods: Upon transfer to ICU, 67 patients (54 males and 13 females) of mean age 63_-+6 years, were randomly assigned to receive 0 (N=22), 5 (N=24), or 10 cm of PEEP (N=21). There were no statistically significant differences in demographic data or preoperative respiratory status among the three groups. All patients were ventilated on the assist control mode with a tidal volume of 10 ml/Kg. The fraction of inspired oxygen (FiO2) was adjusted to keep a PaO2 around 100 mmHg. Mixed venous PO2 and SvO2 were measured at 30 min, 4 and 8 hours after application of mechanical ventilation in the ICU, just before extubation (BE), half hour after extubation (AE), and at 4 hours post-extubation. Differences at each study time were analysed by ANOVA. Results: Mean SvO2 and PvO2 values among the three groups, for all study intervals, are presented in the table. Conclusion: We found no differences (p=NS) in tissue oxygenation (expressed by SvO2 and PvO2) among the three groups, at any study interval, in the early postoperative course of patients following open heart surgery. Intrinsic PEEP (PEEPi), and high elastance and resistance increase inspiratory work load in COPD. CPAP reduces work of breathing by counterbalancing PEEPi. PAV provides flow (FA) and volume (VA) assistance proportionally to patient resistance and elastance and inspiratory effort. We studied the effects of partitioned support (CPAP-FA-VA) on breathing pattern and inspiratory effort in five COPD patients on PAV compared to spontaneous ventilation (SV) and full support (FS: CPAP+FA+VA). Flow, volume, minute ventilation (VE) respiratory rate (RR), inspiratory swing in esophageal pressure (APes), and its integral per breath (PTI/b) and per minute (PTI/m) were measured. Objectives: To evaluate airway pressure fluctuation (APF) during spontaneous breathing in a high compliance CPAP system. Methods: The CPAP system consisted of two 7L weighted balloons in a wedge shaped holder. Ventilating gas flowed from one balloon through a low resistance one way valve into a tracheal tube (ETT) provided with a Pycor CO 2 sensor to monitor rebreathing. The ETT was connected to a piston drive mechanical lung. Expired gas flowed through a low resistance valve into a second weighted balloon, from where it was exhausted through a PEEP valve connected in parallel with the second weighted balloon. We evaluated system performance at V r from 70 to 500ml, at RR from 10 to 120 bpm, while closely monitoring CPAP airway pressure swings. At V v of 400 and 500ml the RR was limited to 60 bpm. For comparison we explored APS of a one 16L balloon CPAP system, the CPAP mode of the Puritan Bennett 7200, and Siemens 300 ventilators, when connected to a healthy adult volunteer breathing through an ETT. Results: The Compliance (Cpl.) of one 7L balloon system was linear over a range from 1.0 to 3.3L, with a Cpl. of 4.0 L/em H20.The Cpl. of the 16 L balloon (0.5 L/em H20) was linear between a volume of 13 and 14.5 L. APF of the weighted balloon system was under 1 em H20 at all V r (except at a V r of 500ml APS was 1.5em H20), while the APF in the 16L balloon was up to 3 em H20. APF witli human volunteers with the two commercially available ventilators in the CPAP mode was about 7 cm H20; while under identical conditions APF in the 16L balloon system was 1.5 emHzO; and in the two 7L balloon system was below lcm H20. Conelusions: CPAP using the two balloon system exhibits lower airway pressure fluctuations than a single balloon system; and is substantially lower than found in the two commercially available ventilators when used in the CPAP mode. Objective: To perform independent lung ventilation (ILV) with individual tidal volume (Vt) set at a value generating a plateau airway pressure (Pplat) < 25 crnH~O and to evaluate the usefulness of the continuous monitoring of endtidal CO2 (EtCO2) as a guide to titrate individual lung Vt during ILV and for the weaning from ILV. Methods: In seven patients, ILV was performed with ttvo ventilators set with the same FiO: and respiratory rate. Each lung was ventilated with a Vt that developed a Pplat < 25 cmH~O. This setting led to a lower Vt on pathological lung (PL). Vt was increased in PL following EtCO~ and PaCO2-EtCO2 variations. ILV was discontinuated when EtCO~., Vt and statical compliance (Cst) were similar in both lungs. Results: One hour after starting ILV (Ti), PL mean Vt was significantly lower than in normal lungs (NL) (224 + 46 ml vs 377 + 766 ml, p<0 001) Two individual behaviours were observed on Tl in PL: four patients presented low EtCO: (range 18 -31 mmHg)and normal PaCOz (range 38 -42 mmHg), while three patients had normal EtCO2 (range 35 -45 mmHg) with high PaC02 (range 44 -61 mmHg). One hour before stopping ILV (T2), Vt, EtC02 and PaCO2 were the same in each lung. The PaO2/FiO: ratio improved in all patients from the beginning oflLV Cst of PL was 526 + 30 % of the normal lungs' Cst on TI and improved to 97.6 + 27 % ofNL's Cst on T2 (p<0.005 vs Conclusions: Setting Vt of PL to a value not overcoming a Pplat threshold does not impair oxygenation and is helpful in avoiding barotraumatism. Measurements of differential EtCO2 and of the differential PaCO2-EtCO2 gradient can be used to titrate Vt allocation during ILV and as a guide for the weaning from ILV. Total respiratory resistance in mechanically ventilated patients exceeds values obtained in normal subjects, due to the added and highly flow dependent resistance of the endotracheal tube (Rett). This can adversely effect the efficacy of pressure regulated modes of assisted ventilation, such as pressure support (PSV) and proportional assist ventilation (PAV). Recent work demonstrates that the influence of Rett during PSV can be overcome by using tracheal (Ptr) rather than airway opening (Pao) pressure to regulate the pressure applied (Intensive Care Med 20:$41, 1994) . The purpose of this study was to see if this approach would also be effective during PAV. Flow, volume, Pao, Ptr, and transdiaphragmatic pressure (Pdi) were measured in 5 intubated patients in which either Pao or Ptt were used to regulate the pressure applied during PAV where volume assistance was varied from 20 to 80% of respiratory elastance. Representative results (mean + SE) are shown below. Compared to spontaneous breathing (PAV 0%), PAV increased tidal volume (VT) while reducing respiratory rate (RR) so that minute ventilation ('~E) also rose. This was associated with a reduction in inspiratory effort, as reflected by a decrease in the pressure-time integral ( [ P) of Pes and Pdi both per minute and per liter ~rE. The effects on breathing pattern were similar for Pao and Ptr regulated PAV. In contrast, the reduction in inspiratory effort was always greater for Ptr regulated PAV. In conclusion, the volume assistance provided by PAV is more effective when Ptr rather than Pao is used to regulate the pressure applied. PAV Methods: Retrospective data analysis of 596 adult patients with normal pulmonary function before operation and uneventful course following coronary artery bypass graft surgery over an 18 month period. We compared assist/controlled mandatory ventilation (S-CMV, 123 patients), synchronized intermittent mandatory ventilation with inspiratory pressure support (S-IMV/PSV, 431 patients) and biphasic positive airway pressure ventilation (BIPAP, 42 patients). Results: Patients ventilated with BIPAP had a significantly shorter mean duration of intubation (10.1 h, p< 0.05) than patients treated with S-IMV/-PSV (14.7 h) and S-CMV (13.2 hi. With S-CMV 39.9% of the patients required single or multiple doses of midazolam but only 13.5% in the S-IMV-/PSV group and 9.5% in the BtPAP group. The mean total amount of midazolam of these patients was significantly higher in the S-CMV group (8.8 mg) than in the S-IMV/PSV group (6.6 mg, p<0.05) and in the BIPAP group (4.3 mg, p<0.05). The consumption of pethidine and piritramide did not differ between S-CMV and S-IMV/PSV but was significantly lower during BIPAP (p<0.05). After extubation the paCO2 patients was highest in the S-CMV group. Conclusion: Ventilatory support with BIPAP reduces the consumption of analgesics and sedatives and duration of intubation. Unrestricted spontaneous breathing as well as fully ventilatory support allow adequate adaptation to the patients requirements. BIPAP seems to be an alternative to S-CMV and SqMV/PSV ventilation not only in patients with severe ARDS but also in short term ventilated patients. _Objectitives: After end-inspiratory airway occlusion we examined the ensuing gradual decrease in tracheal pressure (Ptr) with the following equations proposed by Bates et al. and Hildebrandt: Pv = P'v e'~Cccl~2 +Pst, rs (Bates) [1] where P'tr is tracheal pressure immediately after occlusion, to= is occlusion time, "r 2 is viscoelastic time constant of respiratory system, and P t is static elastic recoil pressure of respiratory system. P~(t) = H 1 -H 2 log t (Hildebrandt) [2] where H~ and H 2 are parameters depending on lung volume, and initial time is 1 s for analytical reasons. Materials & methods: We studied 8 healthy patients intubated, anestethized with propofol, paralyzed with vecuronium, and mechanically ventilated with constant flow (0.5 I/s) at ZEEP for minor surgery. Pressure was measured in the trachea. Flow was measured with a pneumotachograph and volume was obtained by numerical integration. The rapid occlusions were produced by an external valve. The signals were sampled at a frequency of 200 Hz and processed on a PC. The influence of the cardiac artifacts during the occlusion time (4 s) was reduced by a software low-pass filter Kaiser Finite duration Impulse Response of elevated order. Results: The mean (+ SD) coefficient of correlation using Eq. 1 was 0,912 -+ 0.168, and using Eq. 2 was 0.884 + 0.045. The values ofz~ (Eq. 1), however, decreased with increasing the tidal volume (Vt) according to the following equation: "~2 = 1.52 -0.65 V t, Similary, the values of H~ and H 2 increased with increasing V t according to the following functions: H~ = 4.4 + 13 V I and H 2 = 1.15 + 1.88 V t. Conclusions: The behaviour of "% of Eq. 1 suggests that the linear viscoelastic model is not sufficient to further describe the mechanical properties of the respiratory system over the Vt range (6-14 ml/kg) in ventilated patients. Infect this model predicts that "c 2 is constant and independent of tidal volume. On the other hand the plastoelastic model is not sufficient to further describe the mechanical properties of the respiratory system. In fact "r 2 obtained by fitting an exponential for data of Eq. 2, is determined by the time of endinspiratory airway occlusion. Obiectives: According to the viscoelastic model, the viscoelastic pressure of the respiratory system Pv=rs during lung inflation with constant flow e~ is T/ R 1 2 1 wh T lSmS Ira tlmeand R given by:Pv~c.~ = 2d~( -'e-~ )[ ] ere " ' p" tory " 2 and "r 2 are resistance and time constant of viscoelastic unit. In the past, the viscoaletic constants were determinated by performing a series of occlusions at different lung volumes, or a sedes of occlusions at a fixed lung volume achieved with various inflation flows. In the present study we have developed a new method for determining "c 2 and R 2 which requires a single constant flow inflation. Our method is based on determination of Pv~r, during a single breath constant flow inflation, and of z 2 during the ensuing end-inspiratory airway occIusion. Dudng the occlusion the tracheal pressure P~, declines according the following function: Ptr = P'lr e " too= " z2 + e~t.r= [2] where P'~r is tracheal pressure immediately after occlusion, toc c is occlusion time, P,i.rs is static elastic recoil pressure of respiratory system, and ~ is viscoelastic time constant. We first determinated "~2 by analyzing the time-course of Ptr according to Eq 2 and next determining R 2 according to Eq. 1, using the expedmental values of P,i=~, ~ and Ti, as well as "~2 obtained with Eq. 2. Materials & methods: We studied 8 healthy patients intubated, anestethized with propofol, paralyzed with vecurenium, and mechanically ventilated with constant flow (0.5 I/s) at ZEEP for minor surgery. Pres-sure was measured in the trachea. Flow was measured with a pneumniachograph and volume was obtained by numerical integration. The rapid occlusions were produced by an external valve. The signals were sampled at a fi'equency of 200 Hz and processed on a PC. The influence of the cardiac artifacts dudng the occlusion time (4 s) was reduced by a software low-pass filter Kaiser Finite duration Impulse Response of elevated order. Results: The mean coefficient of correlation with Eq. 2 was 0.912. With V t of 7 ml/kg, the mean values (+ SD) of ':2 and R 2 of the 8 subjects amounted to 1.128 • 0.100 s and 3.990 • 0.890 cmH20 I "~ s. With the traditional multi breath method the corresponding values were 0.711 + 0.257 s and 4.445 _+ 1.474 cmH20 I "1 s, respectively. With the T-test the difference between new and traditional "~2 was statistically significant, between new and traditional R2 was not Significant. Conclusions: With the single breath method it is possible to compute ':2 and R 2. The mean values of R 2 with V t of 7 nd/kg, however, was slighUy different than those obtained with the traditional multi breath method. The application of modem principles of respiratory care and mechanical ventilation in ICUs has resulted in increased survival of critically ill individuals with neuromuscular, skeletal and irrevers~le pulmonary diseases. In these chronically ill individunts mechanical ventilation, long term 02 therapy (LTOT) and continuous home care is considered a chronic life supporltng technique that can not be withdrawn after their discharge from an ICU. The aim of this study was to present the results of a Rehabilitation Programme and home care that runs in our ward. TwenW three patients were referred to our clinic f~om ICUs during 1993-94. A specific Rehabilitation Programme designed according to individual's needs was performed. Patients that benefitted from this programme were grouped into the following disorders. 1) Post TB Respiratow Failure 6(26%) 2) Neuromuscular diseases, 3(21%) 3} Undiagnosed SAS 3{13%) 4) COPE) 9(39%) (3 patients had a overlap syndrom). The programme consists of : 1) assessment and mechanical support ff needed of the respiratonJ system with non invasive methods (nasal or via tracheostomy). 2) group and individual respiratory therapy 3) mobilization 4) nutritional support 5) educational classes for the members of the family. Three from the patients passed away (during the year), 11 are under NIPPV during night with or without 02 supply, 13 pts recieve LTOT. Conclusion: The development of a programme for chronically ill individuals in especially designed wards in hospitals and the overall care at home is considered necessary at least in Hospitals with ICUs. A Rehabilitation Programme and Home Care permits the fast but safe discharge of these patients from units of acute medicine that the cost of treatment is high and besides permits beds that are invaluable. We considered that the Rehabilitation Prod'amine and Home Care in our ward is the first performed in Greek chronically ill pts and even though there is no special administxative support we think that the results are quite saltsfactory. Objective: We postulated that the product of the respiratory frequency (f) and the ratio of inspiratory pressure (IP) to maximal inspiratory pressure (MIP) would predict the weaning outcome in deeompensated COPD patients better than either variable alone or other indices previously proposed. Methods: In 28 decompensated COPD patients with difficult weaning, we measured, daily, respiratory mechanics data both during mechanical ventilation and after ten minutes of spontaneous breathing. Then we calculated weaning indices reported in literature and some new integrated indices. According to the results of the discriminant analysis, we considered the integrative index CROP (acronym of compliance, rate, oxygenation and pressure), the rapid shallow breathing index f/VT, the load/capacity ratio IP/MIP, and the following new index: f x IP/MIP. We used receiver-operatingcharacteristic (ROC) analysis by calculating the area under the curve considered as the overall probability of correct classification. Results: main results are reported in the following Objective: To evaluate the reliability of some indices of endurance in predicting the weaning outcome of decompensated COPD patients. Methods: In 28 decompensated COPD patients with difficult weaning from mechanical ventilation (MV) we measured, daily, blood gas analysis, ventilatory and airway pressure pattern during MV, breathing pattern (frequency (f) and tidal, volume (V~)), inspiratory pressure (IP), and maximal IP (MIP) during spontaneous breathing (SB). Thereafter we calculated the following weaning indices: CROP (Compliance * MIP * (PaO2/PaO2) / f), flVT, IP/MIP. Data obtained the day at which the patient was considered ready for a trial of SB on clinical grounds but weaning failed (WF) and those obtained the day of the successful weaning (WS) were compared statistically through the Wilcoxon rank-sum pair analysis. In order to quantify the predictive accuracy for each index with respect to successful weaning we calculated sensitivity, specificity, and diagnostic accuracy according with the standard formulas. Methods : Five patients (64 + 6 yrs) suffering from ARDS (lung injury score > 2.5) for 48 hours or less entered into the study. IRV (volume controlled, decelerating flow, 20 % inspiratory pause, liE = 2/1) was compared to conventional ventilation (CV) (volume controlled, constant flow, no inspiratory pause, IIE= 1/2). These two modes were applied for 6 hours in a randomized order, with the same levels of total PEEP (PEEPt = PEEP + PEEPi), tidal volume (8.0 • 0.7 ml/kg), respiratory rate (20 • 0"bpm) mad Fit2 (63 • 2 %). Measurements (respiratory mechanics, hemodynamics, arterial and mixed venous blood gases) were performed after 1, 2, 4 and 6 hours of application of each mode. RVSUIls : are expressed as mean + SEM and compared by ANOVA. Backeround and methods: Periodic breathing (PB) is characterized by repetitive cyclic variation in minute ventilation. PB is considewxl to be provoked by an instability in the respiratory control. Inintubated, spontaneously breathing patients conventional modes of pressure support ventilation, i.e., triggered inspiratory pressure support 0PS), do not allow patients to breathe with theirinherent breathing pattern. Therefore, PB, if existing, will appear mainiy after extubation. Since our new mode of pressure support ventilation" automatic tube compensation" (ATC) continuonsly corrects for the flow-dependent tube resistance during insNmdon and expiration ("electronic" extubatim), it pemaits patients to maintain their own inherent breathing pattern. Then, ff necessary, tracheal pressure can be additionally supported by volume-proportioead and/or by flow-proportional pressure support (proportional assist ventilation, PAV). (~as~: We report the case of a 70-year-old male patient who was Intubated due to acute respiratory insufficiency after acute myocardial infarction with left ventricular dysfunction. During IPS of 10 mbar the patient showed a regular breathing pattem which became periodic during ATC. In addition, proportional assist ventilation of 10 mbar/L increased periodic breathing in such a way that the typical Cheyne-Stokes breathing pattem occurred (see figure) . Baqkground: The Hering-Breuer reflex (HBR) is characterized by an inhibition of inspiration during lung inflation. This response has been recognized as an important vagally mediated mechanism for regulating the rate and depth of respiration in newborn mammals. In adult man the HBR is considered to be active only at lung volumes well above functional residual capacity, i.e., at tidal volumes above 1000 ml. Assessment of the HBR requires specialized methods such as single breath or multiple occlusion technique. Methods; In the presence of desynchronization between ventilator and patient, which frequently occurs during triggered inspiratory pressure support ventilation (IPS)(see figure) , prolongation of the interval between inspiratory efforts (indicated by negative deflection of the esophageal pressure) due to lung inflation exposes an active HBR. We examined the occurrence of HBR in intubated critically ill patients. Strength of HBR was assessed by the formula: prolongation [%] = ((inspiratory interval of interest -preceding inspiratory interval)/preceding inspiratory interval) * 1(30. Rr162 18 of 50 patients examined showed moderate to severe desynchronization. In 17 of these 18 patients a (re)activation of the HBR was found. The strength of HBR amounted to 134 + 51%. There was a significant correlation between tidal volume and strength of HBR. In contrast to previous reports, an active HBR was shown during lung inflation well below 1000 ml. B pck~round: Triggered inspiratory pressure support ventilation (IPS) is commonly used to support inspiration in intubated spontaneously breathing patients. Despite its usefulness IPS shows some disadvantages which can be deleterious in crificaUy ill patients: -additional work of breathing to be performed by the patient due to the flow-dependent tube resistance -desynchronization between patient and ventilator due to inherent triggering failures of the IPS mode suppression of the patient's inherent breathing pattern -inability to predict successful extubation in difficult-to-wean patients Methods: Based on the known flow-dependent tube resistance our new mode "automatic tube compensation" (ATC) compensates for the pressure drop across the endotracheal tube ("electronic" extubation). Then, if necessary, tracheal pressure can be supported by volume-proportional pressure support (VPPS) and/or by flow-proportional pressure support (FPPS). Results: Hitherto, we have examined 20 patients after open-heart surgery and 50 patients with acute respiratory insufficiency (ARI) or ARDS using ATC with/without VPPS/FPPS. Preliminary results suggest that the new mode avoids additional work of breathing due to accurate compensation of the pressure drop across the endotracheal tube during in-/expiration prevents desynchronization between patient and ventilator allows patients to breathe with their inherent breathing pattern accurately predicts the outcome of extubation even in difficult-to-wean patients due to "electronic" extubation Conclusions: The new mode ATC with/without VPPS/FPPS allows to support ventilation in a more physiologic manner and overcomes the disadvantages of conventional modes of pressure support in intubated patients. Backgound: Cheyne-Stokes respiration (CS) is characterized by regula]; recurring periods of hyperpnea and apnea. In normal subjects, CS may occur after hyperventilation, after arrival in high altitude, or during sleep. It has also been observed in patients with prolonged circulation time due to congestive heart failure, as well as in some neurological patients. There is no report about the influence of sedative drugs on periodic breathing (PB) and CS. Methods: In intubated patients conventional modes of pressure support do not allow patients to breathe with their inherent breathing pattem. Therefore, periodic breathing and CS are rarely seen. Since our new mode of pressure support ventilation "automatic tube compensation" (ATC) continuously corrects for the flow-dependent tube resistance during inspiration and expiration ("electronic" extubation) it permits patients to maintain their own inherent breathing pattem even if pathological, e.g., periodic. Results: Using this new mode of pressure support ventilation, periodic breathing was unmasked in 13 of 37 intubated patients, 6 of which showed CS. In 4 of these 6 patients the occurrence of CS was linked to impaired left ventricular function with increased circulation time. Normal left ventricular and neurologic function was found in the remaining 2 patients. In 1 of these 2 patients CS disappeared after intravenous administration of the benzo-diazepine antagonist Flumazenil (figure). Consequently, in this patient CS was induced by benzodiazepine sedation. Objecti',~s: In contrast to conventional rhodes for pressure supported spontaneous breathing, our newly developed ventilatow mode ,,Automatic Tube Compensation" (ATC) completely compensates for the flow-depandant pressure drop tLPm-r across endotracheal ttLbe (ETT). In the ATC mode, the ventilator supplies a flow V' in order to maintain a constant tracheal pressure P~,,~. To this end, Pk,,= has to be oontinuousiy determined. Since continued measurement of P,,~ by introducing a catheter via the ETT is not reliable, we opted for its continuous calculation socordng to the following equation: P~ = P,,, -APErr, Pw being the continuously measured airway pressure. This also requires the continual measurement .of flow V' to calculata APm-r using the non-fineer approximation: APort = KvV' + K2.W. The constant tube coefficients K~ and K2 are mathematically determined by mesns of a least-squares-fit procadum based on laboratory investigations. Tracheal secretions, however, reduca the omss-saction of the ETT. Consequently, ~ values of KI end K2 are changed rendering the P~,ch calculations inaccurate. Therefore, K1 and ~ have to be pedodcally updated to ensure an a~urete monitoring of Pn,~ and a complete tube compensation under ATC at any time. Background: One of the first steps in weaning patients from controlled mechanical ventilation is to stop muscle relaxation and to reduce sedation. It can take several hours, however, until the patient is able to trigger the ventilator and to breathe spontaneously. During this period, many patients display a sudden increase in peak airway pressure of up to 30%. Patients and methods: To investigate the reason for this potentially dangerous effect, we continuously measured lung and chest wall mechanics in post-operatively ventilated patients. Lung mechanics (airway resistance and lung compliance) was measured using the esophageal balloon technique as described in [1] . Chest wall mechanics (tissue resistance and chest wall compliance) was calculated from lung mechanics and total respiratory system mechanics as described in [2] . Results: We found a decrease of chest wall compliance (Cw) to be the main reason for episodes of sudden airway pressure increase while lung compliance (CL) remained unchanged. The decrease of C w can be inter- Gil Cano A, San Pedro JM ~, Sandar D, Herntndez .1, Carrizosa F, , Herrero A. Emergency and Intensive Care Department, Hospital of JEREZ, SPAIN OBJECTIVE: 1) To determine the incidence of hypoteasion (H) associated with emergency intabatian of mechanical ventilation, and 2) to establish its relaUonship with respiratory mechanics (RM) and arterial blood gases. mechanical ventilation performed in the Emergency Room, in a prospective eans~eative manner, were evaluated. Data collected included patient demographics, diagnoses, blood pressure and arterial blood gas levels before and at~er intabatian, and P_M, including calculated pulmonary end-inspiratory volume above Functional Residual Capacity (VEIc) and calculated dynamic hypetinflatien (DHc). All patients received midazolen and aWaanrinm to facilitate tracheal intubatien and RM measurement. Hypotension was defined as a decrease in systolic pressure higher than 40 mmHg or an absolute decrease in systolic blood pressure below to 90 mHg within 1 hour of intabatian. 14 patients were excluded because met at least one of the following exclusion criteria: preexisting shock or H (8), cardiac arrest (5) .1 There weren't any association between PEEPi or other airway pressures (Paw) and H, but calculated pulmonary volitmes had tendency to be larger in patients with H (p < 0.1). High PaCO 2 before lrasheal intubatian (87.4 4-9 mmHg) with a quickly decrease alter starting mechanical ventilation was a usual finding (p < 0.01) in patients who developed H. Paw. 3) Thexe was a good relatienship between H and high arterial PaCO 2 before traqueal intahatian and its fast "washing" with mechanical ventilation. 4) Because CAO patients had the highest incidence of H, controned mechanicel hypoventilatien driven by PaCO 2 changes and pulmonary volumes monitoring instead Paw, should be attempted in these patients to avoid this cemplication after tracheal intubatiert. Introduction: The endotracheal tube (ETT) and demand valve devices cause an added work of breathing (WOBadd), which is the work necessary to overcome the resistive load of the ETT and the breathing circuit (1). Application of IPS has been shown to partly compensate this added work (1). Since tbe amount of WOBadd is flow dependent, a fixed IPS is not adequate to completly compensate the WOBadd (2). Therefore, ATC has been developed as a new form of assisted spontaneous breathing (3), which provides a flow-dependent pressure support. Thereby, it theoretically should compensate all the WOBadd due to the tube. The purpose of this study was to evaluate the reduction of WOBadd with IPS and ATC for different ETT. Methods: A mechanical lung model (LS 4000, Dr*alger, Liibeck, FRG) was used to generate a constant spontaneous breathing pattern. The LS 4000 was connected to an artificial trachea (AT, 10 cm long, 22 mm ID). The AT was intubated with three different tubes of 7.0, 8.0, 9.0 mm ID and connected to an Evita 2 ventilator modified to provide ATC as an option (Dfager, Liibeck, FRG). Flow and airway pressure were measured between the y-piece and the ETT for four different modes of ventilation: CPAP, IPS of 5 and 10 cm I420 and ATC all with a PEEP of 10 cm H20. The tracheal pressure (Ptrach) was measured in the AT. Total WOBadd was calculated as the area subtended by the Ptrach-volume curve below PEEP. Results: The results for total WOBadd in nd/1 are shown in the figure for the three different ETT: breath/mln, S=success, F=failur% *~p<.05, **-p<01, NS = non significant, F versus S Neveltheless, in 5/26 patients, invasive ventilation was necessary in mean 12.6_+12 hours after beginning of FMPSV. There was no significant difference between the two groups (Success, Failure) in following parameters : sex, age, previous histoly, medical treatment, SAPS1 & 2, clinical signs (RR, SpO2, Heart Rate, Blood Pressure, Glasgow Score...), radiological and echocardiographic findings and standard biological parameters. Only two parameters were related with Failure : 1.a low value of PaC02 on admission until the patients were intubated. 2. an increased level of CPK in relation with an acute myocardial infarction (4/5 cases in the failure group, vs 3/21 cases in the Success group, X~(with continuity correction) : p<.05). Conclusion : FMPSV is a noninvasive, safe, rapidly effective method of treatment in ACPE, which may avoid tracheal intubation. Further studies are necessary to precise if association of ARF and low PaCO2 (<35mmHg) and/er acute myocardial infarction represents an indication of immediate invasive ventilation. Introduction: Since the added work of breathing (WOBadd) imposed by the endotracheal tube (ETs and the breathing circuit is regarded as an important contribution to the total work of breathing, considerable effort has been tmdettaken to compensate for this added work. IPS has been fotmd to decrease the WOBadd imposed by different ventilators (1, 2). Because of the flow dependent pressure drop across the ETF the tracheal pressure (Ptr) should be measured to estimate the total imposed WOBadd (WOBtut) (3, 4). The aim of this study was to assess the circuit imposed work (WOBcirc) and WOBtot (including ETT) for different demand valve ventilators during CPAP and/PS. Methods: A mechanical lung model (LS 4000, Driiger, Lfibeck, FRG) generated a constant spontaneuus breathing pattern. The LS 4000 was connected to an artificial trachea (AT), intubated with an 8.0 nun ET]', end connected to one of four ventilators (Servo 900C and Servo 300, Siemens,-Elema, Sweden; Evita2, Driiges, Liibeck, FRG; PB 7200ae, Puritan Bennett, Carlsbad, USA). Three different modes of ventilator settings were tested (CPAP, IPS 5 and 10 mbar; trigger set at maximal sensitivity, PEEP always 10 mbar). Flow and airway pressure (Paw) were measured between the y-piece and the ETr; tracheal pressure (Ptr) was measured in the AT. WOBtot was calculated as the area under the Ptr-volume curve below PEEP, WOBcirc was calculated as the area under the Paw-volume curve below PEEP. Results: In the Foti G., Patroniti N., Cereda M., Sparacino ME., Giacemini M., Pesenti A. Inst.of Anesth.and Intensive Care-Univ.of Milan -SGH MONZA I Aim of the study was to assess Cpl,rs measurement obtained by the airway occlusion method during PSV. We therefore studied 31 paralyzed CPPV ventilated ALI patients (Lung Injury Score =2.25• that were weaned to PSV. We performed end inspiratory and end expiratory airway occlusions using the hold function of the ventilator (Siemens Serve 900C), first during CPPV and then within the 24th PSV hour. Airway pressure and flow signals were recorded (CPI00 BICORE) for subsequent analysis. An airway pressure plateau was defined as a 0 flow tracing in which airway pressure was stable for at least 0.25 sec. End inspiratory (Pel,rsi) and end expiratory (Pel,rse) recoil pressures were then measured as the mean airway pressure during plateaus. Cpl,rs was computed as TV/ (Pel,rsi-Pel,rse i) Cpl,rs can be adequately estimated during PSV using the airway occlusion method; 2) During PSV inspiratory plateaus are longer than the expiratory ones; 3) The length of plateaus is negatively affected by the respiratory drive. Foti G., De Marchi L., *Tagliabue M., Gilardi p., Giacomini M., Sparacino ME., Pesenti A. Inst.of Anesth.and Intensive Care,-Univ.of Milan *Dept.of Radiology-SGH Monza I We retrospectively compared CT scan and gas exchange findings between a group of patients successfully weaned from VCV to PSV (group S = ii patients) and a group who failed the weaning (group F = 6 patients). We selected 17 ALI patients (LIS=2.5• in VCV mode who had available a chest CT scan performed within 4 days from the weaning trial. A PSV trial was began as soon as the patient reached hemodynamic stability and a PaO2 >80 mmHg, irrespective of Fie2 (PEEP <15 cmH20). Maximum PSV level was < (Pel,rs-PEEp) measured during VCV, where Pel,rs was the respiratory system elastic recoil pressure at end inspiration. PSV ventilation was considered successful if a respiratory rate <40 bpm, an increase in Fie2 lower than 0.2 compared to VCV, a Pace2 increase <20% of VCV value and hemodynamic stability were maintained during the next 48 hours of PSV. If any of these conditions was not met the trial was declared a failure. Interdisciplinary Critical Care Unit, Regional Hospital Lugano-CH *Surgical Critical Care Unit, University Hospital, Geneva-CH Objective: To assess the degree of correlation of cardiac output measured by thoracic electrical bioimpedance and thermodilution in mechanically ventilated patients with different levels of positive end-expiratory pressure (PEEP). Methods: Prospective study with 10 ventilated patients, 7 after head injury and 3 with postoperative sepsis, with normal cardiac output: Simultaneous determination of cardiac output by thermodilution and thoracic electrical bioimpedance performed with different levels of PEEP (0-5-15 cm H20). Results: Cardiac output measured by thermodilution during sequential increment of PEEP did not vary: 7.3 + 2.5 for PEEP 0, 7.4 + 2.7 for PEEP 5 and 6.9 + 1.7 L/rain for PEEP 15. Simultaneously the bioimpedance device recorded a significant increase in cardiac output from 4.4 + 1.3 for PEEP 0 to 6.0 + 1.9 L/mi for PEEP 15. (P < 0,05). Conclusion: Cardiac output measured by bioimpedance cannot replace the invasive thermodilution methods of cardiac measurement output during mechanical ventilation with PEEP. We also isolated a subset (H) of 12 patients who had been hypercapnic (PaCO2>50mmHg) for at least 3 days (range 3 to 60 days) before the end of CV. The PSV trial was started as soon as PaO2 was > 80 mmHg, irrespective of Fie2 and with PEEP < 15 cmH20 and the PSV level had to be < (Pplateau-PEEP) as measured during CV. PaCe2, pHa, base excess (BE) were collected before discontinuation of CV and on the ist day of PSV: 05) . 2) Weaning is more difficult in pts with head injury(p30 (p0,6 (pIO cm H20 (p30 need longer duration of MV (p30 (p60 years than in pts<60 years (p 8 cm Hz0 , Fit 2 > 0.6. A total of 43 patients matched these criteria, 27 males and 16 females with a median age of 44 (17-72) years. Seventeen suffered from severe trauma. CHFJV was started following a median period of 3 (1-22) days of conventional mechanical ventilation. Prior to CHFJV ventilation parameters expressed as median were the following: Fit 2 0.8, Pao2/Fio 2 78, PEEP 12 cm H20 peak airway pressure (PAP) 48 cm H20. CHFJV consisted of high frequency jet ventilation with a frequency of 100 to 300 breaths/minute, driving pressure of 1.8 to 3.5 arm, and inspiration time of 20 to 30 percent, superimposed on the whole cycle of conventional mechanical ventilation with a frequency of l0 to 20 breaths/minute and tidal volumes of 100 to 400 ml. Results: following two days of CHFJV 31 of 43 patients showed an improvement of ventilatory parameters; PEEP could be reduced to < 8 cm H20 in 14 patients, the PAP was decreased with > 5 cm H:O in 20 patients, FiO 2 could be reduced to < 0.6 in 27 patients and finally the median PaO2/FiO 2 ratio changed from 78 to 133. During CHFJV 23 patients died, 4 of respiratory failure and 19 due to multiple organ failure, 6 died within two days of CHFJV. The median duration of CHFJV in survivors and nonsurvivors was 6 days in both groups. Conclusions: our data show that with CHFJV in the majority of patients with SRI who are refractory to conventional mechanical ventilatior" the ventilatory parameters can be improved. Backeround and Obiectives: Although ventilation with PEEP above the inflection point (Pinf) has been shown to reduce lung injury by recruiting previously closed alveolar regions, it carries the risk of hyperinflating the lungs. In the present study we set out to develop a new strategy to recruit the lung during ventilation with small Vt, while maintaining PEEP levels as low as possible. We hypothesized that if the lung was recruited with a sustained inflation (SI) to total lung capacity, recruitment would be maintained as long as the PEEP level was higher than the critical closing pressure of the lung, as observed on the deflation limb of the PV curve (AJRCCM 1995; 151(4) :A432). The purpose of this study was to examine the hypothesis that a strategy using SI and a PEEPPing group 2: PEEPPin~ 2 _Objectives-This report is presenting the results of the clinical study for using EEG examination as a method of the evaluation of patients ability for weaning. Methods: The study incllJqles 42 EEG examinations with Fourier spectral analysis' of 37 patients ~vith respiratory insufficiency and prolonged control mechanical ventilation (CMV). All patients have had a-rhythm of EEG before weaning. We have followed respiratory rate, tidal volume, respiratory pa{tern, end-tidal CO2 and blood gases during weaning. Results: 13 patients had invariable EEG activity or short 13-waves period (till one hour). The weaning of this patients was fast arid sucsessful. Other 24 patients have had a decreasing of a-activity, an appearence of 13-waves for an hour and more, a short episodes of A-and e-activity. After that this patients had gas exchange and respiratory disorders with regression of the weaning right up to CMV. Conclusion: EEG could be used as a method of the evaluation of patients ability for weaning from CMV. Some EEG signs shows the overstrain of compensatory systems before the change to the worse of gas exchange and respiratory pattern. S. Elatrous, P. Aslanian, D. Touchard, D. Corsi, H. Lorino, L. Brochard. Medical Intensive Care Unit, INSERM U 296, HOpital Henri Mender, Cr~teil, France. In Vitro Comparison of flow triggering (FT) systems demonstrated advantages compared to pressure triggering (PT) systems for some ventilators (Puritan Bennett 7 200) but not others (Siemens Serve 300). We studied the two types of systems in two groups of 8 patients mechanically assisted with pressure support ventilation (15 + 6 cmH20). In the first group (PB 7 200) the effort of breathing, assessed by the esophageal pressure time index, was significantly lower with the FT than with the PT (139 + 40 cmH20.s/min -1 VS 158 + 32, p< 0.05). By contrast no significant difference appeared in the second group (Serve 300), as predicted by the bench study despite marked interindividual differences (134 + 55 cmH20.s/min -1 VS 160 + 61, p = 0.1). We conclude that 1) rigorously performed bench studies can predict in vivo effects, 2) mild advantages can be found for the new triggering systems on some ventilators. Objectives: Pressore-volume curves (PV) of the respiratory system is of interest for the determination static compliance (Cs0, lower (LIP) and upper (UIP) inflection points which indicate zones of airway recruitment and overdistension. This study aimed to compare an "automated low flow inflation" method (ALFI) to the reference occlusion (OC) method. The ability of the former method to identify Cst, LIP and UIP was tested in ICU patients. Me,otis: 16 (8 ARF and 8 ARDS) sedated paralysed patients were studied using a Serve 900C ventilator linked to a computer which automatically forced the ventilator to insufflate at a low constant flow a velum up to 1500-2000 ml or a maximum Paw of 50 cm H20 (ALFI). The quasistatic elastic pressure (Pel,qs0 was obtained by subtraction of the resistive pressure of tubing and patient and related to volume for calculation of compliance Cqst. For OC tidal volumes (V0 from 50 up to 1500-2000 ml were followed by a 3 s post-inspiratury pause for determination of static Pal (Pel,st) in relation to volume. Compliance was defined from the linear part of the P/V curves. LIP and UIP were defined from the consistent deviation of P/V data from extrapolated the linear part. ~,~111I~: In ARDS, mean Cst was 27.9 + 3.5 and Cqst 29.7 + 3.9 ml/cm H20 (us), LIPst 5.2 + 5.0 and LIPqst 7.0 + 4.6 cm H20 (us), UIPst 23.1 + 10.8 and UIPqst 26.0 + 5~4 cm H20 (us). Nosocomial pneumonias (NP) are frequent and often unsuspected during ARDS (Bell, !983). In the present study, we evaluated prospectively the onset of NP during severe ARDS (group B of the European Study). Patients and methods: The charts of 15 patients with severe ARDS have been prospectively recorded. A plugged telescopic catheter (PTC) specimen has been systematically performed every 48 hours, for quantitative bacteriological analysis. The diagnosis of NP was defined by a number > 103 colony forming units / ml. Results: For the 15 patients studied, the mean SAPS score (+ SD) was 16+_2, the initial PaO2/FiO2 ratio was 100-&-_35, the duration of mechanical ventilation (MV) was 19+9 days. The mean delay before the onset of the first NP was 8.6+5.6 days (5-12), and the mean PaO2/FiO2 ratio was 110+-28. Respiratory symptoms (purulent aspirates, new pulmonary infiltrates, or gazometric changes) were present in 80% of the patients studied. Alteration of gas exchange was present in 8 of the 15 patients (7 NP) . A new pulmonary infiltrate was present in only 1 NP (10%). An increase of fever was noted in 6 patients, an increase of leukocytosis > 20% in 8 patients, an increase of volume and purulence of sputum in 3 of the 10 patients with NP. The degree ofgazometric worsening (PaO2/FiO2 before NP minus PaO2/FiO2 during NP) during the first episode of NP was 44+17 mmHg. Excluding the bacteriological criteria of NP, the number of criterias of NP present was 1 in 1/10 patients, 2 (5/10), 3 (2/10) or 4 (2/10). Two patients only had a pulmonary colonization (PTC: < 1102 cfu / ml) before the first episode of NP. The incidence of NP is high (53%) during severe ARDS. The first episode occurs in average:at the 9 th day, and is the cause of a severe hypoxemia (PaO2/FiO2 110) . The onset of a NP may contribute to the high mortality rate observed in our patients (93%). Each worsening of hypoxemia during severe ARDS should induce to suspect a NP. respiratory system during mechanical ventilation. The me~hod quantifies the dissipative energy consumption of the respiratory system in terms of energy loss AEk, inefficiency ~k~ and respiratory dissipative resistance Rk~ over a given partition of the tidal volume. The method can be applied in intensive care units with no interference to ventilatory support. It allows for monitoring the combined effects of inhomogeneities, non-linearities and visco-elastic effects, that are subject to change in the respiratory system. The method is studied on pigs~ in the presence of a log-dose response curve of methacholine (MCh) induced disease. In healthy pigs~ we find a mean value of energy loss, AE, of 0.27 • J/l, a mean value of inefflency, ~ of 0.25 ~=0.05 and a mean value of resistance, 7~, of 4.40 • cm H20 s/1. The respiratory resistance, Rk, shows a variation over the partition of tidal volume with ARmax ----3.90 • 0.66 cm H20 s/l. During methacholine provocation~ AE rises more than five-fold up to 1.48 • J/l~ doubles to 0.54 • and T~ increases to a maximum of 22 • cm H20 s/l, with ARmax : 15.1 • 7.0 cm H20 s/1. The variation in Rk becomes more pronounced with higher doses of methacholine. Methods: 10 ARDS patients were prospectively studied. Initially they were ventilated in the AMV (assist mechanical ventilation) mode with the settings prescribed by their primary physician. After stabilization, ventilatory gas exchange and hemodynamic variables were determined. Patients were Then ventilated in the MRV (mandatory rate ventilation) mode with 20 breaths as the target rate. In MRV the target rate is set and the ventilator autoregulates the pressure support level delivered ~o achieve this rate. After stabilization, the measurements done on AMV were repeated. Finally, patients were sedated and paralyzed and ventilated in CMV (control mechanical ventilation) with the ventilatory variables they had during MRV. Measurements done in AMV and MRV were repeated and respiratory mechanics were assessed with the constant flow end inspiratory occlusion method. Results: Two groups were recognized based on their response to MRV. tn group 1 patients responded to MRV by decreasing their V and increasing the T/T T ratio. VE, VO 2, and AaDO 2 decreased while PaCO 2 increased and tda vo ume and CO remained unchanged. On the contrary, in group 2 V, Vr and VE increased; Ppeak and Trr T remained unchanged, PaCO~ decreased while VO 2 and AaDO 2 increased with constant CO, The pressure support level needed to achieve the target rate was much lower in group 1 than in group 2 (19,8-+1.3 vs 29.4_+2.0). Obiectives : In the newly developed mode of ventilatory support ,,Automatic Tube Compensation" (ATC) the ventilator compensates for the flow-dependent pressure drop across the endetracheat tube (ETT) thus allowing ,,e]ectronic extubation". The aim of the study is to investigate whether healthy subjects perceive ATC in inspiration (ATC-in) and in expiration (ATC-in-ex) and whether ATC provides an increase in subjective comfort compared with the conventional assisted spontaneous breathing mode (ASB). Methods : Healthy volunteers (no preceding lung disease, non-smokers, male, 20-40 years)breathed spontaneously through an uncut ETT of 7.5 mm ID via a mouthpiece. The ETT was connected with a prototype ventilator Evita 2 modified by the manufacturer (Drfiger, Lebeck) for ATC. Flow and airway pressure were measured at the outer end of the ETT. Three ventilatory modes, (1) ASB (10 mbarover 5 mbar PEEP), (2) ATCin, (3) ATC-in-ex were selected in random order. Immediately following the transition from one mode to another the volunteers answered by hand sign how they perceived the new mode compared with the preceding mode: ,,better" (+1), ,,equal" (0) or ,,worse" (-1). Inspiration and expiration were investigated separately by presenting 120 mode transitions (in total; including ,,placebo" transitions). Results : The difference between ATC and conventional ASB is perceived in inspiration and in expiration. ATC is positively judged; ASB is nega ively judged. The diagrams show mean values _+ SD of five volunteers investigated up to now. The new mode ATC is perceived as an increase in subjective comfort. Our explanation is that ATC preserves the natural breathing pattern better than conventional ASB. Objectives: To determine the role of cerebral vasoconstriction in the delayed hypoperfusion phase in comatose patients after cardiac arrest. To correlate the results with indices of cerebral oxygenation and the levels of several vasoactive hormones in the jugular bulb. Methods: In comatose patients after cardiac arrest we measured the pulsatility index (PI) of the medial cerebral artery by Transcranial Doppler Sonography. The PI is a reliable indicator of cerebral vascular resistance. We also sampled blood from the jugular bulb and measured cerebral oxygen extraction ratio and jugular bulb levels of endothelin, nitrate and cGMP. The first measurement was done within 4 hours after cardiac arrest and repeated 3, 6, 9, 12, 18 and 24 hours later. Results: We studied 10 patients, 6 females, mean age 64,1+13,7 years. The PI decreased s!gnificantly between th~ first and the last measurement from 1.86 _+ 1.02 to 1.05 + 0.22 (p = 0.03). Cerebral oxygen extraction ratio decreased also from 0.39+ 0.13 to 0.24 + 0.11 (.p = 0.015). Endothelin levels were high, but didn't change during the studied period. Nitrate levels varied in a wide range, but didn't change significantly. However, cGMP levels increased significantly from very low levels in the first measurement to very high levels 24 hours later, rasp. 2.95 pmol/mL (median; 25th 2.48-75th 5.43) and 7.5 pmol/mL (median; 25th 6.2-75th 14.00) (p = 0.02). Eighteen and 24 hours after the first measurement we found a strong correlation between PI and Cerebral Oxygen Extraction Ratio ( r = 0.64, p = 0.05 and r = 0.76, p = 0.01). We.also found 12 hours after the first measurement a significant correlation between PI and cGMP levels ( r = 0.69, p = 0.03). We found no correlation between PI and endothelin or nitrate levels. Conclusion.~; Our results show a high cerebral vascular resistance in the first few hours after cardiac arrest, gradually decreasing during the next 24 hours. This is accompanied by an initially high cerebral oxygen extraction ratio and low cGMP levels, suggesting that the cerebral vascular resistance is induced by active vasoconstriction because of insufficient cGMP levels, leading to a decrease in cerebral blood flow and a compensatory ~ncrease in cerebral oxygen extraction. Objectives: Sudden cardiac arrest is a major cause of mortality in western countries accounting for over half of all cardiovascular deaths. In most cases the mechanism of death is prolonged cardio-circulatory arrest due to ver:tricular fibrillation (VF) preceding final asystole. Recurrent syncopes due to idiopathic VF with good neurological prognosis have been reported in patients with and without cardiac etiology (1,2). In the past measurements of cerebral hemodynamics have been repeatedly done in humans during CPR, but until today no studies of cerebral blood flow velocity (CBFV) have been reported during controlled cardiac arrest in humans not under-going CPR. It was the purpose of our study to evaluate the acute hemodynamic effects of untreated VF on CBFV. Methods: After approval by the local University Ethics Comittee, five male patients aged 34-48 years without evidence of cerebral disease were investigated during VF while undergoing implantation of a Pacer Cardioverter Defibrillator System (Model 7219D; Medtronic| A standard anaesthetic regimen was used (propofol, fentanyl). After implantation of the automated cardiac defibrillator VF was induced by electrical countershock to test effective sensing, pacing, and defibrillation. To measure cerebral blood flow velocities (CBFVMcA) the Doppler probe was placed above the zygomatic arch between the lateral margin of the orbit and the ear and directed towards the M1 segment of the middle cerebral artery (MCA). Results: A total of 12 phases of VF were investigated. Duration of VF ranged from 6 to 26 seconds, with CBFVMc A (mean_+SD, cm sec -1) flow pattern changing from pulsatile to laminar flow immediately after onset of VF. Conclusions: The underlying mechanism of the laminar cerebral blood flow observed during VF in our patients is uncertain, but it may provide insight into the prognosis of patients with idiopathic VF. Theoretically, the laminar cerebral blood flow observed in our pulseless patients may provide a substantial amount of cerebral perfusion even during clinical cardiocirculatory arrest Objective: To investigate whether the intensive care nursing staff can inflate more accurately a specific air volume with the laerdal resuscitation bag when they receive feedback after each inflation about the delivered volume compared to no feedback. Method: 42 ICU nurses were asked to inflate a testlung model 10 times with a specific air volume (600 ml, 800,ml or 1000 ml) under three different conditions (normal, decreased compliance and increased resistance) without and with feedback. We measured the mean absolute difference from the specific airvolume after each ten inflations. Results: The largest absolute difference was found when ICU nurses inflated 600 ml (250 ml). The mean inflated volume for this group was 843 ml. When the ICU nurses had to inflate 800 ml the mean absolute volume difference was 181 ml with a mean inflated volume of 913 ml. Inflating 1000 ml produced an absolute volume difference of 131 ml with an mean inflated volume of 1042 ml. The absolute volume difference decreased when the compliance of the testlung was decreased and even more when the resistance of the used endotracheal tube was increased. When the ICU nursing staff received volume feedback after each inflation the mean absolute volume difference was reduced between the 42 ml and 66 ml for all specific air volumes. 42% of the last 5 inflations with feedback were significantly smaller than 50 ml from the specific air volume (P < 0.05). Conclusion: The majority of nurses overinflated the specific air volumes. The largest over inflation occurred when 600 ml and the smallest when inflating 1000 ml. When nurses were provided with volume feedback the performed significantly better. We concluded that ICU nurses are not able to inflate a specific air volume with the laerdal resuscitation bag without receiving volume feedback. Feedback is desirable in order to reduce the volume trauma. Objectives: A pro_found impairment in systolic and diastolic myocardial function following successful cardiopulmonary resuscitation (CPR) has been demonstrated by using Langerdorff method in rats. In the present study we have investigated post resuscitation myocardial dysfunction in a porcine model of CPR. Methods: Ventricular fibrillation (VF) was electrically induced by alternating current applied to the ep{cardium of the right ventricle in 11 domestic pigs. Following 4 rain of untreated VF, precordial compression and mechanical ventilation was initiated and maintained for 8 min. Electrical defibrillation was then attempted and 6 of 11 animals were successfully resuscitated. Results: Following successful cardiac resuscitation, stroke volume index (SVI) decreased from prearrest value of 1.13 ml/kg to 0.74 ml/kg (p<0.05), and left ventricular stroke work index (LVSWI) from 1.57 to 0.77 mmHg,ml/kg (p<0.05). Both SVI and LVSWI remained depressed for another 3 hours. These decreases were associated with increases in heart rate from 145 bpm to 185 bpm (p<0.05). No significant changes from baseline in mean arterial pressure, mean pulmonary pressure, right atrial pressure and pulmonary artery wedge pressure were observed. PREHOSPITAL RESUSCITATION EFFORTS C. K6ppel. G. Fahron, H. Lufft, A. KrUger, C. Th(Jrk, F. Bertschat, F. Martens Dept, of Nephrology add Medical Intensive Care, Virchow-Klinikum, Humboldt-Universit~t, D-13353 Bedin, Germany Obiective: The success rate of prehospital resuscitation in patients with cardiocirculatory arrest in an Emergency Medical System (EMS) may reach 30 -40% depending on the time of calling the EMS, the distance to cover by the emergency ambulance and the training of the emergency physician and his staff. In the Berlin EMS, which is associated with the Berlin Fire Brigade, the time between alarm and arrival at the scene ranges from 2 -31 min, mean 8 min. Resuscftation is based on the Advanced Cardiac Life Support (ACLS) according to the guidelines of the American Heart Association. If resuscitation efforts fail to restore circulation, they are terminated after 30 -60 min, depending on duration of cardiocirculatory arrest, pre-existing disease, age, absence of an even transient response to CPR. However, there is a lack of practical criteria for termination of CPR in individual decision making. Patients: We report 5 cases of prehospital CPR with primary asystolia terminated after 45 -60 rain of frustraneous CPR efforts including highdose epinephrine and dopamine. Results: After termination of CPR, the ECG monitor remained connected and showed permanent asystolia in all patients while the emergency physician completed his records. Spontaneous resumption of respiration and circulation was observed in these patients after 2 -5 min and CPR efforts were immediately resumed, Nevertheless, 3 of the patients died at the scene, while 2 could be hospitalized with stable circulation. One of them died 3 hours after admission to the ICU, the other survived for 3 weeks in a vegetative state. Spontaneous resumption of circulation and respiration is most likely due to the development of extreme hypercapnia and acidosis, which -at least in some patients -seems to be a stronger stimulant of the circulatory and respiratory brainstem centers than CPR with high-dose catecholamines, Conclusion: Because of the legal and ethical implications of this rare phenomenon, emergency physicians should continue ECG monitoring for at least 5 rain. after termination of CPR efforts. Pulmonary artery catheterezation is used for patient's monitoring [1]. We reported our results on such monitoring in 1969 [F.CoaoBbeB,r.fe6enb~-Kap~MonorM~,1969,N7,p.28-39] .However not all of the received criteria assessments meet demands that are necessary for early diagnosis of critical states. Here we report the data on PO2,PCO2 (mm Rg),SO2,pH levels in femoral [aF) and pulmonary (aP) arteries blood, as well as on summary gas pressure (SGP) calculated from PE=(PO2+PCO2) in mm Hg in aP blood. These data were derived from:i)86 subjects free of cardiovascular pathology according to catheterization data during their spontaneous air breathing (N group in aP blood appears to be a measure of adequacy ratio between PC2 and SGP in aP blood during air breathing; partly its characteristics and variations ranges are presented earlier [2J. In control group it is equal to 1,91• mm Hg. Tests on SGP neither exclude nor substitute conventional (PC2 and PCO2) tests, but rather include them as a part choosing only additive characteristic -pressure. They appear to be a part of general system of human metabolism regulation by pressure (arterial,venous,intracardiac, tissue,liquor,onco-osmotic,etc ietraabdeminal pressure produces perturbations of cardiac, pulmonary, and renal physiology. This most often occurs fonowing eeliotomy for peritonitis or intestinal obstruction; bowel edema and distention prevent wound closure without unacceptable compromise of blood pressure or pulmonary compliance. A variety of temporizing measures have been reported for managing Wounds that cannot be closed: 1) using towel clips to reapproximate skin only, 2)i sewing silastic, Marlex or other prosthetic grafts to the fascia to "enlarge" the peritoneal cavity, 3) using loosely tied retention sutures for partial closure, 4) simply packing the wound without attempts at c~osure. These techniques either traumatize the abdominal wall (complicating definitive closure), expose the bowel to damage, or allow excessive loss of fluid and heat. Since 1989 we have evolved a suturelees technique which permits the abdomen to be partially closed in a quick, safe, sterile, sealed, atraumatic fashion -while providin! decompression of unphysiologic intraabdominal pressure. Methods: Whenever possible omentum is interposed between bowel and the open incision. Viscera are covered by a layer of sterile, non-reactive plastic, placed deep to the fascia and extending we~t beneath the edges. Sump tubes are placed above the plastic and covered in turn by two layers of an adhesive plastic drape which sticks to the skin and seals the wound in all directions, The patients remain intubated and paralyzed. Results: We have used this technique in a total of 27 patients, four of whom suffered from Compartment Syndrome. All of the latter were males and ranged in age from 19 to 51. All four showed immediate physiologic improvement. All four incisions were eventually closed without complication. One Compartment Syndrome patient died 4t days later of multiple organ failure. There were no complications related to the closure technique in any of the 27 patients. Conclusions; 1. Selected patients with Abdominal Compartment Syndrome will benefit from decompression using this temporary sutureless technique. The technique a) is quick, safe, sterile, sealed, and atraumatic, b) minimizes loss of fluid and heat, c) facilitates eventual definitive abdomina| closure. Although M. Brunner M. Mitllncr Objectives: To determine incidence and predisposing factors for cardiac arrest occurring during the first 24 hours after open heart surgery. Methods: The study included patients who, following open heart surgery, had adequate cardiac function and in whom cardiac arrest was not anticipated. All data were prospectively recorded and analyzed. Results: From 12/1993 through 3/1995, 2140 pts underwent open heart surgery at our hospital. Of th~se, 23 pts (1%) (age 65_+9 yrs) had a cardiac arrest during the first 24 hours after transfer to ICU. They were operated on for coronary artery bypass grafting (CABG) (17 pts), valve replacement (VR) (3 pts), CABG and VR (2 pts) and aortic aneurysm (1 pt). The preoperative ejection fraction was 44_+12% whereas bypass and aortic cross-clamp time were 127+70 and 72+42 rain, respectively. Prior to arrest, they had a cardiac index of 2.23_+ 0.5 L/min/m 2 and were receiving 1.3+1 inotropes. Arrythmias leading to cardiac arrest were ventricular tachycardia/fibrilation (10pts) and bradyarrythmia (9 pts). Closed-chest CPR was initially performed on all pts and was followed by open-chest CPR in 12 pts. Eighteen pts (78%) survived to ICU discharge. Causes of arrest included perioperative myocardial infarct (t2 pts, 52%), tamponade (3 pts, 13%), rupture of the proximal vein gra& anastomosis (1 pt, 4%), graft occlusion (2 pts, 9%); no cause was found in 5 pts (21%). Conclusions: Postoperative cardiac arrest in stable cardiac surgery pts is relatively infrequent (-1% incidence) and is associated with a high survival rate following successful CPR. Perioperative myocardial infarct is the most common predisposing factor. Group ~Deptof Anaesthesia and Intensive Care, Semmelweis Univ. Medical School, 2 Buda Military Hospital Intensive Care Unit, Budapest BACKGROUND: When a cardiac arrest occurs in-hospital, the outcome can be improved by a higher quality of basic life support provided by the witnessing health care workers until the code team arrives. This basic life ~pport (BLS) should include the best available method for airway management as well. Since not all medical staff are ready for carrying out endatracheal intnbation, we investigated the effieacy of the use of different airway management methods during BLS. METHODS: We have investigated the efficacy of airway management of 25 doctors and 25 nurses from different hospital wards: internal medicine, department of surgery, trauma, urology and gynaecolagy. Comparing the bag-valve-mask, laryngeal mask and the endotracheal intubafion, we have measured the following parameters: time needs for correct application (sec.), number of incorrect applications (out of ten trial), efficacy of artificial ventilation provided by the device. We used a computerised ALS trainer manikin for the evaluation of the performance. Total performance score was created after the measurement between 0-10. After the first screening we held a 2 x 2 hours training. 8 doctors and 8 nurses were trained for the endotracheal intubation (Group IT1, 1T2) , 8 doctors and 9 nurses were trained to use the laryngeal mask (Group LM1, LM2) . All respondent were trained to use the bag-valve-mask device. 1 day, 1 month and 3 month after the training we have carried out retention study using the same method. RESULTS: We have found that the efficacy of the artificial ventilation using the above mentioned devices were poor before the training. The average after-training performance scores of the groups are presented in the table below. (BLS) should be initiated by the witnessing health care professional. The CPR Study introduced a multi level Code system, which means BLS included sophisticated airway management, early defibrillation and early epinephrine administration provided before the Code Team arrives. Our previous studies confirmed a poor level of CPR performance and a high demand for CPR training among health care professionals. METHOD: We established a CPR Training Course Centre, Where doctors and nurses are being trained for in-huspital basic and advanced life support. 3 x 6 hours of training were held. After the theoretical introduction a step-by-step training method ws used for trainees to be familiar with all sequences of basic and advanced life support. Then we synthetised all separated sequences. Afterwards, a r01e play of rescue groups was taken in simulated situations. We also trained the multi level alarm system fur the in-hospital resuscitations. After the training all respondents had to sit for examination. The quality of performance was scored and compared to our previous results. Semi-structured interviews were carried out before and aider the training among all respondents to collect information about the course. RESULTS: We have found a remarkably high interest among doctors and nurses in our CPR Training Courses. It was very important to use proper equipment for the training: audio-visual training facilities, computerised ALS Trainer manikin, manual and automatic defibrillator units. The evaluation of the examination held immediately a~er the training course showed a significant higher quality of performance than before the training. The self.-eonfidence of the trainees for initiating and carrying out resuscitation had increased. Their overall feeling about the course was positive and 100% responded the course "very useful". 73.6% of doctors and 79.4% of nurses claimed fur regular training facilities with ALS trainers, CONCLUSION: The CPR training for health care werkers is mandatory including the training of sophisticated airway management and use of elad~l~ills~tt~r Wlaa ~en ~r a~ti~atir ~nel r rm~a'*h*nr m~thnd for training will improve the efficacy, the satisfaction of trainees, therefore their compliance for further co-operation will also increase. S 144 Objectives: The effect of reinfusion in emergency surgery and gynecology. Methods: We had an experience of autologous blood transfusion in 22 patients whom was produce t an emergency surgical or gynecological interventions in occasion with break tubal pregnancies (45.5 %), penetrating abdominal wounds with injuries of mesenterial vessels (22.8 %), injuries of the liver (9.1%), blunt abdominal trauma with lien ruption (22.8 %). In 27.3 % patients had the previous somatic pathology. Blood loss volume was 1500-4500 ml, & the reihfuside blood volume was 500-2000 ml, consisting 30-70 % of blood loss. It was needn't to fransuse donor blood in 18.2 % in further but 300-2500 ml of contanined erythrocytes were frasfused for supporting of Hb concentration on the 80 g/L (8 g/dL) rate at the other patients with isovolemie hemodiluttion. Results: The Arterial Blood Pressure fast stabilisation on the perfusion level had noted after reinfusion, excluding the case, when the volume of reinfused blood had conisted just 40 % of blood loss at the patient with massive blood loss. Complications have noted in two cases. One patient with slash wound, injury of arteria gastrica dextra and total blood loss of 4500 ml, has an episode of asystoly, DIC (Disseminated Intravascular Coagulation) syndrome, acute renal failure, and acute pancreatitis that we haven't connected to reinfusion. All the complications were successfully corrected and at thirty first day patient with subcapsular wound of the lien that has happened 14 days before complicated with external rupture of the capsull & massive intraabdominal bleeding, has the hemolytical shock, DIC Syndrome, Acute renal failure developed after reinfusion. He was died. All another have no complications. Posthemorrhagic anemia had corrected rapidly than in case when hemorrange corrected exclusively by donor blood. Conclusions: We consider that simplicity, accessibility, high effectiveness, quite well further results of blood reinfusion, except the case of blood reinfusing that was for time-expired out of blood vessels (more than 10 days in our case) will promote to the wide spreading of this method, especially in emergency surgery, in massive injuries, & in disarters, all the cases of insufficiently of time for selection of lot of donor blood. Objectives: Study of a reaction of the oardioreepiratory system of pregnant women to i/v microperfusion of Clophelinum which is known to eliminate hemodynsmic and endocrine nociceptive reactions and can be used for treating hypertensive syndrome in pregnancy and labor. Methods: The following non-invasive methods were used: capnography, spirometry, oxygenography, indirect Fick principle based on the circle breathing, plethysmography and integral rheography~ 52 functional indices of cardiorespiratory function were evaluated. Results: 74 pregnant women with ~H-gestosis were examined before and after i/v infusion of i00 ml of 0.0001% Clophelin solution, 0.005 mg/kg/hour. Before the treatment intensification of carbohydrate metabolism, hyperventilation with moderate hypooapnia and complete respiratory compensation of metabolic acidosis~ increased alveolar ventilation, decreased alveolar volume, predomination of perfusion over ventilation, hypokinetio type of circulation with dominated load by peripheral vascular resistance to the blood flow was observed in this group of patients. Microperfusion of Clophelin imp~-oved the ventilation/perfusion ratio, ventilatory and gaseous exchange efficiency, resulted in a decrease of congestion in the pulmonary circulation, possibly owing to a decrease of peripheral vascular resistance by 17%, of the heart rate by iO.5%, of the oardial output index by 9.5%. Conclusionm: The resulted type of circulation with a decreased load on the heart both by resistance and volume allowed to improve the cardioreepiratory system function in pregnant patients. Objectives: The injury severity score is a measure of severity of anatomic injuries. ISS is a sum of squares of the highest degrees of the Abbreviated Injury scale (AIS) for each of three most severity injured regions. The purpose of the study is to establish correlation between the ISS values and mortality rate in older, polytraumatized patients. Methods and Results: ISS was determined for 214 patients. The mean ISS value was 27.65 + 17.36 while the median value was 21. Minor injuries were present in 90 (42 %) patients with ISS less than 21, while 124 (58 %) patients with ISS more than 22 had severe injuries. Increased mortality of the older patients was noted in the range 21-30. All patients older than 50 died while 20 % of patients below 50 yrs of age survived, indicationg correlation between ISS and mortality rate in polytraumatized patients above 50 yrs of age. Conclusions: this mode of evaluating severity of injuries may help in triage, determining appropriate level of care and as an indicator of future outcome of polytraumatized patients. Objectives : Tissue hypoxia is a non exclusive cause of hyperlactatemia. Other serious medical situations induce hyperlactatemia. Therefore, lactatemia could be a non specific indicator of severity in patients admitted in emergency unit. The aims of this study were to examine the correlations between lactatemia with the short term survival course prognosis and the unit of hospitalisation; intensive care unit (ICU) or medicine unit, in patients admitted in our emergency department. Methods -Lactatemia was measured as soon as the admittance, in arterial blood sample of patients which needed arterial blond gas. Sixty-one patients were included during 4 months. To assess the statistical performances of lactatemia, sensitivity (Se), specificity (Sp) and accuracy (Ac) were calculated for the threshold determined by the Youden's test (Se+Sp-1). Results : Fifteen patients were admitted in ICU and 46 in a medical unit. Fifteen patients died. A group of 35 patients had a lactatemia up to 2 mmol.l" 1. In this group of patients, 3 had acidocetosis, 3 had asthma, 3 had cerebral vascular ischemia, 3 had neoplasia, 2 had cardiogenic shock, 1 was epileptic, 8 had congestive heart failure, 6 had acute respiratory failure, 2 had septicaemia, 2 had hyperosmolar status finally 3 had medicinal intoxication. Lactatemia was significantly higher in non survivor than survivor ( 5.5• vs. 2.3+1.0, p 0.85 when correlaliou eoet~dent was obtained indixddually. Of the seven ICPe -]CPv studied patients, we observed a cortelafiau ooeffioiont r = 0.47 (P < 0.001) with a regression line y = 7.2 + 0.43x. Corralalmu eoetfieiont was Inwer than 0.5 in all seven patients. Corrdation eoelfieients for levals of ICPv > 20 man Hg, > 25 mm Hg and > 30 tuna Hg with ICPe showed r = 0.89, r = 0.91 and r = 0.97 respectively; and with ICPe r = 0.25, r = 0.13 and r = 0.09. The obtained values did not change during the study. ConclusDns: In our study ICPe was considered a good type of ICP monitoring. /CPe signiticantly infravalorates ICP values. We observed a good correlatinn between ICPc and ICPv values in patients with high inttacramal presanre. Objective: Midazolam is a benzodiazepine agonist widely used for sedation in emergency medicine. Few studies in animals and humans point to a direct analgesic effect of midazolam probably mediated by spinal antinociceptive receptors and/or peripheral benzodiazepine receptors (1,2). In our experience in the Berlin Emergency Medical System (unpublished results) with anecdotal cases of extreme chest pain due to binge drinking but no evidence of acute myocardial infarction or extreme abdominal pain due to peritonitis, acute intermittent porphyria, Peutz-Jeghers syndrome or testicular torsion, we found that small doses of midazolam (2 -5 mg i.v.) were much more effective in relieving pain than repeated administration of high doses of buprenorphine or morphine, which may be associated with a considerable respiratory depressant effect. The dose of midazolam required for pain relief in these patients is non-narcotic and allowed further communication on the character and localization of' the residual pain, which might be very important for the further diagnostic procedure. Patients: Ten patients with abdominal pain due to acute gastrointestinal bleeding, suspected pancreatitis, suspected acute porphyria, and chest pain with no evidence of acute myocardial infarction received first-line midazolam i.v. at an initial dose of 1 mg and were asked how it affected the intensity and character of pain. Results: At the chosen dose of midazolam (2-8 mg), all patients were responsive to detailed questioning on basic orientation, the character, intensity and localization of the pain, and medical history. None of the patients required an additional opiate. All patients stated that the pain was tolerable after midazolam alone. Conclusion: Our preliminary clinical observations suggest that low-dose midazolam might be an alternative to opiates in extreme pain of presumably visceral odgin. Objectives: It is known that severe head injury in elderly patients is associated with higher mortality than in younger patients. It remains however to be clarified whether the preinjury pathology which is frequent among these patients, affects the outcome. Methods: In an attempt to investigate this hypothesis, 79 patients aged over 60 years suffering from head injury, with Glasgow Coma Scale (GCS) of 8 or less, were studied retrospectively. Twenty-six patients (32.9%) had preinjury pathology i.e. diabetes mellitus, arterial hypertension, heart failure, alcoholism, Parkinson's disease etc. (group A) and fifty-three (67.1%) did not (group B). The following data were recorded: mortality in the I.C.U., duration of hospitalisation, incidence of infective complications and neurologic status at discharge. Results: Groups were comparable in terms of mean GCS (6.57 vs. 6.56) and median age (67.5 vs. 67). The incidence of brain pathology in the two groups was the following: Epidural haematoma 7.69% vs. 11.32%, acute subdural! haematoma 30.7% vs. 30.19%, intracerebral haematoma 19.23% vs. 5.66%, subarachnoid haemorrhage 38.46% vs. 39.62%, diffuse haemorrhage 11.54% vs. 13.21%, contusion 26.92% vs. 49.06% and non-visible pathology (normal CT) 3.85% vs. 1.89%. Unilateral pupilary dilatation was found to be 15.38% in group A and 18,87% in group B. The mortality during hospitalisation in the I.C.U. was almost the same: 50% iu group A and 47.2% in group B patients. However, group A patients had significantly more infective complications, required longer hospitalisation and had lower GCS at discharge. Conclusions: The results show that the existence of preinjury pathology does not seem to affect the short-term outcome of elderly patients with severe head injury. It has however an impact on morbidity and perhaps long-term survival of these patients. The assessment of clinical development in Intensive care patients with severe head Injury still remains a problem. To optimize the monitoring of Intracraniel prassure (ICP) we rautlr~dly implant an eplduml measuring device in our hospital. The aim of this study was to prove the correlation of the ICP-values with CT findings and clinical development. During a 12 month period (1993 -9r the ICP was monitored in 23 p~,tients (14 male, 9 female) with severe head Injury by an eplclural measuring device (Epldyn~/$plegelberg| The mean age was 36.9 years (4 -83). The Glasgow Coma Scale at admission was 6.9 (3 -15). In all cases the device was placed wfihln the first 10 hours after admission. The tCP was compared with physical examination, radioidglcal or Intraoperatlve findings and cUnlca! outcome. The average time of measuring was 7. 2 days (1 -19) . The traatment depended on the !CP values recorded. Rising ICP-valuea ~ed to radlologlcal c0ntra!s by CT-scan. In 1 case an Intracranlai hemorrhage was detected and drained. The overall survival rate was 78.3 %. 113 showed a complete resolutl0n, in other 33.3 % psychological residuals like decreased mentatlon, In 17.4 % sensomotorlc residuals like cerebral nerve dysfunction and aphasia, and 11.1% of the Injured remained In a comatous status. In 87 % of our cases the measured values correlated with clinical course and management. In 2 cases (8.6 %) we observed a displacement of the ICP-pevice. There was no ICP induced Infecllon. Istituto di Anestesiologia e Rianimazione, Universit& ,,La Sapienza", Rome, ITALY * Istituto Superiore di Sanit& -Servizio di Epidemiologia e Biostatistica, Rome, ITALY Objectives: Acute renal failure (ARF) can be a severe complication of trauma. The current incidence of post-traumatic ARF is associated with high mortality 1. Identification of risk factors and prevention of this complication could improve the outcome of trauma patients. Methods: One hundred fifty three consecutive trauma patients (age 37.6 _+ 19.6, Injury Severity Score 28.3 +10.9) admitted to ICU were studied. Incidence of ARF was 31.4 % (48/153). ARF was defined as persisteat plasma creatinine >2mg/dl with or without oligoanuria 2. ARF was defined as early when occurring within the first 96 hours (EARF) and late when the onset was after the first four days (LARF). Results: EARF occurred in 31 patients while LARF developed in 17 patients. Age, ISS, and incidence of rhabdomyolysis and acute respiratory failure were not different in the two groups. An higher incidence of Multiple Organ Failure (MOF) and sepsis (76.6% for both) were observed in LARF group, when compared to EARF (25% and 23% respectively). Abdominal trauma was more frequent in EARF group (32% Vs 18%). The GS for EARF and LARF were respectively 8_+4.4 and 9_+4.15 while in the group who not developed ARF (NARF) the GS was 10.5• Conclusions: GS score difference seems suggestive and can be that an abnormal cerebral activity (hipofisary hormones?) may play a crucial role on onset of ARF in these patients. Moreover the frequency of Acute respiratory failure in the group of ARF was higher (91.7 versus 64.5) than NARF group. The early ipoxia in the early phase of trauma, then, may be another crucial point for development organ failure. These are preliminary data. A more exact statistical analysis must be perform to have definitive conclusions. To compare the Active Compression-Decompression Cardiopulmonary Resuscitation (ACD-CPR) with the Standard Cardiopulmonary Resuscitation (S-CPR) in out of hospital cardiac arrest patients. Is a controlled, randomized study. Two groups of patients with cardiac arrest out of the hospitalwere formed. Group I, (ACD-CPR) and group II (S-CPR). For the ACD-CPR groupweusedthecardiopumpdeviceofAmbulnternational. Asfortherest, the ERC (1992) algorithms for ACLS were followed. The Utstein Style (for out of hospitat cardiac errest) was used for listing and evaluating all cases of the study. The CPR was contucted by the crew and the doctors of our Mobile Intensive Care Units (MICU). We studied 146 consequitive patients (75 in group I) and (71 in .group II). Demographics pre-CPR characteristics (e.g. ECG form of cardiac arrest) and procedures (eg Bystanders or second tiers crew CPR, defibrillation, drugs) were quite similar for both groups. The mean arrival time of MICU was 9min. In Group I we recorded R.O.S.C. (Return of Spontaneous Circulation) 17,5%, death 73%, continuation of CPR efforts 9,5%. While in group II, 21%, 69%, and 9,9% respectively (recorded percentage until the admission to the hospital). No significant difference was found in anyofthe short term outcome parameters. No complications related to the ACD-CPR technique, were noted. Not any significant difference between the two methods was proven (from this small evaluated sample). The results of previous clinical studies are controversial (I) . More sophisticated studies proved the superiority, in a certain number of parameters (e.g pressures, flow, etc) of the new technique although there are many difficulties for establishing clinical results. In the pre-hospital setting that is related to many parameters (speed of the intervention, effectiveness of bystanders CPR, education ofparamedics, etc.)the evaluation is even harder. The superiority ofthe ACD-CPR can be proven when it is performed in almost 0 times Increased number of studied patients as w~ll as improvement of the technique could lead us to more established results. Objectives; Infectious morbidity is the major cause of mortality after burn injury, and is due to multiple factors. Trace elements (TE), which are involved in both humeral and cellular immunity, exhibit severely altered status after burns. TE supplementation has been shown to be associated with increased leukocyte counts and shortened hospital stay. The trial aimed at studying the immune responses in severely burnt patients receiving normal TE supplies or early large supplements. Methods: 12 patients, aged 40_+16 yrs (mean_+SD), with burns covering 49+18 % of body surface were studied from day 1 (D1) to D30 post-injury, were randomised in 2 groups (G): G1-control receiving recommended TE supplies + placebo; G2 -receiving in addition large supplements of Cu, Se and Zn from D1 to D8. Enteral nutrition was started within 12 hours of injury in all patients. Immunological parameters: peripheral leukocyte counts, proliferation of mononuclear cells to mitogens, cell surface molecule expression, and neutrophil chemotaxis at D10 and D20. Infectious episodes and micro-organisms were monitored until D30. Results: The patients' characteristics were similar G1 & G2. The total leukocyte counts were higher in G2 between D10 and D20, due to increased neutrophils (significant from D13 to D15). Total CD3+ and CDlg+ cells did not differ, whereas CD14+ (monocytes) were significantly increased at D20. Proliferation to mitogens was significantly depressed in all patients. Chimiotactism was not altered. The number of infectious episodes was significantly decreased in G2 with a mean of 2.0_+ 0.9 infections during the first 30 days versus 3.3_+ 0.8 in the control group (p < 0.03). Conclusions: The large TE supplements for 8 days was associated with a significant decrease of the number of infectious episodes. Supplementation was associated with increases in total leukocyte, monoeyte and neutrophit numbers. Further studies are required to determine the precise mechanism underlying the improved immune defences. Objectives: Evaluate the efficiency of local adsorption (LA) with the use of carbon adsorbents in case of severe burns in expertment and clinic. Methods: Experimental studies on LA were performed on a model of 20% body surface area IIIb-IV burn in 335 rats. A burn eschar was excised on the 3rd day after burn, the wounds were dressed with the gauze bandages (control) or with adsorptive dressings (LA), dressings were regularly changed. Clinical investigations were carried out in the course treatment of 78 patients with severe thermal and radiation Ilia-IV burn. In the dynamics of bum disease some indices of proteometabolism and intoyacation criteria were evaluated. Results: The experiments have demonstrated that the application of LA after early excision of a burn eschar exerts a pronounced normalizing effect on a protein electrophoregram and the activity of proteases and their inhibitors in burned tissues preserving vitality. Thus, by the 14th day after burn infliction the activity of cathepsin D in injm'ed muscles is 6 times lower under an adsorptive dressing than under a gauze bandage (control) (p<0,05), the activity of trypsin-like proteases is 1.5 -3.4 times lower and the antitryptie activity does not differ significantly from the normal level. The cytotoxicity of extracts of burned tissues after the adsorptive dressing application fn vivo and adsorption in vitro is 25-35% and 7-20%, respectively, of the toxicity of control extracts. A similar normalizing effect of LA is oK~rved for an intact muscular tissue and blood serum. The dectron-spin-resonance studies have demonstrated that LA allows to normalize antitoxic activity of liver and functional activity of kidneys. The application of LA in the treatment of patients with severe burns have been shown to localize a region of irreversible tissue changes, accelerate rejection of a burn eschar, attenuate an endogenous intoxication level and, as a result, shorten the time for grafting of a burn wound and accelerate wound heating. Conclusions: Proceeding from the obtained results, we can consider LA as an effective method of localization of a region of irreversible tissue changes as well as of correction of local and general metabolism failures and overcoming burn autointoxication during burn disease. C De Deyne, T Vandekerckhove*, J. Decruyenaere, B. Vaganee, V Vandewalle*, F Colardyn Depts of Intensive Care and Neurosurgery*-University Hospital Gent-Belgium. Jugular bulb oximetry is the first bedside available cerebral monitoring technique providing an estimation of the adequacy of cerebral perfusion. Its routine use in all patients suffering from severe head injury admitted to our IC unit enabled an extensive analysis of all very early cerebral perfusion data in order to evaluate the incidence of abnormal SjO~ data (and their possible causes) in this very eady period after traumatic insult and to search for possible implications as to the emergency management. These very early data were defined as the first 6 hours ICU data and ICU admission had to occur within 12h of traumatic insult. Over the last 2 years, 150 pts with severe head injury (GCS<8) were monitored by jugular bulb oximetry, starting immediately after their arrival at the ICU (mean of 4.8h after trauma, range between 2-9h). in a total of 85 pts (=56.6%), jugular bulb desaturatiens (<55%) were noticed during this early 6h period. In 24 pts (=16%), jugular bulb saturations higher than 75% were observed, whereas 41 pts (=27.4%) revealed no abnormal SjO 2 data (55-75%) during these first 6h. Concerning the periods with too low jugular bulb saturations (n:85), we found the following correlation ; in 49 pts (=57.6%) cerebral perfusion pressure (CPP) was below 70 mmNg, in 36 pts (=42.3%) PaCO~ was below 30 mmHg and finally in 6 pts (=7%) we found primary intracranial hypertension. For the high jugular saturations (n:24) we found a primary intracraniaf hypertension in f0 pts (=41%), and a PaCe 2 level above 40 mmHg in 6 pts (=25%). In all patients we could restore jugular bulb saturation within normal range (55-75%) with the correct!on of the presumed causative factor. We can conclude that ultra early jugular bulb saturation data revealed a high incidence of abnormal values, with a predominance of jugular bulb desaturations, confirming once again the high incidence of disturbed and too low cerebral perfusion within the first hours after severe head injury. These jugular bulb desaturations were especially correlated to systemic causes, as a too low CPP (caused in the vast majority by primary MAP insufficiency, and not by intracranial hypertension) and hyperventilation were the 2 major causes of the desaturation periods. As jugular bulb desaturatione are known to be significantly correlated to a worse neurological outcome after severe head injury, one might improve outcome by an emergency management avoiding these possible causes of jugular desaturation. Therefore, extreme attention should be paid to the maintenance of an adequate mean arterial blood pressure (above 90 mmHg?) even duhng the few time spent at the emergency department. One should be as attentive to the maintenance of normoventilation during this very early period of admission and hyperventilation without any knowledge of ICP or SjO2 should be abandonned. Recently, indomethacine has been proposed for the treatment of therapy refractory intracranial hypertension in pts suffedng from severe head injury (1). Indomethacine, a cyclo-oxygenase inhibitor, gives rise to a significant fall in cerebral blood flow by inducing cerebral vasoconstriction. Therefore, its use could result in a drastic lowering of the intraeranial pressure (;CP) in pts suffering from intracranial hypertension secondary to cerebral hyperaemia and in whom the use of other cerebral vasoconstrictive drugs (barbiturates or hyperventilation) appears insufficient to control ICP. For the last 18 months, we included the use of indomethacine in our therapeutic flow chart for severe head injury management. Pts revealing intracranial hypertension (ICP>20 mmHg) and cerebral hyperaemia (SjO~>75%) and in whom ICP was not efficiently controlled by the combined use of hyperventilation and barbiturates were given indomethacine in a trial to control ICP. A total of 98 head injured pts received treatment for intracranial hypertension over the last 18 months. Six of them met the criteria set for the administration of indomethacine. In 2 pts, no decrease in ICP or in SJO 2 was observed and both pts died due to therapy refractory intracranial hypertension. In the other 4 pts, a significant fall in ICP and in SjO 2 was observed shortly after indomethacine administration. In 2 pts we observed a catastrophic fall of SjO= even below 55%, indicating an extreme cerebral vasoconstriction with the possible risk of inducing cerebral ischaemia. In one of the 4 pts, ICP remained under control without further administration of indomethadne, but he died 3 days later in multiple organ failure. The other 3 pts, needed multiple indomethacine administrations (for 1 pt even during 4 consecutive days) to finally control ICP. In all 3 pts, ICP was finally controlled, but only 1 pt survived. Both other pts died from systemic causes (multiple organ failure in 1 pt, massive gut infarction in the other tat, possibly due to the systemic vasoconsttictive effects of the indomethacine administration). In conclusion, indornethacine might have a role in the treatment of intraoranial hypertension, especially when caused by cerebral hyperaemia. We observed however a poor final outcome and a threatening high incidence of systemic events (multiple organ failure, gut infarction) in those pts receiving indomethacine for ICP control. Therefore, indomethacine in the treatment of intracranial hypertension should be reevaluated in controlled study settings, before its routine use can be considered. Untill recently, intracranial hypertension (ICH) in pts suffering from severe head injury was managed in a staircase approach, with CSF drainage as first therapeutic step, mannitol as second step, hyperventilation as third step, and finally, barbiturates as the last rescue step for therapy refractory ICH. This staircase approach for the treatment of tCH was only guided by the intracraniat pressure, and not by other parameters such as e.g. the actual state of cerebral perfusion of the concerned pt. Jugular bulb oximetry provides us with the first, bedside and continuous available, estimation of cerebral perfueion. Its implementation in a rigourous flow chart, based on as well ICP-as jugular bulb oximetry-data might result in an altered strategy for ICH management. We adopted a '~ugular bulb saturation (SjO~)-guided approach" for ICH management in 86 consecutive pts, suffering from severe head injury (GCS<8). We maintained CSF drainage as first therapeutic step, but the decision for the second step was guided by SjO2 information. Pts revealing ICH and SjO=values above 75%, were treated with hyperventilation, and did not receive mannitol. If ICH persisted, barbiturates were added as a third step. On the other hand, pts with ICH and SjO= vales less than 75%, received mannitol administration as second step. Hyperventilation and/or barbiturates were only added if ICH persisted and if no cerebral hypoperfusion was discerned (SJO=>55%). Our objectives were to prospectively analyze this new therapeuticstrategy, as compared to the formerly used staircase approach of ICH. We managed 86 pts with ICH, with an overall mortality of 13.7% due to therapy refractory ICH. All pts received standard primary care with head elevation, full sedation and normovenfilation. Fer 16 pts, CSF drainage alone was sufficient to control ICE Of the remaining 70 pts, 38 pts received mannitol and 32 pts were hyperventilated as second approach. In the third line, 14 pts were managed with barbiturates, 12 with mannitol and 10 pts with hyperventilation. Finally, barbiturates were used as the final rescue in 14 pts. These results reveal a less frequent use of mannitol as only 50 pts received mannitol, compared to the 70 pts that would have received mannitol using the former staircase approach. Hyperventilalien was used much earlier in the treatment course, as 32 Lots were already hyperventilated in the second line approach, were this was formerly exclusively reserved for the third line approach. Finally, also barbiturates were used much eadier (14 pts received barbiturates as third approach). We may therefore conclude to a important change in the management of ICH, induced by a SjO2-guided flowchart. However, future studies will have to elucidate if this new strategy for the intensive care management of severe head injury will also result in an improved outcome. ObSectives: In a first series of experimental brain injury we investigated the course of brain PO2, ICP and cerebral blood flow after traumatic brain injury (TBI), whilst accordingly there are very few data available and the mechanisms leading to secondary brain damage are poorly understood. Methods: In 6 piglets (14 days old, 3,3-5 kg) of either sex we produced a moderate brain injury (1,5 arm., 20 msec.) using a lateral fluid percussion {FP) device. Complete measurements were made before and 5 min. after brain trauma and after 3, 5 and 24 hours including blood gases, cardiac output (htermodilution), heart rate, EEG, laser doppler flow probe (LDF} and ICP values (Camino), brain temp., pO 2 by a clake type oxygen electrode (Licox) and coloured microspheres for regional blood flow. Results: Immediately after the trauma a typical "Cushing"response to the ICP peak up to 130 mm Hg being highly significant (before mean i0 mm Hg, range 4-12 mm Hg) could be observed: mean arterial blood pressure rose from appr. 85 mm Hg to ii0 mm Hg for 3-5 min. In two animals this was followed by an ischemic period lasting 15 min. Accordingly ICP values gradually returned to starting measures within 3 hours; in the ischemic animals they remained at a level of about 30 mm Hg.-No secondary increase of ICP could be observed, once ICP dropped to starting values within 24 hours. Cerebral blood flow (LDF) fell from mean values being i00 before trauma to appr. zero and recovered to around 50. Brain pO 2 started at mean values of 20 mm Hg (range 15-30 mm Hg) and fell to around zero depending upon the severity of the ischemic reaction. On average values of 15 mm Hg were reached over the time course. Conclusions: With our FP trauma model we can reproduce the well known "Cushing"-response after brain injury; secondary ICP elevations cannot be achieved, although local edema is observed. Direct brain pO 2 measurement seems to be a very sensitive variable for detection of cerebral ischemia and anticipates eventually following ICP elevations by far. PULMONARY ASPIRATION S,Traoaras. V. Sgountzos, P. Agouridakis, M Eforakopoulou, E. Ioannidou. Intensive Care Unit (tCU) of "KAT" Hospital, Athens, Greece Ob!e=ives: The reported mortality rate after pulmonary aspiration is variable in several series. The purpose of this study was to find out the influence of preexisting disease or situation on morbidity and mortality of Intensive Care Unit (ICU) patients with pulmonary aspiration. Methods: Patients who were treated in ICU and had pulmonary aspiration, were studied, Entrance's criteria in the study, all of them obliged, were: 1) Suction of gastric contents from trachea during intubation, 2) Presense of a predisposing factor, e.g. coma. 8) Recent hypoxaemia or new infiltrates in xray. Preexisting disease was recorded and correlated with complications and outcome. Patients with Glasgow coma Scale 3, because of cerebral injury, and patients who died within 3 days from cause other than aspiration, were excluded from the study. Method of statistical analysis: Chi-square test, Results: One hundred forty five patients were studied. The trauma patients were 96 and the non trauma patients 49. From the trauma patients, 77 had cerebral injury and 19 were polytreumatized without cerebral damage. From the non trauma patients, 13 had malignant neoplasms, 14 neurological diseases in terminal stage, 7 old age, 10 drug overdose, and 5 several diseases. Eighty seven from 96 trauma patients (91%) and 45 from 49 non trauma patients (92%) manifested several complications (pneumonia, ARDS, etc), so there was no statistical difference in complications' frequency between the 2 groups (p>0,1). The severity of complications was also proportional in the 2 groups. Eighteen deaths were recorded in the trauma patients (mortality 19%). Only 7 deaths correlated directly or indirectly with the aspiration (7%). In non trauma patients, 32 deaths were recorded (71%). Twelve deaths were recorded in 13 patients with neoplasms, 12 deaths in 14 patients with neurological diseases, 6 deaths in 7 aged patients, 1 death in 10 drug overdose patients, and 1 death in 5 patients with several diseases, The mortality difference in trauma and non trauma patients was statistically significant (p< 0,001). In patients with drug overdose the mortality was significantly lower from the other non trauma patients and the difference was statistically significant (p<0,001). Conclusion: The preexisting disease or situation plays a major role in the outcome of the patients with pulmonary aspiration. The mortality of patients with aspiration seems to be caused by severe preexisting situations rather, that lead to death, than from the pulmonary aspiration per se, which may be a final happening in a predetermined course. Obiectives; The purpose of this study was to compare fluconazole and amfotericin-B in the treatment of fungal infections in severe trauma patients. MethodS: Thirty five severe trauma patients who were treated in Intensive Care Unit (ICU), were studied prospectively. They all developed fungal infections, prooved with blood positive cultures and at least one of the following: fever, positive urine or bronchial secretions cultures, infiltrates in xrays. The patients were separated randomly in 2 groups. The patients of group A (15 patients) received fluconazole 200 rag/day for 15 days. and the patients of group 8 (20 patients) amfotericin-B 50 rag/day for also 15 days. Compaiison's criteria were the clinical responce to treatment (fever etc), the fungal elimination (blood and other cultures), the relapses of the disease, the side effects of drug, and the outcome of the patients. As method of statistical analysis was used the Chi-square test. Results: Nine patients from 15 of the group A (60%), and 18 from 20 of the group B (90%), presented remission of fever (patients of group B had better clinical responce than patients of group A, and the difference was statistically significant, p<0,05). All the patients before treatment had positive for fungi blood cultures. After 10 days of treatment, 3 patients of group A and none of group B had positive cultures. Eight patients (from 13 who had positive cultures of bronchial secretions before treatment) of group A. and 5 (from 17) of group 8. had positive cuttures of bronchial secretions after 10 days of treatment, so positive bronchial secretions were fewer in group B than in group A, but this difference wasn't statistically significant, (p<0,1 and p>0,05): Ten patients (from 12) of group A and 7 patients (from 16) of group B had positive urine cultures, after 10 days of treatment (positive urine cultures were fewer in group B than in group A and this difference was statistically significant. (p<0,05). Two patients of group A and none of group B had a relapse of fungal disease. In group A, no side effects were obsePced, while in group B were observed only minor side effects (small increase of serum creatinine in 2 patients, chills and fever during infusion in 3 patients, and hypokalemia in 12 patients). Three patients of group A and 1 patient of group B died, because of sepsis. conclusion: Amfotericin-B (even i~ short regimen of 15 days), is superior to fluconazole in the clinical and laboratory responce and also in the relapse of fungal disease, Fluconazole is superior to amfotericin-B as it has no side effects. Ob!ectives: Flail chest after thoracic trauma is a serious injury. It is controversial if flail chest by itself orthe concomitant intrathoracic injuries e.g. pulmonary contusion, is the cause of the reported significant morbidity and mortality. In this study we searched the influence of concomitant thoracic injuries in the course and outcome of patients with flail chest. Methods: Eighty five patients with flail chest after isolated chest injuries were studied, For the purpose of analysis, we separated the patients into 4 groups, Patients with isolated flail chest were included in group A, patients with flail chest and hemo-pneumothorax in group B, patients with flail chest and pulmonary contusion in group C, and patients with flail chest and hemo-pneumothorax and pulmonary contusion in group D. Complications from the chest, duration of mechanical ventilation and mortality were compared in the 4 groups. Statistical comparison of results belween groups was made using Chi-square and T-studend tests. Results: The patients were 85. All patients received mechanical ventilation, Twenty eight patients were ihcluded in group A, 19 in group B, 20 in group C. and 18 in group D. Seventy three patients manifested complications from the chest, especially pulmonary infections. There was no statistical difference among the 4 groups as to number of COmplications ( twenty four patients had chest complications in group A, 16in group B, 17 in group C, and 16 in group D. p>0,1}. The duration of mechanical ventilation was not statistically different among the 4 groups (the mean duration was 15,9 days in group A, 16,8 in group B, 16,5 in group C, and 17,5 in group D, p>0,1). There was also no statistical difference in mortality among the groups (six patients died in group A. 4 in group B, 4 in group C, and 5 in group D, p>0,1 ). Conclusion: Flail chest by itself is a serious thoracic damage with many complications, regardless of the presense of other thoracic injuries, which don't contribute to greater morbidity and mortality. The present study investigated the correlation between blood lactate mortality and organ failure in 129 trauma patients admitting between December 1, 1992 and July 31,1993 in the ICU. Road traffic accidents were the most common cause of trauma in this studded population. Brain damage was the main cause of mortality .Nevertheless, 29 of patients died from sepsis and multiple organ failure without significant brain damage and these deaths were potentially preventable. Respiratory failure was the most common complication and was developed in 44 (44%) of survivors and in 26 (86%) of non survivors .We noted low fncidence 5 of renal failure may be do to the early and aggressive itTv'asive hemodynamic monitoring and cardiopulmonary support. As part of our routine case protocol serial blood lactate levels were measured in each patient at least 3 times a day until the valses returned within the normal range or until death. We analysed the blood lactate levels on admission, the highest value and the number of days until the first normal value (5 in the rest 7. 15 patients 20 mmHg at the beginning. ZEEP Ob/ectives. Critically ill patients are transpoded to an Intensive Care Unit(ICU), under conditions, which have not been systematically evaluated. Therefore, we set suite investigate transportation and admission condition of these patients to our department. Methods. We studied 36 patients(16 females), aged (mean-..+-sd) 56.2_ 17.3yrs, which were consecutively (from August 1994 to March 1995) admitted to the ICU, through the Greek National Emergency Transporta~on Service. APACHE II severity score upon admission was 17.4-+6.8 (range 4-31). The following data were evaluated: 1) Number of medical departments, where health care was provided until final admission to the ICU, 2) ambulance transportation conditions, 3) catheters and tubes inserted before admission, 4) vital signs upon admission 5) information provided by referring physician (scored on a 1 to 5 scale: History, electrocardiogram, chest x-ray, laboratory data, drug therapy already administered), 6) comparison of the state of the patient described by referring physicians, to the actual state u pen admission. Resu/ts. One to four medical departments had provided health care before the palient was admitted the ICU (1:22.2%, 2:47.2%, 3:27.7%, 4:4%). Thirty/36 (83.4%) patients were escorted by a physician. Twenty-six/36 (72.2%) were transported on oxyge n, FIO2(mean__.sd): 46-+3%, PaO2: 78.6-+35.2mmHg. Five of the remaining 10, for whom no oxygen was provided, had PaO2: 46.2-+7mmHg. Twelve/36 (33.3%) were intubated and ventilated during transportation. Thirtyfour/36 had a peripheral venous line, 5/36 had an arterial line, 13/36 had a nasogastdc tube, 20/36 had a urinary catheter. Eleven/36 were sedated and 2/36 were paralysed. Three/36 were on inotropes. Vital signs upon admission were: arterial blood pressure, systolic 100.6-+44mmHg, diastolic 57-+23mmHg, heart rate 104-+22 bpm, temperature 36.3 -+2cC. Patient information score was 217--. 1.7. The actual state upon admission was found substantially different, as compared to the description of the referring physician, in 28/36(77.7%) patients. Conclusions. We conclude that several aspects of the Greek National Emergency Transportation Service to an ICU should be reevaluated and further improved, i. e. ventilatory support, adequacy of information provided and accuracy of prior description of the patient's state. A new perspective must be applied for critically ill patients transportation since 78.8% of the patients were evaluated and treated in more than one, medical departments, mostly primary care, before they were finally admitted to our ICU. DCLHb is a human derived hemoglobin molecule that has been cross-linked to stabilize and permit heat pasteurization to remove residual proteins and inactivate viruses. DCLHb is mixed with a lactated electrolyte solution to yield a total hemoglobin concentration of lOg/dL Objective: To present an overview of four recently completed clinical safety studies of DCLHb in the U.S. and Europe, and to discuss the properties, actions and potential indications for DCLHb. Method: Patient populations in the four studies included males and females ranging in age from 18 to 84 years. Dosing ranged from 25mglkg to 300mg/kg. The controlled randomized safety studies were conducted in chronic renal failure patients, surgical patients undergoing total hip replacement or abdominal aorta repair and in hemorrhagic hypovolemic shock patients. These very diverse patient populations allowed safety evaluation of the product in patients who were generally elderly, often hypertensive with some degree of cardiovascular disease, and receiving medications for treatment of other conditions. Results: Over 150 patients received DCLHb in the four:studies. No product related sarious adverse events occurred during the clinical trials. Conclusion: Results from Phase Itll safety studies of DCLHb in patients undergoing chronic renal dialysis, abdominal aorta repair, or total hip replacement and in patients in hemorrhagic hypovolemic shock, indicate that the product was well tolerated in these distinct populations. Although these studies were designed to evaluate safety, the data suggest clinical benefit. Follow-up efficacy trials are indicated. Prehospital emergency services represent the extension of emergency care into the community and constitutes the manpower, communications, transportations and facilities used to provide care for patients outside hospital. One of the main points of the system is how to decide the hospitalization of patients and what kind of facilities to provide : emergency medical service, fire brigade, locat general praclitionner or ambulance officers. Objectives : to realize guidelines for using the prehospital emergency medical service in case of patient'calls outside hospital. Methods : from 1st June 1994 to 14 July 1994, all the calls for emergency care were analysed using a questionnaire of 114 items (origin of the call, responses to the questions of an emergency practitionner, kind of emergency service provided and the issue of the patient). After taking account of the appropriatness of the decision, statistical method used was a logistic regression. Results : 996 calls were analysed. The criteria, for prehospital emergency medical service using, given by the logistic regression were as following : existence of a call for emergency, thoracic pain, dyspnea, seizures, cyanosis, drug intoxication, fall of the patient, fracture, age, the state of consciousness and the neurologic reactivity. The minimal and maximal predictive values of the model given by the logistic regression are respectively 2% and 100%. The performance of the model is 88 %. Conclusion : It seems possible to help medical decision of emergency medicine by using only some easy criteria and a predictive model. (Italy) Objective: To evaluate the incidence of blunt carotideal injury (BCI) in patients admitted to our ICU after head injury. Methods: We reviewed the medical records of all patients diagnosed to have a BCI. At admission, the severity of trauma was assessed either with Glasgow Coma Scale (GCS) and with CT scan. BCI was demostrated by Doppler ultrasography (US) and by angiography (ANG). Results:Since May 1991 to April 1995, 4 patients were admitted to our ICU with BCI (2m,2f, age 29+ 1 3). A history of direct trauma was present in 2 patients. Admission GCS was 15 in all patients, and was associated with hemiparesis in 3 of them; the last became paretic 48 hours thereafter. Two patients had concomitant injuries (a homoIateral clavicular and a controlateral zygomatic fracture, respectively). The initial CT scan was negative in every patient, and showed signs of ischemia after a variable timespan (2-4 days) after the onset of the symptoms. The BCI was diagnosed with US and ANG, which demonstrated a thrombosis of the internal carotid artery (IC). In two patients, an intimai dissection was also present. Three patients were treated with heparin associated with antiaggregating agents and were discharged alive. The last patient was referred to our ICU after the development of a massive hemispheric infarction, and died three days after the admission. At necropsy, the IC thrombosis was associated to an extensive homolateral extra and intracranial venous thrombosis. Conclusions:The presence of focal neurological signs despite a negative CT scan should address the diagnosis toward a BCI, thus implementing the diagnostic workup with US and/or ANG. Tab I: Distribution of l~tients (%) in the 3 groups the outcome were monitorett Results were sabmitted to statistical analysis using a continence table 3X2 in Z2 test. Res.cl~s: Of 43 patients 34 were submitted to thrombolysts and 3 died. The higher incidence of bracb, ar~lhmias (II degree gg p t39e 1 and 2 AV block. 11I degree AV block. AvsB 6.141 210 roroHg and diastolic blood pressure > 110 nunllg were included into the study. Prior to treatment blood samples for determination of plasma renin activity (PRA), angiotensin converting enzyme (ACE), angiotensin II (ANG II) and aldosterone (ALDO) were collected. All patients received 5 rog enalaprilat intravenously. Success of treatroent was defined as a reduction of systolic blood pressure below 180 mmI-Ig and diastolic blood pressure below 95 mmI-Ig within 75 minutes after start of treatment. Results: 35 patients were included in our study, 20 (57%) patients responded successfully to treatment. Mean arterial pressure decreased in responders by 36.5 mmHg and in non-respenders by 12.7 mmHg (p<0.001). Responders and non-respenders differed signii'icantly concerning PRA (p=0.001), ACE (p=0.003) and ANG II (p=0.04). 0.003 0.04 The extent of blood pressure reduction correlated positively with the pretreatment PRA and ANG II concentrations (correlation coefficient for PRA: r=0.43; ANG II: r=0.66). Conclusion: Our data confirm that in patients with hypertensive crises blood pressure response to ACE inhibition is mainly determined by circulatory PRA, ACE and ANG II. As the extent of blood pressure reduction correlates with PRA, ACE-inhibitors in patients with suspected high renin status cannot be recommended, as excessive blood pressure reduction, which carries a considerable risk for further organ damage, may occur. F. Staikowsky, N. Grillon, F.Pevirieri, C.Jedrecy, C. Zanker, F. Michard, A. Haft Medical Emergency Department. Hospital Bichat, Paris Epidemiology of acute intentional self medications-poisoning (SMP) in France is especially known by data of Poison Control Centei,s and Intensive Care Units (ICU). The purpose of this study is pro~,ided characteristics of this problem in a MED for adults. Method: July 1992 to June 1993, files of patients consulting to the ED for SMP have been retrospectively analyzed. Results: 727 patients, 482 women and 245 men, 33.3 + 12 years old (range 15-92) have been admitted for 804 episodes of SMP (4% of all consultations) whose 77 relapses during the period of study. Psychiatric disorders, drug addiction or HIV patients was found for respectively 42.6%, 9.1% and 2,9% of patients. The interval of time between the ingestion and emergency consultation was noted for 43% of SMP (332 + 532 min, ranges 15-4320). The involved products name was known in totality in 89% of cases with an average number by episode of 1.7 + 1 drugs (ranges 1-8). The most often, 1 (52%) or 2 (21%) different products were interfered. The nonbarbiturate psychotropic drugs accounted for 76.7% of the products (benzodiazepines 67%, antidepressants 9.5%, neuroleptics 8%, carbamates 5.8%, imidazopyridines 5.1%, cyclqpyrrol0nes 2.7%). Analgesics and nonsteroidal antiinflammatories represented 6.8% of all drugs, anticonvulsants 3.4%, cardiovascular drugs 2%, antiinfective agents 1.9%, drugs against cough 0.86%, muscle relaxants 0.86% and antihistamines H1 0.5%. The benzodiaz6pines were present in 531 episodes, alone in 316 episodes. In 36.5% of cases, there was a simultaneous intoxication with alcohol. The processing consisted of gastric lavage in 32.5% of cases, activated charcoal in 16.7% of cases, flumazenil in 16.9% of cases, naloxone and acetylcysteine in 3.4% of cases; orotracheal intubation was performed in 12 patients. Admission in hospital was effective for 280 patients, in medical ward (n = 156), psychiatry (n = 63) or ICU (n = 62); no fatal case was recorded. Conelusion: SMP to ED are often benign. The benzodiaz6pines are the most often incriminated but the new anxiolytics and hypnotics (imidazopyridines and cyclopyrrolones) take a growing place. The latsion burn center of athens. Its planning constructive and functional refinements J. Ioannovich, A. Petalas-Vourekus, D~ Serbetis, H. Carsin A 18 bed Burns Unit is under construction following a donation to the General Hospital of Athens. The plan of the Unit, covering a surface of approximately 3.500 m2 is based on the principle of three identical 6 bed satelites which may function totally independent from each other. In the center of the Unit the common facilities are installed, like operation theatres, storage rooms etc. This new modification in the plan of a Burn Unit is presented in this paper. The advantages from the fucntional, administrative and medical point of view are discussed. Tiffs anisotropic conduodon could favour the ocenrence of a circular movement of the impulse that leads to tachyeardias by reentry. Purposes of this work were to study, with the help of epicardial mapping, the influence of a trieyclie antidepressant, clomipramine (C), on the conduction velocity longitudinal (VL) and transverse (VT) to myocardial fiber orientation and on anisotropy (A = ratio VL/VT), and their modificutions by the sodium bicarbonate (13). Method: a plaque of 64 electrodes, positioned on the left anterior ventricular wall of 9 anesthetized dogs, allowed to deliver, thanks to central electrodes, programmed electrical stimulations inducing vcuttienlar complexes, and to collect them. Each entailed unipolar dectrogram was processed by a computer system that drew the isochrones and a map of activation allowing the calculation of V. The C was infused (0.5 mg/kg/min iv) during 75 rain; at T60, dogs received the B until the retuni of QRS to its initial value fro). A lengthening of QRS of at least 30% of its value at TO was demanded before the administration of B. Results: 1 dog was excluded because of an.~nsufficient prolongation of QRS before the administration of B. All values (MAP : mean arterial pressure, I-IR : heart rate, QRS andQT intervals, V) differed significatively (13<0.05) compared to values control frO)except QRS at T65. The B (7 + 6 ml/kg; ranges 2.8 and 20.5 ml/kg) modified no studied dements outside of the ( }RS. TO TI5 T30 T45 T60 T65 T75 A 2,1 + 0,6 2,1 + 0,5 2,1 + 0,4 2,1 + 0,4 2,1 + 0,7 2,1 + 0,7 2,1 +-0,~ Conclusion : the C slowed V L and V T without modify the anisotropy. The B did not modify the V of~conduction while the QRS prolongation was corrected. The C acts as a class I antiarrythmie drug on the inward sodium current during the phase 0 of action potential; the gap junctions have shown to be important in the conduction and an action on the gap junctions such as a modulation of the junctional resistivity, can not be rule out. IS THE DOCTOR A HEROE ? P. T.schies~.he, T. Bauer, M. Seyr Dept. of Anaesthesiology and Intensive Care, AOKH Krems, Austria Objectives: Helicopter Emergency Services (HES) are getting popular more and more. The results concerning outcome are encouraging. However, some recent accidents with dead or badly wounded HEScrew-members have shown the relatively high risk for the crews. Therefore we were interested to eval0ate the motivation of physicians to participate in a HES. This survey was designed to investigate current concerns about safety and motivation of doctors on emergency call. Methods: A questionnaire was sent to 205 doctors of the Austrian emergency system. The survey consisted of multiple choice questions and subjective scoring tables from 1 (--full agreement) to 5 (=disagreement). Overall, 64"/. of the active emergency physicians participated in the survey. Results: 61.1% of the doctors assume the system is basically safe, experienced doctors tended to have less trust in safety. Only 13% would not hesitate to go into action by dark. 14.8 % stdctly refuse night flights to accidents outdoors. Although defibrillations are assumed to be safe dudng flight, only 29% would do it. 52.8% of the doctors would rather stop flying. The most common reasons for 9,uitting were wish of family and fear of an accident. 47.2% Conclusioq: Short transportation times help to avoid trauma related stress, pain and shock-induced organ complications. Therefore the physiologic and economic advantages of HES are undebatable. However, the survey data indicate a considerable concern about safety of the medical personal in a HES. 14 crash landings within less than 10 years with 15 deadcases and 17 badly wounded crew members in a small country like Austda make desire for safe flying conditions understandable. Obiectives: To evaluate the clinical usefulness of Trachlight. Methods: Trachlight is a new device facilitating endotracheal intubation. A stylet with a lightprobe is inserted into the endotracheal tube. Intubation is guided by the light glowing through the neck tissues, thus rendering direct laryngoscopy unnecessary. Intubation using Trachlight was studied in 37 patients (age 21-68 years). The indication for intubation was elective surgery in 21 patients (ASA I-II) and emergency intubation in 16 patients. In the elective patients, anaesthesia was induced with thiopentone supplemented with fentanyl, and intubation was facilitated with vecuronium. The cause for intubation in the 16 emergency patients was dyspnea in 8, cardiac arrest in 2, trauma in,2 and unconsciousness due to drug overdose or seizures in 4 patients. Intubation was facilitated with medication in 12 patients. Results: Of the elective 21 patients, 19 (91%) were successfully intubated. Six patients (29 %) needed two attempts before successful intubation. The duration of intubation exceeded 30 seconds in 8 patients (38 %). Of the emergency patients, 14 (88%) were successfully intubated. Six patients (38%) needed two attempts, and the duration of intubation was more than 30 seconds in 9 patients (56 %). In 54 % of all 37 patients, intubation was assessed as easy. No or insufficient glow, prolonging intubation or necessitating two attempts, was noted in 11 patients (30 %). Oesophageal intubation occurred in 2 patients. Conclusions: Trachlight may be a valuable adjunct for intubation in varoius settings provided that adequate training is provided. A learning curve was found to exist. Objectives: To compare enoxaparin and standard heparin in CAVHD and calculate the value of laboratory controls in the treaanent. Patients and methods: Twenty patients needing dialysis for acute renal failure participated in the study. The main exclusion criteria were massive bleeding or a thrombocyte level < 50 x E9/I. In each treatment the same type (AV-400, Fresenius Ag, Germany) of a polysulfone capillary haemofilter was used. The study scheme consisted of two consecutive four-day CAVHD treatments, one course for each type of heparin. The order of heparin administration was counterbalanced between patients. The standard heparin was given as a continuous infusion aiming at an activated coagulation time between 200 and 250 s. The initial enoxaparin dose was 80 rag every 8:th hour intravenously, but was modified by any signs of coagulation in the dialysis blood lines or bleeding complications. Results: The dialysis treatment was adequate in both treatment modes, with mean blood urea levels 24.3 and 25.2 mmol/l respectively (NS). The bleeding complications were moderate and similar in both treatment modes. The mean life-span of haemofilter using enoxaparin as an anticoagulant was some longer than using heparin (35.7 +31.6 h versus 22.3+26 h, NS). The mean APTT-levcl during heparin treatment was 79s and during enoxaparin treatment 54s (ref. 24 -34s). The mean daily dose of heparin was 422 nag, that of enoxaparin lg7mg. The mean anti-Xa activities were 0.40 U/mi and 0.47 U/mi, respectively, reflecting a better bioavallability of enoxaparin. Conclusions: Both anticoagniation modes were equally effective and well tolerated. The amount of enoxaparin needed for a proper anticoagulation was, however, less than half of that of standard heparin. The changes in APTT level were too slight to make its use possible in controliing the dose of enoxaparin. The use of enoxaparin seems to be rather safe in CAVHD even without laboratory controls. The adv~ucea in the management of computerized data of an intensive care unit have been petalled to the clinical advauces and the increasing sophistication of methods of diagnosis fop the clinical application an therapy. This has led our Unit to design and develop a computational system called TIMBU which is used to help physicians assist patients. Among its various uses, this system has a software for the hemodynsmic control of a critic patient. This program was carried out to get as fast as possible the hemodynamic data of the patients in an intensive care unit. As an example, we can mention that when we load 17 data obtained through direct measurement from the monitors and the lab, the program calculates 18 parameters that guide, intelligently, to the diagnosis and therapeutic behaviour of the hemodynamic problem through screen messages. The validation of this program in the unit of intensive care has demonstrated that its use allows a more efficient handling of the patient with serious hemodynamics and respiratory disorders. Ohieetlve: Traema is a heterogeneotm 'disease' that ecatr~ a~"o~s all age ~oupe with v~ying degrees of severity. This Imerogeneity has made the di~e, trmma, diflkaflt to r The ehn of this stady wa~ to assr the fitaen of SAPS in Ibis popeleties. Methode: In order to compute the ~ probability, a model derived from logistic regression w~ developed. Meam'e8 of calibration (goodaess-of-fit StetislJ.r and di~'riminafion (ROC ou~e) were adopted in developmm~ and validetLon set randomly taken from a database of 10065 pts eeeseemivety admitted in ICU (Arohidia). ~ witho= salm, p~ yom~ am IS yam, with LOS ~horter thma 24 hotam wore exr fa'om thi~ mmly~ir Thi~ model v~s then evahmed on the ~per ~mbgro~ (i.e., trmma pts). If'it did t~t fit the data well ~, new model wm developed rer the logit only on trm=~apm. ReimS: Data were availabte for 8059 pts during aperiod of three .y~m , treama pts were 1156 04.3%), Teats of calibration iadioaled probability model did mot provide m adequate refle~on of the mortality ezperieace in pm with Ireutae, being the observed mortality lower flma the expected (figm'o). A aew model was then variable. This oastomized model fit~ the de~t of trmara pts very well (g 2=-8A7 p>0.25; ROC = 0,94 ). The di:lFerencea between the two modele were evident. Conclusion: This ltudy shows that mortality in Iramna pts is over WcfE~d when ~se~ed by menm of SAPS. However the r mode! meets high standmcd in terms of calibration mid dil~'iminat'~o~ ']"he advaatage Of ~imd models meaas the colleotion of the ~ set of variables for all pm admitted in ICU e~einat the ase of diasma specific ~oring syatex~. ("SL"): EFFECTS ON CARDIOVASCULAR AND HEMOSTASIS SYSTEMS (CVS, HSS) A.Oborin~PH, ~.~yndiuk~PH, B.Kondratsky ~pt. of'""Su~gery and Transfusiology, Research Institute of Hematology, Lvov, Ukraine Objectives: Great interest has been shown recently in the use of HOSs for the initial resuscitation of hypovolemic shock. Methods: The study was carried out in 6 dogs 1-~h HS was induced by Jet momentary hemorrhage (H) from a. femoralls (the bloodloss volume made 29.8+1.9 ml/kg). The treatment was begun after 7.U+O.2 hrs of H. "SL", created on the basis of-sorblt and natrium lactate (1800 mOsm/L) was injected into v. femofalls at the dose of IO.0 ml/kg. Results: It is established that before treatmen-~rterial blood and central venous pressures (ABP, CVP) diminished to 30.0 mm Hg and -0.6+0.2 cm H20 (P4.O01), while heart rate (HR)-increased to 190.0+9.6 per min (P<.O01). By this the indices of ~latelet counts (PIC) and plasma fibrinogen (PF) lowered by 42.2% (P<.I) and 6.4% (P~.05), while fibrin degradation products (FDP) enlarged by 215.6% (P~ .001). After 30-40 min of treatment termination ABP and CVP increased to 98.3+6.1 mmHg and 4.1+O.2 cm H20 (P<.O01), and ~[R diminished to T80.0+6.3 per min (P>.5). At the same time the indTces of PIC and PF enlarged by 36.4% and 2.6% (P>.I), while FDP diminished by 8.2% (P>.I). One of 6 dogs survived. Life duration of the other 5 dogs was 3.9+1.2 hrs. Conclusions: The obtained data are ~he evidence of normalizing influence of "SL" on CVS and HSS, and allow to recommend it as a mean of initial resuscitation of HS in clinic. Oblectives: We prospectively studied 64 ICU patients with severe head injury (HI), which cerebral lesions monitorized with SjO2 through oplJcal fiber and the cerebral flux with TCD. Methods: since january 1990 until june 1994, we collected 152 Ht admitted to the ICU, and 64 of them monitorized with optical fiber in the right jugular bulb and TCD. All patients needed mechanical ventilation related to GCS <__ 8, with CT in admission (classifing lesions according to Marshall and al.) . We related the final results to the evolution of SjO 2 and TCD, with other monitorizing methods like GCS, CT and ICP. ~sults: Conclusions: In patients with GCS _< 8, SjO2 is useful to evaluate the evolution towards vegetative state, still more in cases with CT type II in admission and higher Apache Ill. Elevation of ICP implies an evolutive nsk to brain death and data of TCD is a good indicator of brain death, The complete monitorization of these patients can improve the therapeutic control of this neurologic problem, , (16m,6f) , (m. age: 39+4 years), divided in two groups (A and B) under specific criteria(tremor and/or fever during admission in I.C.U., or not). The Injury Severity Score was >25 in all studied patients. Tbe group A (9 m, 41") had no tremor and/or fever on admisskm, while em group B (Tin, 211 the above criteria were Ix)sitive. Bhx~d samplings were taken 2-9 hours after accident and 10-25 rain. after admisskm in I.C.U. Micro-Eli~ method was used for measuring cytokinc-levcls. Statistic analysis was performed by Studcnt-T test. As control group, 25 healthy people were examined. _Resu!_ts-IL-lct, IL-II~, IL-2 and TNF-tt levels were similar to control group levels in both groups A and B. I!,-6 and G-CSF levels were found increased in both groups (p<0JXJl), while IL-6 levels were statistically significant comparing to group A. in con_tin_skin, during immediate post4raumatic period,proinflamatory cylokines IL-I~, IL-I~ and TNF.-ct, produced in an earlier stage than 11,.6, cannot be detected,whereas 11.-6 was increased significantly, especially in group B. G-CSF was fimnd in increawal levels in both gr(mps, without statistically significant difference between gnmps A and I|. Objectives-l~valantc proteolitic activity, disorders in" eariy, period after combined trauma and p(~.ssibilit, i' of their correction by injection of proteo[ysis inhibitors contrycal and S-fto~:nracil in combination with driving an isotonic snlu~ion of sodLum chloride and polig[ucine. Methods: BiochemicaI studies of proteolitic activity in dogs with limited deep burn and acute bloodloss, . Result:s: In case of deep 5% burn, cornplicated by bloodshed the of blood grows at 6-7 times. It; is the restdt of the pancreas glandischemi demage, caused by the centralised circulation of blood and intensifies the deviations of haemodiaamics and albumin exchange. The degree of endogene intoxication by mean mofecular peptides which are the products of albumin decay reses to 30%, and 77% in 3 hours. In 3 hours after the trauma the-process is accompanied b3! 59,6% lower inhibitory activity of blood, where as at the peak of the trauma it was 14,5~ higher. That proves the nnfavuurahle process of the shock in case a combined trauma. Conclusion: The vein injection of 'proteolysis inhihitotz cnntrycal and 5-fforuraei[ in cumbination with driving an isotonic solution of sodium chloride and p.dligh]cine to refill lhe loss of blood helps to lower at 2 times the profeolitic activity of blood. But it still remains above the initial level. The degree of endogene intoxication lowers at 2 times; [15emodinamics aml albumin exchange stahilised. Objectives: Nimodipine, a known calcium antagonist, has been shown to dispose a beneficial effect on patients with subarachnoid hemorrhage, but its efficacy on traumatic or spontaneous intracerebral hematoma has not been justified. Therefore, we studied the effect of nimodipine on the histopathological changes following an experimental intracerebral haematoma in rabbits. Methods: Twenty-three New Zealand albin rabbits of both sexes, weighing 2-3,5 kgr and at age of 4-6 months were anesthetized and a small burr hold in the left parietal aerea was carried out under aseptic conditions. The dura was opened and 0.1 ml (this volume assuring a normal incranial pressure after Kaufman 1985) of autologous blood was injected into a depth of 3 mm via a needle of 0.36 mm bore. The wound was closed and the animals were left to recover. Nimodipine, of 2,1 mg/kgr of by weight per day was given via a nasogastric tube to fifteen animals for a period of time of fifteen days (GROUP B). Six rabbits were given water and served as control (GROUP A). Both groups of animals weie sacrified on the fifteenth day, their brains were removed and immersed into 10% formalin solution. Tissue sections of 5 ~ were embedded into paraphin and stained with haematoxyline and eosin, Mason and GFAP stain for gliac cells. Results: Two animals died after the surgical procedure, because they developed large intracerebral bematoma. No animal developed neurological deficit except one of group A which manifested a right side hemiparesis. The results of the bistopathological changes are the following: i) the mean -+ SD diameter of the lesions in the group A was 260 --. 26 ~t while that of group B was 76 + 12 ~t (p<0,01) ii) secondary ischaemic neural tissue changes, characterized by the extravasatlon of red cells, the presence of haemosiderin-containing macrophages and Signs of low grade inflammation zpredominated in the specimens of group A and were totaly absent from those of group B. iii) A ring of gliac hyperplasia and a low grade local fibrosis was found, encircling the lesions in the specimens of group A in contrast to those of group B. Conclusions: Nimodipine when administered in rabbits following the development of a non increasing the ICP experimental intracerebral haematoma, prevents the extention and the severity of the lesion. Objectives: To study the efficacy and side effects of adding intramuscular Clonidine (Clophelinum) to analgesic regimen in early management of patients with serious burn injury. Methods: 20 pts with 20-40% BSA second to third degree flame burns (respiratory tact injury excluded) 19 to 61 yrs of age were randomised to study (n=10) and control (n=10) groups. Burn shock was treated with hypertonic saline -bicarbonate solutions (250 mmol/L Na +) 2ml/kg/%BSA for the first 24 hours and 1 ml/kg/%BSA for second day. Analgesia in control group for the first 48 hours was provided by regular 6 hourly intramuscular administration of 20 mg of Morphine sulphate and 500 mg of analgesic -antipyretic Analgin with 10 mg of Diphenhydramine (Dimedrol). From the 3rd day regular administration of Morphine was finished. In the study group 100 Ixg of Clonidine was added 8-hourly for 72 hours and dose of Morphine halved. VAS, Verbal rating scale for sedation (VRS, 1 -5), sleeping time, SpO2, HR, BP, diuresis, vomiting and other complications were comparatively evaluated during patients' stay in ICU. Results: Addition of 300 ~g of intramuscular Clonidine daily allowed to achieve better analgesia and sedation with halved consumption of Morphine. Mean VRS in study group for the first 3 days was 3.1 -3.3 vs 1.3 -1.7 in control group with twice longer sleeping time. There was significantly less tachycardia in study group; dynamics of BP for the first 24 hours did not differ considerably; later, there, was tendency for hypotension in study group without adverse effects on diuresis or other indices of tissue perfusion. Because of high incidence of chronic ethanol abuse among study population 7 pts of control group suffered from psychomotor agitation or delirium, probably as a sign of alcohol withdrawal syndrome (AWS). This made regular evaluation of VAS impossible. In the study group only 1 pt showed sign of AWS. Mean VAS score was in 2.4 -2.9 range for first 3 postburn days. 7 pts appeared excessively drowsy due to Clonidine, but it had no adverse effect on their overall clinical course. Mean SpO 2 values in study group were in 95 -98 % range, among controls 90 -95 %; vomiting was absent in. CIonidine group vs 4 cases among controls Conclusions: Clonidine could be a valuable addition to analgesic -sedative regimen in burns, especially for prevention of AWS and deserves further study in this regard. Hemodialysis -Hemoflltration modifications and/or Intratracheal Gas Insuflation have been recently used for blood gas exchange in several models of respiratory failure. Objectives: Evaluate the combination of CAVH-M and IGI for respiratory support in experimental Acute Lung Injury. Methods: Five mongrel dogs (22-+4Kgr) were Mechanically Ventilated inroom air, Paralysed, Heparinized, Connected with a CAVH-M system (Diafilter-30 polysulphone membrane) and remained stable for one hour (PaO~= 98.8• PeCO2=34-+8mmHg, PH=7.37-+0.07, BP=137-+12mmHg and PAP=15-+2mmHg). ALl was induced two hours after oleic acid infusion (0.07ml/Kgr) into the Pulmonary Artery (Poo~=46.6_+6 -P<0.001, PaCO~-50.2_+10 -P<0.05, PH=7.10-+0.10 -P<0.01, BP=158-+25 -P=NS, and PAP=29_+5 -P<0.01). FiO 2 70% for the next 30 minutes did not significantly altered the b3ood gas abnormalities. Afterwards, pure oxygen applied simultaneously a) through the inlet of the filtrate's compartment of the hemofilter (2L/min) while filtrate and gas were removed from the outlet port (bypass flow 220 ml/min) b) through a thin Intratracheal catheter positioned 2cm above the carina (4L/min). The FIO 2 given through the Ventilator readjusted to 21%. Results Replacement fluids/filtrate during the next four hours were not exceed 0.7L/hour, whilst the blood gases and pressures were improved as follow: CAVH-inlet:PaO.=88. OBJECTIVE. To compare the changes in humoral immunity in trauma patients following massive transfusion of autologous and homologous blood. METHODS. We studied 3 randomised clinical groups of patients each containing 24 patients with trauma and operation of large arterial vessels. The amount of autologous or homologous blood transfused to the patients was exceeding 1 500 ml, while the patients in the control group did not recieve blood or blood products. RESULTS. We recorded most pronounced and characteristic changes on the 1-st and on the 7-th day in the group of patients recieving homologous blood transfusion, i.e. decreased amount of IgG,IgA,IgM,C5 and C4 fractions of the complement system, haptoglobin and significant and sustained rise of circulating immune complexes up to the end of the study period. In the control group of patients the decrease was weaker and lasted only during the 1-st post-operative day; the dynamics of the circulating immune complexes level were almost the same as in the first group of patients. In the group of patients recieving autologous blood transfusion, the parameter values did not change significantly from preexisting levels after the 1-st day, while on the 7-th and on the 30-th day showed a tendency towards aslight rise. CONCLUSIONS. Autologous blood has a favourable effect upon humoral immunity and should be the transfusion medium of choice in cases where autologous blood reinfusion is technically possible. Ivan Petkov, M.D., Rumen Farashev, M.D. and Dimitar Terziiski, M. D. Medicine, Military Medical Academy, 3 G. Sofiiski str.,1606 Sofia, BULGARIA OBJECTIVE. The amount of blood lost during trauma and operation could hardly be forseen and donor blood supplies are not always available in sufficient amounts. Rare blood group types and/or unexpected haemorrhage pose a great challenge to the transfusion therapy and the methods of intraoperative autologous blood transfusion. METHODS. We report a case of a 18-year old male patient with extremely massive intraabdominal haemorrhage ( 7 300 m( blood loss ) during an abdominal aorta reconstruction following a traumatic injury of the abdominal aorta. We achieved a successful reinfusion of 6 600 ml of autologous blood using an original autotransfusion system developed by us ( pat. No 95311/ 11.10.1991) . RESULTS AND CONCLUSIONS. The autotogous blood in the case reported here was the only and the most suitable transfusion medium for the rapid intraoperative compensation of the acute haemorrhage and the favourable outcome of the patient. The post-operative period was smooth and no significant disorders in the clinical course as well as in the laboratory tests ( morphological,biochemical,coagulation and immunological) were recorded. There were no complications during the postoperative period despite the fact that the amount of blood reinfused to the patient was slightly exceeding his own volume of circulating blood. OBJECTIVE. The haemoglobin concentration and the perfusion pressure value could not be the only criteria for the early signs of tissue and organ dysfunction. Because of this, we employed the extensive monitoring of oxygen transport during severe trauma in order to. achieve dynamic evaluation of physiologic compensatory mechanisms and to assess the efficacy of intensive care management. METHODS. We conducted a prospective controlled trial on the blood oxygenation, oxygen transport and tissue perfusion during the first 3 days after the trauma in 20 patients with polytrauma. We used a Swan -Ganz pulmonary artery catheter (Beckton -Dickinson, U.S.A.), Deseret 1000 Cardiac Output Computer (Medical Inc., U.S.A.) and Hewlett -Packard Monitor (Hewlett -Packard, Germany) to measure and calculate all the parameter values. The severity of the injury was assessed using the APACHE II score system. All the patients had scores over 18. RESULTS. The results show a significant decrease in the arterial blood oxygen content and in the arterio-venous difference, as well as an increase in alveolo-arterial oxygen difference and in the transpulmonary right-to-left shunt. The tissue oxygen supply and the tissue oxygen consumption reveal a tendency towards a decrease below the physiologic minimum of adeqate values. The erythrocyte current velocity and the ratio between oxygen transport and erythrocyte current velocity also decrease inspite of the optimal blood rheology. CONCLUSIONS. The dynamics in the parameters values are most pronounced between the 2-nd and the 18-th hr after trauma, which predisposes patients to the risk of developing stable hypoxemia and characterizes this period as the most critical for tissue metabolism and organ dysfunction. Posttraumatic changes in immune mechanisms in lung compartment in trauma were analyzed in AO and DA inbred strains of rats which differ in their immunological reactivity: the former being low responder and lat-~er hiperresponsive. Methods: The levels of TNF-alpha activity in the 24 supernatants of cultured lung lobes and dynamics of cells migration from tissue explants in 6h lung cultures were assessed in AO and DA rats subject ted to severe burn trauma. Results: Increased levels of TNF activity (160+3 pg/ml compared to 50+4.9 pg/ml in control) were found od day 3 following trauma in lung sups of AO rats while no changes in the levels of activity of this cytokine were found in lung-sups od DA rats More pronounced extent and dynamics of cell emigration were noted in DA rats, while almost unchanged in AO rats Sharp rise in PMN percentages 3h following trauma (60-70% compared to rare PMNs in control), followed by increase in lymphocyte numbers at later time points among lung cell emigrants was detected in AO rats. Slower but persistent increase (25%, 3h following trauma and 60% and 50% on days 1 and 3 after trauma infliction, respectively) in PMN numbers among DA lung cell emigrants was detected, which appeared to be activated, as judged by their NBT reduction capacity. Increased percentages of peripheral blood PMNs and increased state of leukocyte aggregation/adhesion were detected in both strains, but different levels of plasma TNF: increased levels in AO rats on days 1 and 3 following trauma, and initially but persistently high levels of plasma TNF alpha in DA rats (4-5 fold higher compared to initial levels in AO rats). Conclusions:Different patterns of local (lung) and systemic changes in cell numbers and cytokine levels implicate differential posttraumatic migratory capacity of PMNs vs. lymphocytes in lungs in AO and DA rats. EARLY DIAGNOSIS OF ACUTE INTESTINAL ISCHEMIA BY COLOR DOPPLER SONOGRAPHY E. Danse, B.Van Beers, P.Goffette, F.Hammer,AAV.Dardenne, F.Thys, P-F.Laterre, M,S. Reynaert, .LPringot Dept of Radiology (ProfB.Maldague) and Dept of Intensive Care ( Prof M,S.Reynaert), St.Luc Univ.Hospital, Brussels, Belgium Ob Emergeny medical squad service is the most important segment in the process of saving the people, in the cases of mass accidents, like industrial accidents caused by the: explosion, fire, chemical poisoning, traffic accident, elemental catastrophes and the war. Because of that, each Emergency medical squad service needs to have in its motor-pool vehicle for the mass accidents/ for provoding at least 100 people, wounded as well as the people became ill/. Objectives: Presentation of such special vehicle, produced by "Zastava-Kamioni" and it's medical-technical equipment. Methods: Descriptive and comparative analysis of the medical and technical characteristics, based on the actual norms/din, 75080, ISO 9001, YUS.../ Results: On the base of doctrinaired requirements of the emergency medical squad in the case of mass accidents, our researches resulted in the following medical and technical characteristics -The vehicles for mass accidents are GVW/with a payload off cca 5-8t, with the fixed, closed body, type: universal van, -Technical equipment aggregates, stretches, anti-fire device, equipment for pitching the tent and for maintaing technical conditions of the work -Medical equipment: Linen bags with complete sets of bandage material, means for the reanimation and immobilization, for the infusion, medical instruments and remedies as well as the tent for lodging at least 50 wounded and sik people. In Federal Republic Yugoslavia, it was proposed 24 such vehicles for the emergency medical squad needs. Conclusion: We suggest to introduce this vehicle in the production range of the ambulance vehicles for saving, especially in the circles where can occur serious accidents. Introduction : Carbon monoxide (CO) poisoning commonly generates central nervous system abnormalities though an important cardiac morbidity and mortality must be considered. Long-term exposure to CO with COHb levels < 30% may be more dangerous than short-term levels of 45-50%. We report a case of an adolescent who after prolonged exposure to CO developed a severe reversible cardiac dysfunction with low levels of bloed COHE C a.ase History : A 15 year old boy was found comatose at home. His mother in the neighbouring bathroom died severn hours earlier of what was later proven to be a CO intoxication. On arrival the GCS was 8/15 and the patient was breathing spontaneously. A postictal status with eventual postanoxic encephalopathy was suspected. A COh'b level of 10% was objectivated. The cardiorespiratory situation quickly deteriorated requiring mechanical ventilation. Chest X-ray showed diffuse bilateral patchy infiltrates. ECG revealed signs of ischemia. Severe left ventricular dysfunction was evidenced by pulmonary artery catheterisation and echecardiography and later by isotopic angiography (LVEF 20%). Treatment was intensified with inotropic support, intta-aortic balloon counterpulsation and oxygen therapy. The clinical course was further complicated by a crush syndrome and renal failure. The patient's condition gradually improved and he fully recovered without any residual lesions (LWF 80 %) Conclusion : Even after prolonged exposure COHb levels can be misleadingly low. High tissue levels of accumulated CO can be associated with coma and fulminant cardiorespiratory failure requiring advanced life support facilities. Introduction : Both neuroleptics (NLP) and tricyclic antidepressive agents (TCA) can induce arrhythmias, prolongation of the QT segment and the PR interval and hypotension. We report a case illustrating that combined overdose of these agents increases the toxicity of each compound and the risk for adverse cardiac events. .C, gse History : A 44 year old male ingested 2500 mg doxepin (SinequanR), a TCA and 3500 mg prothipendyl (DominalR), a potent NLP in an attempted suicide. Upon arrival in the Emergency Department the patient was unconscious (GCS 6/15), breathing superficially, and presenting signs of recent vomiting. Physical examination revealed a taehycardia of 140 b.p.m., an arterial blood pressure of 90/70 mmH4g. ECG showed a brood QRS complex tachycardia. A chest X-ray revealed the presence of an aspiration pneumonia. Laboratory investigation demonstrated increased levels of crcatine phosphokinase, lactate dehydrogenase and aspartate transaminase ; hyperglycemia and leucocytosis were present. The plasma concentrations of doxepin and prothipendyl were respectively 410 gg/L (toxic level 5130 #g/L) and 3900 I.tg/L (no reference). Treatment consisted of mechanical ventilation, gaslric lavage and administration of activated charcoal and IV fluids and antibiotics. A hemodynamically well tolerated veatricular tachycardia developed 1 1/2 h later. NaHCO3 (250 mEq/24 h) was administrated inducing an ectopic atrial tachycardia with a normal QRS complex and prolonged QT. 8 h after admission a normal sinus rhythm was present; the prolongation of the QT segment persisted for 2 days. The patient fully recovered. Conclusion : The treatment with NaHCO~, alkalizing the blood and thus increasing the protein binding of the tricyclic antidepressant molecule, can readily correct the potentially life-threatening cardiac arrhythmias and therefore should be part of the routine treatment of combined TCA-NLP overdose. Ob/ectives: The development of diabetes insipidus (DI) in patients with brain injury is a known negative prognostic sign. The aim of this study was to investigate whether this is also a reliable early prognostic sign of brain death. Methods: This is a retrospective study of 85 patients treated" during a two year period (1-3-1992 to 1-3-1995) in our I.C.U who meeted the following criteria: (1) coma score _< 8 GCS within the first 24 hours, (2) positive brain CT scan on admission classified according to Marshall's diagnostic classification (classes 1-6), (3) normal renal function during the entire ICU stay. For the definition of DI were used the usual Di criteria plus hypematriaemia (serum Na" >_ 155 mEq/L). Survival was defined up to the 30th postadmission day. Conclusions: According to the findings of this study, the development of Diabetes Insipidus in brain injured patients seems to be a highly specific index for brain death (positive predictive value = 0.95). However, further prospective studies are needed for the definitive evaluation of these findings in such patients. Emergency care in Italy, despite all efforts, is still lacking a nationwide organized prehospital care system and, until today, there are only different regional solutions. The majority of these realities imply rather simple ambulance first-aid services without attending emergency physicians and without resuscitation equipment. The emergency medical service (EMS) system in Falconara M., Italy, was implemented in august 1994 by a collaboration between the School of Anesthesiology and Intensive Care of the University of Ancona and the, already existing, volunteer rescuer organisation "Yellow Cross". According to the guidelines pubblished in 1992 [1] the pre-existing equipment of the volunteers was completed with type A ambulances and 1 special equiped motorcar (patient monitor, defibrillator) for ambulance indipendent physician transpur[. A special data collecting schedule was created to memorise every emergency intervention in a computerised data-base. The intraining members of the School of Anesthesiology and Intensive Care provide 24 hour ready intervention. In this report the Authors describe their experience concerning primary firstaid medical interventions. For a preliminary evaluation we considered, retrospectively, 300 consecutive emergency interventions in the time period from novembre 1, 1994 to april 30, 1995. The emergency physicians treated 131 male (44 %) and 169 female (56%) patients, 15 patients died before hospital admission and 75 patients (25%) were treated at home by the ambulance indipendent physician and did not need any further medical treatment. In the same time period 1 year earlier (november 1993 to april 1994) without attending physician the volunteer rescuers transferred all 257 first-aid interventions to near-by hospitals. We conclude that the presence of an attending, iudipendently motorised physician in emergency interventions is essential for the establishment of precise priorities and may be helpful to reduce hospital admissions by ambulance intervention, though reducing primary" health care costs. We have developed the method of liquor filtration which allows to purify the cerebrospinal liquor from blood and its decay products in the subarachnoid bloodstroke. The hemipermeable dialysis membrane was used as a filter, which lets only in water, electrolytes and substances with small molecular weight. The liquor filtration was used for the treatment of 19 patients with the subarachnoid bloodstrokes of different etiology. The perfusion of liquor was performed at the rate 3 ml/min in the recirculatory mode. Its duration was 180 -240 min depending on the bloodstroke intensity. The filtration makes possible the most completely purifying of the hemorragic liquor, the reducing of the content of blood ceils and its decay products 80 -300 times as less. The monitoring of the patient's state during the perfusion didn't revealed the departure from the norm of the main vital part. The liquor filtration technique compares favo-~ rsbly with the routine method of cleaning by the absence of toxical effect of heterogenous solutions on the central nervous system. The filtrstion of the cerebrospinal liquor in the subarachnoid bloodstroke sllows to provide the the early cleaning of liqour, the regression of meningeal syndrome and to improve the patient's state of health. E3TABLI~MCzr 3BD ~ OF REI~IDNAL MEDICAL FIRST-AID Zhoulittoing, ED., Tan Zi, M.D. Dept. of Sargery, The First Teaching t[ospitat, 29 Yejin-l)a-l)ao, Wuhan 430080 fltlNA Objectives: The medical first-aid is the most important task of the public hc atth department. In general, single hospital model couldn't fatty, effective ly rescue mony severe patients who need mergant treatment in the scene. Bub establishing the medical first-aid network, the severe patients can be given the most timely und the most scientific emergent treatment. So that, the suc cessfut rate of the saving wilt be greatly increased. Methods..; Our hospital is a general big hospitaL. Through developing and cons tructlng for more than ten years, the medical first-aid network distributed art over the area under our jurisdiction has been set up. It consists of thr ee units: the medical first-aid unib center comartd and mnagment unit, co m~nlcation and tiaison unit. The principle of the network operation is with oat having to 90 far to mergoncy, specialized emergency and the best merge acy. ResuLts: The results of the network operation were notable. Cmpari~ the to tat successful rate of the saving (91.6~), the successful rate of saving tra ma (93.~), the suscessfut rate of saving shock (98.~) and the successful rate of cardioputmonary resuscitation (52.4~) daring the three years after t he network operated with these before (86.6~), (91]. 4~), (92. ~) and (4ft. 5~), the successful rates after operating were remrk~iy higher ( P=3) were admitted into the study. The mean ISS was 36.2 (16-75). Thirty-six patients required artificial ventilation for at least 24 hours during the ICU slay. Three of them, who had a tension pneumothorax, were submitted to an emergency thoracic decompression on the field by the Emergency Helicopter team. In 7 cases pneumothorax was diagnosed an the initial CXR 14 more patients had a pnx which was identified only on the CT. In 4 cases a large pnx with lung collapse was missed on the CXR. In our group of severe blunt trauma patients, 54% (24/44) presented a pnx that required the insertion of a thoracic drainage. Only one third (7/21) of the pneumothorax could be recognised on the initial CXR, while other 3 were decompressed before performing the CXR. As many as 58% of the cases of clinically significant pnx were missed on the CXR, and a CT performed soon after admission allowed an early diagnosis bringing to changes in the treatment. (As the patients were mechanically ventilated a chest tube was inserted in all these cases). In 4 cases, the initial CXR overlooked a huge tended pnx which was the cause of hemodynamie instability. Conclusion: in patients with severe blunt chest trauma even large pnx can be missed on the initial CXR. Moreover due to the non compliant compressible lung, a 20% pneumothorax which can be recegnised only on a CT, can bring to high intrapleural pressure altering eardiopulmonary function. N. AndOeli6, 0.~osid, M.ZeSevid, M.Risovid, D.Stepi6, D.Djokid B~rga~yC~qterClinicalCaqterafSerbia, Belgrade Cb~ctives:~lis study ~ the use of ~rq]ofol earbired with k~t~ine (aq a~sJgh~ic s@~qt widn inirJrsic armlgesic pro~mities) or with fsqtmTyl,with psrtial aZgmsis an hgenxlyn-a~ic ~ durirg ~ ~ re:~ver~ f~m ~ in hXh ~ of ~ti~. ~: Yali~mial and ~bod:30 a~it p~tie~ts A~ I-II were included in ibis shxly. Patients were rsrd]nly dieided in two ~ns. All D~tie~ts ~me given 5-5 prcpofol bolus doses (o,5 ~gkg) for ird~iQn of ~. ~ia ~s m~sJn~ with an infusion 6 ~ ~ropafol. As sdflitianal were given fan-i~l (o,2 n]g) ~tely before ~ anJ trad~e~ irfoJoation followad by feasted bolus of o,i mg in ~ro4o l.Patients in gr~4o 2 received i~ (an initial bolus dose of 35 rg slowly intcavax~ 8rd 25 mg as infusion over ~0 rain) .Infusions Of pro~fol or ImcPofol with kg~mine ~ stopFsJ 10-15 rain ]:~o~ extuhation.Arterial blood ~ (sistolic arterial blood preassu-re~ZAP,mean ~rTerial blood pr~,d~lic arterial preassure-[ZP a~ h~art rate-~) ~ m~ before induction Of a~ iO, 30 snd 60 rain aftem ~ intutation. Results: Arterial blood preasstre ~s decreases duri~ irn~ction of sn~Wd~sia in hY~ ~n~s,tnt mare in th~ ~ who r~eived fsqtanyl.~ere w~s statisticslly sifnific~ntly difemerme dmir~ m~ Of an~ia. Arterial blood r~easatre and heart rate were stable in the t-..e~min 9-~a4~. All th~,fl-e keta'nire grcqo hsd e~rly :~e~y time. Ctrmlusi~s: ~e ombiretion Of proTxfol wilh keta/ne for irduorion a~d ~ Of sn~sd~esis w~s yell accept~ by p~tierfcs anJ coald he ~ as an alterrstive ~o ccnva~icrsl a~es -d~sia. Objectives : Assess the relation between cytokine or endotoxin release and indices of splanchnic malperfasion after hemorragic shock in multiple trauma patients. ]~r study was approved by the local ethical committee. Trauma patients admitted to the emergency room who met the entrance criteria of more than 1 hour MAP < 60 mmHg or use of vasoactive agents or blood lactates > 5 mmol/1 were selected for study. A nasogastric tonometer (Tonometrics, Inc, Plastimed, France) and a Swan Ganz catheter were placed on admission. PHi, lactates, hemodynamics, plasma cytokine and endotoxin concentrations were measured on admission and at 3.6, 12, 24, 48 hrs. An immunoradiometric assay was used to determine plasma concentrations of IL6 (N<0.03ng/ml) and TNFc~ (N<5pg/ml). Plasma endotoxin concentrations were measured using a chromogenic limulus assay (N<0.1EU/ml)(1 Endotoxine Unit= 100pg). Results : 9 severe multiple trauma patients (Age = 42_+18 yrs, ISS = 40-!-_15, SAPS = 19+'~, mean-+SD) were studied. They received 15+6 packed red cells during the first 24h. Mean duration of collapsus before inclusion was 5.1_+2.9 hrs. Death occm'red in ~tients. ~ pglml, *: ng/ml, Etox : endotoxin(EU/ml), Lact: lactate (retool/l) A significant correlation between initial IL6 level and SAPS was observed. In the early post-injury period phi, SaO2, SvO2, VO2 were significantly associated with ;IL6 release (p<0.05 at Ho, H3, H6). Later a significant correlation existed between lactates and II6 (H6, H24). A peak of TNF was detected at 24 and 48 hrs. It was associated with low phi and low arterial pH of the early post-injury period (p<0.05 iat Ho, H3, H6 ,H12, H24) and with high lactate levels of later period (_>H12). Only the late release of endotoxins (I{48) was correlated significantly with initial !oxygea-delivered parameters. iConclusion : There was a marked increase in IL6 in the early phase of trauma . I16 and TNF release after major trauma iwith hemorragic shock is associated with splanchnic malperfusion, as assess by the ivery low values of phi. Lactates seem to be a later indice. Toxic effects are a well-known complication of an overdosage of prescription theophylline. What is less known is that over-the-counter (OTC) asthma medications contain theophylline, and that in some cases this might cause toxic effects. A case seen by us involved toxic effects from theophylline in an OTC medication and to date is the only published case in the English literaturet The rationale for this study was to delineate the OTC products containing theophylline from whatever data sources available. Hyperthermia frequently occurs in intensive care treated patients and intentional application of whole body hyperthermia together with chemotherapy is a therapeutical access to treatment of malignant disorders. Anaesthetic support is required in either condition. Due to the marked decrease in systemic vascular resistance seen in hyperthermia an additional vasodilatory effect of the anaesthetic is unwanted. The vascular effects of anaesthetics in hypertherm organisms is not known in detail. Therefore, we performed an experimental study to detect the effects of inhalational anaesthetics in whole body hyperthermia. In 30 Sprague-Dawley-rats katheters were inserted into trachea, jugular vein, and carotid artery. For continuous monitoring of cardiac output a flow probe was placed around the aortic arch. The rats were mechanically ventilated with different concentrations of inhalational agents in oxygen. We compared the effects of enflurane, isoflurane, and halothane in stepwise increased body temperature by submerging in a temperature controlled water bath. Results: Isoflurane lowers arterial pressure more than halothane or enflurane. The inhalational anaesthetics lower the cardiac output similarily and independently of temperature. Isoflurane decreases systemic vascular resistance independently of core temperature and the decreasing effect of halothane on the resistance is completely abolished in hyperthermia. Conclusions: The influence of hyperthermia on the systemic vascular resistance is dangerous. This allows no additional effect of the anaesthetic management. In spite of the vasodilating effect of inhalational agents in normotherm subjects, this effect is abolished in hypertherms using halothane. The condition of management of analgosedation in hyperthermia is different from normothermia. Objectives: To evaluate a bedside computer processed cerebral function monitor for assessment of brain wave activity when clinical/visual clues are not present. Methods: Ten ICU patients undergoing neuromuscular blockade monitored with the Aspect 1000 brain wave monitor from January 1 to June 1, 1995. Results: Time to onset and depth of sedation were readily apparent to ICU physicians not specifically trained in EEG reading. Objectives: To determine whether non-depolarising neuromuscular blockade reduces oxygen consumption (VO2) in sedated, apnoeic patients. Methods: Haemedynamic. metabolic and oxygen transport variables were determined in 5 sedated, apnoeic patients with severe acute lung injury. All patients were ventilated using a Puritan-Bennett 7200ae ventilator with integrated 7250 metabolic monitor. Inclusion criteria were; 1) stable cardiorespirator s" status; 2) systemic and pulmonary artery catheters already in situ; 3) inspired oxygen < 80%. Patients were sedated with midazolam or propofol to abolish response to verbal stimuli, and sufficient morphine or alfentanil to abolish all spontaneous respiratory efforts. Following baseline measurements, neuromuscular blockade was induced with intravenous vecuronium, 150 ug/kg, followed by an infusion of 80 ug/kg/h to maintain the train-of-four ratio at 0. A further four sets of measured and calculated variables were obtained at 20 min intervals. Results: Statistical analysis was by repeated measures ANOVA. There were no significant changes in any variable over time. The changes in calculated oxygen consumption (VO2FICK) , and measured oxygen consumption (VO2GAS), and in energy expenditure (EE), are shown in the table. Objetive: To study the effects on coronary hemodyrtamics and myocardiai metabolism of administering propofol during postoperation sedation of patients with normal coronary circulation and good ventricular function undergoing cardiac surgery. Patients and methods: 18 patients (12 women and 6 men) undergoing aortic and/or mi~-a/ valvular cardiac surgery were selected, with an ejection fraction greater than 0.5 and normal coronary circulation. For postoperation sedation propofol was administered in 0.5 mg/kg i.v. bolus, followed by a 2.2 mg/kgth perfusion. All data were registered before administering propofol and after 20 minutes, the patients being hemodynamically stable and a rectal temperature of 34 _+ 0.5 -~ Systemic and pulmonary hemodynamics, and global, as well as regional myocardial blood flow, and metabofic variables were measured. Results: The patients studied were about 56 years old, and the average period of aortic cross-clamp was 77.50 min. The adminstering of propofol caused a decrease in the coronary blood flow (-9%), great curonary vein flow (-23%), myocardial oxygen consumption (-14%), regional myocardial oxygen constanption (-11%), myocardial oxygen extraction (-6%), regional myocardial ooxygen extraction (-10%), while coronary vascular resistances and global coronary vascular resistances did not change. Oxygen saturation increased in the coronary sinus (+16%) as well as in the great cardiac vein (+32%). In no patient were significant changes suggestive of myocardial ischemia objectified. There was also found a decrease in systolic (-23%), diastolic (-20%) and mean (-25%) arterial pressure, systemic vascular resistance (-20%), and cardiac output (-8%). Conclusions: In accordance with the clinical conditions of this study, the administering of propofol is not likely to cause changes in coronary autoregulation, oxygenation and myocardial metabolism. Obietive: Analyse the effects of 0.4% "end tidal" isoflurane (sedative dosage) on the metabolism and coronary hemodynamics during the postoperation period of patients undergoing cardiac surgery. Patients and methods: 16 patients (12 women and 4 men) undergoing aortic and/or mitral valvular cardiac surgery, with an ejection fraction greater than 0.5 and normal coronary anatomy, were selected. After the surgical operation, 0.4 "end tidal" isoflurane was administered for postoperadon sedation. The determination of variables to be studied was carried out before and 20 minutes after administering isoflurane, die patients being hemodynamically stable and a rectal temperature of 34 _+ 0.5 -+C. Systemic and pulmonary hemodynamics, and global, as well as regional myocardial blood flow, and metabolic variables were measured. Results: The average age of the patients studied was 57 83 -+ 8.87 years. During surgical operation the period of aortic cross-clamp was 78.56 _+ 32.09 rain. The administering of isoflurane was followed by a statistically significant drop in coronary perfusion pressure (-26%), coronary vascular resistance (-29%), regional coronary vascular resistance (-29 %), regional myocardial oxygen consumption (-7%), regional myocardial oxygen extraction (-6%) and accompanied by a significant rise in oxygen saturation in the coronary sinus (+16%) and in the great cardiac vein (+32%). Myocardial oxygen consumption, myocardial exU'action of lactate and regional myocardial lactate extraction did not change. In no patient were enzyme or electrocardiograph changes objectified. Systolic (-23%), diastolic (-25 %), mean (-25 % ) arterial pressure, and systemic vascular resistances (-28%) decreased, while cardiac output did not. Discussion: The administering of 0.4% "end ddal" isoflurane, in the clinical conditions of this study, produced a decrease in systemic arterial pressure due to a reduction of systemic vascular resistance without deteriorate cardiac output. At coronary circulation level, has and effect on coronary autoregulation but had no effect on oxygenation and myocardial metabolism. The idea of TIVA implies the realisation of major anesthesia components (los of consciousness, neurovegetative inhibition, analgesia, myorelaxatiou, providing the adequate gas-exchange) through i.v. introduction of drugs exclasively. Aim: Providing for the main TIVA components with minimal side effects of the drugs used, taking into consideration the patients characteristics and the surgery specific character. Methods: 78 anaesthesias have been conducted in patients aged 15 75 years (28 females, 50 males), undergoing planned and urgent operations with the pathology of lower, extremities, perinaeum, small pelvis, hypogastrium and with reserved spontaneus respiration against a background of 100 % 02 insnffladon through mask. Operations lasted from 0.5-1.5 h. Anaesthesia adequacy was assested by constant monitoring: "CARDIOCAP" (NIBR HR, RR, SaO2, T), through glykhaemia level and mimicry reactions. Standart premedicatioo of M-cholinolytics (0.01 mg/kg) and H2-blockers (0.3 mg/kg) on the operational table was sumplemented by administration of 0.5-1.0 mg/kg of Lidocaine, 1.50.0 mkg/kg of Clonidine, 0.5-1.0 mg/kg of Pentamidine by the tachifilaxia method. The premedication adequacy was assessed through haemodynamics characteristics. Sedation: 0.05-0.1 mg/kg of DroperidoI, 0.l-0.15 mglkg of Diazepam and analgesia: 2-3 mkg/kg of Phentanyl, 1.0--1.5 mg/kg of Ketamine were introduced fractionally according to indications. Infusion rate of Ringer-Lactat solution was 5-15 ml/kg/h and depended on the intraoperational blood loss volume and on the patients preoperational condition. The duration of postoperative analgesia was registered. Results: Clinical assessment of analgesia according to this techniques allowed to decrease the anaigetics dosage to the subauaesthetic levels. Smooth stabilisation of haemodynamics (BP) at proper age norms in patients with the initial hypertension by the 30-th min. of anaesthesia as well as the absence of its increase in response to the additional introduction of anaesthetic have been achieved. (HR) had no abrupt changes and remained in the range of 70-80 per rain. Adequate external breathing: decrease (RR) by 2-3 per rain., with SaO2 increase from 9446 % to 98-100 %. Hypoventilation was avoided by respirate ventilator. According to unauthentic data the glykhaemia level had been lowered by 10-t5 % to the end of the operation with the initial moderate hyperglykhaemia of up to 10 mmol/L The cutaneous covering grew warm and got pink colouring. No mimicry reactions. In the postoperative period patients were in the superficial sleep state (3-48) and analgesia lasted 6-8 b. There were no complications due to anaesthesia. Conclusion: Combined using of BZ, opiates, neuroleptics potentiate the i.v. anaesthetics effects allowing lowering of each TIVA component dosage and, as a consequence avoiding their negative influence on respiratory and heart vascular systems. Complex application of adrenergetics (therapeutic doses of Cionidine and Pentamini with using of taehfilaxy effects) permitted to provide for analgetic and neurovegetative components of general anaesthesia under subanacsthetic doses of TIVA main components, and manifestation of hyperdynamic reactions of haemodynamics decreased while using of Lidocaine -the economicaI activity of heart-vascular system. Good level of muscle relaxation was achieved allowing for widening of surgical intervention extent without respirator ventilators and inhalation anaesthetics application. Anaesthesia is easily controlled due to fractional introduction of drugs with quick recovery of CNS functions after anaesthesia. Postanaesthetic analgesia is increased while concurrent opiates doses are decreased. Absence of marced haemodynamic, endocrine and metabolic reactions during the operation and after it resulted in shortening the period of patients staying in hospital. A 64 yo white man was admitted to hospital for dyspnea and a productive cough. He had CABG in past, but no recent cardiac ischemia. Physical exam: decreased breath sounds over right lung. Chest XRay: consolidation of right lung. Admission medications included diltiazem, furosemide (both were continued) and trazodone (which was discontinued). Admission ECG: sinus rhythm, QT 0.44/QTc 0.49 sec, with ST and T wave abnormalities similar to prior tracings. He required intubation and mechanical ventilation for progressive hypoventilation and hypoxemia. Between ICU days 8 and 16 he received haloperidol, 10-44 mg/d (cumulative dose 209 rag) for agitation and delirium. ICU day 11: QT 0.46/QTc 0.57 sec. ICU day 12: for better control of delirium, trazodone " 50 mg q hs was added. ICU day 15: he developed frequent nonsustained ventdcular ectopy. ICU day 16: QT 0.70/QTc 0.74 sec, pHa 7.48, PaCO2 50 mm Hg, PaO2 72 mm Hg, K 4.9 mEq/L, Mg 2.0 mEq/L. Later in ICU day 16 the patient had 3 brief episodes of torsades de pointes, each responding to precordial thump, and finally rhythm stabilized with i.v. lidocaine and magnesium. Haloperidor and trazodone were discontinued. ECG was unchanged and myocardial infarction was ruled out. Next day, ICU day 17: QT 0.42/QTc 0.53 sec. Torsades de pointes, a form of ventricular tachycardia characterized by a twisting QRS axis, is commonly associated with QT prolongation. Haloperidol is used frequently in ICU for control of agitation and delirium, with reported doses up to 1000 mg/day. Over past decade, 9 cases of torsades de pointes with prolonged QT related to haloperidol have been reported. Trazodone may also prolong QT and cause ventricular arrhythmias, especially in patients with pre-existing cardiac disease. In this patient, trazodone likely exacerbated QT prolongation from halopeddol leading to torsades de pointes. Critical care physicians must be aware of this interaction. It is imperative to follow the QT interval for patients receiving halopeddol, especially when another drug also known to prolong QT is added. One must consider discontinuing the drug when QT/QTc becomes prolonged. Objectives: Analgesics and intravenous anesthetic drugs are routinely used in critically fll patients, who often suffer from a secondary impairment of the immune system. Previous in vitro studies have demonstrated inhibitory effects of these drugs on polymorpho nuclear cells (PMN). The potentially important role of endothelial cells (EC), however, was not investigated, since suitable test systems were not available until recently. Therefore a physiologically more relevant in vitro migration assay through cultured human endothelial cell monolayers (ECM) we established. Using this assay system, the comparative effects of fenlanyl, sufentanil, propofol and the known PMN inhibitor thiopontal were tested. Methods: Human umbilical vein endothelial cells (HUVEC) were isolated and cultured on microporous membranes (CyclopererM) until an ECM was grown. PMN from male and female volunteers were separated by standard procedures. ECM and PMN were preincubated with clinically relevant concentratious of thiopental (104 M), propofol (4p_g/ml), the solvent of propoful (intralipid), fentanyl (30ng/ml) and sufentanil (Sng/ml). After preincubatiun (ECM 30 minutes, PMN 15 minutes) with the reslx~tive drug, leukocyte migration towards the chemoatfractant FMLP (1O - 7 M) was measured in a two chamber 24 well system for 3 hours. The migration rate of untreated (untr.) and treated (treat.) PMN through untreated and treated ECM were determined. As a control untreated PMN and untreated ECM were used. Results are given as means from 5 independent duplicate determinations and expressed as a percentage of control (Table) . Statistical analysis was done with Student's t-test. Results: Clinical concentrations of fentanyl, sufentanil and prupofol showed similar inhibitor~ effects as the known PIVIN inhibitor thit e 1 ). 77% Conclusions: For the first time we could show that analgesics and anesthetics exert their inhibitory effects not only on PMN, but mainly on the interaction of PMN with endothelial cells. Moreover, we could shmv a significant suppressive effect of the opinids fentanyl and sufentanil on both EC and PMN. The known inhibitory effect of thiopental obtained in EC-free test systems were also confirmed in our physiologically more relevant assay system. Objectives: to investigate when and how sedation is used in a consecutive cohort of patients admitted in a large sample of Italian intensive care units (ICUs), gathered in a network named GiViTI, representative of the Italian ICUs system. Methods; the study called for a recruitment period of one month, from January 10 to February 8, 1994, Data collection included age and other demographic variables, acute diagnostic broad profiles, severity of illness scores, treatments, lenght of stay and vital status at ICU discharge. As concerned sedation, each patient was observed until discharge or for a maximum period of seven days. Information on all the drugs used for analgesia/sedation, the route and modalities of administration, the timing, dosages and purpose of the administration have been recorded. Results: the study involved the cooperation of 138 ICUs, 128 of which enrolled at least one case. The total sample included 2932 patients. Overall, 60.7% of patients analyzed (t780/2932) received at least one prescription of sedative during their stay. Globally, at least one sedative drug was prescribed to these 1780 patients in 5014 days in ICU. Although over 38 drugs were reported to be used, 10 pharmacological principles accounted alone for 89% of all prescriptions. Opioids were actually used in 33% of prescriptions; Propofol in 24% and Benzodiazepine in 18.3%. As regards the way of administration, intravenous administration was applied in 74% of cases and, followed by intramuscular in 17.3%. Moreover, non-steroidal anti-inflammatory drugs (NSAlD) were used in 19% of patients and neuromuscular blockade agents (NMBA) in 23%. Detailed analysis on certain subgroups (surgical, trauma, ventilated patients etc.) have been also carried out in order to describe the practice of sedation in these peculiar subgroups. Findings will be widely discussed during the presentation. Conclusions: These results should be interpreted keeping in mind how peculiar is the intensive care setting compared to many other less complex settings of hospital care. In conclusion we thought it was important to present the data currently available in the most neutral form, to start moving in a direction which will enable us -by means of more specific and detailed studies, and with the cooperation and involvement of all those participating in the project -to shed light on one of the many aspects of medical practice in the field of intensive care which deserve closer attention. Introduction: The aged run perilously high risks in cardiac surgery: among others, of haemodynamic fluctuations, respiratory depresskm and organ failure. Response to anaesthetics is a crucial determinant for post<)perative complications, none the less being reintubation due to mechanical ventilation difficulties which increase morbidity, mortality and Intensive Cdre Unit (ICU) stay. Objective: We wanted to assess our a,aesthesia window (selection, and a view of the induction -extubation period) for predicting safe and swift awaking, thus: ICU dismissal for the aged. Methods: In 1994, 162 selected patients (pts) (>70y, 62f) followed a regular elective cardiac surgery protocol (Propofol given at precisely designated time intervals). Upon 1CU arrival, they were subjected to an admission protocol. Our predictive criteria for early extubation at 8h included: a) alertness and ready response to commands; b) adequate gag reflex and sufficient protection for respirak)ry tract; c) PaO2 >75 mmHg with FlU 2 <0.4; d) stable pH>7.35 with spontaneous respiration; d) stable haemodynamics without dysrhythmias; e) adequate perfusion and diuresis (>1.(I mL/kg/h); f) mediastinal bfeeding<100mL/h for at least 2h; g) normothermia (core temp>36~ and no shivering). Subsequent reintubation was for: 1) RR>35/min; 2) spontancx)us ventilation for 30 rain with PaCO2>50 mmHg; 3) PaO2<50 mmHg with FIO2>0.4; 4) pH>7.45; 5) heart rate>]20 bm; and/or 6) non mental alertness; and 7) other medical disorders, after which adequate weaning therapy was necessary. Then, successful weaning after 24h was considered: 1) spontaneous breathing without any forrn of mechanical assistance; 2) stability in haemodynamics; and 3) elimination of fever threat. Results: 122 pts (75%) were extubated at 8h without complication; 29 other pts (18%) at 8h but had to be reintubated because they were hypoxic and began weaning therapy; finally, they were all re-extubated by 48h. Only 11 pts (7%) proved problematic. Conclusion: A,aesthesia wimhlw options (selectkm, extubation, reintubation and weaning) predicted quick (times Propofol administration) and safe (rigid criteria) extubation (75%=8h and 18%=24h), exempting pts with developed post-operative complications (7%=extubation<72h) unrelated to al~aesthesia window or ICU protocol. Dismissal and recovery then became an abbreviated question of time. Fifisetll P, Domeneg~i ~, Sforzini I., Veronesi I~, Maconi A.G. *, Breg~ Massone P.P 30h [] IC+PCA request CONCLUSIONS:Using E~aprenorphine, a synthetic,long-acting, ago-antagemist opinid drug as analgesic, in the major Surgery we obtained the best clinic results with association of conttheus infusion of haft dose drug with bohts of PCA in the first 15-20 hours and just PCA in the secmad day after surgery when the patient is less sleepy. In this way we dent have a great sav~g of suppled drug but the major well-belng of patient without ~erious side-effects and quick mobilization; the dosage used don't compromise a good awake of patient: all patients are sleepy but ready for answer, no allueinatian, bradipnea but not less than 10 b/m without ipoxia. Also the patient proffered this kind of truit meut than the traditional at demand. The ward staff feel it useful] and rehabl~ The negative feed-back technology of the Electronic Infuser System makes possible to use it safe in the ward with high drug's concentration too. The infusion rate of low dose of drug assure a continuative analgesic covering ~n the first postoperative periad; the PCA mode involves the patient him-self in the managemenl of therapy and enables him to choose the best way to confront the dll~icuity of postoperative period without call medical stall Using PCA-device we have had no probicm~ no accident. ANALGESIA DURING EXTRACORPOREAL SHOOK WAVE LITHOT RIPSY A .Levit, B.Grinbezg Regional Hospital, Ekaterinbu~g, Russia 0b~ectives: Our task was to compare ~he analgetic effect of Norphin and Tramel. Methods: Study was made of two groups of uro-li~patients aged 25-61. Group A (23 patients) received baprenorphine hydrochloride (Norphin) at dosages of #.52• mg/kg. Group B (30 patients) received tramadel hydrochloride (T~aasl) st dosages of 1.50z0.38 mg/kg. Before the procedure diazepam was administrated i.v. (0.24!0.03 mg/kg). Blood saturation (SpOz), hemodynamics incides (BP, HR,SV,CO,SAP,SVR) were examined and the patients' subjective assessments of snsesthesis quality were analyzed. The Hospital Ethics Committee approved the investigation. Results: When using Norphin HR increased by 17.7% on the onset of the procedure while SAP and SV decreased by 8.%% and 9.6%, respectively (p<0.05). However, there were no reliable CO chsnges. SpOz ~educed by @.2% (p<0.05) and remained lower than the initial one after the procedure was oyez. When administrating Tramsl 50 min. after ste~ting the procedure SAP and SVR increased by ~1.2% and 7.3% respectively. SV and CO decreased insignificantly. Nine patients in Group B saffeting some dlscomfo~t needed additional Tm~msl in~ection. In the course of the whole p~oced~e SpO, was constant and was highez than that in ~he case of Nozphin (p. Four subgroups of IGER's members (having access to an ethical library) worked independautly and submitted their reflexions in a tdmestrial plenary session of IGER in the presence of an external chairman, allowing a synthesis. At the issue a report was writted to be used as a reference for bedside and individual decisions. Conclusions : Constitution of IGER seems to improve ethical management in ICU. The first result of IGER is that it is now possible to began collectively a reflexion concerning therapeutic's withholding and withdrawing in ICU. The work is going on and further subjects will be studied. Objectives: 1) To compare the value of heat-moisture exchangers with bacterial filters (HMEF) and without bacterial filters (HME) in the prevention of colonization of ventilator tubing and ventilator-associated respiratory infections. 2) To asses the temperature and relative humidity of inspired all using both types of heat-moisture exchangers. Methods: 48 mechanically ventilated patients were randomized, to either HMEF or HME. Endotraeheal aspirates, pharyngeal swabs and samples from tubing were collected for bacterial cultures on the 1st, 2nd day mechanically ventilation and weekly thereafter. Temperature and relative humidity were measured in 23 patients (13 HMEF and 10 HME) 3 h and 24 h after placing the HME or the HMEF. Results: Both groups were comparable as regards age, mechanical ventilation period, severity score (SAPS II), leukocyte count, and number of patients with prior antibiotic treatment. From the HMEF group, 10 (42%) ventilator tubing yielded microorganisms in, at least, one sample as compared to 7 (29%) of the HME group; p=NS. The incidence of respiratory infection was similar in both groups (25% vs 17%, p:NS, for HMEF and HME respectively). Among the 16 bacterial species isolated from ventilator tubing in the HMEF group, 7 (44%) were not isolated from pharyngeal swabs. A similar ratio was shown in the HME group (6/15, 40%). Both heat-moisture exchangers were efficacious in keeping a good relative humidity of inspired air (97% • vs 96% • 3.%; p=NS, for HMEF and HME respectively). Relative humidity was significantly higher after 3h of mechanical ventilation in the HME group as compared to HME group (28.5% • vs 26.5% • 2%; p=0.03). Conclusions: Both types of heat-moisture exchangers have the same effect on the prevention of colonization of ventilator tubing. Similar relative humidities are achieved when using either type of heat-moisture exchanger. Results: Tumor and nontumer enhrgements of the thyroidea were present in 85~ of the operated, surgicel adrenal disease in io!, hyperplssle or persthyroid gland tumor in 2~ end endocrine pancreatic tumors in 3%. In the Intensive Oere Unit, these patients wore screened by noninwsive monitoring in 85~ of cases: and invasive monitoring was applied in 15% of ceses.The basic noninvesive methods included: Electrocardiogram with standard end precerdial leeds, percutaneous eutomotlc measurement of systolic, diastolic and mean arterial pressure, measurement of hourly diuresis and body temperature, frequency, hearing capacity and rhythm of one s own breathbng BS well as pulse oxymetry. A special plece in monitoring and control of vital parameters in postoperative period belonged to the nurse, thoroughly trained for enelysis end interpretation of the observed parameters which would be discussed in the paper. It has been believed that the leader sits at the pinnacle of power. Over the years, this has proven to produce frustruation and anguish instead of the expected results. Leaders have not been able to produce the changes they know are essential to their organization's survival with this command-and-control paradigm. Through literature reviews and evaluating leadership styles, one can clearly see the most effective form is that of empowering people to a new level of performance -not ordering it. Changing the leadership paradigm to a manner/style that has been shown to be effective and one of people empowerment shifts the focus to personal responsibility for performance. Removing obstae}es~ stimulating self-directed actions, and determining focus and direction are just a few elements used to create the successful environment of empowerment. With increasing pressure in the health care arena, it becomes critical that a leader's job is to get the people to be responsible for their own performance. Developing ownership, creating an environment where people want to be responsible, being a mentor or coach, and learning faster while encouraging others to do so demonstrates the commitment to effective leadership. This presentation will illustrate the critical components that are achieved when every person in the institution is empowered to perform at a level that is directed toward positive, effective results. Herrera M. (MD) . ICU. Hospital Regional. Malaga. Spain. The systems of veno-vanous continuous haemofiltration (WCHF) have a high cost and a limited life span. In an attempt of lengthening their mean life it has been proposed to accomplish programmed washes of the ~-stems. This practice supposes an increase in nursing workload. In order to evaluate the real efficiency of this practice we have accomplished this study. Material: Prospective randomized study of all the filters of VVCHF used during the last year in our ICU. We have determined two groups of filters, in the first (group A) we accomplished washed in a programmed way, and in the other (group B) only when the alarms of the system suggested a clotting of the filter. For the statistical analysis we used the Kaplan-Meier test for survival analysis. Results: We have studied a total of 24 patient submitted to WCHF during the last year. We used a total of 32 filters with this results. Objectives. Sounding out the nurses about the need to inform patients" relatives and the rigth kind of such information, like a preliminary approach to an information cuality assessment, Methods: We inquired all the nurses of the Intensive Care Unit of an Regional Hospital by an semiestructurated questionary which included personal data: age, sex, contractual relation, professional experience.., and opinion data: do you think to inform relatives is a nurse task?. Which of the next informafions do you think is more important?, Please, write others topics about information you think are relevant. We process the data on Epi-Info estatistical program and use X 2 test to compare the results. Results" From 80 nurses of staff 5 refused to flU the quetionary, and 8 were not available. Of the 67 remaining, 71%were v~men and 29% men. the mean age were 31.51% had an svable contract and 499( eventual, the mean professional experience were of 10 years and 44% worked in the Unit since more than 6 years. The 88% answered that offer information to relatives is part of the nurse activities. We did not find differences with nurses who answered negatively comparing by sex, age, contractual relation or proffesional experience. The three information topics found out like more important were: 1) to inform about patient mood. 2) To inform about happenings from the last visit. 3) To inform about dressing instrument required by the patient, Nurses who answered negatively think that to inform is a Doctors task or that nurses are not competent. Conclusion~ Intensive care unit teams (Nurses, Doctors and auxiliar personnel) should get accord on who and how to inform relatives, We consider the nurses' role on information as unquestionable. Objective: Investigate the respiratory and cardiovascular response after discontinuing oxygen therapy durir~ intr~/]o~pital transport. Desiqn: Fifty-one patients (29 male and 22 female, aged 69+2,5 and 73,912,4 years respectively, ~+SYM) being on 02 therapy were studied prospectively in two consecutive intrahospital transports. Oxygen therapy was continued in the first transport while the second one was performed as usually, i,e, without 02. During transport each patient was monitored by pulse oxymeter and Holter whereas arterlal blood gases were tested just before a~xl aft~-trar~portation. Results: Compared to Daseline, Pa02 and Sa02 were signif~canthy decreased in the case of oxygen discontinuation (p<0,00i). PaCO2 was significantly inur~ds~i only in the subgroup of patients with obstructive lun[ disease (p<0,01) . Heart rate increased in all phases of the transport when 02 administratlon was discontinued. Blood pressure remained stable in either case. The percentage of supraventricu!ar extrasysto!es, ectopic v~r[hicui~r contractions and ST-s6~ment depression was progressively increasing and became very high at the end of transport in the case of 02 therapy discontinuation. Other arrhythmias did not change significantly. Conclusion: Discontinuation of oxygen therapy during intrahospital transport causes severe drop of PaO2 and Sa02, increases the heart rate and contributes to the appearance of arrhythmias which were not present before. Methods:For evaluation of the functional state of brain the complex of methods was used,whieh included electro encephalngraphy ( brain mapping ), rheoencephalography, tetrapolar transtorax rheography. For the estimation of humoral status the level of histamine and serotonine, products of free-radical oxidation,enzimatic markers of ishemic damage of brain and of endogenous intoxication was investigated. Results:92 patients with encephalopathies after resuscitation were observed.Asystolia was as a result of:shock, trauma, asphyxia,poisonings,appIication of drugs, eclamp sia,injury of the heart,diseases of fhe cardiac vessels. All patients with postasystolic syndrome entranced in comafose condition.In the 1 group (reconvalescents) the depth of coma by Glasgo~ Pittsburg"s scale was 23,3+-1,78. The duration of coma was from 30 rain. to 48 hour,average 11,9+-4,Sh.ln the 2 group (the deads) the depth of come was 13,8+-0,86.The artificial lung ventilation was used in all patients:in the 1 group 2,64+-0,92 days,in the 2~ 6,1 +-1,1 days.Apallish syndrome developed in 5 cases,in 5 patients diagnozed <,, plasmofllter PMF-800,with effective area-800 cm,the volume of extracorporal contour-60 ml.Such PPH has no the ~ agressive effect,,, as in cases of application another extracorporal methods. This method was incalcated in our practice recently, so results will be reported in further publications. (2). post-operative cerebral neoplasm (1), post-operative subdural hematoma (1). ICP was monitored via a catheter inserted in the lateral ventricle and values were continuously digitally recorded by means of a bedside computer data acquisition system (MacLab). The fiberoptic tracheobroucosenpe, which guided the procedure, was passed between the nasotracheal tube and the trachea in order to avoid hypoventilalion. The patients had stable baseline hemodynaimcs. Propofol infusion and fentanyl boli were administered to mantain stable mean arterial pressure values. Peak (mean(SD)) ICP duping the 30 minutes pre-Ciaglia procedure (baseline values) were compared with values during Ciaglia procedure, and the 30 minutes p0st-Ciaglia procedure. Data were compared with repeated measures ANOVA. Results: Ciaglia procedure duration was (mean(SD)) 30 (14) Objectives: Transient global amnesia (TGA) is a syndrome caracterized by impairment of short-term memory, inability to form new memories, retrograde amnesia and repetitive queries, without other neurological signs and symptoms. The pathophysiology of TGA is unknown; thromboembolic, epileptic, migrainous and metabolic mechanisms have been suggested. To address some of these issues, we undertook a study of 25 cases of TGA in whom we examined clinical, laboratory data, electroencephalogram, CT of the head, ultrasonography ecodoppler. Methods: 25 patients were included in this study: 9 men and 16 women. The mean age was 64 years. All cases underwent a standard clinical examination, electrocardiogram, routinary humoral tests and x-ray, electroencephalogram (EEG), CT scan of the head, ultrasonography ecodoppler. Results': The mean duration of amnesia was 5 h. 32 m. +/-7 h. 10 m. Hypertension was found in 19 patients (76 %), ischemic heart disease in 4 patients (16 %), hypercholesterolemia in 10 patients (40 %), hypertrigliceridemia in 3 patients (12 %), smoking in 2 patients (8 %), atrial fibrillation in 1 patient (4 %), history of epilepsy in 1 patient (4 %), migraine history was not recorded. CT scans of the head showed multiple small deep infarcts in 4 patients (16 %), a single hypodense lesion in 4 patients (16 %). In 11 patients electroencephalogram was normal (44 %), in 8 patients there were widespread nonspecific electrical changes (32 %), in 6 patients there were focal nonspecific EEG abnormalities (24 %). Conclusion: In our study TGA was more common in women (64 %). We showed a prevalence of hypertension, hypercholesterolemia and cerebral infarcts compared to normal controls. We have demonstrated a higher incidence of nonspecific electrical changes in TGA of lower length, while ischemic lesions in CT of the head were more frequent in TGA of greater length. These data seem to be in agreement with the hypothesis that TGA is a heterogeneous clinical syndrome, consisting of pure, epileptic, and ischemic types. However we did not find any correlation useful in discriminating pure from associated TGA forms. From our study it is tempting to speculate that pure TGA is a rare event, underlying still unknown mechanisms wich differ from ischemic, epileptic, migraineous causes. Objectives: Aneurysmal subarachnoid haemorrhage (SAH) is special condition increasing intracranial pressure (ICP) in various ways. At the other hand cerebral vasospasm and related delayed ischaemic deficit (DID) could answer for the poor outcome. Triple H therapy seems today a basic option to prevent DID, but it may increase the ICP worsening the altered intracranial pressure condition and thereby the cerebral perfusion pressure (CPP). Is there any way to individualise the triple H therapy when it is necessary? Methods: Between Sept. 94 March 95 thirty-seven patients with intracranial aneurysms were operated on within 48 hours following SAH. Five patients were in Hunt-Hess IV at admission. All patients received triple H therapy in a preventive fashion following surgery and were monitored by daily transcranial Doppler ultrasonography (TCD). ICP and CPP was measured in twenty-four cases. Twenty-two of them received lumbar liquor drainage (LLD) and nineteen were administered induced hypertension. The other group was treated by basic triple H therapy. Results: In group with monitored ICP the outcome was twenty-one excellent, one poor, two died (one of them died from extracranial decease). In the other group four had excellent, six moderate, two poor outcome, and one died. Conclusion: According to our recent observation the patients can be divided into two groups of therapy. In group I, the patients with elevated TCD values and either low or high ICP reacted to LLD. We are concerned that haemodilution and slight hypervolaemia should dominate in the triple H therapy. In group II patients having high ICP with TCD and/or symptomatic vasospasm should be managed by the induced hypertensionhypervolaemia dominated therapy focusing on CPP (ICP) and focal neurological signs. Air emboli were detected in lo% (n=12) of natients undergoing coronary srtery bypass craftin~ (CABG). Central nervous system ~ysfunction occured in 23~$ of the nstients with air embnli and in none of those ~ithhout air embo!i. Hvtothermia is the classic form of oro-tect~on used dur~nc ~"~" " ~ ~ ca~.,~moDu] :r, on~_,_.7 bj/oass. The surf~eon sho,;,ed thorough!~: evecnnte air from the heart, but the onesthesio!o[[ist can signifieamt!y influence the outcome by emt!oyin2~ methods to detect and treat air emboli. The changes in head rate are primarily due to alterations of autonomic tone. The heart rate variability (HRV), that express the degree of heart rate fluctuation around the mean heart rate, reflectS somehow the condition of central nervous System. HRV may be measured by a number of techniques. Short-term time-domain variables of HRV are reflect generally the vegal activity. In this study the changes in HRV variables of patients with brain damage, and in addition the changes in HRV measurements in comparison with the clinical evolution were evaluated. Eight patient with brain damage and six normal individuals as control group were studied. A elecrocardiographer with availability of computation the sequence of beat-to-beat intervals for one minute was used. The following variables of HRV were measured: 1) standard deviation (SD) of beat to beat R-R interval differences that reflects the respiratory control, 2)the maximum/minimum (max/rain) interval that reflect variability related to baroreflex and thermoregulation and 3) the coel~cient of variation (CV), The results are shown in the In the patients with brain death and in vegetate state there were virtually no HRV. Increased HRV pattern was found with clinical improvement, the changes of HRV precede of the changes of GCS, We conclude that time-domain HRV could reflects the degree of brain damage, it is good prognostic index of the brain damage and may change earlier than the GCS. Objectives: Cerebral CO 2 vasoreactivity is an important determinant of cerebral blood flow (CBF) and has been shown to be of prognostic value in head trauma (Acta Anaesthesiol. Scand. 1991; 35:113-122) . We wondered whether CO 2 vasoreactivity could be selectively altered in one hemisphere in comatose patients. Methods: 6 patients (5M/1F, age 32-65yrs, Glasgow 4-8) in coma due an acute brain lesion (trauma, hemorrhage, or infection) were studied. CBF was measured bilaterally using jugular thermodilution at PaCO 2 25, 30, 35, and 40 mmHg by increasing PICO 2 with mechanical ventilation kept constant. Normal CO 2 vasoreactivity was defined as an increase in CBF of at least I ml/min.100 g per mmHg PaCO 2. Results: 2 patients had normal CO 2 vasoreactivity bilaterally, 2 patients had altered CO 2 vasoreactivity at both sides, and 2 patients had a normal response at one side (left or right) with an altered response on the other side (dght or left). For the 6 patients left CBF was in mean !7 ml/min.100g lower than right CBF (figure Methods: Following institutional approval 4 piglets (body weight 25:tl .5) were anaesthetized by 2% fluothane. A catheter was placed in the right femoral artery for blood pressure monitoring and a fiberoptic catheter (oxymetncs-3 Abbott) was advanced via the right internal jugular vein to the jugular bulb for SjO 2 determinations. Another catheter with a balloon on the tip was advanced in the right atrium via the right femoral vein. A mean arterial pressure (BP) at 25 mmHg was achieved by appropriate balloon inflation for 10 rain and two groups were cleated: i) the hypoxemic group by respirator disconnection (*) and it) the hyperoxemic group by FiO2=l on respirator (o). Samples were obtained at 0 time (1), 10' min at hypoperfusion (2) arid at reperfiJsion at 1' (3), 3' (4) and 10' (5). PaO2, PjO 2 and oxidative brain stress evaluation was performed from jugular bulb blood. The latter included: i) NO synthase (NOS) and xanthine oxidase (XO) activities by a method based on the oxidation of scopoletin detected fluorometrically, it) NO levels estimated as ONOO-by luminol enhanced chemiluminescence in the presence of 500~tM hydrogen peroxide (H202). Resul'~s: The mean PaO 2 was 34 mmt-Ig for Group I and Methods: We retrospectively reviewed all 411 upper GI-endoscopies, performed in the period January 1992-July 1994 in 301 patients (199 men and 102 women) admitted at the 4 ICU's of our hospital. Results: It concerned 129 surgical, 103 medical, 50 eardiological and 19 neurological patients with a mean age of 57.9 yrs (range: 14-91). In 86%, the endoscopy was performed at the ICU and in 14% at the endoscopy department. In 56% of the cases, the endoscopy was primarily diagnostic, of which 70% was performed for localization of upper GI blood loss. In 44 % the endoscopy was primarily thempentic, of which 89 % was performed for placement of a duodenal feeding canula. Location of the upper GI bleeding was: variees (31%), duodenal ulcer (20%), oesophagitis (13%), gastric ulcer (11%), others (13%) and none (10%). As coincidental findings were noted: cesophagitis (37%), gastritis (16%), gastric deer (14%), duodenal ulcer (9%), duodenitis (8%), oesophageal ulcer (7%) and others (8%). Conclusions: There were marked differences in indications and findings of endoscopy at the different ICU's. These differences reflect an admission bias and differences in populations and treatment preferences. Compared with cardiological and neurological ICU's, substantially more endoscopies were performed at surgical and medical ICU's. In a considerable number of cases, no source of upper GI blood loss could be found endoscopicaIiy. When upper GI blood loss was the ICU admission diagnosis, the main cause was Needing varices, which could be controlled endoscopically in the vast majority of cases. When upper GI blood loss was ndt the ICU admission diagnosis, peIgie ulcer and oesophagifis were the main causes of bleeding. Because of the considerable number of coincidental almom~adities found at endoscopy, there is still room for debate whether antacid medication and/or motility stimulating agents should be given prophylactically at ICU's. Many studies have shown that blood lactate levels in survivors and nonsmvivors of traumatic and septic shock are significantly different. The degree of multiple organ failure is related to the duration of lactic acidosis (1). The aim of this study was to evaluate blood lactate level as a prognostic marker of high risk postoperative patients who may benefit from invasive hemodynamic monitoring and aggressive fluids administration and early inotropic support based on oxygen transport parameters. Methods: 32 patients undergoing elective long term vascular and abdominal surgery (ASA I-BI) were studied. Blood lactate levels were measured after ICU admission. In the case of blood lactate level above 2 mmoltl, measurement was repeated every 4 hours for 12 hours or until normaiisation (blood lactate level less than 2 mmol/1). Type of surgery, length of surgery, amount of fluids delivered intraoperatively and postoperatively, hemoglobin levels, hemodynamic variables, diuresis, postoperative complications, length of ICU stay and clinical outcome were recorded. Because no attempts were made to randomisr therapy or change our standard therapy protocol institutional approval was not required. RebUts: The frequency of postoperative complications was 12,5 % and mortafity was 5,5 % in a group of patients with blood lactate level less than 2,5 mmol/l (n = 18). Frequency of complications (62,5 %) was significantly increased in a group of patients with blood lactate levels 2,5-4 mmol/l (n = 8), mortality was 12,5 %. Mortality (60 %) and frequency of complications (80 %) were significantly increased in a group of patients with blood lactate levels above 4 mmol/l (n = 5). Conclusion: Blood lactate levels can serve as early marker of high risk postoperalivr patients and may predict increased risk of postoperative complications mad ~e death. Objective.~: Investigated practicability and clinical value of the routine measurement of hepatic venous oxygen saturation (ShvO2) after major liver surgery, as ShvO 2 is considered an indirect parameter for splanchthc and hepatic blood flow. Methods: 30 consecutive patients were included in this study after liver resections for primary or secondary liver tumors. 5 patients suffered from liver cirrhosis (Childs A). Immediately after post-operative admission on the ICU a PA-catheter ,was inserted under fluoroscopy via the right jugular internal vein into the hepatic vein contralateral to the resection area. Hepatic venous and arterial blood samples were drawn every two hours. ShvO 2 was correlated to the clinical course, macro hemedynamics, ABGs aug other established lab parameters. Results: In 26 out of 30 attempts the catheter could be placed correctly. In four cases after right hemihepatectomy the left hepatic vein could not be intubated due to a dorso-lateral tilting of the left liver. This is also reflected in a significantly longer time of fluoroscopy for catheterization of the left hepatic vein (12.9 _+ %5 rain vs. 3.7 + 2.5 rain; p < 0.001). The procedure requires a total of between 45 and 75 minutes. Relevant clinical complications were not observed except for short term supraventricular arrhythmias during passage of the catheter through the right atrium. Hemodynamics and pulmonary function could be considered normal in all individuals at time of measurement. ShvO 2 showed a span from 27.4% to 90.0% with a mean of 67.0% -+ 10.8%. The following statistically significant findings could be obtained: (a) Patients with liver cirrhosis showed a significantly lower ShvQ than patients without (53.4% • 5.3% vs. 68.7% • 10.1%; p < 0.001). (b) A negative correlation between ShvO 2 immediately after operation and the duration of intraoperative hepatic vascular occlusion could be observed (r = -0.58; p < 0.05). This correlation could also be seen for the first 12 post-operative hours (r = -0.42; p < 0.01). (c) A negative correlation between ShvO2 and the difference between arterial and hepatic venous lactate levels was found (r = -0.39; p < 0.02). Conclusions: The routine measurement of ShvO 2 appears to be a promising extension of post-operative monitoring after major liver surgery. It is a safe method easily feasible on any major surgical ICU though relatively time consuming. A further validation of this method is necessary in larger studies. Therapeutic recommendations on the basis of ShvO 2 findings cannot be given yet. Methods: In 5 cases after major liver resection, in which abnormally low readings of ShvO 2 suggested an impaired hepatic blood flow, PGI 2 was applied at a dose rate of 5 ng/kg/min. As ShvO 2 can be considered an indirect parameter for hepatic blood flow, the effect of PGI 2 infusion on ShvO 2 was measured. Moreover, the changes of macro hemodynamics and pulmonary function were monitored. Results: Before the application of PGI z mean ShvO 2 for all 5 patients .was 54.1% (47-9 -47-3). In three cases without major structural alteration of the remaining liver tissue the continuous intravenous administration of PGI 2 lead to a sustained increase of ShvO z to an average of 67.1% (65.6 -69,1 ). The postoperative course in these three cases was uneventful. In two cases with compensated liver cirrhosis after hepatitis C no change in ShvOz under PGI 2 infusion could be observed. Both patients died 32 and 45 days respectively after operation in protracted liver failure. Side effects of PGI 2 included a slight decrease of systemic and pulmonary vascular resistances. Consequently MAP decreased by up to 10% as did intrapuimonary right-left shunt increase. In none of the observed patients did these side effects posed a limitation of continuous application of PGI z. Conclusions: In patients without structural alteration of the liver the systemic application of prostacyclin at a dose rate of 5 ng/kg/min could significantly increase an abnormally low hepatic venous oxygen saturation after major liver resections, tn two cases of severe liver cirrhosis a similar increase could not be observed. After first clinical investigations and with the results of recent studies in animal further controlled clinical studies of prostacyclin in the postoperative management after liver surgery appear justified. Any delay in gastric emptying can promote micro-aspiration and give rise to ventilator associated nosoarnnial pneumonia. H2-receptor antagonists have been suspected of promoting pneumonia by changing the gastric pH. In a few tri',ds on humans ranitidine was noted to delay gastric emptying. The aim of this prospective, randomised, blinded study was to evaluate in a ventilated ICU population if there was a difference between cimetidine (C) and ranitidine (R) on the gastric filling index (GFI Conclusion: in this population there was no difference in GFI between C and R; however the age and creatinine were significantly different and could have favoured the C group. Also the very long t/2 could have hidden smaller differences between C and R as has been described in volunteers. Between April 22, 1990 and April 19, 1993 , 102 patients with severe acute pancreatitis were admitted to 16 participating hospitals. Patients were entered into the study if severe acute pancreatitis was indicated, on admission, by multiple laboratory criteria (Imrie score >_ 3) and/or computed tomography criteria (Balthazar grade D or E). Patients were randomly assigned to receive standard treatment (control group) or standard treatment plus selective decontamination (norfloxacin, colistin, amphotericin; selective decontamination group). All patients received furl supportive treatment, and surveillance cultures were taken in both groups. Results: Fifty patients were assigned to the selective decontamination group and 52 were assigned to the control group. There were 18 deaths in the control group (35%), compared with 11 deaths (22%) in the selective decontamination group. (Adjusted for Imrie score and Balthazar grade: p = 0.048). This difference was mainly caused by a reduction of late mortality (> 2 weeks) due to significant reduction of gram-negative panreatic infection (p = 0.003). The average number of laparotomies per patient was reduced in patients treated with selective decontamination (p < 0.05). Failure of selective decontamination to prevent secondary gram-negative pancreatic infection with subsequent death was seen in only three patients (6%) and transient gramnegative pancreatic infection was seen in one (2%). In both groups of patients, all gram-negative aerobic pancreatic infection was preceded by colonization of the digestive tract by the same bacteria. Reduction of gram-negative colonization of the digestive tract, preventing subsequent pancreatic infection by means of selective decontamination, significantly reduces morbidity and mortality in patients with severe acute necrotizing pancreatitis. IECO by sodium hypochlorite (NaCIO) infusion is considered to be a model of microsomal oxidation in liver on cytochrome P-450. Active C10 provides oxidation of toxic metabolic products in the blood and exfused during plasmapheresis plasma, and also hydrophobic to hydrofilic transformation of substanses. Sterile NaCIO in necessery concentrations was obtained by electrolysis of saline (0,85-0,9% NaCI solution) in electrochemical set E~IO-4 (Russin,Moscow). Methods: 1. The NaCIO in concentration 600 ragfl (400-800 ml/24h ) was administred into central veins in patients with extensive peritonitis and endotoxicosis 2-3/t. Erytrocytes resistance to NaCIO, circulating blood volume glycemia and hemostasis were initially estimated. 2. After plasmapheresis exfused toxic plasma was mixed with NaCIO conccantration of I000 mg/t in 10:1 ratio in sterile "hemacons".The effectiveness of plasma detoxication and possibility of its reinfusion were evaluated by determination of albumin effective concentration (ECA 35 g/l), the concanlration of medium molecular oligopeptides (MM 0,2) and other biochemical tests (bilimbin, creatinine, carbomide and so on). Results: 1. The intravenous administration of NaC10 excels detoxicative effect of hemosortion by 17-20% provides effictive presentation of protein components and blood cells and improves the transport function of albumin by 37%. 2. The return of exfused plasma after its purification IECO was 70-80%. Only the remaning 20-30% of deficient plasma were compensated by fresh cryoplasma and albumin solutions. Ischemic hepatitis (IH) is a severe complication in critically ill patients. Acute circulatory failure of multiple etiology can lead to splachnic hypoperfusion and cause acute and reversible anoxic damage. Over a period of 26 mos 12 pts, 8 M and 4 F, mean age 64+6.6 yrs developed liver disease compatible with IH. Eight pts had a documented hypotensive episode (six pts with septic shock and two hypovolemic shock), while cardiogenic pulmonary edema in the absence of hypotension was responsible for IH in the remaining four pts. All the pts had a rapid striking elevation of AST, < and LDH with equally rapid resolution of these parameters to near normal wimin 9 days (mean 6.25). The mean peak level of AST, ALT and LDH was 4340 IU/L (range 2105 to 7500), 3453 IU/L (range 1685 to 5150) and 2868 IU/L (range 1440 to 6960) respectively. Serum total bilirubin levels rose transiently with a moan t:eak level of 1.95 mg/dL (range 1.1 to 2.7), while altered coagulation paran-,ete's (PT> 1.5 times normal) was observed in four pts and clinically significant coagulopathy with fibrin degradation products occurred in one pt (8.3%). Renal impairment (Cr>2.0 mg/dl) was manifest in all pts; six pts developed non-oliguric renal failure (50%) while two pts required hemodialysis. Ten lots required vasoconstrictor inotropes [dobutamine (range 3-10pg/kg/min) and dopamine (range 7-25 pg/kg/min), while replacement of circulatory blood volume was performed in two pts with hypovolemic shock. Eight Lots expired (66.6%), but none died as a direct result of hepatic damage. The mortality rate was higher among pts with concurrent renal failure (75%). It is concluded that: 1) IH is not uncommon complication in the ICU with the prognosis depending on the underlying disease. 2) Clinically significant coagulopathy is uncommon complication of IH. 3) Titration of inotropes is required to obtain optimal cardiac output support and subsequently liver blood flow. It is difficult to ascertain the perfusion of free flaps such as jejunal loops after surgery. Objectives: To assess ischaemia as evidenced by intramural pH of jejunal free flaps used for reconstructive surgery following total pharyngolaryngectomy. Methods: The sigmoid pH tonometer ( Tonometrics Inc.,USA ) was used to monitor intramural pH of the jejunal free microvascular flaps ( pHig ) in 15 patients who underwent total pharyngolaryngectomy. A standard general anaesthetic was given and all patients were admitted to the ICU for controlled ventilation and monitoring. All had similar postoperative care. PHig was measured pre, post-revascularization of the flap and on ICU admission, 4, 12 and 24 hours postrevascularization. Objectives: To classificate the wide spectrum of ITC of ANP into distinct pathophysiological patterns according to presentation and course. Patients (pts) and methods: 52 pts, 34 ~(65,4%), 18 (34,6%) were admitted in the ICU because of ANP and acute respiratory failure(ARF), ilean age:54,3• years. Hean stay in ICU:29,2• days. 38 pts were operated, 15 of them twice. Hean value of Ranson's scale:4,4• (2-7). We analyzed hemodynamic measurements,arterial blood gases(ABG), X-ray findings(XRF), CT-scans and operative records. Results: 5 patterns of pleuropulmonary complications were identified: a)Early hypoxia without XRF -33 pts. b)Early ARDS with typical XRF -5 pts(1 died), c)Early ARF with XRF(atelectasis,infiltrates)-15 pts(9 died). d)Late ARDS with typical XRF-32 pts(31 died), e)Pleural effusions in various combinations with the above patterns -38 pts. Overall mortality rate: 41/52 = 78,8%. Conclusions: l)Frequent X-rays and ABG are important for the classification of ITC of ANP. 2)Even though patterns of classification in ANP are not clearly distinguishable,they facilitate an anticipatory management. 3)Deterioration of ABG and XRF indicates that preventive measures for ARF must be intensified and agressive surgical therapy is required. 4)Delay of surgical therapy is related to worse prognosis(p2500 at T 8 while mean output ALP values increased from 3.66 at T O to 197 at T 8. Mean output K + values increased from 3.93 at T O to >8 at T 8. Histology revealed lesions of ischemic necrosis, more prominent after T 6. Conclusion: Results show that the isolated liver graft presents satisfactory function and morphology at least for a five hour perfusion period in the described extracorporeal circuit. Correction of pH contributed to an increase in bile flow. Between 1982 and 1993 the practice of transplantation has changed drasticaIly in Switzerland -besides Kidneys also hearts, heart and lung, lung, Iiver and pancreas transplantation has started in several centers. Major information efforts have been made, organ exchange rules were set up and a national coordination center was initiated. The aim of this retrospective single center study was to assess the influence of transplantation on organ donation. In the past eleven years 205 organs were donated from 458 potential donors i139 single, 66 multi organ donations) analysis of refusal was evaluated categorized into medical and/or familiar reasons. The number of potential donors increased from 28 (1982) ,to 61 (1992) with a concomitant drastic reduction of donations from 64% in 1982 to 26% in 1992; amounting to a net unchanged number of donations over the last 10 years (1982 = 18; 1992 = 17) . The import and export of donor organs was balanced since the introduction of the national coordination center. In contrast multi organ donation increased from 0% in 1986 to 90% in 1993 despite of the more stringeant selection criteria, In conc]usion the introduction of a full range of transplantation procedures at several new university programs and the increase of multi organ donation has not had the forecasted impact on organ donation despite a sustained informative and promotional campaign, Objective: Monitoring hepatic venous oxygen saturation (SvhO2) provides online information about hepatic-splanchnic oxygen supply-demand ratio [1]. Previously, x~ reported hepatic venous catheterization in patients undergoing orthotopic liver traru~lantation (OLT) [2] . In the present study, we assessed the effects of nitroglycerin (NG), a vasudilator that affects the venous capacitance vessels more than arterial vessels and prostaeyclin (PGI2, Flolan r~, Wellcome, UK), an arterial and splanchnic vasodilator on hemodynamies and hepatic venous oxygen saturation (SvhO2) in human liver transplantation. Methods: With institutional approval and informed consent, 14 consecutive patients, mean age 50-2-_10 years, were studied following OLT. Postoperatively, fiberoptic pulmonary artery catheter was inserted into the right hepatic vein. Timed infusions Of NG at a rate of 0.1 gg/kg/min and PGI2 at 5 ng/kg/min were initiated for a 45 rain period. Each sequence was followed by baseline therapy for 45 rain. Results are expressed as mean=tSD. Statistical analysis was performed using Friedman's-two-way-ANOVA-test, significance was accepted at p<0,05. Results: NG at 0.1 gg/kg/min induced a decrease of mean arterial pressure (MAP) (84_49 [baseline] vs. 75+9 mmHg) and pulmonary artery wedge pressure (PCWP) (8J:2 [baseline] vs. 65:1 mmHg). Cardiac index (CI) (5-41 vs. 4+1 l/rain/m2), oxygen delivery index (DO2I) (655-+108 vs. 618+123 mgnfin) and SvhO2 (74_~12 vs. 69-L-_19%) were decreased (p<0.05). PGI2 at 5 ng/kg/min induced a reduction in MAP (73• nm~. _g) and PCWP (6+1 mmHg). CI (6_+1 l/rain/m2), DO2I (7555:135 ml/min) and SvhOz (81+6%) were increased (!o<0.05). 9 Vasedilatation induced by NG decreased systemic oxygen supply and impaired splanclmie oxygenation. 9 PGI2 increased systemic oxygen delivery in parallel with SvhO2, suggesting a corresponding improvement of hepatic-splanchnic okygenation. 9 Thus, if vasedilator therapy is indicated in th6 15orient receiving liver grafting, PGI2 appears to be advantageous. However, due to its platelct aggregation inhibiting properties, the usefulness and safety of PGI2 in OLT patients has still to be determined. Objectives: To analyze the effect of steroid treatment given to donor on the early function of transplanted kidney. Methods: From January, 1994 until now 56 donors were involved into this prospective study. Every other donor was treated with 30 mg/kg Solu-Medrol one hour before organ retrieval. According to the steroid treatment of the donor the recipients were divided into two groups: Group 1 -steroid pretreatment goup (y~=35), and Group 2 -control group (n=37). The donors and the recipients were treated using the same kidney transplantation protocol Onl~r the adults, and the first cadaver kidney transplanted patients were involved into the study. The daily routine parameters were analyzed pre-and intraoperafive, and on the 0-5th, 14th and 30th postoperative days. Results: We could not show any clinically important differences between the two groups in respect of donor parameters. Preoperative, the patients in Group 2 had slightly lower ereatinin level (819 -+ 244 g.,noN vs.923 -+ 254gmol/1) which persisted into the early postoperative phase. The values of the other examined pre-and intmoperativc parameters were almost the same. During the first 5 postoperative days the patients in Group I needed less diuretics (furosemide and renal dose of dopamine) and their sodium excretion was closer to the physiological range than in Group 2. The other parameters did not differ significantly. The less furosemide need in Group ! pe~isted to the end of the first month. Conclusions: According to our data the steroid treatment of the donors improves the early function of the transplanted kidney in some respects. To prove the real benefit of the donor steroid treatment needs more data and further analysis. Objectives: Severe infections may compromize the outcome of liver transplantation..Determination of new parameters may increase the knowledge of pathophysiologic mechanisms and may lead to changes in postoperative therapeutic management of patients at risk. Methods: Between August 1993 and September 1994, 81 patients with 85 transplants were monitored for cytokines and extracellular matrix pammeters on a daily basis. Serious infections (n=10) included microbiologic evidence and more than 2 secondary organ failures. Patients with cholangitis (n=ll) or uneventful postoperative course (n=37) referred as control groups. Results: 1-year patient survival was 88.9% (72/81): 5 patients died due to serious infections, while 4 died for other reasons. Mean bilimbin, sTNF-RII-, IFN-7-, IL-4-, IL-8-, IL-10-, laminin-and neopterin levels were significantly elevated in patients with serious infections compared with patients experiencing mild cholangitis or with an uneventful postoperative course. A further increase of all parameters was observed in patients who subsequently died; TNF-RI/: 28310_+788 pg/ml vs 20452• pg/ml; IFN-7:466_+57 pg/ml vs 4.4-+1.8 pg/ml; IL-4:214-+35 pg/ml vs 148-+29 pg/ml; IL-8:667-+48 pg/ml vs 251_+26 pg/ml; IL-10:149_+52 pg/ml vs 52• pg/ml; laminin: 3010-+312 ng/ml vs 1263-+ 117 ng/ml; neopterin: 247_+37 nmol/1 vs 96_+19 nmolB for non surviving vs-surviving patients. A significant decrease of sialic acid yeas observed in patients with serious infections; and a further decrease occurred in patients who subsequently died: 455-+31 mg/l vs 685• mg/1. Conclusions: The increase or decrease of various cytokines and extracellular matrix parameters may be indicative for severity of infectiolx Routine monitoring of these parameters may improve current diagnostic tools and poss~ly lead to changes in therapeutic management of patients at ~k. Objectives: Evaluation of the cytokine network after liver transplantation may give some insight in pathophysiologic mechanisms of rejection and may lead to detection of patients at high risk. Methods: 81 patients with 85 transplants were monitored for various cytokines on a daily basis between August 1993 and September 1994. Rejection was assessed by histology in combination with clinical signs of rejection and laboratory investigations. Results: During the first postoperative month, 28 patients (34.6%) developed rejection; 14 patients were successfully treated with methylprednisolone (steroid-sensible rejection), while further 14 patients required additional treatment with FK506 or OKT3 (steroid-resistant rejection). 4 patients subsequently developed chronic rejection. Mean levels of various cytokines and extracellular matrix parameters including TNF-RII, IFN-7, IL-IB, IL-2R, IL-4, IL-6, IL-8, hyaluronic acid and neopterin were significantly higher in patients with steroid-resistant than in patients with steroid-sensible rejection. A further increase of some parameters was observed in patients who subsequently developed chronic rejection; bilirubin: 18.2-+4.1 mg/dl vs 11.2-+1.7 rag/all; TNF-RII: 23374-+798 pg/ml vs 18246_+679 pg/ml; IL-8:1024+-192 pg/ml vs 275-+67 pg/ml; neopterin 148_+37 nmol/1 vs 49-+21 nmol/1; hyaluronic acid: 290_+63 ~tg/l vs 223_+28 ~tg/l for patients with chronic versus patients with acute steroid-resistant ~ejection. Sialic acid levels decreased in patients with acute steroidresistant rejection; and a further decrease was observed in patients who tieveloped chronic rejection: 437_+34 mg/l vs 671_+55 mg/1. ~onclusions: Various cytokines and extraeeUular matrix parameters were indicative of severity of rejction. The extensive increase of bilirubin, TNF-II, IL-8, hyaluronic acid and neopterin may indicate subsequent chronic ection. Monitoring of these parameters may, therefore, lead to changes in immunologic management after liver transplantation. Background : Combined kidney and pancreatic transplantation is being performed with increasing frequency in patients with diabetes mellitus and renal failure, as it offers more chances of success and better results than kidney transplantation alone. Mycotic arterial aneurysm constitutes a devastating complication following pancreatic transplantation. All cases of mycotic arterial aneurysms have been however reported with exocrine pancreatic drainage into the gastrointestinal tract. Intervention : We describe a series of 8 consecutive whole kidney-pancreas transplantation performed at the University of Geneva Hospitals (1500 beds) between December 1992 and May 1994. Exocrine pancreatic drainage into the bladder (EPDB) was performed to improve early detection of rejection episodes. EPDB was hypothesized to reduce the risk of contamination from the gastrointestinal tract and the subsequent possible occurrence of potentially fatal infectious complication. In all patients the dual transplantation was performed through a median incision according to the procedure described by Nghiem. Results : Two out of the 8 patients who received kidney-pancreatic transplant developed arterial mycotic aneurysms 15 and 35 days following surgery. Aneurysms developed at the site of the arterial anastomosis used to rearterialize the homograft. Both patients had peritonitis caused by Candida albicans requiring surgical drainage and intravenous antifungal therapy. Rupture with hemorragic shock occured in both patients leading to graft removal in one patient, and three episodes of lffetreateniug hemorragic shock followed by graft failure and removal 32 days after transplantation in the other. Conclusion : Arterial mycotic aneurysm constitutes an early, lifetreatening complication of kidney-pancreatic transplantation; it mandates graft removal. Although exocrine pancreatic drainage into the bladder consitutes a definitive advantage for caller diagnosis of graft rejection, it does not eliminate the risk for retrograde colonization and subsequent severe infection in our experience. S. Bocharov, I. Teterina, Regional Clinical Hospital, Irkutsk, Russia Acute profound loss of blood can result from the very different injuries and hepato-pancreato-duodenaI operations enter such a rank. Ill-timed and inadeguate correction of operation hemorrage is one of the reasons for postoperation complications, including polyorganic insufficiency. The pathogenesis seems to be very complex. In early stages of bleeding the liquid enters the vessel bed, followed by hypoproteinosis and hematocrit fall. However, as decompensation develops, the fluid leaves the vessel system in the result of increasing postcapillary resistance and lowering col-Ioidnooncotic blood pressure (COP). The resulting hypovolemia causes primarily acute disturbance of central hemodynamics and then of microcirculations and transcapillary exchange. Central hemodynamic failure after acute loss of blood manifests itself through cardiac output lowering and capillary blood flow deceleration. Taking into consideration, that 35 % is critical value for CPV loss and for CEV it is 65 %, we consider arising the level of COP to the immediate task. COP raising allows to normalize transcapillary exchange, which we assess through COP and MCP (mean capilary pressure) gradient. The next task is to make up for globular volume till homeostasis providing level. Considerable attention is given to catabolism inhibition and maximum possible enegry provision. Control over high proteolitic activity of blood and callicreinkinin system activity implies direct proteases inhibitors. Reologic, membrane stabilizing, antihypoxanthine and anticoagulant therapies are obligatory. Virehow Clinic, Dept. of Surgery, Humboldt University Berlin, Germany Regarding a high mortality up to 85 % of fulminant hepatic failure orthotopic liver transplantation seems to be the only promising therapeutic approach in many cases. This study shows experiences from a transplantation center. Between June 1991 and April 1995 39 patients suffering fulminant hepatic failure were admitted to our surgical intensive care unit All patients showed severe liver dysfunction with grade II to IV encephalopathy. After a period of diagnostics and conservative treatment ranging from few hours to 10 days (mean 2.4 days) we reported 22 of these patients as possible organ recipients to Eurotransplant. All of these 22 patients were transplanted within 48 hours, 16 (73 %) of them even within 24 hours. The principal aetiologies were hepatitis B (7), hepatitis C (1), NANB hepatitis (5), mushroom poisoning (amanita phalloides 1). After transplantation 2 patients suffered from initial-non-function and underwent re-transplantation. The one-year-survival rate was 82 %, 5 patients died within 3 months after transplantation due to various reasons. 17 patients were not referred for liver transplantation. 10 of them never met transplantation criteria, improved by conventional therapy and could finally be discharged from hospital. The known reasons for liver failure in this group were mushroom poisoning (4), paracetamol intoxication (4) and fulminant hepatitis A (1). 7 patients suffering from fulminant hepatitis (6) or intoxication (1) were excluded from emergency liver transplantation for various contraindications. 6 of these 7 patients (86 %) died despite conventional intensive care. We don't know if some of the patients in the transplantation group would have survived without transplantation, because whenever we decided on transplantation we could perform the operation within 48 hours. But 9 the good survival rate in the transplantation group (82 %) 9 the 100 % recovery rate in the group, where there was no transplant-indication in our opinion 9 and the fatal outcome (86 % mortality) in patients with contraindications are an encouraging proof of a successful therapeutic strategy in acute liver failure. These results are based on a close cooperation between experienced transplant surgeons, hepatologists and intensive care doctors, using sophisticated laboratory and imaging techniques in a specialized center. Introduction: During brain death patients suffer from multiple endocrinologic disturbances. One of the most important are those related with thyroidal axis. It is well described the Euthyroid Sick Syndrome whose more frequent pattern consist of decreased triiodothyronine (T3), increased reverse T3 (rT3) with normal levels of tetraiodothyronine (2"4) and TSH, This lacking in "1"3 levels lead to a change from aerobic to anaerobic metabolism which results in tissular damage. Objective: 1.To study thyroidal pattern in brain death patients potential organ donors. 2.To avoid organ impairment by administration of T3. 3.To study the hemodynamic and hormonal changes after the administration of T3 in these patients. Material and Methods:Population: 22 brain death patients of any etiology potential organ donors admitted to the Intensive Care Unit. Patients were classified in hemodynamically stable (group 1) and unstable (group 2). Group 2 received a bolus of 0.25p.gr/Kg. and a perfusion at a dose of 2-3.5 p.gr]h of T3. Hormonal assays: Total T3 (TT3), Total 2"4 (TT4), TSH. fxee T3 (FT3), free 1"4 (FT4) and rT3 were determine at the moment of clinical brain death (0 hrs) and in group two these assays were repeted at hours 4, 8 and 12. Results: 22 patients (17 male) with a mean age of 33 years (range 17 to 71yrs.) were studied. The clinical brain death was confirm later with other explorations (EEG, doppler). There were 15 patients in group 1 (68,1%) and 7 patients in group 2 (31,8%). Hormonal pattern: At the moment of brain death TT3 was normal in 21 cases (95,4%) and decreased in I (4,6%); TT4 was normal in 9 patients (40,9%) and decreased in 13 (59,1%); FT3 was normal in 3 cases (I3,6%), decreased in 19 (86,4%); Fl'4 was normal in 19 patients (86,4%) , decreased in 3 (13,6%) .rT3 was normal in 17 cases (77,2%) and increased in 5 cases (22,8%). There were no statistically significant differences in hormonal pattern between the two groups. Only T3 levels at hours 0, 4 and 8 were significant in group 2. In the 19 cases with FT3 decreased, the TT3 was normal in 18 (84%) and decreased in 1 (16%), TT4 was decreased in 11 (57,8%) and normal in 8 (42,1%), TSH was decreased in 1I (57,8%), normal in 7 (36,8%) and increased in I(5,2%) and FT4 decreased in 3 (15,7%) and normal in 16 (84,2%) and rT3 was normal in 14 (73,68%) and increased in 5 (26,3%). There were no statistically significant differences in cardiac index, vascular resistances and pulmonary shunt before and after the administration ef T3. Conclusions: 1. The hormonal pattern most often find in brain death patients was: normal TT3, decreased TT4, normal TSH, decreased FT3, normal Fr4 and normal rT3. 2 . There were discrepancies in the values of FT3 and TT3 3. There were no statistically significant differences in hemodynamic and pulmonary parameters. Objectives: Magnetic Resonance Angiographie (MRA), a non-invasive procedure, provides flow-related information additionly to the anatomy of the vascular system. Measurement of signal intensity and edge detection of vessel structures permits to calculate blood flow velocity and vascular diameters. We examined whether cerebral hemodynamic changes by altering the arterial pressure of carbon dioxid (Pace2) could be detected by MRA. Methods: Following institutional approval and informed consent, 10 mechanically ventilated patients without elevated intracraltial pressure underwent MRA with defined periods of hyper-, hypo-and normoventilation (Pace2: 30, 50, 40 mmHg; arterial blood gas probes; AVL). MRA was performed with a 1.5 Tesla Magnetom (Vision, Siemens). Two different MRA techniques were used: a conventional Time-of-Flight-3D-Angiography (TR: 39 ms; TE: 7 ms; FL: 20 deg; Slab: 56 mm) for vessel diameter detection and a Flash-2D-Gradient-Echo-Sequence (TR: 28 ms; TE: 5 ms; FL: 30 dog) for measurements of blood flow velocity. An axial view parallel to the AC-PC-Iine (anteriorposterior-commissur-line) was used for repeated imaging of identical regions of interest 0tOI) of the proximal part of the internal carotid (ICA) and middle cerebral artery (MCA) as well as of peripheral branches of the MCA and the posterior cerebral artery (PCA). Results: Changes of Pace2 correlated with changing signal intensities, whereby under hyperventilation a decrease of 23,7% (P 0.01) and under hypoventilation an increase of 28.4% (P 0.01) was observed compared with normoventilation. Blood pressures were stable throughout the whole study period, Pace2 dependent changes in vessel diameters were more pronounced in peripheral branches of MCA and PCA. A change from normo-to hyperventilation produced a decrease in proximal vessel diameter of -3.5% (P _< 0.01) and in peripheral diameter of -22.2% (P _< 0,001). A change from normo-to hypoventilation produced an increase in proximal diameter of +6.1% (P < 0.05) and of +21.3% (P -< 0.001) in peripheral diameter. Conclusions: Pace2 related changes of cerebral vessel diameter can be easily detected by MRA without injecting a contrast agent. The results confirm that CO2-reactivity is more pronounced in peripheral cerebral vessels, which are subjected to greater changes in diameter than major basal arteries. Hyperventilation leads to a decrease and hypoventilation to an increase in signal intensity thus reflecting the corresponding changes in blood flow velocity, intensive Care Unit (ICU) of "KAT" Hospital, Athens, Greece, Ob!ective$; The value of bronchoscopy in pulmonary atelectasis of ICU patients is under question The presence of an air bronchogram sign in xrays, which is considered as evidence of central bronchus patency, is referred in several studies as a negative criterion for bronchoscopy, whereas its absence as a positive one. It is also referred that air bronchogram sign correlates with delayed resolution of atelectasis, probably because of obstruction of many periferal airways (not central). The purpose of this prospective study was the evaluation of the air bronchogram sign on frontal chest film as a negative criterion for bronchoscopy and as criterion of delayed resolution of atetectasis, Methods: ICU patients with atelectasis were studied prospectively. They underwent bronchoscopy, Bronchoscopic findings, presense of air bronchogram sign, and outcome of atelectasis were recorded, Correlations were made, between: 1) Bronchoscopic potency of airways and air bronchogram sign 2} Resolution time of atelectasis and broncoscopic potency of airways. 3) Resolution time'of atelectasis and air bronchogram sign, Methods of statistical analysis were the T-student test and the Chi square test, Results:The patients were 23, men 19 women 4, Seventeen patients had atelectasis of whole lung, 3 of upper lobe, and 3 of lower lobe. Ten patients had atelectasis in right and 13 in left lung. Eight from 28 patients had air bronchogram sign in x-ray, There was no statistical correlation between air bronchogram sign and bronchoscopic potency of airways [6 from 8 patients with air bronchogram sign (75%) and 11 from 15 without air bronchogram sign (73%), had bronchoscopic potency of airways, p>0.1], Resolution time of atelectasis didn't correlate statistically with bronchoscopic potency of airways (mean resolution time in patients with bronchoscopic potency 2,29 days and in bronchoscopically closed bronchi 2,33 days, p>0,1). There was also not a statistical correlation between resolution time of atelectasis and air bronchogram sign (mean resolution time in patients with air bronchogram sign 2,25 days, and without air bronchogram sign 2,33 days. p>0). Conclusion~i; The presense of an air bronchogram sign in x-ray of ICU patients with atelectasis, does not coexist obligatorily with bronchoscopic patency of airways and cannot be used as a negative criterion for bronchoscopy, neither as a criterion of delayed resolution of atelectasis. Th. Wertgen Chest sonography (CS) is routinely used in our department to examine ICU patients with clinical symptoms of pulmonary embolism, pneumonia, pleural effusion or unclear chest pain. We perform CS with a sector transducer (4.0 MHZ) and a linear transducer (7.0 MHZ) using ACUSON 128XP/10 c. The sonographic signs of pulmonary embolism and infarction are most well demarcated, mainly wedge shaped and triangular pleural based lesions, more roughly structured, observed with a hyperechoic reflex in the center corresponding to the bronchitic (Fig. 1) . Pneumonia is characterized by homogenously hypoechoic, wedge shaped parenchymal lesions, containing air or fluid bronchograms; they move with respiration (Fig. 2) . Pleural effusions are spaces of various echogenicities, from anechoic to homogeneously echogenic, which may contain floating strands or complex septa, located between visceral and parietal pleuras (Fig. 3) . From march 1994 to april 1995 we did 55 examinations by CS in 34 ICU patients (20 male, 14 female; age from 29-87). Patients Examinations Pulmonary embolism 10 16 Pneumonia 7 16 Pleural effusion 13 19 US-guided thoracic punctions were performed in 7 patients. In two patients we found pneumonia or pleural effusion caused by a lung carcinoma. Another two patients showed a normal CS (diagnosis: inflammation of the gall bladder, inflammation of the myocardium). Conclusion: CS is a very useful method for ICU patients with chest diseases. It takes less time and is less expensive than CTand sometimes of a higher diagnostic value than X-ray. Last but not least CS is invaluable for the ICU patient, because the examination is done save and quickly at bed side and the results of CS are very helpful in diagnoses and treatment. Results : Inter-observer reliability was evaluated as an 83 % concordance. Results of the TEE classification were : class 0 : n = 21 (34 %) ; class 00 : n = 13 (21%) ; class 1: n = 7 (12 %) ; class 2 : n = 8 (13 %) class 3 : n = 12 (20 %). Therapeutic implications of TEE in class 3 patients were : cardiac surgery in 5 patients (two cases of acute mitral regurgitation, two valvular abscesses and one hematoma compressing the left atrium), discontinuation of PEEP in one ventilated patient with an atrial septal defect, weaning of mechanical ventilation in one patient with an atrial septal defect, prescription of antimicrobial therapy in 8 patients with endocarditis and prescription of anticoagulant therapy in 2 patients with left atrial thrombus. The only noteworthy complication was a case of spontaneously resolving supraventrieular tachycardia. Conclusion : TEE is safe and well tolerated, and is useful in the management of ICU patients with shock, unexplained and severe hypoxemia or suspected endecarditis. The aim of this study was to determine whether ultrasound guidance can help interns to improve the results of jugular vein access in ICU. Methods : In a prospective and randomized study, we compared, in 79 patients admitted to the ICU, an ultrasound-guided method (ultrasound group : 37 patients) with an external landmark guided technique (control group : 42 patients). All jugular vein accesses were performed by young interns with an experience of < 5 procedures. Results : Internal jugular cannulatian vein was acI~ieved in all patients in the ultrasound group and in 10 patients (24 p.cent) in the control group (p < 0.01). Average access time was longer in the control group (235 • 408 sec. vs 95 • 174 see. ; p = 0.06) and puncture of the carotid artery occurred in 5 patients in each group (p = 0.83). 32 patients (86 p.cent) in the ultrasound group and 23 patients (55 p.cent) ia the control group (p < 0.05) were cannulated in 3 rain. or less. The cannula was therefore unabie to be inserted within 3 minutes in 19 patients in the control group, with failure of eannulation in 10 of these patients (53 p.cent). Failure was due to thrombosis (n = 1), small calibre of the internal jugular vein (< 4 ram) (n = 5), abnormal vascular relations (n = 3) or cervical irridation (n = 1). Among the 10 primary failures of cannulation, an internal jugular vein catheter was able to be inserted in 4 cases by an experienced physician on the side initially selected and with ultrasound guidance in 2 cases. The catheter was inserted into the contralateral internal jugular vein under ultrasound guidance in the remaining 4 cases. Jugular cannulation was obtained at the first attempt in 26 p.cent in the control group and 43 p.cent in the ultrasound group. Conclusion : Ultrasound guidance improved the success rate of jugular vein cannulation by inexperienced operators in ICU patients. When the internal jugular vein has not been successfully eannulated within 3 minutes by the external landmark guided technique, the authors recommend the use of the ultrasound guidance. In the majority of cases right atrial or ventricular thrombi represent pulmonary emboli in transit. These may be fatal in patients (pts) treated conservatively with anticoagulation only. In literature the incidence of right heart thrombi in pts with proven pulmonary embolism (PE) is said to be in the range of 3-4%. Extremely mobile, long, worm-shaped masses in the right heart cavities carry an especially high early thrombus-related mortality rate which ranges from 40-50%. Current therapeutic strategies favour fibrinolytic therapy with consecutive anticoagulation. We report five cases (4 male, I female, 55-74 years) of right heart and pulmonary thromboembolism. In these pts diagnosis and regression of thromboemboli following systemic intravenous lysis therapy with recombinant tissue-type plasminogen activator (rt-PA) was documented by transesophageal echocardiography (TEE). A submassive PE occured in 3 pts, a massive PE in 2 pts. One patient (pt) had a cardiac arrest. In all 5 cases TEE clearly identified the extensive thrombns formation in the right-sided cavities of the heart and in the central pulmonary artery in 2 cases. All pts were treated with 100 mg rt-PA, 3 pts in a front-loaded regimen over 90 minutes, 1 pt over 120 minutes, and, due to the life threatening situation, in one case a bolus injection as ultima ratio was performed with no intracerebral bleeding complication. Regression of thromboembolic masses after fibrinolytic therapy was demonstrated by transthoracic and transesophageal echocardingraphy after 1 to 15 hours. All pts survived and were put on coumadine, 1 pt developed an intracerebral bleeding with persistent hemiplegia. Conclusions: The use of thrombolytic therapy is highly efficacious for the therapy of pts with PE and concomitant right or ventricular thrombus formation. Transthoracic and especially transesophageal echocardiography are powerful bed-side diagnostic tools for the immediate diagnosis and follow-up of successful treatment in this life-threatening condition. Although widely used, catheterisation of the femoral vein in the groin using "landmark" technique is frequently complicated by accidental arterial puncture. Suboptimal hygiene and patient discomfort are also associated with this technique. With regard to these last two factors cannulation of the femoral vein 10-20 cm below the inguinal ligament would seem an attractive alternative. As "landmark" technique is not possible for the cannulation of the femoral vein in this part of the thigh, ultrasound was used to locate the vessel and the results of this technique were evaluated. Methods: A portable compact ultrasound device (Site Rite,Dymax Corp.) featuring a 7.5 MHz transducer (ultrasound depth 4-5 cm) fitted with a needle guide and a 6 cm screen was used by residents with no previous experience in ultrasound guided cannulation. Patients consisted of a surgical ICU population. Results: In 46 patients 55 catheters were introduced.In 6 cases more than one (2-4) attempt was made and in 3 patients the procedure was unsuccesfull due to the fact that the vessel was situated out of reach of the ultrasound (vessel depth > 4-5 cm), during the 55 procedures one accidental arterial punction was registered. The catheters remained in situ for a mean of 9 days (range 1-22) and were used for volume suppletion, medication, parenteral nutrition and haemodialysis.Co-Ionisation rates compared to those of subclavian catheters in our ICU. In the first 20 patients 3 cases of asymptomatic thrombosis of the femoral vein were seer on CT-scans performed for other indications, in the following 26 patients duplex scanning performed after removal of the catheter yielded another 3 cases of asymptomatic femoral vein thrombosis. Conclusions: Ultrasound guided femoral vein catheterisation 10-20 cm below the inguinal ligament is a safe and simple technique that can easily be performed by residents without prior experience. The incidence and impact of thrombo-embolic complications associated with this technique are still subject to further investigation. Objectives: To estimate the cost of antibiotherapy (AB-cost) in a multidisciplinary 8-bed greek ICU and to correlate AB-cost with total cost of drugs and consumables and with patient's outcome, severity of illness and type of admission. Methods: Prospective data from 137 consecutive patients admitted to the ICU from 1/10/1994 to 30/3/1995 were studied. A tick chart was designed to record all drugs, materials and consumables regularly used for ICU patients, but did not include low price drugs and consumables, which are provided from hospital's pharmacy as stock and were included in a fixed ICU cost calculated for a 12 month period. The chart also contained demographic details and data necessary for the calculation of several Illness Severity Scoring Systems. Obiectives: Over 3 years evaluate the necessary efforts and expenses to implement a CIS in the routine of a 16-bed StCU. Methods: In June 1992 a commercially available, unix-based CIS was installed on a 16-bed surgical ICU. The goal was a paperless documentation at the bedside. After more than 2 years clinical experience two aspects were investigated: What effort is necessary to install and support a CIS, and what is the benefit for patients and personnel on the ICU? Results: The installation and support of a full-fledged CIS requires a considerable effort: (a) The conceptual framework for the CIS has to be defined. This includes the definition of documentation standards, as well as nursing and therapeutic standards, which is the essential basis for the configuration of any CIS. (b) Configuring a CIS, i.e. "fine-tuning" it to the user's specific needs, is always a laborious task. Moreover, constant maintenance is necessary. These tasks require the following personnel: Experienced health care professionals for defining the conceptual framework, 1-3 trained health care professionals for configuration, 1 system administrator. On a single ICU (12-20 beds) these are not considered full-time jobs. (c) Training is best done employing the "train-the-trainers" approach. (d) Beside the necessary amount of man power and money to install and purchase a CIS, administrative and MIS support is needed, especially when interfaces to the hospital and laboratory information systems have to be set up. In general, a CIS needs the commitment of all people involved. Without a really professional approach with a longterm goal any major CIS can turn into an unnecessary but inevitable night mare. After 3 years clinical use and a thorough implementation of a CIS on a major SICU it can be said that full-fledged CIS offers an opportunity to dramatically improve the working environment on an ICU. Moreover, it adds to patient safety, quality of care and cost efficiency in one of the most advanced and expensive areas of medicine. Conclusion: A major investment in man power and money is necessary to install and maintain a full-fledged CIS. A sincere professional commitment to the goals of a CIS is necessary. In exchange, a well configured and well maintained CIS dramatically improves the quality of therapy and care on the ICU. Even return of investment and financial profitability of a CIS seem feasible todayl From the clinical perspective it appears that the users themselves are the central determinant whether a CIS makes a dream come tree or turns into a night mare. Objectives: To establish a relationship between the activities of the staff and the occurrence of auditory alarms on the I. C.U. ard to evaluate confusion between auditory alarms. Methods: Laboratory based studies which investigated aspects of confusion between alarms in current use on the I. C. U. The observational studies were conducted over an 18 month period and examined the frequency and duration of alarms together with the concurrent activites being undertaken by staff on the unit. The laboratory based studies showed that there were enduring confusions between the alarms on various items of medical equipment, for example a ventilator alarm and an E. C. G. monitor alarm. The results of the observation studies demonstrated that alarms are activated when specific activities are being undertaken by staff. sounds could be used in future recommendations for alarms on medical equipment. Suggestions are also discussed for improving and rationalising auditory warnings in the I. C. U. Obiectives: We investigated inferior petrosal sinus (IPS), the lowest affluent to jugular bulb (JB), as a possible source of contamination of samples in JB for monitoring oxyhemogiobin saturation (SjbO2). Pulling back the catheter the oxyhemoglobin saturation usually rises indicating extracerebral contamination (Jakobs en Met al: J Cereb Blood Flow Metab 1989;9:717). Methods: The study was carried out on patients undergoing IPS sampling to differentiate Cushing disease from ectopic ACTH syndrome and to lateralize any resulting pituitary lesion. We studied the value of oxyhemogiobkn saturation high in JB (SjbO2), at IPS (SipsO2) and at mid jugular vein (5th cervical vertebra) (Smj 02) bilaterally. Results: We found significant differences between right SjbO 2 and both right SipsO 2 (p= 0.007) and right SmjO 2 ( p= 0,017) and between left SjbO 2 and both left SipsO 2 (p= 0.01) and left SmjO 2 (p= 0.017) We did not Fred any difference bilaterally. Objectives: We studied various methods of receiving and editing of clinical datas in critically ill patients (different ethiology). 163 patients were investigated in Regional Intensive Care Center. Methods : The following datas were studied : anamnesis, status praesens objectivus ( organs and systems ) ,. clinical and biochemical markers of critical condition , datas of EEG ,Rheography . The medical information complex contained : 8channel electroencephalograph, 4-channel roencephalograph, AD-converter (16 analog inputs, 12 bit resolution, 60 k Hz), IBM 486 DX2, software includes set of routines for spectral EEG analysis, EEG-mapping, correlative analysis, and brain bloodstream REG-monitoring (written in Turbo Pascal 4.0), expert programs for estimation objective and humoral patient status (written in Clipper 5.0) and statistics. There were used following programme-language instruments : Borland C++ 3.0, Nantucket Clipper 5.01, CA-Clipper Tools II. As the methods of statistical processing of dates were used: T-Students criterion , Fisher criterion, methods of correlation analisis, calculation of the regression levels, dispersion analysis, Results : There was created the optimal structure of hard and sofware complex of search steady objective regularity in dynamic of critically ill patients condition. Conclusion : The created system allowed to value effectiveness of intensive care and give us new opportunities in study pathogenesis of systems disorders in critical condition . Over a five year period a Patient Data Management System has been installed which allows individualised patient data to be accurately collected. Using this data a costing system has been developed which ascribes costs thus: 1. Direct Costs -drugs, fluids, consumables, interventions. These are ascribed to individual patients, according to data collected from the PDMS. 2. Indirect Costs -energy, depreciation, admm costs, maintenance etc. These are summed for the year and ascribed as an overhead per patient day. N.B Staffcusts contain art element of both cost types The aim is to make as many costs as possibie 'direct', hence 'activity costs' have been calculated winch comprise staff time, drugs and consumables -these are direct costs. These costs of patient care are then searnlessly integrated into the financial and budget management of the ICU environment. It was found that by calculating costs in this manner 50% of the total cost of ICU are captured within the 'direct' element, and so are able to be ascribed to individual patients. This is much more accurate than simply dividing the total costs of ~CU by the number of patient days. Temporal costs (variations during patient stay) and cross sectional costs (cost differences between admitting specialities) were also noted with interest. Results of the initial analysis of data captured by the system will be presented. Little is known about the resource costs (not simply cash costs) of ICU. Even less is known about individual patient costs, with previous estimates of these costs varying widely. However, if cost effectiveness studies are to be undertaken accurate calculation of individual, group and total ICU cost is an essential, prerequisite, which, via this system of costing, is now achievable. Information about intensive care of cancer patients is limited in the literature, despite the increasing use of such facilities in oncology over the two last decades. In order to determine if and how critical care facilities can be used specifically for these patients, we performed a world-wide inquiry in anticancer centers selecting the hospitals by using the International Directory of Cancer Institutes and Organizations. We mailed a questionnaire to 146 centers and we received 84 responses (57.5 %). There was at least one uncological (i.e. with > 50 % of cancer patients) ICU in 59 (% % An 18-year old woman with Graves disease presents with sore throat, vomiting, diarrhea, sinus tachycardia at 155/minute and a temperature of 40~ Several weeks before, treatment with Propylthiouraeil had been stopped (rash and fever) and replaced by Methimazole and ledide prior to a minor surgery. However, both drugs were discontinued by the patient two weeks before admission. Shortly after arrival in hospital, patient's condition progressed to respiratory failure (upper airway edema), delirium and shock requiring ICU admission, intubation and resuscitation with fluids and vasopressors. White blood count was 1300/mm ~ with 0 neutrophils. Patient's hemodynamic data showed initial hyperdynamic profile followed by low output state with decreased Sv02 (59%) (N 70-80%) and cardiac index (2,37) (N 2,5-3). Echocardiogram confirmed cardiac chambers dilation as previously described in thyroid storm. Lithium carbonate, corticosteroids, antibiotics and Beta-blocker perfusion were given. Plasmapheresis was started. Free T& (N=9,2-21pmo/L) went from 143,6 to 16,6 after the first two pheresis. After a remarkable clinical recovery, sub-total thyroideetomy was done I0 days after admission. In life-threatening thyroid storm, plasmapheresis is a very effective therapy when anti-thyroid drugs are counterindicated. PURPOSE: To compare the reliability of prognostic indexes in crhically iU patients admitted in an Intesive Care Unit (ICU) who had acute renal failure (ARFI and were treated with different dialytic techniques. MATERIAL and METHODS: 1087 patients were included in a prospective study from June 92 to November 94. 220 patients presented ARF defined by creatinin serum leve(s greater than 150 pmol/l and previous normal levels. Patients were divided in three groups. Group I (control) : 156 patients with ARF who did not receive substitutive techniques. Group Ih 21 patients under intermittent hemodialysis (HD) or peritoneal dialysis (PD). Group II1:43 patients under continuous hemodiafiltrstion (HF). The statistical analysis was Chi-square test and analysis of variance. RESULTS: The table shows the results we obtained, We did not find any significant difference betwen the two groups of patients undergoing dialysis. D(fferences were observed only between Group I and the other groups as shown below. We did not find any significant association between the theoretical mortality predicted and the observed mortality according to SAPS in the three groups. Due to exposure to a wide variety of unpleasant stimuli, for example, tracheal suctioning, venipuneture and physiotherapy, most pataents admitted to the ICU will require some form of sedation. This review will describe the suggested properties of an ideal sedative agent for use in the ICU and review the current limitations of some of the available agents from this perspactive. Methods used to quantify the level of sedation, such as the Ramsay Score, Glasgow Coma Score, Newcastle Sedation Score and Visual Analogue Scores, and their deficiencies will be examined. Consideration will be given to defining the optimal level of sedation and the circumstances under which sedation might be varied over the ICU course will be discussed. Preliminary results from an ongoing study examining the role of light versus heavy sedation and ischaemia in a cardiac surgical ICU population will be presented. The pharmacceconomics of ICU sedation will be briefly addressed. Finally, the role that sedation may play in increasing morbidity, pastieuiarly nosocomial pneumonia, in the ICU will be discussed. Objectives : Therapy cost(TC) in ICU patients is a substantial component of total hospital care cost. Estimation of TC during this year, partitioning to various groups of drugs used and attempt to minimise it, were considered practically useful. Methods : In collaboration with the Hospital Pharmacy we were able to have a complete report of aU drugs used for ICU patients (including enteral and parenteral nutrition). Mean APACHE II severity score upon admission was 19.7 and mean length of tCU stay was 6.7 days. Price per drug unit and cost per group of drugs were also available Drugs were divided into two groups: antibiotics (1) cardiovascular drugs (2), gastrointestinal system drugs (3), enteral and parenteral nutrition (4), respiratory system drugs (5), sedative, analgesics and paralysing agents (6), parenteral solutions with electrolytes, vitamins and trace elements (7), anti-inflammatory agents (8), protein substitutes and immunomodulation agents (9), anticoagulative agents (10). Antibiotics were further subdivided into those "freely" prescribed (A) and those whose prescription and administration requires filling of a relevant form (B). Results : !) TC for ICU patients/day was 30.530 Drs ($122). Total TC/patient was 295.195 Drs ($1.108.7). II) Partitioning total TC per group of drugs reveals : (1) 43%, (2) 2.7%, (3) 2.7%, (4) 9.2%, (5) 0.3%, (6) 8.6%, (7) 9.3%, (8) 1.3%, (9) 15.8%, (10) 2.5%. t11) Concerning antibiotics which consist the major cost component, group A and group B contributed by 29.1% and 13.9% to the total ICU TC respectively. Group B were administered to 13.9% of all ICU patients. Conclusions : I) For the above studied patient population antibiotics consist almost half of total TC followed by protein substitutes and immunomodulation agents. II) If TC control could be attempted in the ICU, prescription of beth groups must be reviewed. Appropriate treatment should be prescribed and readily provided to any patient. Clinical significance of routine protein substitution, currently controversial, should be re-evaluated. New antibiotics (third & fourth generation cephalosporins, quinolones, carbaponems) should be prescribed on the basis of strict diagnostic procedures using modern technology available. Rationalisetion of antibiotic therapy will lead to cost control, redistribution of ICU expenses and substantial contribution to infection policy in our country. Objectives: I -To investigate the clinic efficiency of the monitoring of the rSO2 cerebral, in relationship to the stroke prevention, in patient undergoing carotid surgery. 2-To determinate the variations of the rSO2 during the different surgical and anesthetic procedures in these patients Methods: Ten patients undergoing carotid endarterectomy. Precise neurological exploration previously to the surgery and in the immediate postoperative period. Angiography evaluation to the extend of carotid artery disease. Invasive blood pressure, ECG, pulse-oximetry ( pSO2 ) and rSO2 were collected previousty to the induction of anesthesia. The premedication was administered intravenously -midazolam (50 mcgr/Kg) and fentanyl (I rncgr/Kg) -. Thiopental (4 mg/Kg),fentanyl (3 mcgr/Kg) and atracnrium (0,5 mg/Kg) have been used for induction of anesthesia. CO2TE is monitoring al~er the orotraqueal intubation ! The anesthetic maintenance is accomplished with lsofluorane (0,5 -1,5%) and bolus of atracurium and fentanyh The surgical procedure is standard (without arterial shunt during the carotid cross-clamping). We register each 5 minutes: blood pressure, cardiac frequency, pSO2, CO2TE and rSO2. The rSO2 cerebral variate in relation with: the anesthetic induction, blood ~ressure, CO2TE, cross-ulampping carotid and with the modifications of the head position. The maximum decrease of rSO2 cerebral was in relation with the :ross-clampping carotid ( minimal value: 52 ). No patient had neurologic complications and postoperative stroke after carotid endarterectomy were not observed. Objectives: There are more than 8000 anesthesia in Chelyabinsk emergency hospital every year. To 80% patients of it emergency anesthesia is applied. More than 900 patients have ishemie heart disease (IHD), hypertansion (HP) and previos miocardial infarction (PMI). More than 5% of all patients are old patients (OP). The resalts deep noninvasive bioimpedance monitoring (NBM) in surgical patients have been studied by us. Methods: Our NBM system "KENTAVR" includes 21 parameters of cardiac and vessels function. It is realised by monitors in operation theatres and computer network. Moreover we are able to examine surgery patients before anesthesia and perioperatively by using special computers system for cardiovascular reflex control by Fast Fourie Transform (FFT) of 12 parameters simultaneously. Results: 187 pathients extremly needed peryoperative monitoring of hemodinamics. From these 187 patients more 40% had stroke volume (SV) less than 30 ml, 18N -CO less than 2.1/mim/m2, 25% -ejection fraction (EF) less than 65N and 32% -puls bioimpedans microvessels (PBM) less than 10 mOrn. 100 patient had intensive care in special department. 42 out of 187 died. Comparing with survived with these patients before operation HR was larger, SV, CO,EF, PBM and puls bioimpedance aortha was smaller. Much more of these patients were with IHD, PMI, HD, OP. Even with survived patients these parameters decreased the towards the end of operation. Surgery patients had different variability of 12 basic hemodinamical parameters with common tendency to increase power amplitude in low frequency by FFT. Conclusions: Using of bioimpedanee noninvasive parameters allows to have criteria for corrections (infusies, vasodilatators, inotrops and others) and then us the final goal, to have more sucssesful surgery. With survived patients was perioperatively and postoperatively care more intensive. Obiectives: The aim of the study was to compare the pHi with the hemodynamically derived tissue oxygenation indexes as: Oxygen delivery (DO2), Oxygen consumption (VO2), Cardiac index (el), and arteriovenous difference in oxygen [(a-v)DO2]. Methods: 18 patients (15 males and 3 females) with major trauma or major abdominal surgery were studied. On admission, a nasogastric tube allowing pHi measurement was introduced and a pulmonary artery catheter was inserted for optimal hemodynamic management. Each pHi measurement was accompanied with a complete hemodynamic study comprising systemic and pulmonary artery pressures, blood gases, and cardiac output measurements with the thermodilution method. Derived parameters VO2, DO2, CI, (a-v)DO2 were measured according to the standard formula. Hemodynamic parameters were opt• as soon as possible with fluids, inotrepes, and vasopressors according to repetitive hemodynamic measurements. All patients were under mechanical ventilation. After hemodynamic stabilisation pHi and hemodynamic measurements were repeated every eight hours, during a 24-hour study period. A total number of 52 measurements were obtained and compared. Statistics: Results are presented as means + SD, correlations were performed between pHi and the hemodynamically derived oxygenation parameters. A p<0.05 value was considered as significant. Results: Mean values were pHi=7.19+0.1, DO2=984+313, VO2=181+71, C.1= 3.7+ 1.2, (a-v)DO2 = 4.47+1.2. No correlation was found between pHi and DO2, pHi and VO2, pHi and C.I, pHi and (a-v)DO2. On the contrary in 14 patients pHi remained below 7.30 for more than 24 hours despite adequate hemodynamically derived tissue oxygenation parameters. Mortality in this group of patients was very high (85%). Conclusion: No correlation was found between pHi and the hemodynamically derived tissue oxygenation parameters Our data suggest that phi is a better oxygenation indicator than the hemodynamically derived tissue oxygenation parameters, because it is closely related to the patient's outcome. Objectives: The pathogenesis of septic shock and multiorgan failure is believed to be related to tissue hypoxia of the gastrointestinal tract. Therefore new monitoring techniques, preferably organ specific, are required to establish the adequacy of tissue oxygenation. PEEP is used to reduce pulmonary shunt volume and improve blood oxygenation, but is accused to impair splanchnic perfusion. We studied mucosal oxygenation and perfusion on the capillary level in the stomach and the duodenum. Methods: We used the Erlangen microlightguide spectrophotometer (EMPHO ll) together with a specifically designed fibre probe (Bodenseewerk Ger~tetechnik, 0berlingen) in combination with a standard gastroscope. Measurements were performed on 9 ventilated, traumatized patients (ages 17 -75 years), with no evidence of shock or severe infection, after informed consent was obtained from the relatives. All patients were hemodynamically stable without inotropic support. An area of 9 cm 2 was analysed in the gastric corpus, the antrum and in the duodenum. In three patients we simultaneously measured the muc0sal blood flow using a laser Doppler flowmeter ( Objectives: To investigate the influence of Hb-O 2 affinity in the monitoring of SvO~ during improvement of Cardiac Index (CI) in Cardiogenic Shock. Design: To state whether changes in SvO: were associated in changes in actual Pso (P~0) and standard P~0 (Ps0st) 22 consecutive measurements of artero-venous BGA, before an.d after therapy-induced changes in CI, were evaluated in 11 patients (mean age 73-*7 y) suffering from cardiogenie shock, all under mechanical ventilation in PSV modality. Methods: Together the hemodynamic measures, m~xed venous samples were analysed at 37 ~ C using the ABL500 Radiometer for PO2, PCO: and pH, and the OSM3 Radiometer for HbO2%, HbCO% and MetHb%. Psost (i.e. the P~0 at pH=7.40, PCO:=40mmHg and temperature at 37 ~ C) was calculated automatically by the instruments on mixed venous blood as was the Ps0"in vivo" (i.e. the Pso at the patient's value of pH, PCO2 and temperature), using Siggaard-Andersen's computerizated algorithm. Mean time between paired measurements was 6.1 -* 1.2 houm. The data were compared by Anova test for linear regression and t-test for paired samples. Results: A dose linear relationship was found between SvO2 and oxygen extraction ratio (OER), r=0.94,p=0.00000. The improvement of CI (1.41 -* 0.47 to 2.55 + 0.5 L/min/m 2, p<0.0000001) induced a significant increase in SvO~ (0.495 -* 0.131 to 0.636 • 0.060 %, p<.0001). A significant decrease in P50 (32.5 • 6.7 to 27.9 • 2.5 mmHg, p<0.05) without any significant change in P~0st (29.1 • 2.2 to 28.7 • 2.3 mmHg, p=NS) was also found. These data show that either OER or the shift to the left of the oxygen dissociation curve account for increase in SvO2 occurring with restoration of systemic blood flow. The program is intended to help the Intensive Care Unit interne providing him with a practical tool when making decisions concerning patients in a critical condition. In his daily practice in Intensive Care Unit, in this case the interne of the Unit, uses this program for each patient as follows: on the first stage of data collection he should complete the following modules: (1)personal data (2)patient's Pathology (3) laboratory and~ monitor Lug data (4)drugs prescribed or toxic elements ingested. In this way, the system allows optionally the consult with a computerized data base about the drugs prescribed, standardized parameters and techinques performed by the central Laboratory. (5)reference to an antibiotics guide regarding becterian sensitivety in our Unit, whitch ee checked every six month (6) access to de questionnaired APACHE II to load up new data. (7) statistcs about Patient's Admission and Discharge. Results: Once all data collection is finished the system performs the followin duties: (1)detailed drugs interactions, including toxic elements (2)diagnosis starting from the clinical, laboratory and monitoring data. In some cases, it also establishes therapeutic strategies, e.g. a coagulopathy (3) give the l~narmacological incompatibilities between the drugs p~escribed and %he diagnosis established, and (4)perform dosage adjustments based upon the Personal and Pathological data. Objeatve: To assess the power of diseri~,~ion ofa multiperpose severity score (SAI~) when applied to subgroups ofpatieals (pta) according to their lemg~ of ~ay (LOS) in ICU. Design: In order to compute the SAPS probability, a model derived fi~m legible regression was developed. Meaumree of calibration (goodmem..of.fit statistics) end discrimination (ROC cm've and relative Area Under the Cm've) were adopted in develotammtul asd validation set. The whole databue was ~ati~ed in five gronps reeked on LOS as follows: LOS = 2 days, LOS = 3-4 days, LOS = 5-7 da~, LOS = 8-14 days, LOS >15 day~. Area Under the Carve (AUC) was ud~ninted for each 8ro~. S~ing: 25 ImlimlCUs. Patents: Of 10~65 pts comec~ively admired ~ a period of three yeet~ (1990) (1991) (1992) , a total of 8059 was i~leded in this study. Pts without SAPS, p~ yolmger them 18 yearn, p~ with LOS shorter ~ 24 hom'~ were excluded from this maly~is. Iaterventinns: nose Mema'onm~ end result: The logistic model developed gave good remits in terns of calibration md discrimin~on, both in developmental set (Do.s g2: 9.24, P > 0.25; AUC = 0.79i-0.01) and in validation ~t (g.o.g g2: 8.95, P > 0.50 ; AUC = 0.78..+0.01). AUC of each grottp showed a loss in di~zimination (i.e., prediaton) closely related with LOS, being 0.90i-0.01 in pts with LOS = 2 days El 0.59~.02 ia tm with LOS > 15 da~ (figure). Following the present guidelines of integral management, in order to achieve optimization of sanitary resources and better use of facilities, we feel that the setting up of objetives is a key factor in the continuous process of improvement of quality care. Postsurgical intensive care services maintain an interdepent relationship with other hospital services. Within the general plan of the hospital it's of the utmost importance to delegate autonomy to the various depertments and service units in determining and achieving objetives. It's also necessary to establish mechanism for coordination of the activities in order to assure the succes of the program. The objetives cannot be improvised, they must be carried out in a specific manner in the following stages: 1.-Analysis of the present situation (Starting point). Where are we?. Defining objetives and making explicit the activities and methods to achieve them is to anticipate the future; It is of the utmost importance to comunicate said plans to all whom affect by encouraging them to attain the desired results. In the present paper we intend to show the guidelines to follow in carrying out a course of objetives. INTRODUCTION:We presents results related to the quality of life (QOL)of critical patients, from PAEEC Project data. MATERIAL AND METHODS: The PAEEC Project is a multicentre study define the type of patients cared for in Spanish ICUs, and the therapeutic activity provided. Ninety-five ICUs from Spain are taking part. This study analyzes the QOL of critical patients prior to their ICU admission.For the evaluation of QOL a questionnaire designed by our team for critical patients was used, with 15 items grouped in 3 sub-scales: physiological functions (4 items); functional capacity (8 items) and subjective aspects (3 items). QOL is classified in 4 levels: normality (0 points); Slight deterioration (1-4 points);moderate deterioration (5-9 points); significant deterioration (>i0 points). The We present results related to therapeutic activity in critical patients and their age, from the PAEEC Project. MATERIAL AND METHODS: The PAEEC Project is a multicentre study to define the type of patients in Spanish ICUs, and the therapeutic activity provided. Ninetyfive ICUs from Spain are participating. This study analyzes therapeutic activity in the first 24 hours as evaluated by TISS, and related factors. RESULTS: The sample was 9,291 patients, sge 57.91~17.46 years. Severity by APACHE II system was 15.59• points. The TISS score was 19.87• points, distributed as follows: I (39 points):5%.There is a positive correlation between the level of therapeutic activity and severity by APACHE II (R = 0.46, p < 0.001), and a very weak but negative correlation between TISS and age (R = -0.048, p < 0.001), so that an increase in age corresponds to a lower level of therapeutic activity.Patients The multivariate analysis of the relationship between TISS and age took into account: severity, existence of previous history, need for mechanical ventilation, size of hospital, diagnosis and mortality. It indicated that there continued to be a relationship between therapeutic activity and age, so that as age increased, therapeutic activity diminished. CONCLUSIONS: Therapeutic activity performed on critical patients is less in the oldest patients, in whom excessively aggressive procedures are limited. A relational data base management system in the ICU. C. Kotsavassiloglou*, D.Matamis, G. Dadoudis, J. Kioumis, D. Riggos. ICU dep., G. Papanicolaou Gen. Hosp., Exohl, Thessaloniki, and * A' Neurological Clinic of Aristotelian University, Thessaloniki, Greece. Objectives: The introduction of the Information Technology in the I. C. U seems to be unavoidable because of the large amount of produced data and the need for their systematic analysis. Such an information system should be a) easy to use, b) friendly to the user, c) powerful and d) modular. On that basis, we created a patient data management system (PDMS) according to the expectations of the medical staff of an eighteen bed multidisciplinary ICU. Methods: We selected Paradox for Windows V4.5 for the implementation of a relational data base because this program meets the above mentioned criteria. Informations regarding the patients include a) demographic data, b) previous medical history, c)diseases upon admission, d)complications during hospitalization and e) outcome data. The diseases' registration consists of 421 items classified in 15 categories upon the principal system affected. Specific informations about the need and duration of mechanical ventilation, nutrition, renal replacement, right heart catheterization and ICP monitoring are also available. An extension was added concerning ICU infections and related informations about antibiotic-resistant pathogens. All ICU pathogens can be matched to their resistance or sensitivity and cost of antibiotics. The program can perform queries and various statistical analyses based on complex criteria. New modules can be added later according to the future needs and remarks of the users. Results: The program was well accepted by the medical staff and 300 patients were registered as a test. The first analysis of the data related a) observed mortality versus the APACHE II predicted mortality, b) mortality according to the age, gender, pathology aud duration of ICU stay and c) pathology upon admission and ICU related complications. Conclusions: The long term use of this PDMS can be an efficacious research tool. It can be used in retrospective or prospective studies by addition of necessary modules. The first data analysis revealed the Iack of an international diseases' classification system. The development of a worldwide common classification system is essential for the compatibility of the data analysis among various ICUs. This will allow the realization of multicenter trials on a large scale. S. Nanas= N. Sphiris, A. Precates, A. Lymberis, M. Pirounaki, and Ch. Roussos Dept. of Critical Care, University of Athens, Athens, Greece The complexity of the cases submitted to an ICU, the variety of underline disease, tbe severity, as well as the large number of substances administered to each patient constitute obvious the need of support with an easy available DSS. This system will assure the safety of the administered treatment will help to adjust the dose according to the situation of each patient and it will screen for possible interaction and incompatibilities between the administered drugs. The goal of the present effort is the design and development of a software system acting as a decision support tool to physicians of ICU. The application is organised around a Relation DataBase Management System (RDBMS) that consist of: a) all available substances (1.850), b) all generic names of medications available in our country for each substance, c) incompatibilities (2.300 cases) and d) interactions with other substances (50.000 cases). The following figure shows the structure of the RDBMS. y ta~ortaTO~ [ c~rs Using the stored parameters for each patient the dose and the rate of administration of selected substances will be possible to calculate. The continuous monitoring of the treatment for each patient supports the medical staff to make the necessary changes of the prescriptions. The application is currently developing in wireless pen based computer systems which place patients at the centre of "islands of information" located throughout ICU. In conclusion this DSS is a powerful and useful tool for ICU staff because it provides without additionai work to the routine of daily practice, the currently available information for each order concerning drug interaction and incompatibilities as well as treatment monitoring is to obsea~ among 113 critically ill pfdieats, stdJdivided following the diagn~s at the adn~ssio~ the diffmeax:es in the ~ and oxyplx~efic l~mmems bawe~ strvwors [S] and non sumvors INS] and to test the pc~'bih'ty to have soar survival criteria, as earliest as tx~able. Method~ :We made a ~ study on 113 consexa~e ~ilically ill paliffas, subdivided in 3 series following the diastases at the admission: medical pafiea~ (29 S and 23 NS), surgical patients (19 S and 22 NS), a~d poliWauntas ( 14 S and 6 NS). Follow up was done at 20 d,.ays from the admission in ICE. All the patienls were ramitored with a ~ c~eter and 18 laeno:lymmi. "c and o .x.xyphorefic txuamaers va:~e coUected at 7 fin~es (T): at fiae ~draission (T0), at 121x~ars from T0 (T1), at 24 (F2), 48 (Y3), 72 (T4), % (T5) and 120 horus from T0 CF6). In~,h ~ies, for ~y ~ a all the lin'~ n~an and sandaid d~viation was ~ tx~h for S and for NS. Th~ betw~ S and NS tl~ roeaas of ~h porarneter ~e ccmpared tt~ng t-lest and p < 0.05 w~ considered ska~ significant In each series in the T wheae the mast significative diffemx:as ~goeamd bet~en S and NS, we made a txedictive criterion, asamting as predictive indices for stnvival the I:r values, higher or lower than flae treans of the ~rar~ers of aU flae patients, axx)rdhlg to those ones t~iatistically diff~'e~ betw~m S and NS. Fhmlly xse co:weaTxt onaong the 3 series the Nrametees of the st~rs with the analysis of variance, to daserve the lxJsable differealt Irea~ of stY1 hflices, following the diagn~s of admission: :nedkal, angical patient or poll~tam Results: We c~ld not find ~ predictive criterion for politraonaas, perhaps Ixx:ause of the few ntanber of l~fients. For high ri~ saw~cal patieras the following criterion at T2 has a sensitivi .ly of 100~ ,and a ~ecificity of 27.8%: SV1>32.89 nffmin/n~, MAP>92 mmHg, PMAP<27 nmalqg CVP34 g M/m 2, SxO2>67~ DO21>515 mlhnin/m2, O2ER<31%. For lx~dical l~tienls at T5 the following criteric~a has a ser~tivi.ty of 100% and a ~zificity of 36.8~ CVP<7.5 mn~g, SAO2>97%, S,g)2>74~ VO2I<133 ml/nfin/m 2, O2ER<25%, Shunt<19% Survlvops' data of the 3 series ~ signitic~atly differenl~ both for the t~mody~nic a~ for fl~e ox3rphomfic lxLmn~s; moreover we ~ that the vatt~ of hemodynamic mad ox.~ho~tic indices were higher in politrautms. Conclus'ions: Acx~ording to the fftffe~mt patho!o~es, the ~ rnelabo~c needs are diffeten~ so that it is juslified to mash ~ the~alceutic goals, following the type oflmthology. 2hen~ we foru~d for high ~k mrgical pmka~ and for medical patier~s assme, ff mlLslied, a good prognosis while, if n0[ NtlJsfled~ the plINsliclioil ofdl~tth is no[ g(ioct Finally, ab~ high iis~ suPgical palieaats, according to what other Atmhors say, txatWs Sh0~'N~ers ' therapeutic goalsvvould seem inadeqt~te, bec~Jse they need a gear physiologic and themtx~ic elth~ in rdation to the rretabolic needs. Figure 1) . Thus, the smaller European nations had a greater participation than ~e larger ones, with the exception of Norway. A similar result was evidenced for contributions to Intensive Care Medicine (Figure 2 ). These findings can be explained by different submission policies and language banners. However, there was no significant correlation with the gross national product of each country. Conclusion: We conclude that the smaller European countries generally contribute more to international intensive care journals than the larger ones. Objectives: To evaluate the agreement between a new and three old methods measuring Ctp and to assess their reproducibility. Methods: We studied 20 patients ventilated with a Siemens 900C respirator. We measured Ctp by dividing the tidal volume with the increase in airway pressure (Paw), either with the respirator setting used (Ca) or with a fixed setting (Cf). By modifing the inspiratory time (Ti) without changing inspiratory flow, we were able to deliver two series of 10 inflations (100, 200,... 1000 ml) before and after curarisation of the patient. The same volumes were also inflated in paralysed patients with a super syringe. At the end of each inflation a plateau of 3 sec was performed and Paw was recorded. The above three sets of pressure-volume (PV) points were used to reconstruct the corresponding PV-curves ((31, C2, C3 The new method for Ctp measurement without a super-syringe had the best reproducibility in paralysed patients and gave similar results without curarisation in the majority of them. However, agreement between the methods tested was unacceptable for clinical purposes. Further investigation is required in order to improve the accuracy of Ctp measurement in ICU patients. M Kunert, R.Sorgenicht, L.Scheuble, K.Emmerich, H.G01ker Med.Clinic B (Dept.of Cardiology) I Heart Center of Wuppertal/University Witten-Herdecke,Germany Objective To determine the accuracy of activated partial thromboplastin time (aPl-l) and activated clotting time (ACT) studies when samples are drawn through heparinized central venous catheters (CVC). Methods A total sample of 80 paired ACT/P't-/" values was analysed in 40 patients (28 m.,12 f.,66 + 12 y.) for monitoring heparin therapy.All patients had a CVC (Certofix Trio,Braun,FRG) in the internal jugular vein receiving a continous infusion of 20.000 U heparin via the central catheter.ACT (HR-ACT, HemoTec,USA) and aP'I-F (Neothromtin, Behring,FRG) samples were drawn from the CVC using the double syringe technique (removing and discarding 5 ml blood before drawing the sample). These blood samples were compared to ACT/aP'CF blood samples obtained by venipuncture (V.fem.) at the same time, ACT values were analysed directly in the intensive care unit (ICU),aPI-I samples were measured in the hospital laboratory within 30 minutes. Results AC-i -~ pI-F~ CACT/~PI7 r = 0,62) CVC samples 132 88 +34 +22 . V.femoralis samples 1"28 84 +29 +24 p-value n.s. n.s. Conclusion There is no difference in heparin anticoagulation studies drawn from heparinized central venous catheters compared to those obtained by femoral venipuncture,Withdrawing 5 ml blood prior to obtaining the blood specimen is a safe way for eliminating heparin contamination.Not only the aPTT test but also the ACT test is a useful method for heparin anticoagulation assessment in the ICU. Objectives: Evaluation of the delicate balance between filter-coagulation and patient-hemorrhage using heparin as anticoagulant in continuous renal replacement procedures. Methods: From January 1991 through August 1994, we studied filter surviva[ and hemorrhagic complications during 240 filter periods in 78 critically d[ patients, treated with continuous arterio-venous hemo(dia)filtration, with special emphasis on the heparin dose, concurrent use of coumarins, systemic Activated Partial Thromboplastin Tirne(APTr), platelet count, mean arterial bloodpressure and the type of filter used. Results: 141 Filters (59%) were disconnected because of coagulation. Mean survival of Multiflow AN69 filters was twofold shorter compared to survival of FH66 Gambm filters. A total of 48 hemorrhagic complications occurred of which three patients died at APTT values of respectively 39, 48 and 56 seconds. After adjustment for mean arterial bloodpressure, platelet count and the type of the filter, the risk for filter-coagulation decreased 25% (relative risk 0.76, 95%C1 0.68-0.85) for each ten seconds increase in APTT. The risk for patient-hemorrhage increased 50% (relative risk 1.50, 95%CI 1.38-1.64) at an APTT-increase of ten seconds. The occurrence of filter-coagulation and patienthemorrhage was not correlated with the administered dose of heparin. Concurrent use of cournarines had a positive effect on filter-survival, without increasing the overall incidence rate of patient-hemorrhage. Conclusions: The systemic APT]" is a good predictor of the risk for filtercoagulation and patient-hemorrhage. Heparine therapy seems optimal at an APTT between 35 and 45 seconds, although one should realize that fatal hemorrhagic complications still can occur. Objectives: The alterations in vascular tone which are primarily regulated by adreno-sympathetic tone(AST) are compensatory responses in hemorrhagic patients. This study was designed to evaluate the correlation between vascular tone and AST in patients with hemorrhage, Methods: The vascular tone was expressed by volume elastic modulus (Ev) that is defined as; Ev = Ap/(Av/v) (Ap; the arterial pulse pressure, Av/v; the volume change ratio). Ev was measured using a non-invasive transmittance infrared photoelectric plethysmography (TIPP) and a volume oscillometric sphygmomanometer . We prospectively studied 6 patients with hemorrhage. The initial Ev measurement was performed on arrival and repeated for a 48hours duration. As a parameters of AST, serum concentrations of adrenalin (Ad), noradrenalin (Nor), plasma renin activity(PRA) were measured simultaneously. We analyzed the correlation of Ev and conventional parameters to AST by multivariate statistical analysis. Results: Ev values at Transmural pressure 40mmHg on admission and 48hours later were respectively 864.2 + 249.5mmHg, 270.0 +_ 92.0 mmHg (mean + SD). Systolic pressure(Pas) and serum hormones on arrival and 48hours later were respectively, Pas; 96.5_+20.4, 152+18.7mmHg, Ad; 1.21_+1.02, 0.07_+0.04 ng/ml, Nor; 1.60_+1.48, 0.65+0.39 ng/ml, PRA; 26.6_+37.8, 2.5_+2.9ng/ml/hr. The Ev values correlated significantly with Ad (r=0.47, p=0.006, n=33), Nor (r=0.47, p=0.005, n=33), PRA (r=0.38, p=0.032, n=33). By multivariate statistical analysis, Ev correlated more significantly with Ad and Nor and PRA (p=0.00079) than the conventional parameters such as Pas, heart rate and pulse pressure. Conclusions: The alterations of Ev correlates closely with AST. The compensatory mechanism in hemorrhagic patients can be detected noninvasively by Ev monitoring. Obiectives and Method: Autologous oxygenator blood was processed at the end of cardiopulmonary bypass (CPB) by either hemofiltration (HF 60, 1,2 m 2, Fresenius) or by cell washing with a onntinous autologous transfusion system (CATS, Fresenius). Prospectively the blood of 10 patients for each group was processed and then retransfused intravenously to the patient. Besides, volume and time requirements, standard hematologic chemistry, coagulation and complement activation were measured. Results (mean values for oxygenator blood at the end of CPB, and results of concentrate after processing by filtration or washing): Both processing techniques show excellent hemoconcentration of the diluted CPB blood with a good transfusion effect for the patient. Filtration retains all plasma proteins and large molecular weight plasma bound waste products. In contrast, cell washing with CATS significantly depletes plasma proteins and waste products. The newely developped CATS machine gives eonsisinnt laboratory result in a fully automatic continuous processing mode. In conclusion, both filtration and washing are effective for processing CPB blood. Filtra tion yields a highly concentrated whole blood, whereas CATS washing produces a high quality autologous erythrocyte concentrate. Soluble fibrin has during the last years gained interest as a marker for the activation of the coagulation in connection with various clinical conditions, e.g. disseminated intravascular coagulation, deep venous thrombosis and myocardial infarction. Elevated levels of soluble fibrin in plasma can be detected by the chromogenic assay Coaset Fibrin Monomer, relying on the ability of fibrin to enhance the tPA-catalyzed conversion of plasminogen to ,plasmin. Using this test, it has been shown that the level of soluble fibrin can be correlated to severeness of illness in critically ill intensive care unit patients. A revision of the Coaset Fibrin Monomer kit has now been made and the new product, Coatest Soluble Fibrin, is considerably more convenient to handle and gives higher resolution at low fibrin levels. The test is performed by the addition of a buffer dilution of the plasma sample to a microstrip well containing the colyophilized mixture of tPA, plasminogen and the plasmin specific cbromogenic substrate S-2403. The reaction is allowed to proceed at,. room temperature for 15 minutes before discontinuation. The absorbance at 405 nm, measured in a microplate reader, is proportional to the content of soluble fibrin in the sample. The assay is carefully standardized and calibration curves are provided in the kit. The convenient and rapid assay procedure makes the Coatest Soluble Fibrin test well suited for single test analysis in acute situations. Objectives : Blood coagulation abnormalities have been reported in the systemic blood of patients with cerebral lesions. The physiopathology of such events is not yet completely understood. We compare the coagulation profile of blood from the right jugular bulb with systemic blood of patients with head Injury. Methods: We studied 4 patients, who were admitted to our Neurosurgical Intensive Care Unit between January and March 1995 with head injury and no other associated pathology (age 20-60 yrs), a Glasgow Coma Score <= g, no abnormality in baseline coagulation profile and no history of coagulopaties. The patients did not undergo angiography. A one-way 16 gauge Certofix catheter was inserted through the right internal jugular vein up to the jugular bulb. An identical catheter was inserted through a subclavian vein. Blood was sampled from either catheter (a=atrial; j=jugular) 6-10 hours after trauma (T 1) and t 2 hours later (T2 The inddence dpontolx'rative thmmhi~e and haumord~gic complieatiom were assessed in padents treated with indobefen, heparin calcine CAeCa), low mollecolar weight heparin (LMWH) (F.nosheparin) and undergoing hemodiludun, blood predeposhing, intra mad postoperative blood saving. ]'he indolmfon tempota~.Norks platelet aggregation through ,,elective inhibition of the cyclatygenasis and thus atacbldonicadd(1).Tbe n'mimum effect occurs after 3 hours from the fast administration and is still present after 24 hours. ~-979 patients, mean age 62---11 yrs., weight 68---10 Kg were studied. 321 (32.8%) were male and 658 (67.2%) female. 858 onderwent 713 hip prosthesis (7 previously plate and screw removal) 145 hip revim'un (10 stem, 33 cop and 102 stem + cop), 121 tutal knee prosthesis, in the 1st Anaesthesidogy DepL from 14-1992 to 30-6-1994. As for antithromboembolic ptephylam, apart from hemodihitiun 668 pts were with treated indobufen 0ndo), 199 with heparin ealdum CAeCa) and 112 with low mo!lecular weight hepam (LWr,91). As the slightest clinical and/or imtmmental suspidon of deep vein thrombosis (DV'I') or polmonary umbolism(PE), a phlebogram or sdndgram were respectively carried out. -The inddence of homologom transhisiom was significandy lower (p=0.000l) in the padeats treated with indobufen (4.256) compared 7.'ith HeCa (14.5%). The con~gency table shows statistical signifleance for the use of HeCa in patients with vein deficiency in the lower limbs, past DVr and/or PE, coronary heart disease (CDh'), while there is no correlation for renal, cardiac or liver defidency, obesity, systemic hypertemion, atrhythmy, diabetes, chronic bronchitis and rheumatoid arthritis. By comparing the postoperative cumplications with the risk factors, there ks a highly significant correlation (p=0.000l) between CDH and thrombotic and humord~agic complieatiom (PE, death, he~atoma, die use of hum_ologous blood). Thee data show that hep~in, preferred in patients with C'DH, roost likely for leagal-tuedical reasons, did not have the de~'ed effect. Conclusions -The stastisfical aar~ais shows ~nifieanfly different efflea~ (pro0.0001) between the therapies (see table) : it can be seen that in patients undergoing autotramfusiun and hemedihidon, indobufen produo~ a lower incidence of haemotrhagic complieatiens compared to HeCa and LMWH and is more effective in the prevention d ~c complications at clinical e~idence. The duration of i~toperadve hospital stay is signi~cantlylonger for patients transfused with homologous red ceils and treated with HeC,7.a (13.7-+1.4 days) and LMWH (13.5+-1A days) compared with Indo(ll.8_+1A days). One of the main causes for postoperative complications in major orthopaedic surgery is postopemtive bleeding with local effects in the operation site (hematomata, pain and delayed mobilization) and/or systemic and subsequent cardiodrculamry repercussions that are sometimes severe. The aim of this study is to assess the possibility to apply a new system of monitoring, control and saving postopemtive blood loss from the drainage. The BT797 Recovery Dideco (MArandola, Modena-Italy) ~ used since it is the only apparatus capable of doing this. The apparatus consists of a pressure transducer, adjustable from -100 a +50 mmHg, which activates a peristaltic pump connected m drainage robes. The BT797 Recovery display shows hourly bleeding in the first 8 hours, total bleeding, time passed since the start of monito~g and subsequent salvage and the aspimtioo pressure on the drainage robes; the latter is inserted at -10 mmHg and then modified according to bleeding/minute, g BT797 Recovery also has an alarm that sounds automatically if.' blood loss is more than 300 ml/hour; air is in the circuit; the batteries are running low. Materials and Methods: 191 pts were studied (53 m and 138 ~), aged 63.5-+11.lyears, basal hemoglobin 13.1-+ 13(range 7.8-16.6)g/all, treated from 1st January, 1992 to Mst December, 1994 in the 1st Service of Anesthesia and intensive Care unit of our hospital. The patients underwent the following surgical treatment: total hip revision (132pts), cup revision (~ipts), stem revision (13 pts), total knee revision (2 pts). The average dumtion of the operations was 173-+58 min. Intranpemtive monitoring and blood salvage was applied to all patients. Genera! anesthesia was used on 57 pts. and integrated (epidural analgesia + light general) on the remaining t34. Anttthromboembolic prophylaxis consisted of external pressure bandage, isovolemic hemodilution with iodobufen in 131(68.6%)pts., calalc heparin in 35(18.3%)pts., low molecular weight heparin in 25(13.1%)pts.; 1 pt did not give a predepoalt of blood, 4 gave 1 unit, 45 pts 2 units, 110 pts 2 units, 31 pts 4 units. The data obtained was statistically analysed using contingency tables and ANOVA. Results: Average intmop salvage was 420-+345 ml, average postop salvage was 420-+265 mI the average intra+postop 909+-460 ml. Average postop loss was 677-+359 ml. The global incidence of postop complications was: h~natomata 5.2%, DVT 1.1%, pulmonary thromboembolism 1,196, myocardiac ischemia 0.5%, acute myocardic infarction 0.5%, respiratory deflciecy 2.6%, arrhythmia 2%, cystitis 0.5% there were nn complications in 86.4% of pts. Postop bleeding over 300 ml in under 60 minutes (with bleeding alarm activation) occurred in 30 pts (15.7%). This sta~tically correlates only with the type of operation performed (more frequently in total hip revision p=0.034) and with a significant decrease (p~0.003) in the pruthrombic activity detected about 8 hours after the operation. This bleeding, also made the alarm sound, calling the attention of staff who could act accordingly, by making the drainage pressure positive and incre~sthg the tension of the external pressure bandage. Conclusions Postop monitoring, control and blood loss salvage combined with predepoalting and intmop salvage has enabled allogenic transfusions in 16% of cases to be avoided in operations with high postop blood loss like hip or knee revision. The usefulness of the system can be seen by the fact that in the 30 patients with so much bleeding to set off the alarm, there was no significant difference in the incidence of allotransfusions and complications. References 1)Borghi B., Bassi A., de Simone N., Laguardia AM., Fonnaro G. An injury of the brain may result in various disorders of hemostasis caused by the release of • into the circulation through a damaged blood-brain bar tier. Disseminated intravascular coagulation(DIC) is one of these disorders. It is a freguent but relatively rare ly diagnosed complication of subaraohnoidal haemorrhage. The aim of this study was to evaluate some parameters of both blood coagulation and fibrynolisis in patients with SAH.In addition one wanted to find out wh~ther potential changes correlated with the pa• condition in the acute phase of SAH and whether they influenced the course of this disease. 60 patients with SAH were studied. In 17 of them SAH was due to closed eraniocerebral injury and in the rema ining 43 resulted from vascular malformation. The following parameters were evaluated:the prothrombine time,the activated partial thromboplastin time, the thrombine time,level of factor V,fibrinogen degrada tion products and fibrin monomers. The results let us show the presence of OIC in 3 patients with closed craniocerebral injury and in 14 with vas. cular malformation despite the lack of clinical symptoms The tests in 5 posttraumatic patients and in 6 patients from second group showed incomplete DIC.On admission patients with such changes in measured parameters were in poor condition.The course of the disease and the effe cts of treatment were also worse in these patients. The results showed ihal in patients with SAH complex disorders of both coagulation and fibrynolisis occur, and they depend on clinical condition of the patient. They also influence the course of the disease. Methods : Charts of all patients admitted with D.I.C. over a ten year period (85-94) were reviewed. Diagnosis of DIC was based on the association of fibrinogen < 2 g/1 -platelets <150 109/1 -FPD > 40 ~tg/ml in the 24 hours of the admission. Results : 40 patients -mean age 29+6 y -SAPS 8+_5 -gestanional age 35_+5weeks -the two first conditions associated with D.I.C. were placental abruption (35 %) and preeclampsia or eclampsia (22,5 %). Bleeding episode was present in 22 pts (55 %) and surgical treatment has always been necessary. 26 pts (65 %) were given packed red ceils (12+10 u) and fresh frozen plasma (9+8 u). 6 patients were given platelets packs. Heparin was never administered. 6 pts required mechanical ventilation and two patients hemodialysis. All the 40 patients survived. Correction of prothrombin time (P.T.) and fibrinogen (F) was quick (P.T. at T24 h 69~2 % -F at T36 h 2,60+1,5 g/I). But platelets count remained low (plat. at T48 h 80 +42109/1) -no difference was observed in patients who received platelets. Conclusion : Prognosis of critically ill O.P. is good. Blood loss is the main complication. Correction of hypovolemia and anemia with concomitant surgical treatment are essential. The administration of coagulation factors or platelets is still under discussion. Objectives: To evaluate the effects of antithrombin III I AT-III) and a protease inhibitor, gabexate mesilate FOY), on the coagulation and fibrinolysis in disseminated intravascular coagulation (DIC). Methods: After the approval of our institution and consent from patient's family, 40 patients with a DIC score (1988, Japan) more than 5 points (DIC or having a risk for DIC) entered this study. They were randomly divided into two groups, FOY (i-2 mg/kg/h for 7 days or more) treated group and no FOY group, each of 20 patients. Platelet count (Plt), fibrinogen (FEn), AT-III fibrin degradation product (FDP), D-dimer (DO), fibrin monomer (FM), thrombin-antithrombin complex (TAT), plasmin-plasmin inhibitor complex (PIC), and prothrombin time ratio (PTR) were measured before the start of treatment (at admission) and I, 3, 5 and 7 days after the admission. AT-III at 1500 units for 3 days was administered if the AT-III at admission was less than 70%. Finally the patients were divided into four groups: Group A, FOY (+) and the AT-III ~ 70%; Group B, FOY (+) and the AT-III < 70%" Group C, FOY (-) and the AT-III 70%; Group D, FOY (~) anffthe AT-III <70%, each of 7 patients, to match the patients for backsrounds. All parameters, DIC score and survival rate In a month following treatment were compared among the four groups. Results: The AT-III and Plt from day 3 to 7 were significantly higher in Groups A and C than in Groups B and D. The FDP, DD, TAT, and PIC after treatment decreased significantly from the baselines in Groups A and C but not in Groups B and D. The Fgn and FM were not significantly different among the four groups. The PTR decreased in Groups C and D but increased in Group B. The DIC score decreased significantly in Groups A and C than in Groups B and D. Survival rates were 57%, 43%, 71% and 57% in Groups A, B, C and D, respectively, although not significantly different. Conclusions: In patients with DIC or a risk for DIC, FOY had no expected effects but AT-III had suppressive effects on the coagulation and fibrinolysis mechanisms. A PROGNOSTIC FACTOR ? Carbon monoxyde intoxication is a classical complication of inhalation injury. Carbon monoxyda is also physiologically produced during the heme metabolism: heme is conversed to bi]irubin by the hemeoxygenase which is an intracellular stress protein. 20 ICU patients (pts) were studied prospectively for APACHE II score and carboxyhemnglobin (HbCO) arterial level to assess if HbCO level could be correlated with the severity of the pts. Objective: To evaluate a new technique of non-surgical tracheotomy. Patients: 55 adults, mean age 54 years and 11 children, mean age 18 months (2 me.-5 yrs). Method: Through a needle inserted in the trachea, a guide wire is retmgradely pushed out of the mouth and attached to a special device formed by a flexible plastic cone with pointed metal tip joined to an armoured tracheal cannula. This device is then pulled back through the oral cavity, larynx and trachea, and outwards across the neck wall by applying traction on the wire with one hand and counterpressure on the neck wall with the fingers of the operator's other hand. When the cone and 2/3 of the eannula have emerged, the cannula is cut off from the cone, straightened perpendicular to the skin, rotated and advanced caudally to its final position. Results: Endoscopic control facilitates and improves the safety of all manoeuvres. The pointed cone easily pierces the tissues, and the cannula is extracted without difficulty since it has the same outer diameter as the cone. Tissue adherence around the cannula is absolute thus preventing local inflammation. The time in apnea required for dilation and cannula placement does not exceed 60 see., and it is well tolerated because within safety limits in patients hyperventilated with oxygen. Only one case of bleeding occured in a patient on dialysis with severe coagulopathy. Autoptic findings in subjects who died due to progression of primary disease showed a very regular stoma with an almost complete lack of hematic and flogistie infiltration in recent tracheotomies. .Conclusions: Translaryngeal tracheotomy (TLT), by virtue of its greater inherent safety and lower tissue trauma than percutaneous techniques, can also be carded out in infants and children, a severe test bench for any tracbeotomy technique. Further specific indications are recently stemotomized patients, since TLT is associated with a low rate of infection, and short term tracheotomies after laryngeal surgery, to prevent obstructive complications. References: Fantoni A., Translaryngeal tracheotomy, APICE, Ed. Gullo, Trieste, 1993, 460 . BACKGROUND: Inhalation of NO has been shown to reverse hypoxic pulmonary vasoconstriction 1, to reduce pulmonary pressure in pulmonary hypertension of different origin and to improve gas exchange. In putmoflary embolism, pulmonary hypertension is caused by mechanical vascutar obstruction and by reactive vasoconstriction. The effects of inhaled NO in putmonary embofism has been partiatly studied' The purpose of this study was to investigate and determine the effects of NO inhalation on pulmonary hemodinamica and gas exchange in a hypoxic canine model of pulmonary embolism. METHODS: Two groups of adult mongrel dogs were studied: group 1 (control} 5 dogs and group 2 (NO inhaled) 6 dogs. Both groups were anestesized with tiopental, mechanically normoventilated with an hypoxJc mixture of 02 and N~ (F[Q2 0,16) and instrumented (Swang-Ganz catheter, femoral artery catheter) Pulmonary embolism (PE) was induced by Fisher's method s. NO inhalation (80 ppm) in group 2 was started 15 rain. pdor to PE and kept constant throughout the experiment. NO inhaled concentration was analyzecf by chemiluminiscence technique. Pulmonary artery pressure (PAP), central venous pressure and sistemic arterial pressure were continuosly recorded. Cardiac output, artedat PO~ (Pan2) and mixed venous PO~ were measured in both groups under hypo)dr conditions, before PE and 5, 15, 30 and 45 rain. after PE. Pulmonary vascular resistance (PVR) and gas exchange (PaO21FiO:~ ratio), were calculate using standard formulas. Data were process and analyzed with non pararnetdc test, and reported as mean -SO and statistical significance was considered if p < 0,05. : NO produced an increase in arterial oxigenation (PaO2/FiO~ ratio) and reduced PAP before PE induction in group 2. After PE we found no significant difference with .respect to the time eour.se of PAP, PVR and gas exchange between beth groups throughout the experiment. probably, the severe mechanical obstruction produced in pulmonary embolism masked the small effects of NO inhaled. Obiectives: Blood volume measurement would be useful in critically ill patient management if it were easy to perform. This is not the ease and current methods are based on radiolabelled red cell dilution. Inhalation and uptake of a known mass of carbon monoxide (CO) gas and measurement of earboxyhaemoglobin increase can give results accurate enough for clinical use. This requires a rebreathing system providing oxygenation and carbon dioxide removal, yet complete retention of all carbon monoxide administer&l, and so most authors hand ventilate with a bag and Waters soda-lime canister, adding oxygen as necessary. We aim to popularise this method by; i)Design of an automatic CO administration system driven by the ITU ventilator and ii)Writing of software for a portable computer to perform all necessary calculations 9 Method: We show the computer is use estimating the CO dose required and later estimating the blood volume. We also show the new gas administration system. This is a fully closed circle attached to a "bag in bottle", driven by the ventilator. The novel feature is the mechanism by winch driving gas (set to 100% 02) spills automatically into the circle, balancing O5 uptake by the patient, yet allowing no CO loss. Conclusions: This equipment is easy to use, reduces human error and allows optimum ventilator settings to remain. The operator merely administers the volume of CO determined by the computer and takes blood on two occasions. Carboxyhaemoglobin measurement is easy to perform, thus there is a cost saving also. With our modifications use of this technique may potentially become more widespread, The video demonstrates the method in use in our ITU. -10 (25%) underwent conventional surgical therapeutics. 9" (90%) with resection of tracheal stenosis with end-to-end anastomosis(RTS). I (10%) With broncoscopic dilatation. One patient died and the others still have stable patency(SP) without continued treatment. -29 (72,5%) have received endoscopic laser ablation with or without calibration tubes. 17 of them (58,6%) are receiving continued endotracheal treatment until now. 12 (41,4%) have SP wihout continued treatment. -I (2,5%) endoscopic laser therapeutic case turned to RTS and is having SP. Conclusion: Conventional surgical aproach has been progressively replaced in our Hospital by endoscopic laser ablation and silicone calibration tubes. This study suggests that these technics are effective and could be the elective treatment for iatrogenic stenosis. Obiectives: Hemorrhagic disorders due to thrombocytopenia and thrombocyiopathia remain one of the most serious complications during long-term extracorporeal membrane oxygenation (ECMO) in patients with severe acute respiratory distress ~drome (ARDS). In the presented study, nitric oxide (NO), kwown as a potent endogenous platelet antiadhesive, disaggregating and antiaggregating compound, was evaluated for its possible antagonistic effect on platelet trapping when added to the gas compartment of membrane oxygenators (MO). Meti~ods: Two parallel separated extracorporeal circuits, consisting of heparin bonded hollow fiber oxygenators (Minimax, Medtronic, Carmeda Eioactive Surface), tubing systems, low pressure reservoirs, and roller pumps were prepared. For each measurement, a pair of circuits was simultaneously filled blood from the same volunteer. Low-heparinized fresh warm blood was obtained from four healthy volunteers, who had no drugs for at least two weeks. The gas inlets of both oxygenators received dry gas (21% oxxygen, 5 % carbon dioxide, 84 % nitrogen); gaseous NO (20ppm) was added to the gas of one of the oxygenators (NO-MO), whereas the other one (MO) was used as control. After 270 minutes NO gas was switched off, so that the NO-MO received no more NO, and NO was added to the gas inlet of the membrane, which had no NO before_ To assure iutracircnit volume stability, drawn blood for measurements was replaced with saline, and platelet counts were corrected for dilution by hemoglobin values. The mean of four platelet counts (Coulter Counter) of each timepoint (start, 30, 90, 150, 210, 270, 330, 390 , and 450 minutes) was used for statistical analysis (paired sample t-test). Results: In the NO-MO platelets remained at 96 + 3,2 % (percentage of baseline value, mean -+ SD) until 270 min. In contrast, platelets of the MO continuously decreased after start and were significantly lower after 150 minutes ( 96,4 9 +3,5% vs90_+3,1%(p<0.05);210min. 95,9-+4,5%vs84,5_+2,2%(p< 0.05); 270 min. 82,7 _+ 2,5 % ( p < 0.05). After switching of NO gas to the MO, further decrease of plateleta was stopped and platelets remained at 81,4 +_ 4,5 % until termination of circulation. Platelets of the former NO-MO decreased slightly after cessation of NO gas to 92,6 _+ 5,4 %. Conclusions: These data indicate that gaseous NO significantly attenuates platelet trapping in hollow fiber oxygenators, when added to the gas compartment. This might be a new therapeutical approach for membrane oxygenator induced thrombocytopenia during long-term ECMD. Objectives: Nitric oxide (NO) plays a pivotal role in regulation of vascular hemostasis. Several studies elucidated the antiadhesive, antiaggregating, and disaggregating properties of endothelially synthesized NO to platelets. Additionally, agonist-induced NO production in platelets by the L-arginine-NO pathway was found as a negative feedback mechanism after platelet activation. Although NOplatelet interactions were intensively studied by several investigators, no data exist, about changes in platelet surface molecule expression in NO-modulated platelets measured by flow cytometry using monoclonal antibodies (MoAbs). Methods: P-selectin (alpha-granule-membrane protein, GMP-140, CD62P) and glycoproteiu 53 (GP53, lysosomal protein, CD63) are expressed only after platelet activation and degranulation. Activation was quantified in thrombin (0.4 U/ml) and ADP (0.1 raM) stimulated platelet rich plasma samples (PRP). Blood was obtained from healthy volunteers (n=7), who had no drugs for at least 14 days. For evahiation of NO-modulated activation, the spontaneously NOreleasing compound SIN-I (0.1 mM) (3-morpholino-syndonimin-hydrochlorid) was added in parallel prepared samples prior to the addition of agonist. Platelet surface molecule expression was evaluated with MoAbs directed against CD41a (GPIlblIIa, fibrinogen-receptor, phycoerythrin(PE)-conjugated), CD62P (FITCconjugated), and CD63 (FITC). Only CD41a-positive signals were gated in sideangled light scatter, and assayed for activation marker expression (defined as percent of gated population). Results: Basal P-selectin expression was 1.6 + 0.7 %, and increased to 75.2 _+ 12.2 % after thrembin-activation, and to 26.7 + 5.3 % in ADP-stimulated samples. Addition of SIN-1 attenuated P-selectin expression to 34.0 -4-193 % in thrombin (p<.001, two-tailed paired t-test), and 5.2 + 2.2 % (p<.001) in ADPactivated platelets. Basal GP53 expression was 1.7 _+ 0.5 % and increased to 63 .0 + 6.4 % in thrombin, and to 7.7 _+ 3.4 % in ADP-stimulated samples. With SIN-l, GP53 expression decreased to 346 _+ 10.4 % (p<.001) in thrombin, and 3.0 5:1.4 (p .001) in ADP-stimulated samples. Conclusions: These data implicate, that NO leads to a significantly reduced activation of surface molecule expression in thrombin and ADP-stimulated platelets. In addition, flow cytometry might be a useful tool for studying modulation of platelet activation by NO or NO-releasing compounds. INTRODUCTION: Acute cadmium poisoning is very rare. On initial presentation may mimic metal-fume fever, but acute inhalation cadmium toxicity may produce fatal chemical pneumonitis. CASE REPORT: We present a case of acute fatal respiratory failure secondary to cadmium-fume irthalation. A 53 year old patient was trasferred from another hospital with acute respiratory failure presumably due to pneumonia. The last days before he had had commom cold symptoms. He had been cutting with a welder during one hour without any respiratory protective measure. Three hours after exposure he developed progressive dispnea and was admitted to hospital. With presumtive diagnosis of respiratory infection, antibiotics were begun, however be failed to improve. All microbiological studies were negative. Chest x-ray showed bilateral diffuse infiltrates. On seventh day he needed intubation and mechanical ventilation and on 10th he was admitted to our ICU. Antibiotics were stopped and new microbiological studies were performed including brochoalveolar lavage and virologic studies. All results were negative. He developed progressive hipoxemia and hipercapmia and finally, multiorganic disfunction syndrome. He died 19 days after exposure. The metal he had been working with was a 10% cadmium alleation. Blood cadmilam concentration 15 days after exposure was 0.34 mcg Cd/g Cr, and urine cadmium concentration was 16.9 mcg/L. On postmortem examination, tissue cadmium concentrations were: blood 175 ng/ml, liver 823 ng/g, kidney 3571 ng/g and lung 1143 ng/g. These values confirm that cadmium was the cause of the fatal respiratory illness in this patient. CONCLUSION: This case evidences the considerable hazard of acute poisoning after inhalation of eadmium-fume and stresses the need of appropiated safety measures against metal-fume poisoning. Aim : Lactic acidosis is considered the hallmark of cyanide poisonirig. However, the relationship between plasma lactate and blood cyanide levels has not been determined. The aim of this study was to determine the significance of plasma lactate concentration (PLC) during the course of cyanide poisonings. Methods : The patients were included according to the clinical suspicion of pure cyanide poisoning at the time of presentation. Fire victims were excluded. Serial blood samples were collected before and after intravenous hydroxocobalamin (HOCo). Blood cyanide concentration (BCC) was measured colorimetrically. PLC was measured enzymatically. Results : 8 patients were studied. On admission, PLC ranged from 4.8 to 53 mmol/L, and BCC from 12.7 to 256 gmol/L. Mean systolic blood pressure was 80 • 56 mm Hg, mean arterial pH 7.34 • 0.14, mean anion gap was 29.7 + 7.7 mmol/L and mean PaO 2 32.2 • 27.0 kPa. Three patients died. Before antidotal treatment, there was a significant correlation between PLC and arterial pH (p = 0.008), anion gap (p = 0.008) and BCC (p = 0.016) but not with heart rate, PaO2, PaCO 2 and blood glucose, or blood pressure. During the whole course of the poisoning, a PLC _> 7 retool/1 was a sensitive and specific indicator of a blood cyanide concentration > 40 ~tmol/1. Sustained catecholamine administration reduces the correlation coefficient. Conclusion : Baseline measurement of PLC allows assessment of severity of acute cyanide poisoning. Thereafter, PLC may be used to assess the adequacy of antidotal treatment, more especially in patients not requiring sustained infusion of catecholamines. Aim: the aim of this case report was [o study the correlation between the plasma lactate levels and several clinical, biological, and toxicological parameters serially measured during the course of a cyanide poisoning treated with a high dose of hydroxocobalamin. A 63-year-old male ingested potassium cyanide leading to cardiac arrest. CPR was performed prior to hospital arrival where the patient received 10 g hydroxocobalamin. SBP rapidly returned to normal allowing withdrawal of epinephrine. The patient remained comatose and died from brain injury 12 days after the ingestion. Methods Plasma lactate and blood cyanide levels were measured serially. Blood cyanide levels were measured using a colorimetric method.~ Plasma lactate levels were measured using an enzymatic method. For correlation Spearman rank correlation test was used. Results. Initial plasma lactate and blood cyanide levels were 53 mmol/L and 256 gmol/L, respectively. There was no overall correlation between SBP and either blood cyanide or plasma lactate levels. Similarly, there was no overall correlation between arterialvenous oxygen saturation difference with either blood cyanide or plasma lactate levels. In contrast there was a strong correlation between blood cyanide and plasma lactate levels (R=0.976, P<0.0001). The time-course of the blood cyanide concentrations was described by a mono-exponentiai decay (R2=0.968) with a blood half-life of 1.14 h. Similarly, the time-course of plasma lactate levels was described by a mono-exponential decay (R2=0.986) with a blood half-life of 3.94 h. Discussion. In this case of acute human poisoning, SBP was a much poorer indicator of continuing cyanide effect both before and after antidotal treatment, than was lactate production. This suggests a potential clinical role for following serial plasma lactate levels as a marker of the evolution of cyanide toxicity. Aim : Cyanide (CN) poisoning in fire victims is frequent and rapidly fatal. In a prospective study we tried to assess the clinical tolerance of a high dose of hydroxocobalamin (HOCo) administered at the scene of the fire in fire victims suspected of CN poisoning. Methods : Inclusion criteria : Soot in mouth or sputum ~ any degree of neurological impairment. Exclusion criteria : children, pregnant women, burns of total surface body area > 20 %, multiple trauma. Protocol desigrl following examination and the collection of a blood sample in dry heparin, a 5 g dose of HOCo (10 g in case of cardiovascular collapse) was administered intravenously over 15 min. The systolic blood pressure was monitored before and after the administration of HOCo, and one hour later. Results : there were 28 females and 22 males. The mean blood CN concentration was 83 • 73 pmol/1. The mean blood carbon monoxide was 3.2• 2.1 mmol/1. Nineteen fire victims eventually died. Among the non-CN-intoxicated patients (blood CN < 40 ~mol/1), there was no significant change in arterial blood pressure. In the 33 CN-intoxicated patients (blood CN > 40 gmol/1) a significant increase in blood pressure was observed both immediately (p < 0.001) and 1 hour later (p < 0.001) after the admistration of HOCo. No allergic reactions were observed. Conclusions : In fire victims with cyanide poisoning, the administration of a high dose of hydroxocobalamin was associated with an improvement in systolic blood pressure. Hydroxocobalamin is well tolerated in fire victims without CN poisoning. Objectives: Tricyclic antidepressant (TCA) overdose can lead to serious complications including cardiac arrhythmias [ 1] . Because of the known risk of early deterioration and the implication for management, emergent evaluation is essential. We determined the diagnostic usefulness of the electrocardiogram (ECG) in TCA poisoning. Methods: Retrospective study of all patients with TCA intoxication (pos. ,toxicology screening in urine and/or pos. history) in a 800-beduniversity hospital from 1989 through 1994. The severity was graded with mild= no symptoms or agitation; medium= disorientation, somnolence, tachycardia, or convulsions; and sever~ coma, significant arrhythmias or death. We analysed the first ECG after admission with a special emphasis on QRS-and QTc-intervals and the terminal 40ms frontal plane QRS-vector (tQRS), which, was reported to lie typically between +130 and 270* 798+310 915+453 5609• The best correlation with severity grade was found with QRS-and QTc-duration (p=0.0001), the TCA-dose (p=0.0003) and HF (p= 0.027); tQRS did not correlate. 2 patients died (5.7%). Conclusion: QRS-and QTc-prolongation in the admission ECG, and the reported dose of ingested drugs are useful predictors for severity of poisoning due to tricyclic antidepressants. We did not find additional benefit in determining the terminal 40ms frontal plane QRS-vector. Objectives: Since treatment of amphetamine poisoning is usually symptomatic and often associated with a fatal outcome, a search for specific drugs to help the amphetamine-intoxicated victim is sorely needed. Methods: We report a case of a suicidal ingestion of large amounts of the amphetamine-derivative 3,4-methylenedioxy-ethamphetamine (MDEA) and heroin (diacetylmorphine) and present the hypothesis that the two drugs produce opposing clinical effects. Results: A 25 year old caucasian male was admitted to the emergency ward because of acute-onset confusion. At presentation, he was agitated and showed increased muscular rigidity. He had taken 40 tablets of "Eve" (MDEA, approx. 4 g) and 12 g of "Smack" (heroin) by oral route approximately 2 h before admission. Because of rapidly progressive tachypnea and exhaustion, the patient was intubated and ventilated. The serum concentration of "Eve" on admission was 1400 ng/ml (lethal range 950-2000 ng/ml). Trace amounts of cocaine and substantial amounts of heroin (115 ngtml; mean value in heroin-related deaths: 190 ng/ml) were also found in the serum. The patient was successfully weaned from the ventilator by day 4 and recovered without persistent neurobehavioral disturbance. Despite high serum levels of both drugs, the patient did not present with the classic signs and symptoms normally seen during intoxication with these drugs. Amphetamines in general, and MDEA in particular, have opposite clinical effects to heroin or diacetylmorphine. None of these were however present in the case presented despite the high ingested doses and the serum levels in the lethal range. Conclusions: The fascinating fact that, apart from the respiratory depression, none of the clinical signs reported after massive overdose with these two drugs were present, might be attributed to the opposite pharmacological effects of MDEA and heroin. We believe that the patient unwittingly saved his own life by the oral coingestion of both MDEA and heroin. Our clinical data raise an interesting point about the pharmacological treatment of acute poisoning with amphetaminederivatives. Introduction: The acute attack of AIP still carries a significant risk of mortality of around 10 %. A succesful outcome depends on early diagnosis, removal of pricipitating factors and provision of intensive supportive therapy. Objectives: Twenty one patients (20 females, 1 male) with documented AIP were seen over a 10-year period in the University hospital. 1 patient was in clinical remission and 20 were with the acute attack of AIP, Among them 4 with respiratory paralysis were required artificial lung ventilation and 4 -assistant ventilation with PEEE Pathologic treatment during the attack was normosany, Adenil, androgenes, glueosa, Riboxin parenteral and enteral nutrition via nasogastric tube. Symtomatic treatment -pethidine, propranoton, antibiotics, bronchoscopia. Methods: Intermittent phasmapheresis was performed on 15 patients. The following measurements were peformed: level of porphobilinogen (PBG) in the wire and delta-aminolevulinic acid in the blood. Hematological and routine chemical evaluations, hepatic, hemodynamic and respiratory function. Results: After plasmapheresis the median PBG excretion (normal range 1-2 mkg per/1 kgr creatinine) fill from 188 mkg on admission 140.8 mkg, then on 3-5 day raise to 193 mkg and then during treatment with normosong and Prasmapheresis lowest level was 32.9 mgk. Fatalities occured in two females during attacks with proforma cerebral involvement and 13 patients attained clinical remission. Conclusion: After therapy with plasmapheresis normosong we found that there was consistently reduce the urinary excretion of PBG and shortening the duration of the acute attack. Objectives: Pigs has been reported to present with a higher pulmonary arterial pressure (Ppa) and stronger pulmonary vascular reactivity than many other species, including man. Aim of the present study was to compare pulmonary vascular impedance (PVZ) before and after embolisation in weight-matched adult dogs and minipigs. Methods: We investigated PVZ spectra in 6 anaesthetized and ventilated (FiO2 0.4) minipigs and 6 dogs. After baseline measurements the animals were embolised with autologous blood clots to reach a Ppa above 35 mmHg. Results: Flow (03 and Ppa matched PVZ data (mean-+SEM) are shown in the table. [Zo = 0 Hz impedance (Z; {dyn.sec_em-5}); Zl = first harmonic Z; Zc = characteristic Z; Z1 phase = first harmonic phase a@e {radians}; Fmin = frequency of PVZ the first m{n~mam; *, f p at least < 0.05 between dog and minipig, and before v~. after embolisation respectively]. Before Case Report: A 53-yr-o]d woman affected by legs recurrent thmmbophlebitis, was admired in Medmine Department for tach.~pnea, chest pain, tachycardia and cyanosis. Before starting two-dimensional transesophageal echocardiography (TEE) to confirm the suspicion of pulmonary embolism, she suddenly had ventricular fibrillation. Resuscitation and defibrillation were readily performed. When sinus rhythm was reinstituted she was in superficial coma with preserved corneal and light reflexes: right hemiplegia, poor perfusion and h~posphygrma of the left arm. TEE showed dilation of rigth ventricle (RV), incomplete occlusion of pulmonary arter~ (PAl at it~ hifurcation, severe tigth-to-left shunt through a patent foramen ovate, paradoxical embolism with incomplete occlusion of left subclavian artery Mechanically ventilated with Vt=800 ml, RR=12/mm, FiO2=l, the patient had pH=7.28, PAO2=57 mmHg and PACO2=45.1 Systemic BP was 130/80 mmHg and HR=80 b/min with low dose epinephrine (0.12 g/Kg/min) A thrombolytic infusion (rTPA: 100mg/2h) through a peripheral vein was started TEE imaging and clinical status 3 hours later were unmodified. A new rTPA infusion was performed through the pulmonary hole of a Swan-Ganz catheter with the tip close to the embolus. One hour later PA pressure decreased from 46/30 mmHg to 36/25 mmHg, ETCO2 increased from 26 to 30 mmHg and SaO2 improved from 89% to 96% Three days later the parietal, spontaneously breathing and with normalized TEE scans of RV and PA, was transferred to Rehabilitation Service to perform physical therapy. Conclusions: Massive pulmonary embolism in a patient with patent foremen ovale, paradoxical embolism and refractory hypoxaemia was unaffected by systemic rTPA infusion, while intrapulmonary rTPA administration dramatically improved gas-exchange, hemodinamics and the general conditions of the patient. The presence of a large rigth-to-left _atrial shunt and the rapid rTPA metabolism could likely explain the effectiveness of its intrapulmonary administration in front of failure of systemic thrombolysis. Introduction. Cardiogenic shock during massive pulmonary embolism (blPE) is due to an acute increase of right ventricle (RV) afterload and possibly RV ischemia causing a failure of RV pump function. The rec~;mmended therapeutic strategies are: xoIume augmentation ~n ~rder m }ncrease RV pre-h~ad, adrenergic drugs to increase t'ontractillly and maybe coronary perfusion, fibrinolytic drugs to delermine clot lysis. There have been several reports of noradrenaline (NA) as a useful drug in this setting for its sluing ~z, but also 1~, properties. Case report.An obese 75 },ears old woman was transferred to our ICU for tetanus. She was given the usual antibiotic and immunoglobuline therapy. l'wo thoracic epidural catheters were put in place at different levels and replenished with marcaine qid. A continous infusion of sedation (diazepam § was started together with mechanical ventilation. Curarization ~,as given occasionally. Fraxiparine 0.3/die was used for prophylaxis of thrombotic disease, On day 8th at 11.00 a.m. she started to be hypoxic (Sa02 90%), tach3,tardic l1 l(I b/rain.), her blood pressure(RP) dropped frum norma~ values to 6r mm/hg, the central venous pressure (CVP) raised [rom lb to 27 mm/hg and the end tidal CO2 was 7mm/hg lower than one hour before. The physical examination of the chest revealed a clear bilateral ventilation and the chest X-ray was normal apart from an elevation of the :tiaphragm as compared to the previous. An E.C.G. showed sinus tachycardia, right bundle branch block and a possible inferior necrosis (which was already present on admission). A trans-thoracic echozardiography was performed which showed "an acute overload of the right centricle wilh remarkable dilatation. Tricuspidal regurgitation ++. Paradoxical movement of septum. Small left ventricle with normal wall kinetics". The cardiac enzymes were later shown to be normal. An acute massive pulmonary embolization was assumed m be present.. A bolus of streptokinase 750 x I(I 3 U. was given fonowed by a continous infusion . Two liters of colloids were also given in a sh~rt time, Two hours later the patient was still deeply hypotensive, hypoxemic and anurir(BP 54/32 mm/hg, CVs 23 mm/hg, Spo2 90%) despite a cominnus infusion of dobutamine 20 fag/kg/min and adrenaline 0.5 ~tg/kg/min. At this stage a bolus of aoradrenaline 20 ,g was given followed by a cnntinous infusion of 0.05 !*g/kg/min. An immediate improvement of the hemodynamics was noticed and one hour later the BP was 149/77 mmhg, the CVP 24 mm/hg, the SaO2 100% and a brisk diuresis started. The hemodynamics kept stable and weaning from vasoactive drugs was achieved within two days. One month iater the patient was discharged home in good conditions.. Con c I u sio n.NE administration may help to restore RV coronary flow and ;~ump function during MPE. Aeute putmonary t~omboembo~sm [FfE) cou16 be mamfesLated with either respiratory or cardiovascular syndromes or both. the arm of the study was to establish leading respn'atory symptoms, frequency and form of the roendganographic (rig) changes as well as blood gas disturbance degree in acute PTE with dommam respiratory disease appearance. The study includes retrospeotive analysis of I 14 PTE patients (Pts), 63 males (average age 47,7 yrs) and .q females (average age 53,2 yrs). They were admitted at university, olinie" with suspection ofpleuropnlmonary disease, including PTE. Final diagnosis of PTE was based o~ evident risk factors in 94,7% of the eases (deep venous thrombosis, surgery, trauma, imobilisation, malignancy ere), acceptable clinical, rtg, sdntigraphic and laboratory findings, as well as deep veins examination by Dopple~-sonographie and radioisotopic -~enogmphy. Respiratory symptoms appeared in all cases: sudden pleural pain (79%), dyspnea (64%), hemoptysis (49%), cough (39%) with association of two or more symptoms in 93%. Chest xrays findings were abnormal in 92% with diaphragmal elevation (74,2~ lung opaeilies (69,5%), atelectasis (48,5%), plemal effusion (35,2%), main pulmonary brancah asimetry (22,8~ oligemia (19%), heart shadow changes (10,4%) and pulmonary arteries "cut off' (6,6%). The association of two or more abnormalities was found in 92,1% while normal chest x-rot was found in 8~ of the cases. Hypoxemia with PaO2<10,4 kPa was found in 64,4% followed with hypocapnia and respiratory alealosis in 34,6% In 27,8% of the gas exchage analysis were within normal limits. Among cardiovascular symptoms short syn~cpa appeared in I0,5%, ECG changes-St Q3T3 type in "~1,6%. Results show high frequency of positive ~g findings in PTE Pts that is opposite to oppinion that chest x-ray in acute FIE is the most ofran normal. Leading symptoms are pleural pain and dyspnea, while hemoptysis were found in a half of the study group. Blood gas changes were present in two thirds of the cases. Kakkar, in his classic work ,clearly demonstrated the efficiency of low doses of heparin in prevention of deep vein thrombosis (Lancet 2:669,1971) .After this first study the application of heparin prophylaxis became more and more diffused until to be considered a routine in many surgical departement.Actually application of blood saving technique induces postoperative hemodilution effect. In that condition prophylaxis routinely applied seems a nonsense and can be at risk for postoperative hemorrhage. Methods: To analize this problem we compared 100 patients arrived in our intensive care unit (I.C.U.) in.1980: (Group A) with 100 arrived in 1994: (Group B) .Every patient was operated for major abdominal surgery.In each one we considered the hemoglobin (Hb) value,hematocrit(Hct), and coagulation pattern (C.P.) at the arrive in I.C.U. and 24 hours later. The patients was also divided in those receiving heparin prophylaxis (I) from not treated patients (II) Results:The application of blood saving technique clearly appears from the Hb and Hct level wich have a mean value of 11,4 +/-1,8 (Hb) and 34 +/-2 (Hct) in Group A while in Group B mean value are 9,7 4-/-1,2 (Hb) and 29 +/-2(Hct).Patients of Group A (II) are the only one where a pathologycal C.P. with statistical significance has been demonstrated.In this Goup we got four cases of evidence of venous thrombosis and one of pulmonary embolism.In patients of Group B(I) we encontered the incidence of two cases of severe hemorrhage despite the absence of statistical significance in C.P.modifications. Oxygen desaturation during broncho-alveolar lavage: role of oxygen saturation monitoring in prevention of acute respiratory insufficiency G. Galluccio, B. Valeri, S.Batzella, M. Di Lazzaro*, Servizio di Endoscopia Toracica, Ospedale Forlanini, Rome, Italy * Servizio die Anestesia a Rianimazione, Osp. Forlanini The broncho-alveolar iavage is a diagnostic procedure employed in interstitial diseases of the lung. It requests the introduction through the working channel of a fiberoptic bronchoscope, after occlusion of a segmentary bronchus, of aliquots of saline solution at 37 C, subsequently gently reaspired, in order to remove cells and proteins from ELF (Endoalveolar lining fluid), which is related to interstitial medium. Bronchoalveolar lavage induces deep effects on pulmonary function: -Lowering of the alveolar surface of exchange; -Shunt effect, depending on the perfusion of non-ventilated districts; -Increased pulmonary arterial pressure, due to hypoxic vasoconstriction; -Decrease of lung compliance. In this report the Authors present the result of oxygen saturation monitoring in a group of patients with interstitial lung disease, who underwent diagnostic broncho-alveolar lavage. In most patients with severe interstitial involvement, the lavage performed without supplement of oxygen induced a severe fall in the oxygen saturation during the late phase of the procedure. If supplementary oxygen was delivered during bronchoscopy, since its beginning, only slight modifications of the curve were detected. In patients without thickening of interstitium, in whom the lavage was performed in order to obtain material for bacterial or cytologic examination, no modification of oxygen saturation was observed in standard procedure. As conclusion the Authors strongly reccomend monitoring oxygen saturation in patients with radiologic evidence of interstitial involvement also in patients with no evidence of dyspnoea. G. Galluccio, B.Valeri, S.Batzella, M. Di Lazzaro*, Servizio di Endoscopia Toracica, Ospedale Forlanini, Rome, Italy * Servizio die Anestesia a Rianimazione, Osp. Forlanini The treatment of choice in patients with alveolar proteinosis consists of pulmonary lavage. This procedure requests the introduction, through the working channel of a fiberoptic bronchoscope, segment by segment, of aliquots of saline solution at 37 C, subsequently gently reaspired, in order to remove the proteins deposited in the alveolar spaces. The method is very similar to that used in bronchoalveolar iavage, a diagnostic procedure used to obtain cells and substances from ELF (Endoalveolar lining fluid), which is related to interstitial medium. As known, bronchoalveolar lavage induces oxygen desaturation, because of shunt effect. Understandably, one lung lavage has remarkably more deep effects on pulmonary function than bronchoalveolar lavage, for the amount of fluid introduced, the length of the procedure and the conditions of controlaterai lung. In this report the Authors present the result of oxygen saturation monitoring in a patient who underwent pulmonary lavage for alveolar proteinosis. In the lavage performed without supplement of oxygen a severe fall in the oxygen saturation was observed during the late phase of the procedure. If supplementary oxygen was delivered during bronchoscopy, since its beginning, only slight modifications of the curve were detected. As conclusion the Authors strongly reccomend the subministration of supplementary oxygen in pulmonary lavages, also in patients with excellent respiratory conditions. A. B. Dublisky prof., M. R. Isaakjan ass., V. A. Zasukha, S. M. Vinichuk prof., V. P. Tserty ass. prof., Chair of Anaesthesiology, Resuccitation and Medicine of Catastrophes, Neurology of Ukrainian State Medical University, Kiev, Ukraine. Objectives: detection of plasmophoresis's influence of results in treatment of ishemic insult. Methods: We ve investigate 25 patients with ishemic insult, treated with reverse plasmopheresis in complex treatment. After primary infusive therapy we took 400 ml of patients' blood and separated it within 15 min with rotation frequensy of 2000/rain. After separation of erythrocytes from plasma, the latter has been returned to patients. We made 3-4 procedures during 3-4 days. Hemoglobin, hematokrit, time of blood coagulation were determinated. The brain blood flow in internal carotid arteries, regional volum brain blood flow and total brain bIood flow were evaluated with tetrapotar chest rheography and tetrapolar rheoencephalography. Obtained date were comparised with control group after traditional treatment. Results: It was found that after reverse plasmopheresis the hemoglobin and hematokrit levels decreased significantly in studied patients' plasma (from 140 + 3.2 g/l to 120 _+ 2.3 g/1 and from 44 + 2.1% to 35 _+ 1.8 % respectively). The time of blood coagulation by Lee-White has increased by 2-2.5 times (up to 10-12 rain). The level of brain blood flow has been increased significantly after reverse plasmopheresis in comparison with control group. The following tests of brain blood flow have been increased: a) the total volume brain blood flow from 480.7 + 34.6 ml/min to 625.4 _+ 35.4 ml/min (p < 0.05); b) the regional brain blood flow from 52.2 _+ 2.8 ml/min to 87.1 + 6.2 ml/min (p < 0.01); c) the brain blood flow in internal carotid arteries from 166.1 _+ 12.2 ml/min to 206.3 + 14.6 ml/min (p < 0.05). Conclusions: The use of reverse plasmopheresis in complex treatment of patients with ishemic insult aIIows to improve rheological blood patterns, helps to increase volume brain blood flow. It results in quicer reparation of neurological functions. Objectives: a prospective evaluation of the efficacy of continuous infusion of verapamil in reducing the incidence of postoperative atrial fibrillation after pulmonary surgery. Methods: A total of 199 consecutive patients, 100 on verapamil, 99 on placebo was included after lobectomy or pneumouectomy. A loading bolus of verapamil (10 mg over 2 minutes) was followed by a rapid loading infusion (0.375 mg/min) for 30 minutes and finally a maintenance infusion (0.125 rag/rain) for 72 hours. Results: A mean plasma level of verapamil of 150 ng/ml was obtained only after more than 24 hours. Atrial fibrillation occurred in five out of 78 patients who tolerated the verapamil infusion, and in 15 out of 99 patients on placebo (p = 0.08). Verapamil infusion was not tolerated in 22 patients because of hypotension or a heart rate of less than 50/min, within 6 hours of the start of the therapy. When atrial fibrillation occurred, the ventricular response, mean _+ SD, was not significantly slower during verapamil infusion (132 + 22) compared to placebo (147 + 20). Conclusions: Because of its frequent side effects and the only modest efficacy verapamil should not be considered for prophylactic therapy of atrial fibrillation after pulmonary surgery, and is probably not a good first choice for slowing the heart rate in case of rapid ventricular response once atrial fibrillation has occurred in these patients. Results: Study of haemostasis in these patients has showed deep disturbances of blood coagulation. Fibrogen level has reduced to 0.62 + 0.03 g/l, fibrinogen and/or fibrine degradation products concentration have enhanced to 0.40 _+ 0.04 g/l, monofibrin soluble complex concentration to 0.08-+ 0.04 g/l, blood plasmin level was enhanced to 34.0 + 0.2 mmol/1, plasminogen proactivator level was also enhanced to 153.0 + 0.60 ram, plateletes aggregation has decreased to 52 %. After plasmopheresis aggregation was decreased in 1.6 times. It has been connected with decrease of fibrin and/or fibrinogen degradation products level and level plasmin in 1.7 times, and plasminogtnt activator level in 4.6 times. At the same time we have observed increase in total antifibrinalitic activity of blood in 1.3 times. Activity of activators plasmine and plasminogene proactivators has decreased in 1.2 times and in the same time activity of activation inhibitors and antiplasmines has increased in 7 times. Conclusions: Plasmapheresis leads to considerable improvement of a general condition and reduction of the haemorrhagic syndrom's sings (controlling of gastrointestinal haemorrage, reduction of intensity of subcutaneons haematoma). Evaluation of continuous cardiac output (CC0) monitoring based on thermodilution technique in 35 critically ill patients. Methods: Cardiac output (CO) was monitored continuously using a modified pulmonary artery (PA) catheter, on which a heating filament is located and by which energy is transmitted to the circulating blood. A microprocessor calculated CO by a new algorithm. Standard bolus thermodilution technique (10ml of ice-cold saline solution) was used to compare CC0 with intermittent bolus cardiac output (IC0) measurements. The following subgroups were prospectively studied: I. heart rate (HR) >120 beats/min, 2. cardiac output >10 i/min 3. cardiac output <4.5 i/min, 4. rectal temperature >39.0~ and 5. PA catheter was inserted for more than 4 days. Results: A total of 404 pairs of IC0 and CC0 measurements were obtained from the 35 patients. Bias (ICO measurement minus CC0 measurement) of all measurements were 0.03• i/min and the 95% confidence limits (mean difference• were -1.01/1.06 i/min. Also in the subgroups, CC0 measurement agreed closely with ICO measurement (C0 >10 i/min: bias=0.16• i/min; CO <4.5 i/min: bias=-0.17• i/mln). Elevated temperature and prolonged lay-days of the PA catheter did influence agreement of CC0 measurement with IC0 measurement neither (>39~ bias=0.09• i/min). Conclusions: Monitoring of CC0 using a modified pulmonary artery catheter with a heated filament has proven to be accurate and precise also in the critically ill when compared with "standard" intermittent bolus thermodilution technique. This method enhances our armamentarium for more intensive monitoring of these patients under various circumstances. Background: The number of patients who need coronary artery surgery WAS) grows every year. Most of these surgical operations are with extrar eircuiation (ECC). Since January 1994, this surgery is made without ECC in selected patients in our hospital. This technique is exceptional in Spain. This type of surgery has proved useful in patients requiring revascularization of the left anterior descending, eireunflex or right coronary artery (not for grafting the pos~tefio~r descending branch}. blethods and Results: Since 1994, 30 patients aged 54 to 77 years (mean 66 years) underwent CAS without ECC. The mortality in programmed surgery was 0%. No patient was reexplored for hemorrhage. The mean values of some clinics parameters v~ere: a) blood requeriments: 2 units per patient, b) need of mechanical ~entilation: i3,6 hours, c) postoperative bleeding: 900 cc, d) days at ICUI 2,5. We used the student % t test or Fisber~s exact test to compare these results with the mean values of surgery with ECC: a) blood requeriments 4 per patient (p<0,0001), b) need of mechanical ventilation: 29 hours (p<0,0001), c) postoperative bleeding: 1300 cc (p<0,004), d) days at ICU: 4 (p<0,001), e) programmed surgery mortality: 7% (p<0,05). Conclusion: Our limited experience shows that this surgery is an alternative in the treatment of coronary disease, especially for aged patients with associated pathology and in Jehova's witness. The need of mechanical ventilation, days at ICU, blood requeriments and morbi-mortality were fewer than surgery with ECC. To study the hemodynamic and antiarrhythmic influence of ACE-inhibitor enalapril in acute myocardial infarction (MI). Methods: Holter ECG monitoring, heart rate variability analysis, echocardiography (3 and l0 days after beginning of the treatment), stress-echocardiography and stress ECG (8-10-th day after the onset of MI). Enalapril was included into the treatment of 42 pts with MI (study group), with normal or increased blood pressure, from the 1-st day of the disease. The data were compared with 30 pts treated without enalapril (control group). Results: Silent ischemia during stress-test was registered in 6 pts of the study group and 8 of control group, the arrhythmia episodes during stress test -in 5 and 8 pts and episodes of silent nocturnal isehemia -in 7 and 12 pts correspondingly. Enalapril importantly attenuated the hypertensi~re re~aetioh %0 stress test. In 10 pts of the study group the number of perifocal hypokinesis zones decreased; in the control group it didn't change. The quantity of ventricular extrasystoles in the patients of the study group decreased by 25%; the heart rate variability indices improved as well; in the control group the character of ventrieulir arrhythmias, heart rate and its Va]~i~bili%y didn't change significantly. Conclusions: The inclusion Of enalapril into the treatment of MI is a Useful t0ol to improve hemodynamie parameters and decrease the incidence of ventricular arrhythmias. Objectives: To stUdy left ventricular (LV) systolic function in the patients with acute myocardial infarction (AMI) before and after peroral captopril test. Methods: The original echocardiographic parameter of LV contractility, "coefficient of effective systolic function" (CESF), was proposed in the study. CESF is calculated from LV stroke volume (SV), obtained from Doppler aortic flow in LV outflow tract and LV end-diastolic diameter (EDD): CESF =SV/EDD. The study included 60 patients with AMI, who had local LV dyskinesia and global LV systolic dysfunction (EF<45%). Besides CESF, the ejection fraction was calculated before and after administration of 25 mg eaptopril (on the fifth day of AMI) by methods of Bullet and Simpson. Results: The dynamics of these parameters, as well as heart rate (HR) and mean blood pressure (BP), is shown in the tabte. Before cal~topril EF (Bullet) 32.12 • 2.51 EF (Simpson) 35 . Introduction: The COLD system is a monitoring system for measurement of right (COpa) and left (COart) ventricular cardiac output, cardiac function index (CFI), fight ventricular ejection fraction cRVEF), fight ventricular cnddiastolic volume (RVEDV), intrathoracic blood volume (!TBV), global enddiastolic volume (GEDV), lung water (ETV) and excretory liver function (PDR). Patients and methods: 41 pts have been monitored by the COLD system. Above mentioned parameters are measured by thermal dye dilution and a fiheroptic femoral artery catheter. COpa, RVEF and RVEDV measurements additionally were compared to measurements by the Baxter EXPLORER. :::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: ;;;K;;;;I CoV (%) EXPLORER ! 6 13 ! 1 [ GEDV, ITBV and PDR showed a significant decrease dufing the first 12-24h after the operation, CFI and RVEF si~canfly improved after 48k wheras ETV showed a i~ in the early postoperative phase and fell to normal ranges at 48h. Comparison of COLD/EXPLORER m~ements sb0wed good correlations. Discussion: Concerning m0~toring of ri,ght ventric~ar function COLD and EXPLORER can he seen as equal. RVEF gives an ar report about the performance of the right ventricle without use O f echocardiography. Measuring ITBV and GEDV ~ improve ~gement and con~ol of th.e volume status, monitoring ETV helps preventing lung edema. PDR shows good corre|ati0n to liver blood chemistry and is bedside avai|ab|e. Thus the COLD system offers additional parameters for comprehensive m~nitofing of pts. ~e~ ~c surgery. Obiectives: To evaluate the influence of an a!'~ered cardiac function on the cardiovascular response to the increase in oxygen demand induced by an increase in core temperature. Methods: This preliminary study included 12 adult critica!ly ill patients monitored by arterial and pulmonary artery catheters in whom thermodilution cardiac index {CI) and arteria! and mixed-vef)ous blood gases measurements could be obtained before and after an acute change in core temperature of at least 0.5~ (max 60 rain apartL The patients were separated in two groups according to their cardiac function: 4 patients had an impaired cardiac function as defined by a history of cardiac disease and an ejection fraction below 45% and 8 patients had normal cardiac function. Results: Individual data are shown in the figure. In contrast to the control group (continuous line) in which C! increased without changes in oxygen extraction (02ER), the Q2ER in patients with impaired cardiac function (dottled line) increased without changes in CI. Conclusions: The increase in oxygen demand associated with changes in temperature is met by an increase in C! in patients with unaltered cardiac function and in an increase in O2ER in patients with altered cardiac function. Temperature should be taken into account in the assessment of the adequacy of cardiac output in patients with impaired cardiac function. Objectives: To define the hemedynamic and metabolic response to physical therapy(PT) in relation to the type/level of sedation and the cardiac status in ICU patients. Methods: We studied 34 mechanically ventilated ICU patients (64• years) in stable hemodynamic status (no change in vasoactive treatment for at least 2 hours), separated in 4 groups: Group 1 = deep sedation, cardiac dysfunction required dobutamine (N=5)r Group 2= deep sedation (barbiturates), unaltered cardiac function (H=lO), Group 3= moderate sedation, altered cardiac function (H=7) and Group 4= moderate sedation, unaltered cardiac function (N=12). Complete hemodynamic data, arterial and mixed venous blood gases, respiratory gas analysis (Metabolic Cart CCM, Medgraphics) were obtained at baseline (2x) and twice (q.10 min) during leg mobilization. Data were analyzed by ANOVA. Calcium channel blockers were used in complex preoperative preparation of 59 hypertensive surgical patients. Patients were allotted to 3 groups based on their hemodynamic profile: hypokinetic: ejection fraction (EF)<0.6, 29 patients; eukinetic (EF>0.5),I6 patients and hyperkinetic (EF>0.6),I4 patients. The most noticable change in hemodynamics was in the hypokinetic group: EF and cardiac output (CO) were significantly decreased (p<0.001) while systolic arterial pressure (SAP) (p<0.05) and peripheral resistance (PR) (p<0.01) were elevated. The results showed that in hypokinetic patients on nifedipine EF (p<0.00t) stroke volume (SV) (p<0.0l) and CO (p<0.001) were increased while PR(p<0.0t), SAP(p<0.001) and diastolic arterial pressure(p<0.05) were decreased. Eukinetic type patients also showed an increase in EF,albiet to a lesser extent,than in the hypokinetic group. Increased SV and CO(p<0.01) were observed in eukinetic patients though this was to a lesser extent than in the hyperkinetic group. In the hyperkinetic group of patients nifedipine had no effect on the aforementioned parameters except for a decrease in SAP(p<0.0 I). Nifedipine increased EF in all hypokinetic patients. Comparative results show that isoptin was less effective than nifedipine in decreasing peripl~eral vascular resistance and had a depressive effect on the myocardium. It can be concluded that the action of calcium channel blockers normalizing the circulation in the hypertensive surgical patient depends on: the condition of myocardium, the patients hemodynamic profile and their pharmacological properties. They were most effective in the hypokinetic group. Zalo/nthinos E., Daniil Z. Zakynthinos S., Armaganidis A., Kotanidou A., Nikolaou Ch..,Roussos Ch. Critical Care Department, University of.Athens, Evangelismos Hospital, Athens, Greece. Introduction : Surgical is the optimal treatrnent for Ioculated effusions and the preferable procedure when multiple bands are seen in the pericardial sac by echo. Patients : 6 palients, 4 post cardiac surgery, 2 uremic (3men, 3 women) with large pericardial effusion and clinical or echocardiographic findings of tamponade or both. These particular patients displayed numerous linear echo-dense bands and s~'ands crossing the pericardial space (in one of them a Ioculated effusion compressed the left ventricule). One had aPTT increased, four were mechanically ventilated. Technklue : A 8Fr polyurethane catheter with end and multiple side holes over 18 Ga needle was echo-guided to the ideal site (fluid abundant and closest to the transducer). The catheter was attached to a close system with a Heimlich valve for continuous drainage (pneumothorax kit). Subcostal entry was selected in one patient and chest wall in five. The patient's position was changed every hour at least. (We believe that the small changes in the position of the catheter and the mechanical breaking of the bands in relation with the movement of the heart assist the pericardial fluid to remove). Results : In all cases only a small quantity of fluid was withdrawn in the first minutes(30-70ml) with some clinical and echo-findings improvement. The fluid was bloody or serosanuginous with high protein content (Ht=15% ,Protein 5,1gr/dl) in all cases. In first 24 hours the mean volume of fluid removed was 550ml (350to 720ml). In that period echo showed no residual fluid. The catheter remained within the pericardium 1 to 3 days .. No complications are mentioned. Conclusion : Cardiac tamponade due to hemorrhagic high protein pericardial effusion in uremic and postcardiac surgery patients,, as it is revealed by echo dense bands, can be faced by 2-D echo guided perieardiocentesis. A 8-Fr Polyurethane catheter with multiple side holes, attached to a Heimlich valve was effective to evacuate the pericardial fluid. No catheter was occluded though heparin infusions were not used. Multiple changes of the patient's position may be fundamental. This 2-D echo guided pericardiocentesis performed in in~nsive care unit seems to be useful , safe and quick technique. Determining the best inotropic drug represents a very serious problems. The use of more selective and potential inotropic and vasodilatative drugs does not always lead to improvement of hemodynamic parameters in patients with low cardiac output syndrome. This paper presents patients with ACBP who need an inotropie support after extracorporeal circulation in first 24 hours. The patients were divided into Dobutamin et Dopamine groups. The heart rate (HR). mean sistemic arterial pressure [MAP), central venous pressure (CVP). and termodilution cardiac index (CI) were measured. The measurements were without using inotropic drugs, and then using them after 5 rain, 30 min, and finally with one hour rate, within first 24 hours. The statistical analysis shows that both drugs lead to an increase in HR in the first hour of the application. The final effect of dobutamine is no change in HR, whereas the effect of dopanime is very significant increase in HR. Thus. an absence of taehyeardie response selects the dobutamine as a better choice. Backeround: Pulmonary vascular eadothelium possesses major metabolic functions, which when altered contribute to the development of serious pathologies such as ARDS. One such function is the conversion of angiotensin I to angiotensin II, catalyzed by angiotensin converting enzyme (ACE), located on the luminal surface of the endothelial cells. ACE activity has been extensively studied in animals in vivo, by means of indicator-dilution techniques, providing: i) under toxic conditions, an early index of lung injury, and it) under normal conditions, estimations of dynamically perfused capillary surface area (PCSA). Objectives: To validate the use of these techniques in matt: i) for pulmonary endothelial function assessment, and it) for PCSA estimation. Methods: ACE activity was estimated in ten adult haman volunteers, with no pulmonary medical history and normal pulmonary artery pressures, undergoing cardiac catheterization for coronary artery disease assessment. Single-pass traspulmonary hydrolysis of the specific ACE substrate 3Hbenzoyl-Phe-Ala-Pro (BPAP; 30p.Ci) was measured by means of indicatordilution techniques, and expressed as %metabolism (%M) and v=-hi(1-M). BPAP was injected as a bolus i) into a main pulmonary artery, and it) inside the right atrium, to assess ACE activity in one and both lungs. We also calculated A,~,/I~, an index of PCSA. Pulmonary plasma flow (Fv) was determined by thermodilution. Fp in one lung was estimated as 0.5xF v. Results: Similar values of %M (69.6+3.8 vs 68.9• and v (1.29• vs 1.25• were observed in both and one lung respectively. A~K~ decreased from 4185• ml/min (both ltmgs) to 2122:~175 (one lung). Conclusions: i) Pulmonary endothelial ACE activity and thus pulmonary endothelial function may be assessed in humans by means of indicator-dilution techniques, it) our data denote homogeneous pulmonary capillary ACE coneentratious and capillary transit times in both haman lungs, iii) the 50% reduction of A=~/K~ in one lung suggests that this procedure can be used to quantify PCSA in man. (Supported by the Fonds de la Recherche en Saute du Quebec and the National Health System of Greece). Objective: Verify whether antioxidant activity is higher in reperfused than in no-reflow myocardium after i.v. thrombolysis for acute myocardial infarction (AMI). Methods: 37 patients with AMI were included. Blood for estimation of catalase (Cat), glutathione peroxidase (GPx) and Mn-superoxide dismutase (SOD) was drawn before initiation of i. The mechanism of myocardial cell defence against free radicals is probably identical in both reperfusion and no-reflow phenomena. Therefore, antioxidants cannot be used as reperfusion markers. Objectives_ To evaluate the precipitating factors of hypothermic phrenic nerve injury following CABG with LIMA. Methods: Fifty two consecutive patients (8 females), with a mean age of 59+8 (mean +SD) years were studied. During the ischemic arrest time topical hypothermia was obtained in al~ patients wffh ice slush and no cardiac insulation pad was used. All patients received a LIMA graft, with or whithout additional vein grafts. Supramaximai, bilateral phrenic nerve stimulation was performed percutaneously preoperatively and whithin 24 hours postoperatively. Square wave stimuli of 0.1 msec duration were applied at the posterior border of the sternomastoid muscle. The compound muscle action potential of the diaphragm was recorded, using surface electrodes on the anterior chest wall. The time interval from the application of stimulus to the onset of diaphragmatic activity, phrenic nerve conduction time (PNCT), was measured. Values exceeding 9.75 msec were considered as abnormal. Besults: Preoperatively, all patients had normal (mean+SD) PNCT, 7.69• msec for the left nerve and 7.98• mseo for the right nerve. On the first postoperative day, right PNCT was normal in atl patients (7.93• msec) , whereas left PNCT was normal in 45 patients (7.86• msec) and abnormal in 7 patients (incidence 13.5%). In 6 patients the left phrenic nerve was inexcitable and in 1 patient left PNCT was prolonged (10.50 msec). Comparing patients with normal and abnormal PNCT there was no difference in age, gender, number of grafts used, aortic cross-clamp and bypass time. However, patients with abnormal PNCT had a lower preoperative ejection fraction (45• vs 53• p=0.03). Moreover, in all of them LIMA was dissected from its origin ligating all upper arterial branches, which provide the blood supply to the left phrenic nerve, whereas in those with normal PNCT the small vessels originating from the upper 2 to 3 cm of LIMA were preserved (p=0.001). ConclusiojEL~ A hypoperfused left phrenic nerve seems to be more susceptible to hypothermic injury during CABG with a LIMA conduit. Objectives: To test if necessary interventions on systemic vascular resistance (SVR) along with preset pump flew (Q) during CPB could adversely affect autoregulatory response and cause VO 2 shifts. Methods: We studied 26 males (554-7 yrs) who underwent CPB for cardiac surgery. At o oesophageal temperature 27-28 C we set pump flow at 2.1 I.m~2.min -1. When MAP was higher than 85 mmHg we calculated VO 2 by using Fick equation. Then we infused sodium nitropruaside (SN) to control MAP at 55-65 mmHg for 10 min and we calculated VQ 2. Without changing the SN infusion rate we set Q at 2.5 I.m'2.min "1. Ten min later we measured VO 2. We took VO 2 changes into consideration if greater than 15%. Statistical analysis using students-t-test for paired data and analysis of variance was used as appropriate. Results: Depending on the biphasic VO 2 response to SN infusion during low and high Q we classified pts in four groups (table). I. VO 2 increases with SN and increases further during high Q unmasking hypoperfusion and supply dependency. II. VO 2 increases with SN but the addition of high Q results in systemic shunt. III. VO 2 increase during high Q proves that vasodilatation can turn flow insufficient. IV. VO 2 does not change with any intervention. The small number of pts and the wide standard deviation did not allow any statistical significance. Conclusions: CPB is an interesting model for the behavior of microcirculation. Intervention on SVR and Q can improve or impair effective regional oxygen delivery, resulting in either better perfusion or systemic shunt. VO 2 monitoring seems necessary during CPB. Preoperative cardiovascular optimization (OPT) to CI > 2.8 L/min/m 2, 8 _< PAOP < 18 mm Hg,and SVRI __< 30 mmHg/Ll/min/m 2 decreases cardiac events (EVENTS) and mortality (MORT) in peripheral vascular surgery patients (PVS). Objectives: to determine if OPT to the same endpeints decreases EVENTS in patients undergoing abdominal aortic aneurysm repair (AAAR) and to study the r predictive value in PVS patients. Methods: 44 AAAR patients and 41 PVS patients were admitted to the S CU monitored with e PA and arterial catheters and treated to achieve OPT. Patients underwent surgery independent of success of OPT Data included demograph cs, incremental risk factors, laboratory and hemodynamic data pre, intra, a~nd postoperatively EVENTS, and MORT. EVENTS included arrhythmias requiring treatment or prolonging the SICU stay > 24 hours, a ST depression > !mm or T wave inversion, an acute Mr defined by a new Q wave > 0.03 sec or CPK-MB > 5%. Results are presented as means _4-. SD. OPT was achieved in 32 of 41 (78%) and in 36 of 44 (82%) in the PVS and AAAR group, respectively. EVENTS did nat differ between groups 9 of 41 (21,9%) and 12 of 44 (27,2%) in the PVS and AAAR group, respectively (p>O.05). MORT was 0 of 41 (0%) and 1 of 44 (2.2%) in the pVS and AAAR group, respectively (p > 0.05), While there was no difference in endpoints of OPT between patients with and with.out EVENTS in the AAAR group, there was a significant difference in CI between patients with and without EVENTS in the PVS group. Of note, 8 of 9 (89%) patients who developed EVENTS in the PVS group had a CI < 2.8 in contrast to 4 of 12 (33%)in the AAAR group. The positive and negative predictive value were 89% and 97% in the PVS and 50% and 75% in the AAAR group. Conciusione: f. the endpoints of OPT used for PVS patients cannot be ~sed to reduce EVENTS in AAAR patients; 2. pVS patients who have net achieved OPT are at extraordinary risk of perioperative EVENTS; 3. preoperative card ovascu ar OPT in AAAR patients makes no difference in cardiac related EVENTS, Background : Comparison of the right and left filling pressures (CVP/PcwP ratio) is considered as a useful diagnostic clue : the normal ratio is _< 0.6; ratio >_ 0.8 may suggest right ventricul~ infarction while equalization of the CVP and PewP is a classic sign Of tamponade (1). However after cardiac surgery, many conditions (diastolic dysfunction, pulmonary hypertension, positive pressure ventilation) are susceptible to modify the '*normal" CVP/PcwP ratio. Material and method : We determined CVP/PewP ratio in 100 consecutive patients (pts) after uncomplicated cardiac surgery (78 coronary artery bypass grafts; 22 valvular replacements) Measurements were made before and after tracheal axtubation. Results :Cardiac Index : 3.2 _+ 0.7 1/minlm~; laotate: 144 + 68 rag/I; CVP range : 4-17 rnmHg; PewP range : 2-16 mmHg. Mean CVp/PcwP ratio before extubation is 0.94 (95 % confidence imerval : 0.86-1.02) and after extubation, 0.93 (95 % confidence interval : 0.83 -.1.03), (NS, paired t-test). In 25 % of the pts, CVP was higher than PewP. There are no correlation between the CVP/PcwP ratio and C! before (r = -0.04) and after extubation (r = -0.09) nor between the CVP/PcwP ratio and mean pulmonary arterial pressure (MPAP), before (r = 0.08) and after extubation (r = -0.13), Discussion : Cardiac performance is adequate according to CI and lactate. However the CVP/PcwP ratio is markedly higher than the "normal" (_< 0.6) ratio. This difference is not related to mechanical ventilation because the ratio is similar before and after extubation, nor to pulmonary hypetaension because of absence of any correlation with MPAP, Post-CPB diastolic dysfunction of the right ventricle could be an alternative explanation. In this group of pts, increased CVP/PewP is not associated with any impairment of cardiac performance (absence of correlation with CI), Conclusions : CVP/PcwP ratio as high as 1 within a large range of CVp (4-17 mmHg) and PcwP (2-16 mmHg) may still be considered as normal after cardiac surgery. This emphasizes the limitations of the hemodynamic monitoring after cardiac surgery (in comparison with echographic technics). Careful analysis of the morphology of the CVP and right ventricular pressure curves (x descent, y descent, dip-plateau) is mandatory rather than relying on the quantitative assessment alone. Reference : (1) ntensive Care.-University Hospital -M~laga (Spaink Introduction. Fibrinolitic Treatment (FT) permits the treatment of acute myocardial infarction (AMI) addressing the etiology, thereby eading to mproved ventncular function and a marked reduction m mortality. The main clinical oroblem is the reduced time of application. Delay in hospitalization, which can be from 50 to 120 minutes, is potentially the most avoidable delay. Method. To reduce delays in hospitalization, the following was carried out in two chases. 1 Audit: Analysis of the time lapse from onset of symptoms to start of FT. showed that during "(he period 1 June to 31 December 1994, 79 patients with chest paros were treated within a 0eriod varying from 30 minutes to 6 hours from onset of symtoms. Ages ranged from 33 to 86 (average 64,4), oelng 49 males and 30 females. They were glveD initial ECGs to determine ST mcreases suggesting AMI. Median t~me for this orocedure was l0 m.. 204 potentia AMI patients were then admitted to the Coronary Unit, 103 [)atients, under age 75 with no contraindications received FT The median time apse from admission to Corona-y Care and administration of FT was 15 minutes (12. 17), -he total median delay was 58 minutes ~40-I h.45min,~ Delays n start of this procedure are grouped as follows: Extra-hosDita delays (from onset of symtoms to arrival at hospital) Diagnostic delays (from hospital arrival to ECG). Treatment delays (from diagnosis to FT). 2 Objectives: protocol of procedure to implement a fast-track method. A protoco was drawn up with the object of reducing diagnostic delays to 8-I 0 minutes and treatment delays to less than I 5 minutes Results. Following rmplementatlon of this protocol in January 1995, 30 FTs were glven, with an over all average delay of 24 minutes. This fast-track method did not reveal any Inappropnate FT or any increase m complications, Conclusions: Detailed study of the various times taken for diagnosis ane treatment of AMI patients, showed up weaknesses in the system and improvements througn the protocol based on performence orocedures which led to a 59% reduction in the start of FT Background: The importance of the early use of thrombo!ytic agents in acute myocardial infarction (AMI) is based in the better remaining ventrictJlar function and smaller mortality rate because of the greater reperfusion and sma!ler infarction size, Therefore, it is very impodant to apply this treatment to the maximum number of patients without thrombolytic contraindicati0n, and within the minimun period of time. The "thrombolytic fast track" implementation allows to optimize the time to administrate thrombelytic agents avoiding multiple delays~ Methodology: We anal!ze the application of thromboly0c agents to 73 patients with suspect of AMI from the begin!ng of September 1994 until the end of February 1995. In this time there are two different periods, during the first 4 months thrombolytic agent were admin!strated at Intensive Care Unit (ICU), and during the second period we carried out a protocol of quick detection and thrombolysis therapy in susceptible patients at the emergency room in order to reduce the time to treatment. Ma!n results are shown in the faffewins de ay h=hours m=minutes The implementation of the fast track does not need supplementary personal or equipment but a protocelized approach and training of the personal involved The main problem detected was the usual attendance overload of the emergency department that makes difficult to follow many structurated actions. Conclusions: Pratocqlized changes in the management of AMI can significantly reduce the detay in the administration ef thrombolytic agents. It is not necessary to eomplet the procedure iq the emergency department, as the use of bolus schedules allows to begin the treatment in this area and to transfer the patient to ICU afterwards. ELECTIVE CARDIAC SURGERY. B Calvet, F Ryckwaert, P Trinh Duc, P Colson. Anesthesia -Reanimation, Hopital Arnaud de Villeneuve, Montpellier, France. Obhectives: The study was aimed at analysing the incidence of renal dysfunction following cardiac surgery and its prognosis (acute renal failure, post-operative morbidity and mortality). Methods: Two hundred and thirty seven patients (aged from 28 to 90) were consecutively operated on for elective cardiac surgery and retrospectively included in the study. Patients with preoperative infections and operated on in emergency were excluded. Each patient had preoperative invasive cardiac investigation with angiography and calculated ejection fraction (EF). Anaesthesia, cardiopulmonary bypass (CPB) and cardiac arrest management were similar in all patients. General body temperature was reduced to 28 -30 ~ C. Renal dysfunction was defined as a 20 % increase from baseline of serum creatinine. Demographic data, ASA, treatments, pre-operative creaUnine level, CPB and clamping (AXC) times, intra and postoperative use of inotrope, serum lactate level before surgery, at the end of CPB, at the time of admission in intensive care unit (ICU) and on post operative day one and APACHE score were compared in patients with or without renal dysfunction using ANOVA test for repeated mesures and X2 when appropriate. Data are expressed as mean +__SD. P value less than 0.5 was considered statistically significant. Results: Thirtytwo patients (13,5 %) suffered from renal dysfunction. Age, serum lactate level at the end of CPB, at admission in ICU, at POD1 and APACHE level at admission in ICU, intra-operative use of inotropes were statistically different in patients with or without renal dysfunction (p<0,05). Mortality rate was statistically different in patients with or without renal dysfunction(~,2,5 % and 0 %, respectively, p=0,001). Incidence of acute renal failure following renal dysfunction was 6,2 % (2 patients required hemodialysis). Conclusions: Although our cdteria for defining renal dysfunction were very sensitive, the incidence of renal dysfunction following elective cardiac surgery was lower than communly accepted in the litterature (1). However renal dysfunction appeared significantly associated with a poor prognosis. Reference: 1-Settergren G, Ohqvist G Current opinion in Anaesthesiology 1994, 7:59-64 R 66; 159, 200 Tzelepis, G. 112, 127, 142 Late complications were observed in 14% of cannulations: Local infection in 2 (i,7%), catheter displacement by the patient in 4 cases (3,5%), catheter displacement during nursing care in 5 (4,4%) and malfunction in 5 cases (4,4%). Conclusions: Central venous catheterizations are followed by immediate and late complications in almost the same percentage ACUTE POISONING WITH AMPHETAMINES (MDEA) AND HEROIN: ANTAGONISTIC EFFECTS BETWEEN THE TWO DRUGS Methods: After institutional approval and informed consent, 33 selected patients (65_+9 years) undergoing peripheral vascular surgery (n=17) or carotid endarterectomy (n=16) were investigated. Patients included had either documented CAD (n=18) or two or more (n=15) dsk factors (age >65 years, smoking, diabetes meltitus, hypertension, hypercholesterolaemia >240 mg/dl). 12-lead ECG recordings were carded out preoperatively, on ardval in the postanaesthetic care unit, and 20 h, 48 h, 72 h, and 84 h postoperatively. ECG recordings were analysed by an independent blinded cardiologist for signs of PMI (new ST segment depression >0.2 mV and/or new T inversion). In addition Results: 22 of the patients investigated developed ECG-documented PMI, 86% occurdng in the immediate postoperative phase. Troponin I levels >1.6 ng/ml were found in 19 of these 22 patients Thus, comparing a cardiac troponin I cut-off level of 1 ng/ml with intermittent 12-lead ECG recordings, we found a sensitivity of 86% and a specificity of 91% Methods: Demographic, clinical and ECG data were analyzed. 53.8% of patients were male; 46.2% female. CAD was the most common underlying cardiac disease (85.7%) and 58.4% underwent open heart surgery. 69% received Proeainamide for supraventricular and 31% for ven~cular arrhythmias. 27% received a loading dose. Maintenance was provided by IV route in 36.8% and by PO in 63.2% (40.8%SR end 22.4% IR). 40.4% of patients were obese Right ventricular function following cardiopulmonary bypass: Is important the mode of myocardial protection We underwent this study in order to examine its safety and usefulness in pts with trustable coronary conditions (unstable angina UA The mean age for Group A was 54 • 16 years, for Group B 64 • 11 years, and for Group C 59 • 13 years. A history of previous myocardial infarction was present in 6 pts of Group A, in 3 of Group B and in 54 of Group C. Three pts in Group A, 5 in Group B and 15 in Group C had previous coronary artery bypass grafting. The median time between the onset of symptoms and A was 5 days (2 -19) for Group A We used a continuous fixed intravenous A infusion at a dose of 140 The SN was 100% in Groups A and B, 98% in C, and SP 100% for Group A, (fixed defects included) and 50% for Groups B and C. There was no difference of side effects among groups: Chest pain (I pt -Group A, 2 pts -Group B, and 8 pts -Group C), transient hypotension (1 pt -Group C), headache (5 pts, Group C), dyspnea (1 pt -Group A), while ST depression was seen in 2 pts of Group B and in 2 pts in Group C. The rate of A infusion was decreased to 70 7/kgr/min in one Group B pt due to development of chest pain S228 FIVE YEAR FOLLOW UP OF HUMORAL IMMUNITY IN PACED PATIENTS Athens Polyclinic Hospital, Department of Cardiology Athens, Greece Author index A Abiad Ch 133 Bertschat, E 141 Betbes6 75 Blanch, L1 177 del Nogal Saez 93 E1-Meneza 220 Nolla, J. 98 Nolla-Salas 24 Pilz~ U Puig de la Bellacasa E 38 Scarpa, N. 77 217 van de Wetering Objectives: Only 50% of patients suffering from acute Guillain-Barr@ syndrome (GBS) respond promptly to established therapies like plasma exchange or intravenous immunoglobulines. In contrast to serum, cerebrospinal fluid (CSF) of GBS and CtDP patients contains enriched portions of antiexcitatory factors(I) and cytokines (2) able to induce pronounced conduction block (3). To reduce or remove such pathologic factors we introduced a technique with direct access to the subarachnoid space. Methods: With informed consent we lumbally inserted 18 G catheters in 24 GBS-and 22 CIDP -patients under sterile conditions. Some of them had not responded very well to established therapies. 30-40 ml of CSF were withdrawn and retransfused by a bidirectional pump (Flofors) after passing newly developed filters (Pall). Daily filtrations with several cycles were performed (200 -300 ml) over one week. Results: The 24 GBS patients improved after 19 days (median) for one grade (according to the GBS-Scale from the GBS Study Group) . The ventilator dependent patients were weaned after 16 days (median). Patients not at all treated before (16/24) responded better than patients that had been pretreated (8/24) with plasmaexchange or intravenous immunoglobulines. 18/22 CIDP patients drew benefit from treatment, 10 stabilized Iongterm. Conclusions: CSF-Filtration is a relatively save and well tolerated additional procedure. The costs are considerably lower (1/3) than those for plasmaexchange or intravenous immunoglobulines. References:(1)WSrz Aet al: CSF and serum from patients with inflammatory polyradiculopathy have opposite effects on sodium channels. Muscle Nerve (1995) . (2) Clinical observations were made in 24 patients admitted to the clinic. They were in coma associated with acute alcohol intoxication.Standard evaluations (ECG-monitoring, electrocardiography, neuromonitoring, studies of acid-alkali condition, biochemical and toxicologic investigation of blood and urine) prior to and following the treatment conducted were undertaken in all the patients.To correct irreversible impairement of functions twofold laser blood irradiation by means of ALOK-01 apparatus, the exposure within 20 minutes, was carried out.The data obtained confirm more rapid coma withdrawal of the patients, reconstruction of the heart and central nervous system electrophysiologic indeces, reliable reduction in complications compared with the control group. OBJECTIVE: To know the actual incidence of the Critical Illness Polyneuropathy(CIP). SETTING: Fourteen intensive/critical care unit beds, in 550 bed University Hospital, covering 345.000 inhabitants (majority rural area). The ICU patients are medical, surgical and coronary, excluded the neurotrauma and neurosurgical. DESIGN: A conseculive and prospective study. All the patients admitted during three months, from January lth to March 31th 1993, were eligible (patients with admittance diagnosis of polyneuropathy were excluded ). METHODS: Patients with APACHE II score >10, at the admission and six days after admissions were included into the study protocol. Diagnosis of sepsis, MOF, and all the drugs administered days before were recorded. A complete neurological exam, by a neurologist, in absence of ssdatives and muscles reliant (7th, 25~ and 60th days after ICU admittance) was made. We evaluated the nerve and muscles function with and electromyography study in all patients, at same days. In some paeents with CIP we performed a nerve biopsy. RESULTS: From 285 patients ( APACHE II score: 12.82) admitted in the ICU, 16 (5.6%) enter the study protocol. Seven (2,45%) had an axonal polyneuropathy(CIP), three very severe. Only four of the patients with CIP had pathologic clinical exam. APACHE II score: CIP vs non-CIP was 22.6 vs 16.6. The incidence of CIP by diagnosis (CIP/diagnosis) was: Sepsis, 5/9 and MOF, 6/11. CONCLUSIONS: 1. -We think that it is necessary to define the "critically ill" for some score, before designing a study to know the incidence of this syndrome. 2. -We think that the incidence of the CIP is lower that the latest papers say. Objectives:Acute pancreatitis(AP)is becoming a more important problem among the elderly as the population ages. The increasing presence of gallstone disease,as well as the use of certain drugs,may also contribute to the occurrence of pancreatitis. Methods:All patients(>60 years)admitted to our medical department over an eight year period were included.Pancreatitis was confirmed by biochemical tests and imaging techniques.Scores were developed using Ranson's criteria and a Multiple Organ System Failure(MOSF)index . Overall, 103 patients were evaluated; 21(23%)had pancreatitis of unknown etiology . Results:(1)Patients with pancreatitis of ~nlqnown etiology were sicker and had greater morbidity(48% vs 22%),mortality(24% vs 8%),and longer hospital stays than p~tierf~ with pancreatitis of known cause.(2)The best predicto~of severity and outcome was the MOSF index and not Ranson's criteria;the higher the score,the greater the associated disease,the worse the outcome.(3)Curlously,no difference existed in associated medical conditions between patierts withknown and ur6~own causes of pancreatitis. Conclusions:Greater organ dysfunction exists in patients with pancreatitis of unknown etiology, even though age and associated medical conditions do not differ . The application of the total enteral nutrition in the burns disease has minimized the complication rate and consequently increased the survival rate of children and adults. Time of initiation, composition, duration and way of administration are very important in obtaining the optimum beneficial effect from the treatment and diminishing the complication rate and side effects. The above features will be discussed in view of our experience in 240 cases. TA Buckle?,, RA Freebalm, C GomersalL G Joynt, R Young. TG Short. Department of Anaesthesia and Intensive Cm+e, Prince of Wales Hospital. The Chinese University of Hong Kong, Shatin, HONG KONG INTRODUCTION: Gastric mucosal pH (pHi) monitoring has been proposed as a relatively noninvasive index of the adequacy of aerobic metabolism in the gut. To examine the accuracy of gastric intramucosal pit measurements as a function of time and as a function of the catheter itself to determine whether the measurement error between catheters is clinically acceptable. Patients with a gastric tonometer (TRIP TM, Tonometrics, Worcester. MA) insitu for >3 days were studied. Following informed consent two new tonometers were inserted equidistantly & correct position was confirmed radiographically. Measurements of intramucosal gastric pH were then performed over a 36 hr period. Eight -ten measurements were made in each of ten critically ill patients.Percent differences between the two new catheters were 1.6% ie at pH 7.3 _+0.12 (95% limits) and between old & new catheters were 2.2%, ie pH 7J3 _+0.16 (95% limits). CONCLUSIONS: The results suggest that the function of the tonometer deteriorates over time and that the absolute values of pHi m~ not ~ufficiently accurate. However as a trend monitor pHi may be useful in the clinical setting. Despite a continuous decline both in li'equency and severity of gastro-intestinal stress-lesion/-bleeding (GISB) due to both improvement in preclinical support and in intensive care medicine, patients with cerebral lesion are still considered at high risk for developing GIS8. Therefore the question arises, whether m> specific (}lSB-prophylaxis besides general and neurological intensive care, specific pharlnaeothcrapy or even the combination of two specific drugs reveals any protective efli~ct on frequency and severity of GISB.This pntspcclive randomized study has been perfornted in 173 patients snfrering t'rttna head-injury/cerebral lesion and with a Glasgow-Coma-Scale on admission (GCS:,)of <9. According to randomization the patients have been grouped as tbllows: h analgesia/sedation (n=37); Ih analgesiaJsedation plus pirenzepine 60 mg/day (n=54); .[Ih anatgcsia/sedalkm plus sncraltate 6 x [ g/day (n=47); IV: analgesidsedatkm plus pirenzcpine 60 mghlay plus sucralfate 6 x 1 e/day (n=35). Slalislical analysis has been performed by CHl:*tt~sl. rank correlatinn and unpaired t-test; statistical significance has been set with p <0.05. 28/173 patients (16.2 %) developed GISB. Although the mean GCS~-value (x -+ SD) did not reach significance between patients with and without GISB (5.61 + 1.65 vs 6.12 -+ 1.65). a significant inverse correlation between GCS:, and the incidence of GtSB (Rs~ = 0.89) has been shown. The frequency of GISB among the groups is as follows: h 18.9 %; lh 18.5 %; llh 17.0 %; IV: 8.6 % (CH1 -~ =1.94; not signilicant). No GISB-induced blood translusion or mortality, respectively, could be demonstrated. Survival rate between the groups did not differ significantly (Chi-" =5.86; P=0.1186) and reached an overall-value of 75.1%.Drug-specific GlSB-prophylaxis -administered either as monotherapy (pirenzepine, sueralfate) or in combination of these two specific-drugs -reveals no additional significant influence on the incidence of GISB in patients with cerebral lesion compared to no specific prophylaxis besides the general trauma-/disease-specific intensive care measures. Critical Care Dpt, Evangelismos Hospital, Athens University Scho~" of Medicine Objectives: The correlation of longterm presence of nasogastric tube (NGT) to gastroesophageal reflux (GER) is still in question. In case of positive correlation, PEG should represent an alternative to tube feeding in patients unable to be fed orally. Therefore, we investigated: i) the correlation between NG and GER and ii) the effect of PEG on GER. Methods: A 24-h esophageal pH-metry was performed in 40 patients in recumbent position at 30 ~ who had a NGT for more than 10 days and were on sucralfate for gastric mucosal protection. The tip of the pH-probe was lied 5 cm over the esophagogasttie junction, confirmed by x-rays. Patients who presented a percentage of GER-total (i.e. with a pH less or more than 4) (GER-T) more than 3%, underwent ~t PEG. The presence of a creseent-notch on the esophagogastric junction persisting on inspiration and the grade os endoseopic and histologic esophagitis (scale=0-3) was noted. Two pH-metrles repeated on 48 h and on 7 days post-PEG were compared to the pre-PEG one, with the followin~ parameters taken in consideration: i) % GER-T, ii) number of GER-total per hour (No/h GER-T) and iii) the duration that pH was less than 4 (TpH<4). In case ot GER persistence at the pH-metry on ?th day post-PEG (Group II) another endoscopy was performed, while patients with reduced GER (Group I) were considered as esophagifis-free.Results: 23 out of 40 patients presented a GER-T>3%. Eleven out of 23 Group I Group (n=6) 1I ( Objectives: The aim of the present study was to compare the performance of a specially modified version of a photo-and magnetoacoustic (PA/MA) gas analyzer (Br~)el & Kjaer, Denmark) with a conventional quadrupole mass spectrometer (MS) (Innovision, Denmark) in inert gas rebreathing (RB) tests such as determination of functional residual capacity (FRC), pulmonary capillary blood flow (PCBF) and lung tissue volume (Vtc). Methods : From simultaneous readings of inert gas concentrations with the MS and the PA/MA analyzer during RB experiments a comparison was made of the PCBF, Vtc and FRC values. The RB tests were performed during rest and exercise (0,50 and 100 W) in ten healthy subjects. Results: The differences (mean +/-SD) between simultaneous estimates of rebreathing parameters were the following (PA/MA -MS) for pooled data, PCBF: 0.18 +/-0.38 I/min, Vtc: -33 +/-108 ml and FRC: 0.028 +/-0.048 liters. Conclusions: Smell but significant differences were found between the estimates of PCBF, Vtc and FRC using the MS and PA/MA, respectively. Reference: P. Clemensen, P. Christensen, P. Norsk, and J. Gr~nlund. A modified photo-and magnetoacoustic multigas analyzer aplied in gas exchange measurements. J Appl Physiol 1994; 76: 2832-2839. Objectives: Because transcranial Doppler (TCD) has been proposed to explore cerebral CO 2 vasoreactivity in brain injury (Stroke 1992;23:962-6), we compared this technique with the Kety-Schmidt reference method to assess cerebral vasoreactivity in comatose patients. Methods: 17 mechanically ventilated patients (age 30-81 yrs, Glasgow 3-10) in coma due to acute brain injury were investigated during stepwise changes in PaCO 2 (25, 30, 35, and 40 mmHg) by increasing inspired PCO 2. Middle cerebral artery velocity (Vm) was measured by TCD. After insertion of a catheter in the ipsilateral jugular bulb, cerebral blood flow (CBF) was determined by the Kety-Schmidt method, using the inhalation of 10% N20 through the inspiratory line of the ventilator. For each patient a cerebral CO~ vasoreactivity index was calculated as the slope of linear relationship between Vm or CBF and PaCO2. Objectives: After cardiac surgery the fluid shill, between interstitial and intravasal space may be marked. This is due either to the intraoperative volume loading by the extracorporeal circulation or the increased postoperative diuresis. Therefore, infusion of a large amount &fluids is necessary during the first postoperative hours. It still remains unclear which of the substances at disposal is the best for this purpose. Aim of the present study was to compare the different fluids with special regard to postoperative bleeding and rheological behaviour. Methods: 93 patients undergoing CABG-surgery were investigated and randomizedly distributed to three different groups of postoperative volume replacement to stabilize the mean arterial pressure at 80 mm Hg. 1. ringer's solution, 2. 3.5% gelatine solution, 3. 6% hydroxyaethylstarch (mean m.w. 70.000). We evaluated the following parameters within intervals of 30 min: arterial and central venous pressure, heart rate, postoperative bleeding, urinary output, volume replacement. Results: There was no statistically significant difference between the groups with regard to urinary output and bleeding. In spite of larger amounts of fluids necessary in the ringer treated group patients of this group showed symptoms of hypovolemia. Hematocrit was increased in the ringer patients. This was statistically significant. Introduction: Pulmonary wedge pressure (PCWP) and central venous pressure (CVP) are frequently used as parameters for cardiac preload, although it is known that both are poorly correlated to the cardiac index (CI). It has been claimed that intrathoracic blood volume (ITBV) measured with the thermal dye dilution method reflects cardiac preload better than PCWP and CVP. We studied the correlation between ITBV and CI in a mixed population of critically ill patients. Methods: In 17 consecutive patients (6 sepsis/SIRS, 2 acute heart failure, 3 ARDS, 6 transjugular intrahepatic portosystemic shunt) monitored with a pulmonary artery catheter, ITBV was measured on regular intervals using the Pulsion COLD Z-021 system (Pulsion, Munich, Germany). CI, PCWP, and CVP were recorded simultaneously. Results: A total of 1Ol measurements was made. PCWP and CVP did not correlate to CI, nor did APCWP or ACVP correlate to ACI. ITBV was correlated to CI in a non-linear fashion (F -139, DF = 99, p < 0.001, (Figure) ). AITBV was correlated to AC1 in a linear fashion (r = 0.76, F = 134, DF = 99, p < 0.O01). A rapid and efficient circulatory support system may save a patient in cardiogenic shock. Left heart bypass with percutaneous and transseptal placement of the aspiration canuia simplifies the circuit and avoids the need for an oxygenator. We assessed this preclinical set-up in 5 anaesthetized pigs using a centrifugal pump with a 14 F arterial catheter and a 16 F left atrial aspiration line. Animals were supported for two hours at a mean flow of 3.1 liter (3'680 RPM), a mean hematocrit of 29% and low heparinisetion (ACT double baseline). Hemodynamic and laboratory samples were taken at baseline (A), 10 minutes (B), one hour ( Pulmonary hypertension (PH) usually involves obliteration and loss of functional pulmonary microvasculature. The microvaseular endothelium normally acts as a major metabolic organ, converting angiotensin I to angiotensin II via the angiotensin-converting ectoenzyme (ACE). It is unknown whether the loss of functional vasculature and altered pulmonary blood flow seen in PH will affect lung ACE metabolic activity. We therefore estimated pulmonary vascular ACE activity in 9 patients with PH of various causes: 2 primary; 1 post atrial septal defect closure (ASD); 2 chronic thromboembolic (TE); 1 anorexigen; 1 IV drugs; 2 collagen disease. Single-pass transpulmonary hydrolysis of the specific ACE substrate 3H-benzoyl-Pbe-Ala-Pro (BPAP) was measured and expressed as % metabolism (%Me0. We also calculated an index of peffused functional capillary surface area (Amax/Km). All patients with PH had an abnormality of %Met or Amax/Km, or both. As compared to 9 control humans (mean %Met = 71.4% _+ 11.3% S.D.), the mean %Met in PH patients was 54.3% _+ 14%. The %Met in PH patients correlated inversely with cardiac output (r=0.74), possibly reflecting more complete BPAP hydrolysis with longer pulmonary transit times. Amax/Km was markedly decreased in PH (1663 + 536 ml/min) as compared to controls (4225 _+ 1018 ml]min), consistent with a significant loss of functional capillary surface area. Patients with collagen disease, ASD and anorexigen-induced PH had the most marked abnormalities. In conclusion, patients with pulmonary hypertension have decreased pulmonary endothelial angiotensin converting enzyme activity, likely due to a loss of functional or perfused pulmonary microvaseulature. Supported by the Funds de la Recherche en Same du Quebec and the National Health System of Greece. Objective: To investigate adrenocortical function in patients with ruptured aneurysm of the abdominal aorta (RAAA). Studies investigating adrenocortical insufficiency in critically ill patients report an incidence ranging from 20% to less than 1%. This may in part be explained by difference in methods used (single cortisol measurement vs short ACTH stimulation test) and populations studied (heterogenous groups of patients with great individual variation in underlying disease as well as duration and severity of illness). Methods: We investigated the adrenocortical function in 32 patients with (RAAA).A short ACTH stimulation test (Synacthen test; 250 ug 1-24 ACTH iv) was performed at 0800 hrs within 24 hrs of admission. Plasma cortisol was measured before (cort basal) and after stimulation (cort stim). A plasma cortisol level > 0.55 umol\L Before or after stimulation was considered normal, Severity of illness was assessed using APACHE II. Results: Of the 32 patients investigated 6 died and 26 survived. Mean cort basal in nonsurvivors was significantly (P<0.O04) higher than in survivors; 1.03 (range 0.72-1.29) vs 0.69 (range 0.24-1,14). This difference between nonsurvivors and survivors was also present for cort stim but lacked significance; 1.30 (range 0.96-2.25) vs 1.00 (range 0.57-1.53). While 8 patients showed a cort basal < 0.55, no cort stim <0.55 was found. There was no significant difference in mean age or APACHE II score between survivors and nonsurvivors; 70 vs75 and 19 vs 21. Conclusions: Single plasma cortisol levels were inadequate to assess the adrenocortical function in the patients studied, Judged by a short ACTH stimulation test, our investigation in patients with RAAA showed no adrenocortical insufficiency. Mortality in RAAA is associated with elevated plasma cortisol levels. Obiectives: Mortality in acute myocardial infarction (AMI) prinicipally depends on hemedynamic impairment. Thus, patients (pts) with elevated pulmonary wedge pressure (PWP) present high in-hospital mortality. However, the complete right heart catheterization is laborious, so the central venous pressure (CVP) alone is frequently used to assess the severity of AMI. The accuracy of CVP in estimating pts with AMI was tested in this retrospective study. Methods: 131 pts. aged 68+14 years, admitted to our CCU from 1992 to 1994 with their first AMI, were inctuded in this study. All had undergone right heart catheterization because of overt or suspected heart failure. Swan-Ganz catheters (7F, 85cm, Abbott, IL, USA) had been used, Every treatment had been temporarily interrupted l h before the calheferization. Based on ECG findings the pts were retrospectively divided into 3 groups. In group A we included 54 pts with anterior AMI, in group B, 30 pts with inferior AMI, and in group C, 47 pts with inferior and right ventricular AMI. The initial values of CVP and PWP were considered for the linear regression of the PWP variable on CVP and p<0.05 was accepted as statistically significant.Results: in g~oup A, the CVP and PWP vaiues were 8+3 mmHg and 14_+7 mmHg respectively. Despite the signifanf correlation (p<0.01) between the two variables, it was not possible fo predict the exact value of PWP based on CVP value, 19 pts (35%) presented CVP>8 mrnHg and 16 of these (84%) had PWP_>15mmHg. In group 3, the CVP was 11_+8 mmHg and the PWP, 13_+6 mmHg. Significant correlation (P<0.05) between the two variables also existed, however it was impossible to predict the PWP value. 6 pts (20%) had CVP>8 mmHg but only 2 of these (33%) had PWP>15 mmHg, Similar was the relation between CVP and PWP in group C (P<0.01). CVP averaged 12+6 mmHg, and PWP, 17_+8 mmHg. 33 pts (70%) had CVP>8 mmHg and 25 from these (76%) presented PWP>15 mmHg,Conclusions: A single measurement of CVP in AMI does not ensure an accurate assessment of PWP. Because every pt with AMI needs optimal values of PWP in order to prevent pulmonary congestion or manifestations of low preload, the significance of complete right heart catheterization becomes apparent. In patients (pts) with advanced HF the need and the prognosis for heart transplantation (HT) can be predicted from Vo= max. Indirect measure of functional capacity with the six-minute walk test can also predict smvival in moderate HF. To predict Vos max from indirect astinmtions of functional capadty such as 6-1~q~/, pulmonary and heart function tests, and to assess the prediddve value of the above parameters in HF pts survival. We evaluated 35 pts (age 48+12 yeats NYHA class: 12 II, 15 HI, 8 IV) with HF for PiT. They underwent a pmgmmive exercise test on cycle ergometer for Vo2 max determination, a 6-MW, a right heart catheterization and a spirometry and Dlco estimation. Introduction: Brain death causes myocardial impairment by mechanisms that are not well understood yet. The aim of this work was to assess the echocardiographic features found in these patients from the clinical onset of brain death to somatic death, Methods: Seven brain dead patients were studied (patients" relatives refused to allow them to be used as donors). Mean age was 23.5 (18-32) years old. Four of the patients were female, None of the patients had any history of cardiac disease. Transthoracic echocardiogram (echo) and electrocardiogram (ECG) were obtained at the onset of clinical brain death and were repeated every 24 hours until somatic death. We We detected severe diffuse hypokinesia (EF<50%) in 2 patients and mild hypokinesia in 3 others (EF 50-60%). Systolic function was strictly normal in only 2 patients. Corrected QT interval (QTc) in ECG was 54.6_+5.5 msec (normal range 38-43 msec) just before somatic death (B). Conclusion: In patients with brain death we observed a significant increase of left ventricular mass due mainly to IVS "hypertrophy" without any important change in the dimensions of the left ventricle. To our knowledge, this finding has never been reported before and its importantance in heart transplantations may be of particular interest. PREDICT RIGHT VENTRICULAR OUTCOME. L. Jacquet, R. Dion, P. Noirhomme. M. Van Dijck. M. Goenen Cardiothoracic Intensive Care Unit, St-Luc Univ. Hospital(UCL) We have registred: heart rate (HR), blood pressure (Bp), pulmonary artery pressures (PAp), central venous pressure (CVP), pulmonary capillary wedge pressure (PCWP), pulmonary and systemic vascular resistances (PVR, SVR), right ventricle end-diastolic end end-systolic volume (REDV, RESV), right ejection fraction (REF), right sistolyc ventricular work (RSVW) and cardiac output (CO) using a thermodilution thechnique and a microprocessor (model REF-1; Baxter-Edwards Laboratory); duration of CPB and aortic clamping, and the requirements of haemodynamic support after CPB.Results: In the C group an increase post-CPB of the Fc (62 + 13 95.8 + 18.4, p < 0.01) was produced without significantly changes in the REDV, RESV, REF, RSVW neither CO. In the W group, HR increased from 58.5 + 8.8 to 79.9 + 8.6 (p < 0.01); REDV was reduced from 227.7 -+ 76 to 139.14 _+ 36.4 (p < 0.05); RESV was reduced from 142 • 68.7 to 82 + 35.3 (p < 0.05). There were not changes in the other haemodynamyc parameters. There was a trend (no significantly) to an increase of REF in the W group (40.07 + 9.7|• 4.7) compared with the C"group (39 • 10.9($ 38.3 • 8.5) post-CPB. The need for haemodynamic support was similar in both groups.Conclusions: The warm, continuous, anterograde-retrogade myocardial protection has obtained a decrease of preload, HR, and a trend to an increase in the REF, making an improvement in the right ventricular global performance when is compared with the classic form of cold myocardial protection. Objective: To evaluate the effect of dobutamine on gastric mucosal pH (pHi) after coronaly artery bypass surgery. Design: Prospective study in a university hospital intensive care unit (ICU). Subjects: 20 elective cardiac surgery patients. Interventions: Dobutamine was infused at 4 ug/kg/min for 3 hours immediately after admission to the ICU. Hemodynamics were measured every 30 minute periods until 3 hours and again 2 hours after stopping dobutamine. Results: There were no significant differences in mean gastric pHi between the groups but mean phi decreased in both groups during the study period. Oxygen delivery and consumption both increased during dobutamine infusion but decreased to the control group level after stopping the dobutamine infusion. Lactate levels did not change. Baseline 90 Objectives: The aim of the study was to evaluate the usefulness of a low dobutamine dose in conjunction with intraaortic balloon pumping and mechanical ventilation in cardiogenic shock. We studied 21 patients 58.9-+ t4.4 years of age suffered of post infarction cardiogenic shock characterized by a systolic arterial pressure< 80mmHg, urine output< 20 ml/h and mental confusion or purpueral signs of low output, non responded to dobutamine infusion up to 9 pg/kg/min. All patients underwent mechanical assistance by the Intra-aortic Balloon Pump (IABP). Five patients were additionally placed on mechanical ventilation due to blood gases disturbances. The end points in our study were: reversion of cardiogenic shock, improvement of patients survival or both on the 15th post infarction day and 6 months later. Results: Three patients refused IABP treatment and 0/3 survived on the 15th day. On the 15th day 3/13 supported by the IABP and 5/5 that underwent mechanical ventilation plus IABP were alive (p <0.01). On the 6th month 2/13 supported by the IABP and 5/5 that underwent mechanical ventilation plus IABP were alive (p<0.01). Conclusions: In conclusion, the combined use of mechanical ventilation and IABP assistance in severe cardiogenic shock might improve survival. Obiectives: The study was aimed at analysing predictive factors of Swan Ganz pulmonary catheter (PC) requiremen t during elective cardiac surgery according to the need of sustained inotropic support after surgery. Methods: Three hundred patients (aged from 27 to 85; 89 females and 211 males)were consecutively operated on for elective coronary artery bypass surgery (CABG, n=179), valvular replacement (VR, n=98), combination of both (VR-CABG, n=15), or others (n=6) and retrospectively included in the study. Each patient had preoperative invasive cardiac investigation with calculated ejection fraction (EE). Anaesthesia, cardiopulmonary bypass (CPB) and cardiac arrest managements were similar in all patients. PC requirement was estimated from the need of either dobutamine, adrenaline, dopamine or enoximone use during the first 48 hours after cardiac surgery. Demographic data, ASA and NYHA classifications, preoperative EF and treatments, type of surgery, CPB and aortic cross Clamping (AXC) times, and postoperative incidence of complications Were compared in patients with or without inotropic support using either Student's t test or X2 with continuity correction when appropriate. Results: Seventy4hree patients (24.5%) required inotropic support after surgery. AXC .and CPB times, mean stay in ICU were significantly longer in patients with inotropie support (p<0.001). Type of surgery, preoperative EF, and NYHA classification are the first 3 significant factors related to inotropic support (p<0.005). Most patients operated on for double-VR or VR=CABG required inotropic support (57 and 60%, respectively). Postoperative mortality was higher in patients receiving inotropic Support (8,2% vs 0,9% 'overall mortality, p=0.003). Conclusions: Since PC insertion is most.often justified because inotropes are required, these results suggest that elective rather than routine systemic PC insertion could be helped by considering several but selected preoperative factors. Background: Cardiovascular depression due to anaesthesia, old age and major gastrointestinal surgery is becoming an increasingly frequent challenge .to the anaesthesia-surgOry team. Deliberate preoperative manipulation of haemodynamics and Oxygen transport parametres towards prede~t~mined optimal values may prove to be effective "in reducing 12 morbidity ~nd mortality in high risk surgical patients,. A new concept Of using conlimaous perioperative measurement of cardiac'output to obtain and maintain supranormal oxygen delivery (DO2I) is presented. Methods: Continuous measurement of cardiac output is a relatively new form of on-line monitoring, in which trains of impulses are emitted from a thermal filament mounted on a pulmonary artery catheter. Computer software recognizes patterns generated by minute changes in blood temperature and ealOalates cardiac output every 30-60 seconds. Cardiac 3 output and mixed venous blood Oxygen saturation are displayed graphically on line. In tins TM study cardiac Output Was measured continuously by Vigilance cardiac outpu t compl/ter (Baxter). Preoperative haemodynamiC optimization was performed with the goal of increa-2 1 sing DO2I to at least 600 ml/min/m accordfing to Shoemaker's algorithm . This was.done by infusing colloids (albumin or hydroxy ethyl starch (HAES-steril| until the desired DO21 was reached. Infusion was stopped if cardiac output ceased to increase with infusion, if there were signs of pulmonary oedema or if wedge pressure reached 18 mmHg. Vasoactive or inotropic drugs were infused if the desired DO21 was not reached by infusion alone. Anaesthetic technique included continuous thoracic epidural and isoflourane anaesthesia. Expected mol:bidity and mortality rates were calculated by the "POSSUM" score aasing preoperative clinical and paradinical estimates of organ function as well as surgery characteristics 4. Materials: 15 ASA group Ill-IV patients with a mean age of 75 years (range 60-92) and a mean Weight of 67 kg (range 36-93)) scheduled for major abdominal surgery were included. Results: 2 patients were excluded because DO2I could not be raised at all. Mean DO2I was increased from 488 ml/min/m 2 (range 384-610) to 688 ml/min/m 2 (range 490-967). Mean volume of preoperativdy infused colloid was 954 ml (range 0-2000). During surgery 1230 ml (range 5002300) of colloid was infused. Mean length of surgery was 150 minutes (range 60-300). Mean blood loss was 920 ml (range 04800). Expected mortality and morbidity rates ("POSSUM") were 56% and 92%, respectively, whereas patient follow up upon discharge or at death revealed mortality and morbidity rates of 8 % and 54%, respectively. Conclusion: Based on experience from the present study, continuous measurement of cardiac output has proved to be a valuable tool for perioperative optimization of DO21 in ASA group IlI and IV patients during major surgery. However further studies including a greater number of patients are necessary to confirm the promising preliminary findings. We studied the hemodyn~c effects of three different combinations of pOsitiv inotropic .agents, vasodilators, diuretics and AV-filtration (AV) in 16 patients (pts) With severe left heart faille (left veutrieul0x filling pressure (LVFP) >25 mmHg) due to acute myocardial infarction. Hemodynamic measurements (intravascular pressures (LVFP), thermodilution (Cardiac Index (CI)) were made before (control) and after each therapy. In 11 Furosemide (F) + D0butamin (D) + Nitroglycerin (Ni) reduced LVFP and a small increase of CI occurred. In 6 of these pts :(Group A) Nitroprusside (hip) instead of Ni increased CI significantly, in the other 5 pts adding of Amrinone (A) resulted in a pronounced increase of CI. Group C (n=5): The combination of Ni and AV reduced LVFP but did not increase CI which was achieved by AV+D+Ni. In order to optimize the treatment of acute heart failure a combination of inotropic agents, vasodilators, diuretics and AV-filtration should he used guided by hemodynamic monitoring. Arias Jr, Miragaya D, SandarD, San Pedro dM ~, Herndndez d, Valenzuela 12 . OBJECTIVES: To evaluate the variation in nomdrenaline (NA) plasma concentrations in patients with acute myocardial infarction (AM1) after thrombolytic therapy with noniltvasive reperfusion criteria (clinical, electrocardiographic and enzymatic), in relation to infarct size and location.METHODS: 42 consecutive patiens with AMI, from October 1, 1994 to February 28, 1995, admitted within 6 hours alter onset of symptoms, undergone successfull systemic thrombolysis. 21 of them were anterior (Group A) and 21 inferior (Group B) . Noradrenaline plasma levels at 0 (NA1), 60 (NA2) and 240 (NA3) minutes after admission were compared with CK-peak plasma levels by linear regression. Differences were tested for significance by Student-t-test for paired and unpaired values. NA plasma concentration was measured by high-presssure liquid chromatography. p< NS 0.01 NS means 4-SEM (Normal limit for our laboratory: NA < 37/0 pg/ml; CK < 170 U/I ) CONCLUSIONS: 1. The NA plasma levels at admission (NAI) are more increased in anterior than inferior AMIs, probably in relation to infarct size. 2. The decrease in NA is more evidence in AMIs with anterior location. 3. This decrease is probably due to the major efficacy of thrombolytic therapy in AMIs with anterior location. Arias Jd, Miragaya (Group B) , probably due to certain degree of t~cg'rfueion. 3. There is not significant variation in NA in conventional treated AMI (Group C). V.Suchanov, A.Levit, P.Trofimov, ICU, Regional Hospital, Ekaterinburg, RussiaObjectives: Our task was to improve the technique of preservation of platelet rich plasma. Methods: 38 patients scheduled for multiple cardiac valve replacement in 1994 were divided into two groups: group I (10 patients) -without PP; group II (28 patients) -PP was performed preoperatively. The first PP was made ten days and the second -3 days before the operation. PRP was preserved by cryoconservation. Our technique of cryoconservation is distinguished by the speed of freezing (17-18~ and absence of DMSO. This made it possible to preserve 90 % functionally active platelets during 20 days. The PRP was transfused back after heparin neutralization. The Hospital Ethics Committee approved the investigation.Results: The blood loss through the 1st p. o. d. was significantly greatest in the group I (725 _+ 97 ml) and all the patients required transfusion of the donor blood (480 + 112 ml) whereas the blood loss in group II was 489 +_ 75 ml and olny 12 patients required the donor blood. The number of platelets on the 1st p.o.d, was 107 _+ 12. 109/L (group I) and 153 + 19. 109/L (group II), p < 0.05.Conclusions: Our technique of PRP cryoconservation makes it possible to avoid the crystallization phase during freezing of PRR Thus the infusion of PRP may improve hemostasis after open heart surgery and limit the use of the donor blood. In-hospital outcome of women suffering an AMI is generally considered worse than that of men, but it is still debated whether female sex is per sea negative prognostic factor or is merely associated with other negative determinants of prognosis. The purpose of the present study is to evaluate the independence of the association between female sex and mortality (in the 567 patients of the Swiss Centers) and in the 36381 patients randomized in the ISIS-3 trail Mortality rate in women was 14.8% (1421/9600) compared to 9.1% (2417/26480) in men; in Switzerland: in-hospital mortality for women was 15.5% (20/129), for men 7.1% (31/438).The table shows the results of ISIS-3 in terms of odds ratios and their 95% confidence intervals either after unadjusted analysis or after adjustment for age, known to be the major confounding variable when prognosis of women after myocardial infarction is considered, and for all the available clinical and epidemiological characteristics collected at trial entry: These observations suggest that there is a small but independent effect of female sex on short-term mortality after acute myocardial infarction. (27) and bubble (13) oxygenators a, ere used. Anaesthesia was balanced and pts were extubated 12 to 18 hrs after CPB. Pts were monitored with Swan-Ganz catheters (SGC) for 24 hrs after CPB. At that time Qs/QT was calculate( according to )be standard shunt equation. After the SGC had been removed, an estimated shunt was calculated. Measurements of Qs/QT were performed: before induction of anaesthesia (1), after induction of anaesthesia (I[), 5 mins after CPB (III) 2 (IV) and 6 (V) hrs afiter CPB, 30 rains after extubation (VI), 24 hrs after CPB (V[1) and on the 2nd, 3rd, 5th, 8th and 13tb postoperative day (PD) (VIII, 1 X, X, XI, XI1, respectively). Analysis of data was performed by two-way analysis of variance, p < 0.05 being regard as significant.Results: The figure shows the values for Qs/QT expressed as means + SD. There was a significant increase in Qs/QT above b~setine throughoul the whole investigated period except on the 13th PD. QS/QT reached maximum at 30 rains after extubation (VI). Objectives: Many stndies have shown advantages of membrane oxygenalors over 9ubbie type oxygenators. The aim of this study was to evaluate the influence of 3x3'genator type on pulmonary shunt (as/aT) after coronary surgery. Methods: 40 patients (pts) gave their informed consent to the study which was approved by the University ttuman Research Committee. Pts were divided into two groups: A1 (n = 27) with a membrane o~genator and A2 (n = 13) with a bubble oxygenalor used during cardiopulmonary bypass (CPB). ths were monitored with Swan-Ganz catheters (SGC) for 24 hrs after CPB. At that tfme OS/OT was calculated according to the standard shunt equation. Alter the SGC had been removed, an estimated shunt was calculated..Measurements of OS/QT were performed: betore induction of anaesthesia (I), 30 mins after extubation (11), 24 hrs alter CPB (111) and on the 2nd, 3rd, 5th, 8th and 13th postoperative day (IV, V, VI, VII> VIII, respectively). Analysis of data was performed by one-way analysis of variance, p < 0.05 being regarded as significant.Results: The figure shows the values for QS/QT expressed as means _+ SD. Os/QT was significantly greater at 30 rains after extubation (II) in A2 group. The difl'ereuce between the two groups was no more significant from 24 hrs after CPB (III) to the end of the investigated period. ! i * p < a.0s betw~n ~o~ Conclusions: Membrane ox3'genation during CPB is accomplished by reduction in blood cellular destruction and less alteration in blood. The results of our study show the influence of oxygenator type on value of QS/OT only after extubation (12 to 18 hrs after CPB). The difference in QS/QT disappeared 24 his after CPB and since that time the oxygenator type had no influence on Qs/QT. It may be of particular importance in patients with severe forms of cardiopulmonary disease who are at risk of higher postoperative morbidity and mortality. OBJECTIVES: Hypomagnesemia has been reported with a variable prevalence (20 to 61% ) in ICU patients. Magnesium deficiency can induce a number of climcal symptoms (primarily cardiovascular and neuropsychiatric) but can also be clinically silent (10-65% are asymptomadc), METHODS: We measured whole blood ionized magnesium (LMg++) in 74 patients on admission to the ICU, using a NOVA 8 electrolyte analyzer (NOVA Biomedical), containing an iMg++ electrode. Blood was collected in syringes with dry heparin (Radiometer QS 50 ). Normal range of iMg++ was found between 0.45-0.55 mmot/L (healthy volunteers). RESULTS: For the entire population, we found a 61% prevalence (45/74) of hypomagnesemia (Figure 1) . Among the surgical patients, the prevalence was highest after cardiac surgery (85%) and after thoracic surgery (80%) and was lowest after neurosurgery (8%). Hypomagnesemia was also common in patients after liver transplantation (LVTX) or with hepatic failure (100% for both groups). CONCLUSION: Our findings confirm that hypomagnesemia is common in acutely ill patients, especially in those after cardiothoracic surgery or those with liver disease. Nevertheless. it is difficult to define the associated factors with sufficient specificity, so that measurements of iMg++ are warranted to diagnose hypomagnesemia. Hepariu influences platelet function and may lead to thrombocytopenia called heparin-associated thrombocytopenia (HAT) regardless of the dose and route of administration. Additinnal venous and/or arterial thrombosis may lead to life-threatening complications. The incidence of so-calied heparin-associated thrombocytopenia and thrombosis (HATT) ranges between I-5%. HATT is confirmed by a heparin induced platelet activation assay (HIPA). Results: From 11/93 to 11/94 1146 consecutive patients of our icu were reviewed retrospectively. All patients were treated with heparim The incidence of HATT was 1% (12). In all cases diagnosis was proven by a positive HIPA. 2/12 patients died. In 3/12 HATT could be confirmed before severe thromboembolic complications occured. 4/12 patients developed a deep vein thrombosis (DVT), 2/12 DVT and pulmonary embolism (PE), 2/12 DVT, PE and arterial thrombosis (AT) and 1/12 a DVT, PE~ AT and a sinus thrombosis. Conclusion: The incidence of HATT in a r series of 1146 pts. is 1%. Presence of thrombocytopenia and thrombosis of the great 'vessels is associated with a significant mortality (2/12). Computed tom0graphy (CT) and transthoracic/transesophageal echocardiography (TTE/TEE) are important tools in diagnosing and monitoring the extent of cenlraI venous and arterial thrombosis. A. Cabral MD, M. Shahla MD C. Meneses-Oliveira MD and JL Vincenl MD.PhD. Department of Intensive Care. Erasme University Hospital, Brussels, Belgium OBJECTIVE: To determine extreme hemodynanuc patterns in cardiogenic shock. Although ~.~xdiogenic shock is characterized by a low cardiac index (CI), high systemic w~,scular resistance index (SVRI), and high cardiac filling pressures, some patients may develop art atypical pattern. We reviewed the hemodyuamic pattern of 73 patients with cardiogenic shock, as defined by an initial Ct below 2.5 l/rain/m: in the presence of myocardial dysfimction attributed to ischemic heart disease (N=26), heart failure (N=7), valvulopathy (N=2) or recent cardiac surgery (N=38). After exclusion of 10 patients with concurrently suspected/documented infection, this study included 63 patients, of whom 23 (36.5%) survived. Treatment of shock included dopamine (N=43), dobutamine (N=56), norepinephrine (N=18) and epinephrine (N=23). 39 patients with arterial hypertension (AH) and initially law plasnla renin activity (PRA) had been studied. In all patient changes of arterial pressure (AP) after single administration of Enap was studied. Nypotensive reaction wiht deereasin E of average AP about 20-25 mm Hg aYter single drug administration observed only in 4 patients. EZAP monotherapy accomplished during one week with 20 mg daily dose. Hypotensive effect observed in 5 patients including ones which were susceptible to single ENAP administration. After that first stage of therapy all patints began to combinate ENAP with Hypothyazid in dose of 25 mg per day~ After week of treatment such drugs combination lead to veritable AP lowering in 3 addition patients. In the remaining resistant to such drug combination patients was add Corinfar in daily dose of 40 mg. This new drug combination permits to lower AP in 23 patients. Subsequent discontinuation of ENAP administration to such patients aid not connected with increasing of again.Therefore the most of the patients with AH and law PRA(78,7%)did not susceptible to ENAP therapy and ENAP and Hypothyazid combination. On the contrary-combination of Corinfar with Hipothyazid was effective in 59% patients with AH and low PRA. METHODS: In 35 patients with cardiogenic shock due to ischemic heart disease (N=26), heart failure (N=7) and valvulopathy (N=2), hemod31aamic data including measures of intravascular pressures, cardiac output and mixed venous gases were collected at regular times intervals, at least 3 times a da?. All measurements were obtamed in a relative steady state and in the absence of severe anemia or hypoxemia. Treatment of shock included dobutamine (N=30), dopamine (N=24), norepinephrine (N=I2) and epinephrine (N=7 Objective: Based on our previous studies of the function of isolated liver grafts, this experimental protocol aims at developing a novel extracorporeal liver support circuit, with an incorporated pig liver. Methods:The graft liver was obtained from pigs weighing 15-20 Kg. Under general anesthesia the aqimals underwent total hepatectomy,following cannulation of the portal vein, the infrarenal aorta and the infrahapatic vena cava and peffusion wit h 4 It of heparinised R/L solution at 4~ The circuit consisted of the graft liver connected to a fluid reservoir and a centrifuge pump. Ten healthy pigs weighing 30-35 Kgr were connected to the circuit as follows: The rt carotid artery was connected to the portal vein of the graft and the rt jugular vein was connected to the fluid reservoir, through the centrifuge pump. The fluid reservoir collected the outflow from the graft's suprahepatic inferior vena cava. The cystic duct of the graft was ligated and the bile.duct cannulated for bile collection and measurement. Bridges were adapted to the circuit to bypass the graft liver when necessary, in cases of by pass blood perfusing the graft was oxygenated through a bubble oxygenator. Mean total priming volume of the circuit was 600 ml. Temperature was maintained at 38~ and portal vein pressure at 16 (12-20) mmHg. The flow was 0.5-0.7 ml/gr of graft liver mass per minute. Observation period was 8 hours (T8). Results: Results of the hemadynamic and metabolic monitoring of the recipients [MAP (T0=124mmHg , T8=118mmHg), HR (T0=177, T8=201), RAP (T0=11mmHg , T8=19mmHg), PAP (T0=26mmHg, T8=31mmHg), PCWP (T0=14mmHg, T8=15~mHg), SVR (T0=1940dyn'sec/cm '5, T8=2190dyn'seclcm~ PVR (T0=206dyn.sec/cm o, T8= 354 dyn.sec/cm ,'~), CO (T0=4.631t/min, T8=3.61t/min), DO 2 (T0=662ml/min, T8=261.6 ml/min), VO 2 (T0=118ml/min, T8=111ml/min), O2ER (T0=17.8%, T8=42.5% ), pH (To= 7.48, T8=7.39 ), pO 2 (T0=292mmHg, T8=371mmHg), pCO 2 (T0=28mmHg, T8=30 mmHg), pVO 2 (T0=47mmHg, T8=36mmHg), SvO 2 (T0=84%, T8=66%), BE, Na, K, Ca ++, lactate, osmolality, AST, ALT, PT, APTT, revealed hemodynamic and metabolic stability of the animal. 02 consumption, CO 2 production and tissue oxygenation of the graft were also studied. Conclusion; The described circuit proved to be safe and well tolerated by healthy animals but its value for temporary liver support is currently being estimated, in a surgically induced experimental fulminant hepatic failure modal. Introduction: Prosthetic materials like silikone, dacron, teflon e.tc. produce auto immune responses and may even trigger clinical syndromes like Scleroderma, Sjogren, SLE el.c. In our study we followed the evolution of humorial immunity parametrs for up to five years in a cohort of paced pts with implanted metallic and silicone materials. Method: 24 paced pts (mean age 55+-13 yrs) without clinical or laboratory findings of malignancy or immune disorders were included. We measured the immunoglobulins, the complement, the auto antibodies and the proteins involved in inflammatory reactions every 6 months. The initial and final mean values are shown in the Obiectives: HSP, a systemic leucocytoclastic vasculitis and anaphylactoid purpura can be accompanied by abdominal pain and life-threatening intestinal bleeding. Recently we could disclose, that these patients develop severe FXIII-deficiency and immense haemorrhagic oedema of the intestinal wall. By the following case report we will demonstrate and discuss the importance of FXIIIdeficiency for pathogenesis, therapy and outcome in HSP. Case report: A 41 year old man developed typical skin manifestations of HSP following an episode of severe (biliary ?) pancreatitis and percutaneous draining of a pancreatic pseudocyst. Two days later he had a paralytic "ileus with immense hemorrhagic wall-oedema and massive dilatation of the small bowel. He got fever up to 39.5 ~ and developed severe gastrointestinal haemorrhage (blood transfusions necessary). The coagulation data disclosed a severe FXHI-deficiency (activity 34%), whereas Quickvalues, platelet count and ATIII-level were found to be within the normal range. Elastase was markedly elevated. Substitution of FXIII to normal levels leeds to the cessation of bleeding symptoms and abdominal pain, later resulting in a restitutio ad integrum. Conclusions: HSP with intestinal involvement is a life-threatening vasculitis, in which careful and frequent examinations of the coagulation system, especially of FXIII are necessary. Detailed analysis of the coagulation data suggest, that the severe FXIIIdeficiency is due to a specific degradation by proteolytic enzymes (like elastase) as well as consumption within the immense haemorrhagic oedema of the intestinal wall. Knowing these facts, even most severe cases of HSP with intestinal involvement can be successfully treated by substitution of FXIH. A 49-year-old woman presented a 3 year history of occasional self-limited episodes of weakness, generalized edema and o!!~aria. The immunologic testing showed no~nnai levels of complements, Clq inhibitor, and serum chemistry values, between or during a attack, She was not treated. She was a~mitted to the hospital with symptoms including nausea, vomiting, weakness and ol!guria. On examination, the patient presented facial and g~neralized edema. The systolic blood pressure was 60 mm Hg, pulse 140 beats/mir~ute, hematocrit 0.59, seln~n protein 46 9/i, and se~um albumin 23 q/l. An leG-kappa pa[apFotein was demostrated (7.82 g/l) and urine was neaative for puotein. C~'stalloid and colloid don't increased the blaod pressure but resulted in anasarca, with a total of II lit[as of in~ravenous fluids. Therapy wink flozen plasma, 1.000 units of Clq inhibitor, cortlcosteroids, annihistwnines and antifibrinolytic agents was uns~iccessfull. The a~minist~ation of dopamine, norepineph~ne and epinephrine was inefective. The patient died at the 48 bores, Only a few cases have been reported, all had igG paraprotein, The pathophysio!o~] is urd~no~n% but is possible that the paraprotein may be zesponsib!e for the increased capillary pe~leabilityo Despite efforts to res~scinate the patients during an acute attack, the syndrome is often fatal. The variable course of systemic uapiliary leak syndrome and the unpredictability and self-limited nature of attacks cloud assessment of therapeutic inte~-vention. The purpose of the present work is to provide some information about the nursing care and results from our experience in continous arteriovenus hemofiltration (CAVH).CAVH is an extracorporeal technique, especially applicable in the critically ill patients, for disturbances, and for the control of azotemia.We used this method in 30 critically ill patients 16 men and 14 women ages from 32-74 who had sepsis -ARF 10 congestive heart failure 8 postoperative multiple organ failure 8 and polytrauma 4.This method was applied to these patients from 24 to 168 hours. 20 % of the patients recovered completely their kidney function, 50 % improved their kidney function and 30 % died.We concluded therefore that this method was very effective for the critically ill patients to whom it was applied, but it requires excellent and continuous nursing care; Under the above mentioned circumstances the method works effectivelly. An animal model with rats undergoing a dialysis procedure was designed to test the hypothesis that recovery from ischemic acute renal failure (AIRF) may be affected by the type of membrane used in hemodialysis. Male Sprague Dawley rats were allocated to 2 groups: in group I, (n=48) AIRF was inducted by bilateral renal artery clamping for 60 rain. Group H (n=48) rats underwent a sham procedure. In each group, rats were dialyzed twice (4th and 8th day) with either a Cuprophan (Cupro), a Hemophan (Hemo) or a PAN (AN69) minidialyscr or stayed nondialyzed (no HI)). Renal function was monitored daily by measuring urea and creatinine values and by two single shot inulin clearances on the days following dialysis. Additionally hemolytical activity of complement was determined. Inulin clearance on day 5 was reduced significantly but there was no difference in the degree of decrement in glomular filtration rate (GFR) between dialyzed and undialyzed rats, nor between the dialyzed animals with different membranes (GFR: no HI): 0.78_+0.54; Cupro: 0.84_+0.9; Hemo: 0.82_+0.28; AN69: 0.77_+0.31). The evaluation of renal function by day nine revealed significant recovery for all AIRF-groups compared to day 5 (p<0.001), irrespective of wether they underwent dialysis or not, or the type of dialysis membrane. Complement activation could be detected in all dialyzed groups but no statistical differences between the animal groups dialyzed with different membranes were noticed. Our findings refute the hypothesis that in AIRF exposure to complement-activating cellulosic membranes impairs the recovery of renal function in rats. CHANGES Patients: 150 patients who underwent first cadaver kidney transplantation in our unit between January and December in 1994 were involved. The recipients were divided into 3 Groups: Group I." non functioning graft (n=27); Group II: delayed graft function (n=59), Group Ili: good graft function (n=64). The grouping criteria were: a/haemodialysis in the fiI~t 5 postoperative days, b/diuresis in the I st postoperative day, c,' scram crcatininc difference between the 1st postoperative day and the preoperative level. All of the parameters were involved into the exarainatio, which we measllre in our every, day practice. Results: The preoperative haematocrit level differed significantly between Group I. (0.36) and Croup II. and III. (0.31 and 0.30, p< 0.05). Intmo!0emtive significant differences were found between the different groups in systolic blood pressure (Group I. 110 Hgrmn, Group II. 140 Hgnnn, Group III. 165 Hgmm, p<0.05), mean arterial pressure (Group I. 66 Hgmm, vs. Group II. 83 Hgnun p<0.05, vs. Group III. 116 Hgmm p<0.001), and pulse-amplitude and rate-pressure product too. The second warm ishaemic time in Group III. was significantly shorter than in the other two groups (Group III. 39 inin. vs. Group II. 43 rain. p< 0.05, vs. Group I. 49 rain. p< 0.00!). The rejection rate was higher in the first 5 days in the patients with non-functioning grafts (Group I. 53% and Group II. 24% vs. Group III. 12 %) . The other examined parameters have not differed significantly. Conclusion: According to our results the success of the kidney transplantation is mnitifactorial. The most important factors of this relationship are: the perioperative fluid-balance, the maintenance of adequate perfusion blood pressure during the operation, good surgical technique and immunological problems.