key: cord-0005854-2u0ug8po
authors: nan
title: Acute respiratory failure I. Morophology and pathophysiology
date: 1983
journal: Intensive Care Med
DOI: 10.1007/bf01692912
sha: 84619941e500024ddda4407eecaef67fe123f932
doc_id: 5854
cord_uid: 2u0ug8po

nan

The design of the alveolar septum displays a number of features that favor the maintenance of a dry air-blood barrier in the interest of gas exchange. Whereas gas exchange occurs in relation to a 2-compartment model -alveolar air space and capillary blood space separated by the tissue barrier -the control of fluid movements is understood on the basis a 3-compartment model, with the cellular sheets of epithelium and endothelium separating three fluid spaces: plasma, interstitium, and alveolar lining layer. The capillary endothelial cell lining is continuous, but the intercellular junctions are leaky, allowing smaller solutes to be easily exchanged between plasma and interstitial fluid. In contrast, the alveolar epithelium is made of a mosaic of squamous type I cells and cuboidal type II cells; the epithelial intercellular junctions are tight, preventing simple solute exchange between interstitium and alveolar lining layer. In the interest of optimizing barrier structure for gas exchange the type I cells are of vast expanse, highly complex in their architecture. As a consequence they are very vulnerable and their repair upon damage is difficult. Regeneration of alveolar epithelium must occur from type II cells as stem cells. An important question is how the alveolar septum prevents fluid from accumulating in the tissue space. It appears that the compliance of the alveolar septal interstitium is restricted by design: (I) by eliminating an actual interstitial space on large parts of the surface due to fusion of epithelial and endothelial Oasement membranes; (2) by the myofibroblasts in the actual interstitial space. Moreover, any fluid accumulating in the septum is rapidly drained away towards the juxtaalveo]ar fluid sumps and lymphatics. Efficent gas exchange in the lung requires a short diffusion distance between alveolar air and capillary blood over a wide contact area. To this aim the tissue mass of lung parenchyma is extremely reduced: the capillary bed is covered by a thin tissue membrane and supported by a flimsy connective tissue network. Obviously, these delicate structures are easily deformable, and the dimensions of structural elements important for gas exchange, i.e. the alveolar surface area, the thickness of the blood-gas barrier, and the capillary volume are profoundly affected by the interaction of four mechanical forces: i) tissue forces; 2) the microvascular pressure; 3) the alveolar pressure; and 4) the surfaces forces. Within the lung volume range of normal breathing the surface forces have a strong impact on the free surface area, since alveolar walls appear to be negligible mechanical components except as platforms for surface tension. Compared with fluid filled lungs normal air filled lungs reveal a reduced free alveolar surface area by folding of the alveolar septa beneath the surface lining layer. Abnormally increased surface forces further reduce the surface area. The peripheral airspaces become unusually wide, and are separated by bulky tissue masses of piled-up septa of "collapsed" alveoli. Among other features these structurefunction relations may explain impaired gas diffusion, increased right-to-left shunts, and increased alveolar dead spaces in lung diseases affecting the properties of alveolar surface lining layers only. The acute stage of ARDS is characterized by an interstitial and alveolar edema due to focal alveolo-capillary barrier destruction regardless of whether the primary action site of the noxious stimulus may have been alveolar space or capillary lumen. While epithelial lesions can easily be visualized endothelial defects are sparse although occasional large amounts of red blood cells and fibrin strands in the subendo" thelial and interstitial space are strong evidence of temporary capillary leaks. Different cell repair mechanisms, a fast one for the endothelium and a time consuming two stage repair process for the type I epithelial cells, account for this observation. The thickening of alveolar septa by proliferation of type II epithelial cells and an accumulation of cells and fibers in the interstitial space, intraalveolar fibrosis, changes in extraalveolar vessels and structural rearrangements of peripheral air spaces are distinctive features of the subaeute stage of the disease. Its duration does not necessarily correlate with the severity of structural alterations. Cases of severe functional impairment may go along with extremely reduced capillary volumes and a very low oxygen diffusion capacity as estimated by morphometry. Rapid improvement of gas exchanging and ventilatory lung function may reflect a partial restoration of the geometry of peripheral airspaces rather than a reversion of fibrotic lesions. The pulmonary edema of the ARDS is due-to loss of integrity of pulmonary capillary walls. ARDS complicates pulmonary and systemic infections, hypoperfusion states, inhalation of toxic gases including oxygen, aspiration of damaging liquids, intravascular coagulation or embolism, platelet and leukocyte aggregation, immunopathies and allergic reactions, and hypoxia, all of which may lead directly or indirectly to endothelial cell damage. Many humoral factors are involved (in vitro and in vivo animal studies, very few human studies) : activated complement fractions, prostanoids, leukotrienes, proteases, peptides and coagulation and fibrinolysis products. We have studied complement activation, thromboxane and prostacycline metabolites (TX% and 6-oxo-PGF I ), and immunoreactive and enzymatlcally active thypsin in plasma from 50 ARDS patients. Abnormal C~ consumption (as measured by C3a/C ~ ratio) and elevated plasma Cm -like activity ~as~measured by a leukocyte a~regation assay) were associated with, respectively, 84 % and 86 % of cases of ARDS. Both tests were more sensitive indicators of complement consumption than assays of total hemolytic complement activity (CH50) or total C 3. Sequential samples on both sides of the pulmonary circulation showed initial pulmonary clearance followed by liberation of C~ -like activity. Abnormal TXBe plasmatic values were ~und in 58 % of ARDS patients, during consecutive days of evolution; values up to 8000 pg/ml were measured. Trypsin values were enhanced (> 70 ng/ml) in 87 % of ARDS patients, particularly in septic cases; repetitive high trypsine values correlated with high mortality rate. Our data confirm the participation of the complement-leukocyte cascade in the increased pulmonary capillary permeability of ARDS. Selective sequestration of polymorphonuclear neutrophils (PMNs) in the lung seems to play a central role in early ARDS. Sequestered PMNs m~y damage lung tissue either directly or by releasing mediators such as lysosomal proteases (f.e. elastase). We studied the possible role of PMNs and of elastase on lung function by infusing elastase in normal and in agranulocytic minipigs. METHODS: Among 16 minipigs studied (anaesthetized and ventilated) 4 were pretreated with dimethylmyleran which depleted granulocytes ccmioletely. 3 pigs without pretreatment or elastase infusion for control. Elastase was infused (330 U/kg-h) continously for 3-4 hours in the elastase and in the agranulocytic group. ~itiple determinations of cardio-respiratory functions were repeated in short intervalls before and during the treatment. Finally, morphological studies were performed. RESULTS: The control group revealed steady state conditions during the whole experiment. However, elastase caused typical effects beth in the normal as well as in the agranulocytic pigs: I/anediately after onset of elastase pulm.art.pressure (PAP) increased but only temporarily. Oxygenation (AaD02) deteriorated after one hour when PAP again had returned to baseline. Pulm. diffusing capacity decreased slightly. In the non-pretreated pigs elastase infusion caused a significant sequestration of PFi~s in the lung. But, as the effects on lung function were nearly identical in the normal as well as in the agranulocytic group, this could primarily be an elastase effect independent of the presence of PMNs in the lung. CONCLUSION: Elastase, released by PMNs sequestered in the lung, may (interacting with other systems?) play a role in early ARDS. Elastase itself seems to induce pulmonary P~ sequestration by neurophil activation.

The posttraumatic-hypovolemic shock results in pulmonary changes which predominantly affect the cellular and subcellular levels. Morphologically 2 distinct stages are present, the "lung in shock" within the first hour and the "shock lung syndrome" which manifest itself after 24 to 48 h. The early stage "lung in shock" is characterized by a multitude of changes in the microvascular system: massive leukostasis, polymorphonuclear granulocyte (PMN) degranulation, and endothelial cell swelling. The purpose of the study was to compare the ultrastructural findings of the lung, liver and skeletal muscle with respect to the granulocytes (PMN) and the endothelial cell damage. In many phases, the endothelial cells are considerably swollen in the lung and skeletal muscle. But whereas many polymorphonuclear granulocytes present in the alveolar capillaries of the lung, only a few isolated granulocytes are seen in the vessels of the skeletal muscle tissue. Leucostasis in the lung was also confirmed quantitatively using llIIn-oxine labeled PMN; in the liver it was present to a much lesser extent, while there were no changes versus the pre-shock period in skeletal muscle. The granulocytes might be responsible for the onset of lung damage after shock, acting on the endothelial cells (reactive oxygen species). One needs to consider other causes of analogues in striated muscle (hypoxia?) .

M.K.Sykes, Nuffield Department of Anaesthetics, Radcliffe Infirmary,. Oxford, 0X2 6HE, United Kingdom.

The pulmonary circulation is a low pressure, low resistance system which accommodates the marked increases of cardiac output associated with exercise by recruitment and distension of the pulmonary vaseulature.

The required distribution of perfusion is governed primarily by the relationship between alveolar, pulmonary arterial and left atrial pressures, blood flow being least in the non-dependent zones and maximal in its dependent zones, whatever the body position.

Regional blood flow is reduced by a reduction in regional lung volume (which narrows the extra-alveolar vessels) and by regional underventilation (which leads to a reduction in alveolar P02 and increases in alveolar Pco2).

The reduction in regional flow from hypoxic pulmonary vasoconstriction (HPV) isprimarily governed by alveolar P02.

However, in collapsed regions of lung the mixed venous P02 determines vascular tone.

HPV is, therefore,accentuated when cardiac output is reduced and decreased when output is increased.

This may account for the well known direct relationship between shunt and cardiac output. An increase in shunt may also Occur when HPV is inhibited by drugs, raised pulmonary artery or left atrial pressure or hypothermia, or when drugs cause vasoconstriction in the oxygenated areas of lung.

PHYSIOLOGY OF PULMONARY CIRCULATION AND RIGHT HEART PERFORMANCE. A.Versprille, Deptm. of pulmonary diseases, Erasmus University, Rotterdam, The Netherlands. The cyclic changes in heart performance during artificial ventilation have to be ascribed to mechanisms with short time constants because they occur and recover within the period of a ventilatory cycle. Therefore, nervous and humoral control as well as cardiac contractility changes are rejected as mechanisms explaining the cyclic events. Guyton's theory, that venous return (Qv) is controlled by central venous pressure (Pcv), also counts for the cyclic events during artificial ventilation. Pcv is mainly dependent on intrathoracic pressure (Pth), although a decrease of Pcv,tm (transmural) occurs during insufflation: the decrease in venous return due to the increase in Pcv causes a decrease in filling pressure i.e. preload. Also transmural pulmonary artery pressure (Ppa,tm) decreases during insufflation when PEEP is below about 6 cmH20. This fall in afterload might contribute to the decrease in RV-filling pressure. Above a PEEP of 6 cmH20, Ppa,tm at peak insufflation increases progressively, when PEEP increases, undoubtedly due to a progressive obstruction of the pulmonary circulation. The fall in venous return to the left ventricle (LV) can be ascribed to the fall in RV-output. The time delay can be explained by a squeezing effect on the pulmonary circulation during early insufflation. Plved,tm fell during late insufflation and early expiration concomitantly with the fall in pulmonary venous return. This fall also caused a fall in transmural aortic pressure. Cardiac compression and ventricular interdependence were rejected as mechanisms controlling heart action during artificial ventilation.

LV FUI~CTION DURING HECHA[~ICAL VEI~TILATION. J.L. Robotham, University of Texas Health Science Center at San Antonio, San Antonio, Texas. Conventional wisdom suggests that the dominant hemodynamic factor during IPPV is the inspiratory reduction in systemic venous return and hence a secondary reduction in LV output. The purpose of this paper is to question this simplistic approach to analysis of the LV during IPPV in critically ill patients. Under pathologic conditions, (e.g. congestive failure) other factors will theoretically become increasingly important based on a physiologic analysis of IPPV.

Variables to be examined which have been found to influence LV performance during IPPV include: a) inspiratory expiratory variation in RV and LV stroke volume b) influence of respiratory rate (from I0 breaths per minute to I0 per second) and the influence of the I-E ratio c) changes in pulmonary vascular volume, e.g. hypovolemia or pulmonary edema d) end expiratory lung volume, i.e. PEEP e) changes in tidal volume and its effects boLi~ on pulmonary vascular volume and mechanical compression of the heart by the lung f) changes in heart volume on mechanical heart-lung interaction g) the increases in inspiratory pleural pressure acting to effectively reduce LV afterload h) respiratory changes in abdominal pressure which may influence not only venous return but also LV afterload and regional distribution of cardiac output i) regional changes in LV geometry produced by IPPV with and without PEEP, and j) the influence variation in right heart volume during the respiratory cycle on LV performance.

From analysis of the above, it is concluded that under pathologic conditions, multiple factors other than changes in systemic venous return may significantly affect LV performance during IPPV. It is hypothesized that when better understood, IPPV may be able to assist LV performance in critically compromised patients. To identify the mechanisms responsible for decreased right ventricular (RV) performance during mechanical ventilation with high levels of PEEP in ARDS pts and the effects of mormalizing cardiac output by volume expansion (VE), we studied I0 such pts. RV volumes (EDV-ESV) were calculated by both thermodilution and radionuclide studies at O, 15, 20, 25 Conclusions : In ARDS pts, PEEP decreases RVSV by imposed limitation on EDV (due to increased intrathoracic pressure) despite a higher after load. RV contractility tends to increase (left ward shift of the RVESPt-V relationship) probably due to increased catecholamine levels.

After VE, SV returns to prePEEP values in relation to increased preload and sustained increase in contractility.

NEW METHODS FOR ASSESSING GAS EXCHANGE. F. LEMAIRE, HOpital Henri Mender, F -94010 CRETEIL C~dex.

The oxygen method (Berggreen, 1942) is routinely used in the ICU to evaluate venous admixture. However, values obtained by this method may vary according to the inspired FiO 2. New methods have emerged to provide reliable assessment of gas exchange. The Krypton m 83 scinti-gra~ can be used to evaluate the distribution of regional VA/Q. This short lived isotope, highly soluble in gas and liquid phase, provides scintigrams of ventilation and eerfusion when inhaled and in.fuse.d intravenously. The two scans can be compared and VA/Q ratio can be evaluated zone by zone. This method best used when regional alterations of ventilation and/or perfusion are suspected (pulmonary embolism, emphysema, unilateral pneumonia) and evaluate the role of changes in body position on gas exchange.

The six inert gas method, proposed by W~gner et Evans in 1974, assesses the abnormalities in VA/Q distribution, irrespective of their topography. The high sensitivity of this method is its major advantage, since the use of highly, insoluble gas (SF6, ethane) can detect zones with VA/Q ratios of lower than 0,01. The infusion of the most soluble gas (acetone, ether) provides a measurement of dead space. Additionally, this method is not dependent on Fi% and oxygen diffusion. It can be applied to patients wi~h mechanical ventilation in the Icu, without changing the FiO 2. So far, this method has been mainly directed at evaluating the effect of mechanical ventilation, with and without PEEP, anesthesia, decubitus, pulmonary edema, ARDS. However, technical difficulties require highly specialized laboratory, and prevent its routine use. The method offers a sophisticated aooroach for understanding the complex VA/Q relationshi.o in lung disease. Since the introduction of High Frequency Jet Ve~ tilation(HFJV) (f=1,5-SHz,V~VD)other HFV systems has been designed. In High Frequency Pulsation (HFP) better elimination of CO o is managed by transversely flowing current o[ fresh gas and in using a circle with CO2-absorber(f=5-9Hz), V~V_).In Forced Diffusion Ventilation(FDV) 2jet nozzles are located at the distal end of the tube (f=upto25HZ, .In the field of High Fr~ quency Oscillation(HFO) we developed the so called Pneumatic BFO(f=up to 12Hz). Indications for HFV are based upon our clinical experiences in 137 patients: 1.Postoperative ventilation in cardiac patients (n=45). Gasexchange could be maintained well during controlled ventilation and the weaning period. Conteraction to hemodynamics depends on f,I/E-ratio and primary pressure. 2.Intraoperative application of HFV (lung surgery) (n=36) is favourable (extensive resting position of the lungs, reduced gas loss through pleural leakage).3.Intensive care patients(n=56). In ARDS better PaD 2 in comparison to conventional ventilation could be established in using high frequencies( 5Hz) long I/E ratio and high primary pressure. Bronchopleural fistulae were managed with FDV(Iow VT) , the combination with IPPV faciliated CO 2 elzmination. In normal lung function low frequencies (I,5-5Hz) short I/Eratio and low primary pressure (1,5-2 bar) achieved satisfactory gas exchange with low Pairway" In patients with brain damage we observed a arop in intracranial pressure. HFP acts as a advantageous method of augmented ventilation (flail chest) and is a suitable method for breathing therapy.

DIFFERENTIAL VENTILATION 151 BILATERAL LUNG DISEASE. G. Hedenstierna, Dept. of Clinical Physiology, Huddinge Hospital, Stockholm, Sweden Anaesthesia and acute respiratory failure impede vnetila-Lion of dependent luna units and perfusion of non-dependent ones, creating considerable ventilation-perfusion (V/Q) mismatch. General PEEP can improve V/Q but it cannot restore it to normal. To imarove matching, ventilation must be distributed in proportion to regional blood flow. This can be accomplished by I) placing the subject in the lateral position, 2) ventilating each lung in proportion to its blood flow (differential ventilation) and 3) applying PEEP solely to the dependent lung to ensure even distribution of inspired gas within that lung (selective PEEP). Differential ventilation with equal distribution of the tidal volume between the lun~d a selective PEEP of I0 cm H~O to the deoendent lung resulted in equal distribution o~ perfusion between the lunqs in ana~tized lung-healthy subjects, suggesting "optimum" ~/Q matching. Using this ventilator setting as a rule of thumb in patients with acute, severe, bilateral lung disease, arterial oxygen tension was impr~an average of 45% compared with that during general PEEP, with no reduction in cardiac output. It is concluded that differential ventilation with selective PEEP can offer considerable improvement in gas exchange in acute, bilateral lung disease. We studied the effects of high frequency oscillatory ventilation (HFOV) on gas exchange, hemodynamics and basic pulmonary mechanics in dogs with normal and stiff lungs.

In a group of 9 healthy animals, HFOV maintained normal gas exchange but cardiac output decreased and pulmonary and peripheral resistances rose.

In another group of 5 dogs, HFOV was found to increase functional residual capacity (FRC), mean airway and pleural

pressures. This effect was more accentuated at high oscillatory frequencies.

The FRC was measured over a 16 minute period and found to increase around 60% during the first minute, after which it remained stable.

The effects of HFOV were also studied in 8 other dogs which were first lavaged with saline to remove some surfactant and produce stiff and unstable lungs. HFOV produced a higher FRC than positive endexpiratory pressure even though mean airway pressures were similar.

In these animals A-aD02 gradients were lower during HFOV but C02" elimination was somewhat decreased.

We conclude that although HFOV can achieve gas exchange in normal and stiff lungs, the cardiovascular function may be impaired, probably due to its effect on lung volumes.

L.Gattinoni Istituto di Anestesia e Rianimazione, Univez-sit& di Milano, Milano, Italia Extracorporeal CO 2 removal (ECCO2R) with low frequency positive pressure ventilation (LFPPV) has been introduced in clinical practice in 1979. Most of the experience with this technique has been gained in Milano (22 pts) even tought few other ptsrecently underwent LFPPV-ECC0_R in West Germany, U.S.A. and Australia. All t~e pts had severe ARDS of various etiology: politrauma, viral and bacterial pneumonia, fat embolism and septic, toxic shock lung. The relatively small number of patients does not allow to suggest any indication on etiological basis, even tought an higher rate of success has been reached in some etiological group (Shock lung, fat embolism, virus and bacterial pneumonia). To date the entry criteria are based on functional data: i) blood gases values under defined ventilatory conditions, as proposed in the American extracorporeal membrane lung oxygenation study; 2) total s~atic lung compliance lower than 30 ml cmH 0-, when measured at 10 ml kg-I lung inflact~on.

These criteria still entail a mortality rate close to 100% with conventional treatment. Controindications to the treatment remain systemic degenerative diseases, severe head injury and pulmonary active bleeding (including hemothorax). tic transmitters and by injury currents occuring in the border zone of an acute myocardial infarction. ~oreover 9 acidosis favours the development of ectopic focal activity by inducing early af~crdepolarizations in Purkinje fibres.

The production of reentry requires shortening of the waves of excitation either by accelerated repolarization or by slewing of the propagation of impulses or by both. The slowing of impulse conduction might be due to the generation of slow responses in depolarized fibres exposed to sympathetic stimulation. Moreover, an appropriate delay of conduction may also result from the electrotonic spread of excitation across damaged inexcitable tissue. A non-homogeneous state of excitability of the heart forms another condition indispensible for reentry because it enables unidirectional conduction to take place. Nonhomogeneous excitability is normally encountered during the early phase of recovery in the cardiac excitatory cycle (vulnerable period) and is strongly increased by frequent extrasystoles.

CLINICAL PHARMACOLOGY OF ANTIARRHYTMIC DRUGS. F. Follath. Division of Clinical Pharmacology, Oepartment of Medicine, University Hospital, Basel, Sw~zerland. An exact knowledge of the pharmacodynamic and pharmacokinetic properties of antiarrhytmic agents is essential for their optimum use in intensive care patients. The electrophysiological mechanisms of action, as well as possible effects on sinoatrial,atrioventricular or intraventricular conduction, myocardial contractility and peripheral resistance should be taken into account when a particular drug is selected to treat acute ventricular or supraventricular arrhytmias.A further decisive factor for therapeutic success is the choice of an adequate dosage regimen Potentially effective drugs may fail because the therapeutic serum drug levels are not reached and/or maintained during a dosage interval.Drug elimination rate in acutely ill patients is often altered by cardiac,hepatic and renal failure or by interacting drugs administered simultaneously. Individualization of antiarrhytmic drug dosage is therefore necessary to obtain maximum benefit and to avoid unwanted side-effects. This task can be facilitated by serum drug concentration monitoring. Rapid and reliable techniques for measuring lidocaine, procainamide, quinidine and disopyramide concentrations are now available. The practicability and utility of this approach will be illustrated by data on lidocaine treatment in patients with ventricular arrhythmias following myocardial infarction.

DETECTION AND DOCUMENTATION OF ARRHYTHMIAS -R. Ritz, Intensive Care Unit, Dep. Internal ~ed.,University of Basle

The computerized arrhYthmia-monitoring system allows i) recognition of ectopic beats, for a potentially prophylactic treatment, 2) instantaneous alerting to the first occurrence of life-threatening arrhythmias, including documentation of the preceding rhythm, 3) detection of a need for special procedures, e.g. pacemaker, by announcing even short periods of "omitted beats" (sick sinus), and 4) checking of therapeutic management, e.g. testing of antiarrhythmic drugs in chronic coronary heart disease or pacemaker function. In patients with acute coronary heart disease the value of antiarrhythmic treatment of the so-called "warning '! ventricular ectopic beats to prevent ventricular tachycardia or fibrillation is still questionable. As long as such arrhythmias are treated routinely in most coronary care units, worldwide, the reliable detection of ectopic beats is essential. The conservative methods of continuous ECG observation on a scope of up to eight patients at a time is no longer a justifiable task for well trained and highly skilled CCU personnel, for two reasons: l) Even persons specialized in rhythm recognition miss up to 60% of ectopic beats during more or less continuous observation of an ECG display screen. 2)ECG reading is a typical taskforthecomputer. An investigation of the specificity and sensitivity of our arrhythmia-monitoring system (EP 78220) still revealed a considerable number of false-positive alarms, disturbing the nursing staff unnecessarily, but ahighsensitivityofmore than 90% in the detection of ectopic beats (n=10.400) represents an increased level of security for the patients. Histographical presentation of the trend of the arrhythmias facilitates the overview, while TV display in all rooms of the CCU and the proven reliability of the system give the nursing staff more freedom in patient care. Our nurses would now certainly not like to have to live without the system. Many antiarrhythmic drugs have been studied on their efficacy in preventing primary ventricular fibrillation in acute myocardial infarction, but only lidocaine given in rather high dosages has been shown to be effective in one randomized controlled study. In view of the rather high incidence of side effects this latter study and the equal mortality rates in both groups, prophylactic antiarrhythmic treatment in all patients admitted to the CCU because of acute myocardial infarction is questionnable. However, recent uncontrolled observations on lidocaine have indicated a very low incidence of primary ventricular fibrillation and one case controlled study showed a lower hospital mortality in lidocaine treated patients. Moreover 20% of untreated patients will develop recurrent attacks of ventricular fibrillation which may lead to extension of infarction and higher mortality rates. Other studies have suggested that effective lidocaine prophylaxis during 48 hours can be achieved with an acceptable incidence of minor side effects and a low toxicity provided that the following measures are undertaken: I. adequate loading dosages of 200 mg given in 10-20 minutes, 2. continuous administration using automatic infusion pumps in dosages averaging 3 mg/minute, 3. carefull monitoring of side effects and abrupt discontinuing when they supervene, 4. reducing dosages by at least 50% in shock, heart failure or hepatocellular disease. It may be concluded that prophylactic antiarrhythmic treatment with lidocaine is preferable than placebo, provided that adequate measures are undertaken. MASSIVE PULMONARY EMBOLISM : DIAGNOSTIC ASPECTS. P.Even, D.Safran, G.Dennewald, M.Stern, P.Reynaud, H.Sors.H~pital Laennec, 75007 Paris, France. 500 patients with pulmonary embolism (PE) were studied by heart catheterization, perfusion lung scan (PS) and angi~ graphy (PA) and some by ventilation scan (VS) or digitalized substraction angiography ; 40% had massive PE (> 60% vascular obstruction-Miller index). In spite of new soph~ sticated investigation methods the initial diagnosis remained desperately erratic and delayed (I-90 days) with an eqal number (60%) of false positives or false negatives eliciting dramatic consequences : recurrent PE and undue and hazardous anticoagulation or surgery. The main erro~were : a) in non-massive PE, acute or chronic lung diseases (pneumonias, atelectasis, cancer, asthma, pleurisy) b) in massive PE, heart, abdominal, septic or nervous diseases (myocardial infarction, cardiac tamponade, dissecting aneurysm, hypovolaemic or septic shocks, acute peritonitis, liver or sub-phrenic abcesses, acute cholecystitis or pancreatitis, stroke). To improve the diagnosis performance we have reexamined the sensitivity and specificity of clinical and laboratory findings rather their frequency order. We propose the following statements : I) always think to PE but never accept it without proof, 2) there is no benign PE but only benign first PE, 3) there is no unic massive and lethal PE but only multirecurrent PE, 4) a suspected PE has to be urgently proved or disproved and energically treated whatever its importance, 5) the decreasing order of specifity is : PA (to affirm) = PS (to eliminate) > Chest X-ray >> phlebitis, associated symptoms (pain + dyspnea + hemoptysis), etiology, EKG, hemodynamics, >> isolated sympto~ , blood gas, enzymes, lung scan (to affirm): 6) VS gives rarely more information than PS, 7) recurrency has to be proved by PA and not by PS, 8) PA has to be performed as freque~ tly as possible to establish the diagnosis, specify the site and size of the clots and control the venous iliocaval axis. Fibrinolysis is not indicated in minor embolism where there is no haemodynamic disturbance and is ineffective in chronic thromboembolism. Its use is limited to acute massive embolism (with or without shock) when it is the treatment of choice. In such patients accelerated resolution has been demonstrated (as compared with anticoagulant therapy) in a number of studies.

Improved survival has not been demonstrated and would require larger studies than have been practicable. Fibrinolysis is contraindicated when there is an increased risk of bleeding or where haemorrhage would have a high risk. Fibrinolysis may be contraindicated in supra-massive embolism where only embolectomy provides rapid enough reversal of the haemodynamic disturbance.

Fibrinolysis becomes less effective when the duration of embolism extends beyond 48 hours.

Bleeding complications are no more common than with anticoagulant therapy.

Whatever the initial treatment there is little if any late mortality or morbidity in survivors of massive embolism who received energetic treatment.

ACUTE RIGHT HEART FAILURE IN MASSIVE PULMONARY EMBOLISM. A. Bloch, O. Guinand, H6pital de la Tour, Meyrin-Geneva, Switzerland. Massive pulmonary embolism (MPE), obstructing more than 50% of the pulmonary circulation, leads to an increase in pulmonary artery (PA) resistance, PA pressure and right ventricular (RV) work. This will lead to acute RV failure (acute cor pulmonale) and decrease of pulmonary flow and cardiac output. Echocardiography has revealed several abnormalities in patients with MPE. I) Dilatation of the PA and right ventricle. The size of the PA appears to correlate well with the mean PA pressure. 2) Decrease of the left ventricular dimen~o~ Additional inspiratory decrease in left heart volume may lead to failure of aortic valve opening and therefore to paradoxical pulse. 3) Abnormally high ratio of right/left ventricular enddiastolic diameter. This index correlates well with the angiographic severity of the embolic obstruction in patients without prior cardio pulmonary disease. 4) Abnormal (paradoxical) diastolic motion of the intervenhricular septum. 5) Abnormal motion of mitral or tricuspid valve. Echocardiographic association of a large right ventricle, a small left ventricle and a paradoxical motion of the septum is very common in MPE and allows to explain the physical findings. Dyspnea, tachypnea and tachycardia are usually present in MPE. Other findings may indicate acute cor pulmonale: accentuation of the pulmonic closing sound, RV diastolic gallop, elevated jugular venous pressure with large A waves. These abnormal signs and symptoms Will progressively disappear with the spontaneous or therapeutical removal of pulmonary embolic obstruction. Dept.cf Thoracic Surgery, Karolinska hospital,Stockholm, Sweden. Chronic pulmonary embolism with pulmonary hypertension and cot pulmcnale can,after intense functional and anatomic investigation,be treated succesfully with TEA of the pulmonary artery. Perfusion scintigram is for screening a simple and informative procedure.Pulmonary angiogram is accurate for anatomical localisation,and angiogram of the bronchial arteries important to assure a ~ell functioning periferal circulation of the pulmonary vessels. Heart catheterization is necessary to evaluate the degree of pulmonary hypertension,which is the most important prognostic variable for succesful surgery, the better prognosis the lower pressure. The cases of three young women with chronic pulmonary embolism with systolic pulmonary pressure of three different levels about 25,80 and125 mm Hg respectively were all operated with TEA at clinic of Thoracic surgery,Karolinska hospital,Stockholm,with results corresponding to preoperative pulmonary pressure. The importance of early diagnosis before severe pulmonary hypertension has developed is stressed and the value of adequate diagnostic methods is illustrated. Operation is to be done with cardio-pulmonary by pass stand by. With proper investigation and treatment in cooperation with surgeons,cardiologists,physiologists and radiologists the prognosis f~ chronic ~ulmonary embolism with pulmonary hypertension and cot pulmonale encouraging for many patients,where although optimal pharmacological treatment,pr0gressive heart and pulmonary incompetence develops. The universally accepted classification of Dissecting Aortic Aneurysms defined by De Bakey et al (1965) will be described and two others mentioned. The role of diagnostic imaging methods, particularly radiology, ultrasound and computed tomography in determining the presence of a dissecting aneurysm of the aorta and in demonstrating its site and extent, including involvement of branches of the aorta, will be discussed and illustrated. A review of 15 Gases of surgically treated dissecting aortic aneurysms will be presented. Reference: De Bakey et al (1965 

AORTIC DISSECTION: NATURAL COURSE AND MEDICAL MANAGEMENT. P.C. Baumann, Medical Intensive Therapy Unit, University Hospital, CH-8091 Zurich, Switzerland

The mortality of aortic dissection is very high and the death rate is approximately 83% one month from date of onset, if the course is not altered by treatment (I).

The prognosis is influenced by the localization of the dissection. Without surgical treatment it is extremely poor in patients with proximal dissections (type A or I/II by the DeBakey classification) with early rupture into the pericardium being the main cause of death. Medical management alone should not be considered unless there are serious contraindications to surgery.

Patients with distal dissections (type B or III) should be treated medically in the acute phase, if there are no major complications such as bleeding, obstruction of major arteries or continuing enlargement and pain.

If there is no hypotension medical therapy usually consists of a peripheral vasodilator (e.g. sodium nitroprusside) to lower blood pressure in combination with a beta-adrenergic blocking agent to reduce myocardial contractile force. Blood pressure is kept as low as possible. A diuretic and medications for pain and sedation may be added.

From 1975 to 1982 123 patients with acute aortic dissections (80 type A, 43 type B) were admitted to our Medical Intensive Therapy Unit. Therapy in the acute phase was purely medical in 64% with type A and in 79% with type B. Death rate within 4 weeks was 84% (A) and 29% (B). Death rate in those who were operated in the acute phase was 34% (A) and 56% (B).

RECRUTMENT OF PERSONNEL FOR I.C.U.E. Staub, Bern. This report concerns our experience in the I.C.U. in the Inselspital in Bern. In an I.C.U. recognized for specialized education some regulations exist concerning the minimum number of nurses, as well as the ratio between nurses in training and I.C.U. staff. We could only achieve satisfactory results after greatly increasing the number of trained nurses. The change from one I.C.U. to another is a great problem not only for trained I.C.U. nurses but even more so for I.C.U. student nurses, who have to complete their education in a second unit. Since unfortunately supply is less than demand, departing I.C.U. nurses with "capacity certificates"often must be replaced by untrained nurses in the I.C.U.The number of candidates for I.C.U. training courses is sufficient to satisty demand, but not enough to allow a selection of highly capable I. C.U. nurses.

PRACTICAL EDUCATION OF NURSES J. Grosgurin-Robert, Soins Intensifs de Medecine, Gen~ve. A group of head.nurses working in ten different size swiss hospit~sdevelopped a programme to achieve common practical goals which could be inserted into the educative programm of intensive care nurses. This programme will be presented.

PROBLEHS OF A TRAINING UNIT. M. HOESS, UNIVERSITAETSSPITAL ZUERICH, 8092 ZUERICH I would like to indicate the problems and trends of developement: I. Overstrain by the increase of dangerously ill patients, 2. Physical and psychical strain on the nursing staff, 3. Increasing technicalisation of the intensive care To point I: The nurses starting their intensive care training are often unable to cope with the hi3h demands in supervision. The only answer to these problems is a thorough and gradual introduction. I would like to consider in detail the introductory concept of the UniversitMtsspital Z~rich. To point 2: In the future the shift work will confront us with serious problems. It has to be possible to find specific solutions for special units. The continuous monitoring requires a maximum of concentration and sen~of responsibility. The daily confrontation with other peoples destiny must not turn into ones own fate. The constant confrontation with live and death requires stzong personalities. To point 3: The technical developement has to be critically observed by the staff since a continuos supervision and control of the equipment means security for the patient. Technology is only ment to serve as an aid to do the daily work. 1~e should open our intensive care units to the public so that critical persons too may get an insight of the security described by patients who have been in an intensive care unit. To ~ive the whole staff the possibility to take part in the developements, direc~ves if necessary definition of position, should be worked out. 0nly he who accepts the challenge can look forward to the future.

Afin de pouvoir am@liorer et unifier sur le plan national la formation en soins intensifs et r~animation, l'Association suisse des infirmi@res et infirmiers (ASl)en collaboration avec des m~decins repr~sentant 5 Soci~t~s m~dicales (soci@t~s suisses d'anesth6siologie et de r~animation, de chirurgie, de m@decine interne, de m~decine intensive et de p@diatrie) ont ~labor@ un r~glement et un programme de formation qui entra en vigueur le ler janvier 1973. Une Commission pour la formation en soins intensifs et r@animation fut nomm@e. Elle comprend I0 membres : 5 infirmi~res repr6sentant I'ASI et 5 m@decins repr@sentant chaqUe soci~t@ m@dicale sus-mentionn6e. En outre, sont revus r@-guli~rement: la reconnaissance et contr61e des centres de formation, d~signation des experts aux examens, ~tude de proposition concernant un changement du r~glement. Les modifications propos@es par la Commission sont @tudi@es par les 5 Soci6t~s m6dicales concern~es et par I'ASI. Le r~glement d6finit : les conditions d'admission, le but, la dur6e, le d@roulement et le programme de la formation th~orique et pratique, les examens finaux. L'infirmi@re ayant achev@ sa formation avec succ~s obtient le certificat de capacit@ en soins intensifs et r6animation.

STAFFING AND EDUCATION OF NURSES IN OXFORD. C.E. Church SRN.DipN(LON)Nursing Officer ITU/CCU John Radcliffe Hospital, Oxford, England. The Intensive Therapy Unit (ITU) in Oxford is classed as a general Unit for there is a separate Coronary Care Unit and paediatric ITU. The ITU is set up with 8 beds, 2 of which are high dependancy and medically the Unit is run by anaesthetists. There is a small commitment to cardio-thoracic surgery, otherwise general medicine,surgery, trauma and renal cases are admitted. To cover these 8beds there is a total of 47 trained nurses plus 2 Student nurses in basic training; for a period of 1 month, and post basic course nurses.These are people undertaking National recognised Courses in various specialities; Intensive Care, Accident and Emergency Nursing and Anaesthetic Nursing. There are 6 ITU Course members, of which we have a minimum of 2 and maximum of 6 at anyone time.The other 2 Courses never send more than 2 members each at anyone time. A part from these various other people from different areas come for a period of orientation ie Renal Course, theatre assistant trainees.Of the 47 permanent team, 6 are Sisters, 31 are Staff Nurses (3 yr.general training) and I0 Enrolled Nurses (2 yr.general training). The Unit is covered with two-eight hr. shifts overlapping during the day and an I1 89 hr night.There is a minimum of 8 nurses per shift plus basic and post basic learners.The average lenght of stay of staff is about 7 month,that means a continual recruitment programme, stepped up just before the Summer as this is when the main drop occurs 70% of the permanent team hold a nationally recognised ITU nursing certificate, the other 30% are made up of people with little or noexperience in this speciality, consequently the teaching input has to be high, and the overlap period in the afternoon is an ideal time to do lectures.Education of all staff not only Course members and learners, but experienced staff must be an ongoing process to train and maintain motivation.All staff, commen-cing employment undergo a 3 day general hospital orientation.Staff within the Unit then work with experienced members of staff for the first two weeks and have lectureson basic monitoring and ventilating equipment technique. Following this,those staff programme involving lectures and clinical teaching on basic ITU knowledge and skills ie Care of Ventilated patient, from both staff nurses and Sisters. Intravenous additives and arterial line assessments are taken by all staff before they are able to administer drugs or take blood specimens from arterial lines.Staff are predominantly trained by the bedside with one to one teaching which is ideal for gauging knowledge and adaptibility of the learner. Impromptu lectures are given by all staff including Doctors, depending on work load on the Unit.Available each day are lectures for staff to attend; general hospital sessions,lectures from all disciplines specifically for ITU nurses,XRay conferences; to short talks given by the staff nurses themselves. The nationally recognised ITU Course is run in Oxford and whilst they have a clinical tutor attached to the Course,the Sisters are very much involved in teaching members, and all Course lectures are open for all Unit staff to attend. The Health Authority periodically run Courses mainly for Sisters to extend their managerial skills ie Art of Assessing, Interviewing Skills and every 6 months all staff are seen, to be given career guidance and counselling, so they and their seniors are aware of their progress, capabilities,new avenues to be explored,and where they are going in the future. Since no satisfactory rules exist on how to staff adequatly an I.C.U.,a group of nurses and physicians was created in order to attempt to establish objective criteria for staffing the I.C.U.For that purpose a protocol was created to produce data concerning the time involved in different nursing care procedures,administration and education.

PROBLEMS AND SPECIFICITY OF AN INTENSIVE CARE UNIT IN A REGIONAL HOSPITAL.F.Payot R.H.,H6pial R#gional,2900 Porrentruy, Switzerland. The organization and topography of an I.C.U. in a Regional Hospital are presented. The following problems are considered :-Guidance for the team,-personal responsability of nurses facing emergencies,-difficulties in recruiting staff nurses for remote and non-urban areas.

NURSING CARE PROBLEMS IN PATIENTS TREATED BY LAPAROTOMY L.Douchamps-F.Turc-P.Meier-S~ins Intensifs de Chirurgie, Gen~ve. Presentation of a patient treated with laparostomy for several weeks and the specific problems, their implications on nursing care. Aims, actions,evaluation of the treatment. Recent advances in medical technology have increased our ability to treat many illnesses included those involving previously fatal physiologic probl~ms. New questions have come about when to use an how to determine value of these new developments. A classification of patients based upon the acute severity of their illness would be very helpful not to predict survival for individual patients, but to provide a precise estimate of risk of death for groups of acutely ill patients and, perhaps, a precise estimate of the staff requirements.

Two methods can be used to built a S.I~ : the discriminant analysis and the subjective methods. In all cases the SI will be used by other hospitals only if it has reliability, Validity criteria, data requirements, mathematical consistency.

For ICU patients special SI are currently used, such as coronary prognosis index, or the indexes for burns. Several general SI for all patients have been proposed such as. the complication impact index, the condition index score, but rarely employed. Only two indexes till now are used by several hospitals in several countries : the TISS system and the APACHE.

The TISS system (therapeutic investigation scoring system) of D. CULLEN is not strictly speaking, a S.I. Nevertheless it can be used for a same technologic level as an out come prediction for groups of patients.

The APACHE (Acute Physiologic and chronic health evalua tion) has been proposed by W.A. KNAUS and others : it is reliable, valid and has been used to compare different ICU in USA as well as in a french US comparison. Further research should be directed at mathematical and statistical techniques to maximize the amount of observation that can be obtained from a lesser data set, The study was undertaken to identify dimensions or motivational orientations, that underlie nurses' participation in a critical care continuing education nrogram (CE). ~.~ile the need for continuing education in the health professions is generally accepted, the value of legally mandating specific numbers of hours of CE has been debated, with few states adopting such requirements.

The increasing investment of health care dollars in nursing CE requires a continuing evaluation of the objectives of these programs and their impact on patient care. Previous research, viewin~ participation as motivated behavior rather than a phenomenon in itself, identified motivational dimensions of adults participating in education programs as goal-oriented, activity-oriented, or learning-oriented. Subsequent research has identified other motivational orientations.

It was hypothesized that critical care nurses would differ in motivational orientation from the general nursin~ nopulation due to a DerceDtion of greater or quicker obsolesence and therefore a greater need for more frequent or even continuous updating. The study sampled participants in a two day critical care symposium sponsored by a professional organization which focused on both clinical content and personal/ professional development.

93 questionnaires (32% of the participants) were returned. The Education Participation Scale is a self report instrument in which learners rate the influence of 56 motivating factors in their decision to participate in a CE program. Varimax orthogonal rotation was used to analyze responses, findin~ eight factors meaningful.

Chronbach's alpha method was used to determine internal consistency and Pearson product moment correlation was used to analyze relationships of factors to demographic variables. While the orientations identified indicate that participation is primarily for professional reasons, job competence and credentials acquisition did not "load." No correlations with personal data characteristics were identified.

Subsequent study is currently in progress to afford extended data with emphasis on critical care clinical foci.

Though the interdisciplinary character as well as the fact that Intensive Care Medicine (ICM) eventually might require full time engagement become evident more and more, besides Australia the formation of a (sub)specialty has only been decided upon in outlines in the USA (4), first certification not to be awaited before July 1984 (5). In most other countries (cf. subsequent discussions) there are still but negotiations if any; Austria did include ICM obligatory into the 4 yrs postgraduate scheme for anaesthetists by Sept. 1981. -A similar discomposure is to be observed as to core curriculum (necessary knowledge and skills, not to speak of attitude and experience), special training in emergency (and disaster) medicine, and to prior qualification. Concerning time of training a rising tendency cannot be ignored: Programs of more than 1 yr's duration were offered in 38% for 198o compared to 66% of all USA fellowships in 1982 (2,3). -We (i) have tried to develop a "step by step" training scheme adapted to individual needs reaching from introduction tc training~pecialists. According to scattered experiences "basal" training does improve general inhospital acute care substantially. As ICU nurses are trained now largely on a 1-2 yrs basis unsupervised medical activities or appointments in ICM without equivalent special training are hardly justifiable by today. The term "Intensive Care" usually stands for the practice of active medicine in the Department of Surgery, Medicine or Speciality. The name "Critical Care" (or Reanimation"b in France, defines similar medical practice usually performed in a Department set up to receive a more acute or polyvalent pathology. In reality, there exists a confusion between the vernacular and the facts : "Cardiac critical care", "Hematological critical care", "Surgical critical care"... This ambiguity persists in the teaching and the professional training conditions. "Medical Critical Care" (R~animationMedicale Polyvalent~ is a subject taught far a period of three years in certain Universities, although not leading to a Notional Degree. Anesthesiology is one of the great specialities which provides intensive core and/or critical care. It is recognized by the University. There are three years of training at the end of which o degree in "Anesthesiology-Crifical Care" is obtained. This recovery is mostly employed in traumatology or post-surgery, and often polyvolent.

The qualification for "Medical Critical Care" may be obtained without the specialized exams if the University Hospital residency boards'are passed and tWO years in the approved departments falfilled. Anesthesiology -Critical Care residents may obtain on equivalency of the National Degree in "Anesthesiology -Critical Care" and are qualified in Medical Critical Care after three semesters in the approved Departments of Anesthesiology and four semesters in Critical Care and Intensive Care Departments.

Thus there is major discrepancy between the theoretical and clinical qualification of recovery experts. From 1984 a reform will allow specialists to undergo more homogeneous training.

Hillman KM, ICU, Charing Cross Hospital, London, UK.

Intensive Care Medicine is a specialty which encompasses the acute aspects of everyon e else's specialty. The expertise necessary to treat critically ill patients comes from adequate training of the doctors and constant exposure to the patients' problems. Anaesthetists, Physicians and Surgeons are not specifically trained in Intensive Care Medicine, and because they are specialists in other branches of medicine cannot spend most of their time caring for seriously ill patients. Thus one either becomes proficient in one's parent specialty (Anaesthetics, Medicine, Surgery) or in Intensive Care Medicine. Intensive Care Medicine has become too complicated and demanding to be treated as a hobby. It is a specialty requiring cer tain Anaesthetic knowledge and skills, such as applied physiology, rapid reaction to problems, ability to perform a wide range of invasive procedures and familiarity with monitoring and support equipment. It also requires certain Medical knowledge and skills, such as a broad knowledge of conventional medicine, and the ability to interpret diagnostic data. However, further specialised knowledge about bacteriology, radiology, surgery, resuscitation, and administration is required. More importantly, knowledge of acute organ failure and multlorgan failure is essential for an Intensive Care Physician. This is poorly understood by most conventional physicians and anaesthetists. Furthermore, the specialist Intenslvist must ideally be familiar with research principles and current journals, as much of Intensive Care Medicine is being defined now. It doesnlt matter who practises Intensive Care Medicine (other than a committee); as long as he or she devotes most of their clinical time to working in an Intensive Care Unit.

Care Medicine, Academy of Postgraduate Medical Education of the GDR, Berlin, GDR Anaesthesiology/Intensive Care Medicine is one of 32 specialities licensed in the GDR. The medical curriculum at the universities schedules this field in the first, fourth and fifth year of studies in which the first year is given to an interdisciplinary lecture series on "Emergency Situations" and ',~irst Aid". The fourth and fifth year is reserved for lectures, seminars and practical courses of 30h re, and 45 hrs per year, respectively. The body of knowledge in Intensive Care Medicine forms a substantial part of the State Examination in Medicine. For quality reasons the postgraduate training in amy speciality follows nationwide standards. The standard for this speciality were worked out in close cooperation between the Chair of Anaesthesiology/Intensive Care Medicine of the Academy of Postgraduate Medical Education, and the Society of Amaesthesiology and Intensive Therapy of the GDR. All these standards are subject to final approval by the Ministry of Health. The training takes place in departments and hospitals which are especially appointed for that purpose, and the period of training is 4 years for all specialities. Certification as a specialist is only achievable by passing a final examination before the Board of Examiners of the Academy. After qualification a system of continuing education follows to keep the level high during all the professional life. "]IqTENSIVIST" : TOWARDS A RECYXINIZ~3 SPECIALITY JL.Vincent, Department of Intensive Care, Eramne Hospital, Free University of Brussels, Belgium.

Intensive care was born frc~ several specialities, including internal medicine, anesthesiology, surgery and pediatrics. The will of these primary specialties to keep critically ill patients within their own deparhaent has impeded the development of multidisciplinary intensive care units. These specialities have also feared that recognition of intensive care medicine as a distinct entity would anloutate their own cfml0etence. Hcm~ver, a multidisciplinary approach to an unique "intensive care" improves both conprehension and management of critical states. The introduction of large autoncmous departments of intensive care has been facilitated by the construction of new hospitals. Nevertheless, critical should be better defined throughout the world. Responsabilities in intensive care, coronary care, neonatology and even emergency medicine can be very different not only from town to town but even scmetimes frcm hospital to hospital within the same to~n. In our 31-bed medico-surgical department, internists, anesthesiologists and surgeons who are in rotation represent 8 of the 16 manbers of the medical staff. They are eager to spend this training period and very concerned about Instruction in our discipline. Each patient has one "intensivist" directly responsible for his care, in good relationship with consultants and primary physician. Specialization in intensive care medicine is rendered essential by the coniolexlty of ccmlorehension and management of critical illnesses more than by purely technical aspects of their care. We suggest specialty in intensive care medicine to be recognized to those who have had a two year period of training in a deparlxnent of intensive care, after full training in one of the primary specialties. SHOULD INTENSIVE CARE MEDECINE BE AN INDEPENDENT SPECIALTY? J.P.Gardaz,P.M.Suter,Department d'Anesthi@siblogie,H6pital Cantonal Universitaire,GenOve. The structures of intensive care medecine(ICM)vary widely from country to country.ln France,Japan and South Africa, for example,there are university departmen~of critical care medecine headed by full professors in ICM.It has become a primary speciality of its own in Spain and Australia.The United States has opted for a subspecialty in critical care medicine rather than a primary discipline,whereas others countries like the United Kingdom,Switzerland do not recognize ICM officially. These differences encountrered in the academic and medical structures reveal the complexity of the problem. Intensive care medecine is a "multidisciplinary"endavour crossing traditional departmental an speciality lines and it is conceivable to create autonomous departments of multidisciplinary ICM runned by primary specialits in intensive therapy.Although we recognize that there is a need for special knowledge and skills on the part of those providing patient care in intensive care units,we do not support the specialty of"intensivist" as primary specialty for the following reasons:First ICM physicians should have a solid training in Internal Medicine and Anesthesia(or Pediatrics for pediatric ICM)before training and practicing ICM.Second, at the present time there is no specific training programme,no uniform career structure nor academic future ensured in the majority of countries of Western Europe. Furthermore intensive care medicine as primary specialty can be hazardous,because, outside intensive care units,there is no professional alternative for intensivists who whish or have to change their medical activity at a later time.This point is of particular importance because good ICM needs dynamism, fresh ideas and a great physical and psychological commitment, factors found usually in greater quantity in people at 30 than at 60 years of age. In Intensive Care Medicine (ICM), a rather young branch of medical sciences, almost day b~ day sets of problems are encountered, some of them "shouting with urgency" for rapid answers. Systematic research efforts in this field, however, are not yet taken for granted. A cumulative survey for 198o/$2 comparing 3 leading anaesthesiological (A: Anesthesiology, BJA, Anaesthesist) with the same number of Intensive Care periodicals (I: Critical Care Medicine, Intensive Care Medicine, Intensivmedizin) revealed a striking difference in research activities: 64.9% of papers in (A) vs. only 31.4% in (1) were devoted to scientific investigations (studies related to morbidity, epidemiology, assessment of economic and personnel requirements not being taken into consideration). In our own institution 56.8% of ICM-papers referred to research programs. -Besides that a wide scatter in intensity of interest is to be seen: Subjects like cardiovascular and/or pulmonary pathophysiology etc. covering 45% of all publications are opposed to poorly if not insufficiently (<1%) treated topics like e.g. renal function or immunology. -In order to enhance interest and to overcome the abovementioned gaps being eventually also detrimental to ICUpatients organisational (access to experimental departments, research programs (i), and computing facilities; provision of biomedical and laboratory technicians working IN the ICU etc.) as well as other motivating measures (editorial policy of journals, improvement in career structures etc.) should be promoted. Much of our current knowledge and clinical practice in Intensive Care is based on animal experiments and conventional medical principles and may not be relevant to treating critically ill patients. As an Intensive Care specialist, I have tried to document the three most common important problems I face each day; on the basis that this is where our research should have priority. My first is getting enough money to run the ICU, and being able to justify to my colleagues and the community that my treatment makes a difference and that the difference is worthwhile. Because this is one of the hardest areas to define and from which to gather meaningful results, it does not mean that we should not collectively attempt it. It is imPortant now and crucial for the future. My second problem is how to assess, treat and predict outcome after cerebral damage -from whatever cause. I often feel I am working without hard facts or guidelines, and too often with"personal experience" and instincts. The third major problem I have in clinical practice, and not necessarily the least important, is how to prevent or treat septicaemia. I feel current recipes temPorarily improve the "numbers" but do not substantially affect outcome.

Our research should be linked to our clinical challenges. ICUs treat a small number of patients and we have little control over the type of admission. The patients often have complicated multiorgan diseases. To answer the questions I have posed we need large numbers of patients, changing only one variable, and Using controls. To achieve this we need multi-centre; national and international controlled trials. Furthermore we should all collect the same information about our patients; preferably on common computerised record-keeping systems. This would expedite our research and give us meaningful comparisons on cost, outcome, workload and prognostic information.

In the intensive care unit, both the large variety of critical illnesses and the availability of sophisticated material and motivated personal foster clinical investigation. Spectrum of research activities is very broad, going from cardiovascular, respiratory or metabolic studies to search for technical developments, establishment of prognostic indexes or discussion of ethical problems. Studies can focus on gross clinical abnormalities as well as on precise cellular or biochemical mechanics underlying pathologic states. Multidisciplinary approach to intensive care and collaboration with specialists in basic research should be encouraged to improve the quality of the investigation. Systematic collection of datais essential to yield comprehensive clinical studies. Some forms of treatment, such as fluid challenge or PEEP administration, should be standardized. In our department, a minimum daily set of laboratory data is defined, including arterial lactate concentration, colloid osmotic pressure and urine osmolarity. A simple, mobile cart has been designed to perform routine hemodynamic studies at intervals in patients with invasive monitoring. Clinical practice raises questions that need an experimental support to be answered. Clinical investigation in the intensive care unit should be closely linked to an animal laboratory. Difficulties in the development of reproducible experimental models of acute organ failure should be challenging rather than dissuasive. We have set up a dog laboratory for hemodynamic studies of acute states, for which fonding has been essentially supplied by external support, in a spirit of positive collaboration between the private industry and the university. (RNV] are increasingly used to evaluate patients (pts) with acute myocardial infarctdoe (ANI). These techniques are non invasive and may be applied to critically ill pts at the bedside without appreciable extra risk or discomfort. Approximately 80 percent of the pts with transmural or nontransmural AMI can be detected by Tc-PYP if the imaging is performed 24 to 72 hrs aTter the onset of the symptoms. Tc-PYP has been utilized in localizing the site and deter mining the extent of AMI. The "doughnut" pattern is associated with a large incidence of subsequent congestive heart failure and death. T1-201 performed within 24 hours of AMI has been useful for recognizing, localizing and roughly sizing areas of infarction and surrounding ischemia, The extent of TI-201 defect in an initial image may have important value, For separating high-risk and low-risk subgroups of hemodynamically stable pts with AMI. Short and long term prognosis after ANI appears to be determined essentially by residue] ventricular function. RNV provides the means for assessing the evolutionary changes in regional performance after AMI. In pts with complicated AMI and cardiogenic shock a combination of T1-201 and RNV should be considered to identify those with massive LV damage from those with residual ischemia or right ventricular damage. Ouantitative evaluation of regional LV motion after pharmacologic or surgical treatment is pts with AMI may provide useful information regarding the efficacy of a variety of interventions.

PATHOPHYSIOLOGY OF ACUTE RENAL FAILURE. K.Thurau, Department of Physiology, University of Munich, Munich, FRG In order to reprocess the chemic'al composition of body fluids, the kidneys deliver a large volume of filtrate from plasma to the tubular epithelium, which then forms the final urine by selective reabsorption of filtered water and substances and the secretion of substances from the blood into the urine. Both processes, filtration and reahsorption/secretion, are powered by different energy sources, filtration by myocardial metabolism and reabsorption by tubular cellular metabolism. Normally, the large reabsorptive capacity of the tubules allows the kidney to deliver 100-150 ml fihrate/min to the tubules which have inherent discriminatory activities in the selection of filtered constituents for reabsorption or excretion. This high turn-over of extracellular fluid serves the important function in the kidney of continuously reprocessing the plasma. It contains an inherent danger -that of severe fluid loss if the tubular reabsorptive mechanism should fail due to nephrotoxins or renal hypoxia. There is evidence that the primary site of action of nephrotoxins and hypoxia is located in the tubular cells. The pathobiochemistry of the kidney in acute failure includes defects in the membrane-bound properties for fluid reabsorption, such as changes in permeability and active salt transports. These initial cellular defects may produce secondary phenomena such as activation of the tubulo-glomerular feedback, the intrarenal renin angiotensin system, a fall in GFR, tubular occlusion, vascular constriction etc. The foremost "secondary phenomenon" appears to depend critically on the type of renal insult, its severity and time course.

A rational treatment of acute renal failure, aimed at treating the causes rather than the symptoms, should restore the disturbed celIuIar functions. Since, at present, there is no specific cellular treatment known, dialysis and careful fluid and metabolic balance constitute the only possibility of supporting the patient until spontaneous healing occurs.

EARLY DIAGNOSIS OF ACUTE RENAL FAILURE. H. Favre, B. Ody, Departments of medicine and radiology, Hopital cantonal universitaire, Geneva, Switzerland. The causes of acute renal failure (ARE) have to be early recognized in order to put the patients into one of the 3 following categories : prerenal, renal or postrensl. The two first situations could be separeted on the basis of several urinary indices among which the most efficient is the fractional sodium excretion (FeNs). This parameter classifies correctly 99 % of the patients. FeNa allowed, in our personal series, an early diagnosis of prerenai or renal ARF in 50 patients at the time where serum creatinine was rising from 120 + 30 to 250 + 50 umol/1. Limitations to this index concern mostly patients with nonoliguric ARF associated with salt retaining states. Some of these patients could be considered as prerenal while they are actually renal. In these peculiar setting, measurements of enzymuria could provide an accurate diagnosis.Radiologieal investigations have to be done in all patients in whom an obstruction of the upper urinary tract is suspected or could not be ruled out. Nowadays, the method of choice to demonstrate pyelocaliceal dilatation is sonographic examination. Its specificity and sensitivity is very high. Rare false positive results are due to a very recent obstruction or to large calculi whereas false positive are due to prior dilatations or to constitutive large cavities without obstruction:at the time of ARF. In cases of dilatation, sonography must be completed by intravenous (antero or retrograde) ureteropyelography to determine the nature and the location of the obstruction. Conclusions. Use of FeNs and sonography provides an early diagnosis of the causes of ARF and permits the choice of the appropriate therapy which prevents a further degradation of renal function and may change the outcome of the patients. Exceptional causes of ARF such as acute glomerulonsphritis or vascular obstruction need a different approach based on clinical ground. Therapeutic modalities in acute renal failure J.P. Wauters, Division of Nephrology, University Hospital, Lausanne, Switzerland. The classical treatment of acute renal failure (ARF) i.e. peritoneal dialysis or hemodialysis once uremic symptoms appear, has now been largely replaced by a more aggressive and individualized therapeutic approach.The importance of an adequate caloric and protein intake has been better understood. At least 35 Kcal/kg and 0.5 gr/kg of protein should be given daily. Therefore the management of fluid and electrolyte balance is not always possible by conservative means. It is presently admitted that once oliguric ARF is established there is no further reason for diuretic or mannitol administration. Among the substitutive therapies, peritoneal dialysis became easier through the experience gained with the CAPD technique. Hemodialysis can now be performed with the use of bicarbonate in dialysate,sequential ultrafiltration and/or volumetricaly monitored ultrafiltration during dialysis. Other substitutive therapies have been proposed : I) hemofiltration can be usefull when cardio-vascular instability precludes the use of conventional dialysis; 2) plasmapheresis by centrifugation or filtration techniques has been recommended when ARF is due to acute intoxications by poisons which are largely bound to proteins; 3) spontaneous femoral arterio-venous hemofiltration appears as an easily applicable bedside technique allowing to overcome simultaneously the fluid balance and uremic problems, without the continuous need for dialysis equipment and staff. Despite those advances, the complication rate during ARF remains high and is still responsible for the persistently high mortality.

PERSISTANCE OF THE FETAL CIRCULATION. D. ~ULIN, Soins In tensifs P~diatriques, Clin. Univ. Saint-Luc (U.C.L.), iO Avenue Hippocrate, B-1200 Bruxelles. In 1969, Gersony, Duc and Sinclair reported the fatal out come of two term neonates who had severe hypoxemia within the first 24 hours of life. Carefull investigations and post-mortem examination had failed to demonstrate any underlying abnormalities. In order to explain these observations, the authors suggested that a suprasystemic pulmonary vascular resistance may have create a right to left shunt through the foramen ovale and/or the ductus arterio sus and therefore decreased dramatically the pulmonary perfusion. This hypothesis was thereafter applied to various situations of neonatal cardiopulmonary distress with severe hypoxemia and where no cardiac defect could be detected. Different conditions have been demonstrated to participate in increasing the pulmonary vascular resis tance I. increased blood viscosity 2. inadequate development of the pulmonary vascular bed 3. pulmonary vasoconstriction and 4. extrinsic or intrinsic pulmonary vascular obstruction. At birth, the term neonate has functionnally closed fetal shunts ; he also has very reactive and mu$culated pulmonary vessels. This makes him particularly prone to develop persistent fetal circulation, even more so as pulmonary vascular smooth muscle can be further hypertrophied by chronic hypoxemla or premature ductal closure during fetal life. In assessing the neonates who pre sent severe hypoxemia resistant to optimal oxygen therapy~ echocardiography seems to be the most appropriate procedure. It helps i. to rule out a structural cardiac defect 2. to objectivate concomitant cardiac dysfnnction and 3. to document pulmonary hypertension, pulmonary hypoperfusion and/or right to left shunt. Therapy should first try to suppress any factors favoring pulmonary vasoconstric~ tion. If inefficient it may be completed by the earefull use of vasodilator drugs and/or respiratory alkalosis.

EXPERIMENTAL MODEL. M. Solca, T. Kolobow, R. Funagalli, P. Arosio, A. Pesenti, L, Gattinoni. Ist. Anestesiologla e Rianimazione, Universit~ Milano, Italy, and Lab. Technical Development, NIH, Bethesda MD, USA. Fetal lambs are widely regarded as a useful experimental model of human RDS. We studied 115 animals between 125 and 134 days gestation (tern 147-150 days) to determine the optimal pulmonary nanagement for RDS prevention. In 70 cases we performed a cesarean section, exposed the head of the fetus, and intubated the trachea, leaving the fetus still connected to the nether through the intact umbilical cord and placenta. The lungs were inflated to a pressure of 35 cm H 0 for 5 2 secs for 3 times; we then measured the total compliance injecting known quantity of gas in the lungs and recording the resulting pressure. Thereafter 32 animals were immediately delivered and treated with state oF the art ventilatory care, while 38 underwent apneic oxygenation while still on placenta, until total conpllance -i-i exceeded 0.5 ml (cm M20) kg . Overall mortality rate was 0.31, totally for RDS, and was not different between treatment groups. Survivors had a total co| markedly hi~her than non-survivors --1 (0.28 ~ 0.11 vs. 0.12 ~ O.Ok ml (cn H20) kg , P < 0.001): this suggested a key role of pulmonary nechanlcs in insurgence and severity of RDS in premature fetuses. ~e were then able to dellneate a descriminant value of total compliance between the two groups: -i -I 2 0.205 ml (c~ H20) kg (~., 37.14, P~O.O01; predictive index 87%). It carries a ~ error of 1.7% and a ~ error of 25%: in our study none of the non-survlvors had a higher compliance, and only 19% of survivors had a lower value. Thls implies that we can predict within 2% of error which animal is at great risk for developlng RDS, and subsequent death, with a very simple and safe procedure: inflation of the lungs and measurement of total compliance. were tr@ated with mechanical ventilation with optimal PEEP determined by P/V curves of total respiratory system. 37 P/V curves were performed by slow constant thoracic inflow according to a previously described method. (J.C. MATHE Int. Care Med. 1982 -8 -246 Abst) . The mean optimal PEEP were 9.25 + 3.2 cm H20 (4 -17). Optimal PEEP were applied before H48 in 23 patients and after H48 in 7 patients. Mean exposure to Fi02 > 0.4. were signifiantly reduced in the first group (32.1 + 18.1 hours (5-84) vs 128.4 + ~5.2 hours (60-]92) p < O.OOI). 5 pneumothorax were observed and no bronehodysplasia. 2 imtraventricular hemorrhages (IVH) and 1 periventricular hemorrhage were investigated by means of 18 cranial ultra sonography and 3 venfiricular punctures. 5 patients died (2 IVH, i nosocomial infection). The mechanical ventilation with optimal PEEP in IRDS is a method well toIerated which reduces exposure to Fi02 > 0.4. and provides a successfull treatment for the most severely impaired patients. 29o,1982) . In order to evaluate the effect of nifedipine on pulmonary hypertension induced by infusion of oxidants and lipids, we carried out the present experiment, using isolated rabbit lungs ventilated with 5% CO 2 in room air and perfused st 50 ml/min with Krebs Hensleit buffer in a nonrecirculating manner.

Pulmonary hypertension was induced in 12 rabbits by infusing 69 mM tertiary butyl hydroperoxide t-Bu00H (mean ~Ppa = 13.41 torr ~ 3.52 SE) or a lo% Intralipid a infusion (mean Ppa = 13.66 9 3.13 SE). While reaching a plateau in the pulmonary artery pressure and still infusing the pressor drug, Nifedipine 2o ~g/kg were given. Significant decreases in pulmonary artery pressure (mean Ppa 7.54 i 1.33 SE) were noted regardless of the challenge. Results were evaluated by analysis of variance which showed a significant inhibitory effect by Nifedlpine of the vasoconstriction caused by both agents (p< o.ool).

We conclude that Nifedipine can inhibit pulmonary vasoconstriction induced by oxidants.

HANDLING AND RELIABILITY OF THE TRANSCLTfANEOUS PCO~ In the ~ measurements, the electrode must have a relatively ~igh core tenloerature (44-45UC) in order to achieve a sufficient hyperaemia and it has to be changed to different places every 2 to 4 hours. Some good experiences with lower temperatures for tcPcO2-measurements have been reported (i) which allow for a long-time monitoring without the risk of skin burns 6 We have examined how different tesperatures (41 and 44 C) influence the calibration stability, the repreductibility, the response time and the correlation to the Pco 2 in blood samples. The measurements were carried out on premature infants in serve-controlled incubators. We used two industrial prototypes (Draeger and Radic~eter) which were calibrated at 40 and at 80 Torr, fixed simultaneously on the thorax and conloared values with the Pco 2 in capillary or arterial blood sanloles. Except for the expected slower response t//ne there was no disadvantage at 41UC. With both temloeratures the tcPco 2 was largely increased for both prototypes ccnloared to the Pce2 in blood samples. There were strikingly large interi~dividual differences. This makes a general inset built compensation difficult. Since the introduction ofpositive pressure ventilation into clinical medicine, the parameters used in mechanical ventilation tended to mimic those seen in normal respiration. While in most instances this proves to be an effective method of managing respiratory failure, there are still many patients with severe parenchymal lung disease who have a hypoxemia, refractory to any manipulation of the respirator. (changing I:E ratio, wave forms, end expiratory pressure). With the introduction of HFV, a new era has opened up in ventilator management of patients with severe respiratory disease. High frequency positive pressure ventilation (HFPPV) includes all ventilators that can cycle at rates between 60 and 120 (facility available on some modern conventional mechanical ventilators). High frequency jet ventilation (HFJV) includes a group of ventilators that use a high pressure gas source ventilating at voltages at or below dead space at rates up to 5 HZ. High frequency oscillation (HFO) uses rates of 1-40 Hz. generated by high frequency sine wave. These technic have features and characteristics which are unique but the theory is essentially the same : patients may be ventilated at volumes substantially less than dead space, with lower mean airway pressures than seen with conventional ventilation. There are now nt~nerous studies in the literature to show that both HFJV and HFO are extremly efficient in terms of CO 2 elimination with low mean airway pressures. However, in order for this technique to make a transition fram a physiological curiosity to a therapeutic advance, it requires that it can be shown to iuprove patients with severe respiratory diseases (refractory hypoxemia secondary to either alveolar or interstitial lung disease). As oxygen transfer occurs during both inspiration and expiration, ,the major problem remains to open alveoli and keep them open throughout the respiratorycycle. HFV is an efficient way to achieving gas exchange in both normal and diseased lungs.

Preliminary results fram its use in hyaline membrane disease and ARDS are encouraging, but a large multicentre trial is necessary to prove its advantages over conventional ventilation in these diseases. There is no doubt that it is a major advance in the treating of brcncho-pleural fisulae, and in laryngeal and tracheal surgery where substantially lower moan airway pressures are of undoubted benefit. The proponents of HFV also advocate its use in thoracic and neurosurgical procedures where a motionless operating field would be of great assistance to the surgeon. Before widespread introduction into clinical practices, several major mechanical and safety features have to be ironed out. Htwnidification is a problem ~n to all types of high frequency ventilation. The ht~nidification systems currently in use are not entirely satisfactory, but there does not seem to have been a problem with mucociliary clearance in htmlan studies. The second major problem is that of patient safety. The very high gas flow used in both HFO and HFJV can lead to a rapid buildup of pressure within the lung if the gas outflow beeches obstructed. A rapidly responsive blc~off system needs to be incorporated to relieve the pressure within the system, if the gas exit beccn~s occluded. HFV stands on the verge of beccming a major advance in respiratory care. A well controlled study is urgently needed before it can be clearly demonstrated that it is superior to conventional ventilation in patients with severe lung disease. HFV will never replace conventional ventilation as the standard form of ventilator, but the day is probably not too far off when every ICU will have one or two high frequency ventilators to cope with patients With severe parenchymal lung disease. and must be well known by all physicians who care For neongtes. The clinical picture is usually a combination of strider, obstructive dyspnea, dysphonia, and swallowing disorders. But less typical features are possiblep such as apnea, cyanotic attaks, bradycardia or cardiac arrest. This presentation is a retrospective study on g80 new borns and infants with congenital laryngeal anomalies seen at our pediatric otolaryngology department from 6 Is, 1974 to 5 31th 1982. The diagnosis was : idiopathic congenital laryngeal strider (laryngom~lacia) in 348 cases (51 ~), 15 % of witch were severe, with respiratory distress and/or Feeding difficulties ; laryngeal paralysis in 161 cases (24 ~), among witch 56 were bilateral and 105 unilateral; laryngeal incoordination in 23 cases ; laryngeal stenosis in 79 cases (12 %), among witch there were 58 subglottic stenosis, 18 webs and 3 atresias ; subglottic hemangioma in 45 cases ; ary-epiglottic and hypoglossal cysts in 15 cases ; laryngo-esophageal cleft in 8 cases and bifid epiglottis in one case. For each o~e os these anomalies, clinical features, radiologic evaluation, endoscopic findings and treatment are reviewed. colonization with multiple organisms is frequent. 59 children incubated or tracheostomized, aged 1 day to lO years, and having X-ray evidence of pulmonary infiltrate were studied ~ith the technique of uncontaminated peripheral airway samples (UPAS) described by E.B. Matthew (Grit Csre Red 1977 ; 5 : 76 -81) . 41UPAS were performed,no complication was observed. 29 UPAS showed no growth though one pathogen was recovered from 17 trachea] samples. In 28 patients pulmonary course was fsvourable without antibiotic therapy. In i patient false negative UPAS was observed. In 12 UPAS pure culture of a single organism was obtained, in 8 cases th e psthogene recovered from tracheal samples was identiesl, in 2 cases it was different, in 2 cases tracheal samples were sterile. The organism's role in producinq pneumonia was confirmed by its isolation in 4 patients from blood or empyema fluid. UPAS in children is s single, safe and reliable technique. It shows that most pulmonary infiltrates in children with endstrseheal tubes have not a bacteriological etiology and do not need antibiotic therapy. Positive UPA5 gives s specific etiological diagnosis ; however its interpretation is more difficult when trseheal pathogen is identical. Nasopharyngeal specimens were collected from december 1981 to march 1982 in 85 hospitalized children (2 weeks to 22 months old, 45 under 3 months) with lower respiratory illness to study respiratory syncytial (RS) virus antigen by an indirect immunofluorescence technic (IFI). Collected by_nasopharyngeal aspiration, the specimens were flushed with phosphate-buffered saline 1,5 ml, then centrifuged at 1500 rpm for ten minutes three times. The cellular pellet was used to prepare smears, which were air dried and fixed in acetone for ten minutes. The slides were stained with anti RS antibody (Wellcome), then with antiimmunoglobulin fluorescent conjugate (Wellcome). A slide was considered positive when several epithelial ceils showed a granulous cytoplasmic fluorescence. Result could be obtainable 6 to 12 hours later. Thirteen children with no respiratory illness were used as a control group ; all samples were negative. 37 of 59 samples of children with wheezing associated respiratory infection were positive . Among lO children with bronchitis without wheezing, 5 were also positive. Three children had evldenced apneas of whom two had positive specimens. Overall RS antigen was found in 44 out of 72 children (61%) suffering of lower respiratory illness. A correlation with levels of total and anti RS virus immunoglobulins present in serum and in secretions is under investigation. This study demonstrate that RS detection using IFI is a convenient rapid diagnostic test to precise the etiology of infant viral respiratory illness and to decrease antibiotics prescriptions. We have now extended this approach as a rapid diagnosis technic of RS and others potential viral agents of respiratory illness in children. The bacterial quantification is difficu!t in small samples. The comparison between cultural characters of tracheal secretions and pulmonary biopsy may be an reliable approach for the diagnosis of the pulmonary infection. This study included 22 neonates and infants who died from different diseases. They had been intubated few hours to 57 days. The tracheal secretions were collected for culture twice a week ; the last sample collected on the day of death. Pulmonary tissue specimens were obtained by needle puncture immediately after death. Pathology has been known for 19/22 cases. All positive cultures on the day of death were monobacterial. -12 pulmonary culture were negative ; among them, 7 patients had no bacteria in their tracheal secretions and one germ was found in 5 patients. In these 12 cases, no histologic sign of infection has been reported. -10 pulmonary culture were positive ; among them, in 7 cases the same bacte ria was found in tracheal secretions (with 5 times, histological signs of pulmonary infection] and in 3 cases a different germ was isolated. The bacteriological concordance between pulmonary tissue specimens and tracheal secretions was present 14 times. In 8 cases no concordance was observed. This study allows to conclude that : -absence of germ in tracheal secretions is a favorable factor, statistically likked with an absence of pulmonary infection ; -presence of germ should be followed by clinical, biological and radiological investigation looking for an pulmenary infection. Their absence with presence of a germ correspond to a colonlsation of the tracheal tube.

ADULT RESPIRATORY DISTRESS SYNDROME (ARDS) AND INFECTIONS.

J. Pfenninger, University Childrens Hospital, Berne, Switzerland

The ARDS is a complex sequela of shock, trauma, sepsis, near drowning and many other insults. In a series of 20 children suffering from ARD~, selected on strict criteria, 7 had a septic process most often of intraabdominal origin l leading to ARDS. 8 children of the whole series died. The cause of death consisted most often in an unresolved basic medical und surgical problem, complicated by multiple organ failure and septicemia.

Thus, septicemia plays a major role in the genesis and mortality of ARDS.

The third aspect of ARDS and infections are the bronchopneumonias.

The main reason for frequent bronchopulmonary infections lies in by-passing natural defence mechanisms by prolonged nasotracheal incubation, exposing the lower airways and lungs to exogenous and endogenous microorganisms.

In our series these were most often Pseudomonas, Klebsiella and Enterobacter.

The infections complications of ARDS can be minimized by using strict aseptic techniques, monitoring of the bacterial flora of the patient and the ICU, and a carefully selecte~ antibiotic treatment, if indicated. Attention should also be paid to adequate nutritional support which may help to improve the host defence mechanisms. Four infants less than two months of age were admitted in the Pediatric Intensive Care Uni~ with severe impairment of general condition and respiratory distress, caused by CHLAMYDIA TRACHOMATIS (C.T) pneumonitis. Diagnosis could be quickly achieved by a specific micro-immunofluorescent method. Three infants required assisted ventilation but in all cases the treatment with erythromycin terminated shedding of C.T. Since the course of the illness is usually considered to be mild, it is of interest to be aware of this aspect of natally acquired C.T. infection. AZEMA-S. CHANTEREAU-J.C MERCIER-Y.BOMPARD -PARIS 3 patients classification systems developped for adult population since 10 years has been tried for pediatric patients. I) The Clinical Classification System CCS (I), a bedside assessment of patient stability and care needs dividing the patients in 4 classes of severity. II) The Severity Score (2) a weighted sum of 36 physiological measurements obtained from the patient's clinical record. We will propose a simplified score with only 14 variables. III) The Therapeutic Intervention Scorin~ System TISS (I) consisting in 84 interventions ranging 0-4 points.

The following informations can be obtained by collecting these dat~ : a) Classification of the patients and identification of inappropriate admissions in the ICU. b) Evaluation of results and comparison between different institutions, which otherwise, would be difficult because of heterogenous diagnosis and severity of illness : almost all the deaths are among patients class IV. In the i~ture, multicenter collaborative studies Should also ta-ke into account the organ system requiring the most attention -respiratory, cardiovascular, etc... c) a gaide for the decision of discharge by daily evaluation of TISS (3). d) evaluation of nurse patient ratio in the ICU, and definition of medical staff. For example : class IV patients require a I/I nurse/patient ratio. e) definition of the number of beds needed in a country from a Correct classification of patients who should be admitted in ICU.

All this aspects will be illustrated and discussed with personal data and with data from adult and pediatric literature (4,5 In this report we study 130 mechanically ventilated children observed during 1982. All had bacteriological tracheal or uncontaminated peripheral airways samples, weekly in acute, or monthly in chronic diseases. Among these children : -75 were intubated ; 21 younger than I year (mean age 4,6 ~ 2,5 months~ 54 older than I year (mean age 6,8 ! 4,8 years) -55 were tra~heostomised : 7 younger than I year (mean age 6,3 I 2,9 mQnths) and 48 older than I year (mean age 11,7 -6,5 years). Results are the same in the two groups of age. Among intubated children, we only treated 12 original pulmonary infections (Haemophilus 6, Pneumococcus 2, Staphylococcus 4) and 2 nosocomial infections (Staphylococcus). The airways of infants are usually colonized with a variety of Gram -organisms (Serratia, Pseudomonas Klebsiella, Acinetobacter) we only treated them by preventing retention of pulmonary secretions. Among tracheostcmised patients 6 had alternate antibiotics treatment for bronchopulmonary dysplasia. All these children have chronic diseases (neurological or pulmonary) and are ventilated all year long. During 1982, we treated 13 children for Staphylococcus, 6 for Pneumocoocus and 18 for Haemophilus. Conclusion : Adequate humidification of inspired gases, regular position changes, chest physiotherapy and frequent tracheal suctioning often avoid antibiotic treatment during mechanical ventilation. We only treat Gram + organisms and hemophilus. These infections often occur on an epidemic way. Patients who have suffered severe injuries and those with active sepsis occurring as a complication of either accidental trauma or of surgical operation show many similar metabolic features.

In these patients, for instance, the distribution of metabolic subs~rates in the plasma and tissues is altered. The changes in the carbohydrate and lipid substrates are clearest after accidental trauma but similar changes occur in sepsis, although the pattern is more variable. Basically, the stores of both carbohydrate and fat are mobilized so that it might seem that these patients have a choice of fuel oxidation. In fact, these very ill patients oxidize predomlnently fat. This occurs even when they are presented with extra glucose during parenteral feeding. T~mechanism of this metabolic effect is not understood but it may be related to those in the endocrine system following trauma and sepsis.

There are widespread changes in endocrine function in these patients. Starting with the activation of the adrenal medullary-sympathetic nervous system to give an increased catechelam]nergic effect, there arechangesin thevasopressin, growth hormone, ACTH, cortisol and Tnsulin concentrations in the plasma. The variations in the last two are particularly complex and, again, these are best seen after accidenta] trauma. Given the initial role of the adrenal medulla and sympathetic nervous system in the mobilisation of the stores of carbohydrate and fat the subsequent changes in cortisol and insulin maybe importantindetermining the metabolic features of this group of patients. The presence of insulin resistance after both injury and sepsis may be particularly significant. There is growing evidence that the decrease in tissue sensitivity to insulin is due to a post-receptor block and this may be important both in deciding the fuel oxidized by these patients and for the abnormal protein metabolism wiaich is also present.

CATABOLISM & NITROGEN SPARING IN THE CRITICALLY ILL PATIENT . Herbert R. Freund, Department of Surgery, Hebrew University -Hadassah Medical School, Jerusalem, Israel. The severly injured or septic critically ill patient sufferes a major catabolic insult leading to increased proteolysis and nitrogen loss, accompanied by a modified carbohydrate and fat metabolism. This special metabolic set-up of the critically ill patient dictates a modified nutritional support regime: i. Decrease calorie / nitrogen ratio. 2. Do not supply calories in excess of need. 3. Increase part of fat in total caloric intake. 4. Increase amino acid intake. 5. Increase amount of branched chain amino acids in total amino acid intake. 6. Decrease fluid load. 7. Apply metabolic monitoring if possible.

ENERGY REQUIRIK,~/~TS AND ITS SUPPLY Yvon A. CARP~NTIER, H6pital Saint-Pierre, Universit@ Libre de Bruxelles, Brussels (Belgium).

Accurate measurements of energy expenditure can be performed in man by direct or indirect calorimetry. The values obs by such measurements are very often below those found in the classical literature. In the absence of precise measure/rents, the energy expenditure of various types of patients is usually overestimated.

Hypercaloric parenteral nutrition with high glucose intake can induce severe metabolic coniolications in critically ill patients : hyperglycaemia due to insulin resistence, hyperosmolarity, essential fatty acid deficiency, but also respiratory distress due to high CO 2 production and increased incidence of liver dysfunction.

During the hypermetabolic phase of critically ill patients, endogenous and exogenous fat appear to be a preferential fuel for -at least -some tissues.

Adjusting the total calorie intake to the patients energy expenditure and introducing exogenous lipids in the TPN regimen significantly decreases the incidence and the magnitude of metabolic side-effects. The most important metabolic variations typical of a septic ill patient are as follow: glucid related metabolic perturbations, increase of body metabolism and rise of oxygen consumption, increase of protein catabolism and ureogenesis. These changes are the outcome of centrally controlled neuroendocrine incidents. They mobilize in fact an excess of endogenous substrates. These three kinds of disturbances escort in fact every trauma or important stress, but for the septic patient, they are characterized by their persistence and by the fact that unhappily most of the time, zhe evolution turn co be an uncontrolable multiple organ failure whose mortality is very high. Before infusing the necessary substrates to spare the essential proteins for a good body function, one must try first to suppress or inhibit the condition who maintains the s~ress and it's catabolic reaction what means ~o control the patient milieu, his adaption to stress and the pathological conditions responsable or whorsening stress and infection. The treatment must aim to perfuse and bring enough calories to cover the needs without causing suplementary work to the body by means of lipogenesis, glucogenogenesis and ureogenesis in ~articulary for the failing liver. In practice mos~ of the times one can only infuse a minimum of 125 to 150 g of glucose, i00 to 500 ml of 10% Intral~pid and i0 ~o 20 g nitrogen as a rich in branchchained aminoacids solution. Only repeated calorimetric measurements will allow co quantify the metabolic monitoring and to document the global metabolic care. Diet plays an important role in the treatment of diabetes mellitus and should be considered when parenteral nutrition is used. A complete nutrition solution with a calorie content consisting of 44% carbohydrate, 17% l-amino acids and 39% fat is recommended. In uncontrolled diabetes mellitus parenteral nutrition should be started after water and electrolyte imbalances is corrected and blood glucose level is decreased below I0 mmol/l and beta-hydroxy butyrate below 3 mmol/l. This can be obtained by a low dose insulin regime and rehydratation with a glucosefree electrolyte solution with 90 mmol sodium, 60 mmol chloride, 25 mmol potassium, 10 mmol phosphate, 0.5 mmol calcium, 1.5 mmol mangesium and 43 mmol malate within 1000 ml H20 and an infusion volume adjusted to the value of the central venous pressure. Total parenteral nutrition of acutelly ill diabetic patients containing 2.9 g carbohydrates, 1.1 g l-amino acids and 1.1 g fat with a caloriecontent of 26 kcal (110 kJ)/kg KG/d has been under current investigation. The results showed that the blood glucose level is maintained well by a continuous intravenous insulin dosage of 3 + 0.6 IU and no deviation of osmolality and serum lactate occured. The blood gas analysis remained unchanged. The triglycerides concentration was 0.8 + 0.12 mmol/l at the beginning and reached a steady state-of 1.3 + 0.26 mmol/l. The beta-hydroxy butyrate concentration ~howed a transient increase from 0.6 + 0.05 tO 0.9 + 0.22 mmol/l and afterwards decreased to the normal range. The intravenous infusion of a standard amino acid mixture caused no derangement of the plasma amino acid concentration during adequate supply of insulin. Therefore it can be concluded that insulin treated patients with diabetes mellitus do not need a special amino acid mixture. renal, and nutritional support are at the disposal of the intensive care units, the failure of multiple organs (MOF) has been identified not only as a disturbed function of some organs but as a syndrome-like entity with an expected mortality of more than 50%. Characterized by its progressive nature MOF may end in a complex reduction of nearly all organ functions with an at last untreatable cardiopulmonary insufficiency. The controlling factors in MOF are not well understood. Neuronal, hormonal, and other influences change the metabolic state and -as for the nutritional aspect in a wider sense of the word-one general conclusion can be drawn: in MOF cells, tissues, and organs are more or less unable to provide energy and protein synthesis in as sufficient amounts as they are needed and exactly there, where they should be used. From a practical point of view organ failure in intensive care units is connected with modalities of nutrition in different ways: it is a fact, that an incomplete nutrition or a state of malnutrition always means a risk for an additional organ failure. It is evident too, that a failing organ troubles the metabolic homeostasis so, that a normal "classical" nutrition is dangerous. Therefore nutrition therapy in MOF means to find a way between the necessity of avoiding symptoms of malnutrition by substituting enough substrates and between the knowledge, that this may lead to metabolic disturbances followed by a worsening of the state of illness.

ACUTE RENAL FAILURE (AF). P. FHrst, Inst, for Biological Chemistry and Nutrition, Univ. of Hohenheim, Stuttgart. A major contributing factor to the high mortality observed in AF is the net breakdown of body cell mass. Nutritional therapy in AF is thus, mandatory. Modern principles imply frequent dialysis or ultrafiltration as early as possible after the patient has become definitely oliguric, thus allowing an adequate nutritional supply by the parenteral route. Energy should be provided as much as any intensive care patient, i.e. 10.5-12.5 MS/day or considerably more, if the patient has a marked hypermetabolism as for instance in severe burns. At least 200 g should be provided as carbohydrate in order to decrease glueoneogenesis of ketoacids. To cover the increased energy requirements the use of fat emulsions are of special value since they provide the patient with a high amount of energy in a low volume of fluid. Fat emulsion should, however, not be given unless the need of carbohydrates is satisfied. Parenteral administration of amino acids should be given to all patients with AF of more than 2-3 days duration who cannot take protein by mouth. The administration should not start until the uremic state is controlled by adequate dialysis. The optimal proportion between essential amino acids and non-essential amino acids with a daily amount of 30-60 g is to emphasize. Electrolytes and water soluble vitamins should be provided as for any intensive care patients. Patients in whom the problems are restricted primarily to the kidneys usually survive the acute phase, if adequately dialysed, even when the renal function has not returned.

There is yet no conclusive evidence existing that the use of parenteral nutrition in AF will increase survival when associated with sepsis or other severe organ failure.There are reports, though anecdotic, that vigorous nutritional therapy with glucose, insuline, fat and amino acids combined with frequent dialysis or ultrafiltration therapy promote survival even in high-risk patients. Certainly, it is justified to make every efforts to prevent these patients from starvation and internal canibalism. The classical view of protein wasting in severe illness as developped by Cuthbertson in the 1930ies has been that protein breakdown results from substrate mobilisation to cover energy demands. Today we have more detailed explanatiormfor the pathogenesis of disturbed protein dynamics in catabolic states. The general view is that negative nitrogen balance in moderate trauma results from diminished protein synthesis without major increase in protein breakdown; adequate nutrition may restore in part the impairment of protein synthesis. In contrast, severe trauma is associated with accelerated protein breakdown out of proportion to energy requirements. Protein synthesis may be enhanced provided that adequate nutrition is supplied. The overall metabolic alterations in severe catabolic states are enhanced protein breakdown, ongoing fat oxidation, relative inhibition of glucose oxidation, and failure to increase hepaticproduction of ketone bodies as alternative fuel. This maladaptation is, at least in part, explained by elevated stress hormones, such as glucocorticoids and epinephrine. They result in enhanced protein breakdown and fat mobilisation, they induce insulin resistance and thus hyperinsulinemia. Results are presented demonstrating inhibition of hepatic ketogenesis by insulin; glucose utilization may be impaired due to elevated free fatty acid levels. Therapeutic measures include avoidance of stress, catabolic hormones and administration of insulin as a potent anticatabolic agent. In addition, adapted nutrition should cover the altered requirements of carbohydrates, fat and amino acids.

SICK SYNDROME A.G. Bffrger Thyroid Research Unit, Geneva, Switzerland

Among the known iodothyronines only T3 is nowadays collsidered to be biologically relevant, Contrary to general belief, T3 is not a secretory product of the thyroid, it is mainly produced by monodeiodination in the liver, kidneys and brain. It is now well established that in illness this pathway is blocked while the inactive pathway of T4 degradation via rT3 remains intact. Hence in disease serum T3 can be as low as in severe hypothyroidism and serum T4 values can also be markedly reduced. Clinically it is however easy to exclude primary hypothyroidism (but not secondary hypothyroidism) by a serum TSH measurement which will remain normal even if both, serum T3 and T4 are markedly decreased, the low serum T3 and total and free T4 levels being therefore diagnostically useless. Are these hormonal changes adequate adaptations to the sick state or do they indicate a partial deficiency in thyroid hormones ? The majority of indices favor the hypothesis of an adequate adaptation : kinetic studies have shown that the low serum T4 levels are in part explained by an increased metabolic clearance rate, the production rate being only moderatly reduced. The effect on the target tissues have been studied in the model of the starving rat : there is evidence for a partial resistance to T3 with nuclear receptors clearly decreased which is not typical for hypothyroidism. However in rats with renal insufficiency T3 can have some beneficial effects on enzymes.

A definite answer to the question raised will therefore have to awaide a better understanding of thyroid hormone action.

HYPOTHYROIDISM IN THE CRITICALLY ILL -PART OF THE DISEASE OR EPIPHENOMENON? P.R. Bratusch-Marrain, I. Medizinische Universit~tsklinik, Vienna, Austria Systematic evaluation of endocrine systems during critical non-endocrine illness is severely hampered by the heterogeneity of diseases and the variety of applied treatments. Yet,interpretation of endocrine function tests in severely ill patients may be mandatory to uncover underlying or accompanying endocrine disorders.

In the present study the impact of non-thyroidal disease upon the pituitary-thyroid axis was evaluated. Out of 335 consecutive patients admitted to an intensive care unit 20 patients suffering from severe non-endocrine diseases (septicaemia, fulminant hepatic and renal failure, acute pancreatitis, polytrauma, cerebral haemorrhage) were found to have serum total thyroxine (TT4) levels in the hypothyroid range (below 4~g/dl). Serum concentrations of TT4 (2.3~O.2~g/dl), triiodothyronine (TT3, O.23+O.O3ng/ml), and thyroxine binding globuline (iS.5~l.3~g/ml) were reduced, but were above normal for reverse T3 (0.43+0.06 ng/ ml). The response of TSH secretion to iv TRR was Tound to be either normal, lowered or absent. Primary hypothyroidism was excluded, as no enhanced TSH response was observed in any case. Although decreased TT4 levels may be due to increased thyroid hormone degradation it appears that associated impaired TSH responsiveness to TRH may result from illness-related inhibition of pituitary TSH release. Even though free T4 serum levels may also be decreased in some of the patients at the summit of their illness, the metabolic response to thyroid hormone deficiency is to be expected with considerable delay. Although the finding of decreased thyroid hormone levels is not rare in intensive care patients, it represents an index of poor prognosis. Differentiation between this "low-T4 syndrome" and true hypothyroidism depends essentially on clinical symptoms and course of disease. If there is no clinical evidence of hypothyroidism, thyroid hormone substitution may not only lack a rational basis, but due to their cardiotropic action could also endanger some of these patients. Insulin has been used in the treatment of heart failure almost since its discovery. Such intervention has always been controversial.

Insulin therapy has, of course, been combined with glucose and pomassium as GIK) ro prevent hypoglycaemia and hypokalaemia. Such combination has led, pari passu, no difficulties in defining the exact mode of action of the therapy since each constituent has potentially beneficial effects in the failing heart.

Sporadic reporrs of clinical benefit from the use cf GIK led to therapeutic trials of the treatment in a wide variety of clinical situations -ranging from its use after myDcardial infarction no its use before open-heart surgery. Unfortunately, highly varied doses of G, I ~nd K have been used, and it is not surprlslng that clinical results have been equivocal.

Further, the responses sought by the various workers have also differed markedly, ranging from suppresslon of dysrhythmias to increased cardiac output.

Few reports exls[ which relate dose to response, yet it is exactly this information which is needed by the clinician planning therapy.

There is, however, a noticeable trend in recent reports ~o increase the doses of insulin used, and results appear encouraging, particularly after myocardial infarction and in relation ~o myocardial failure following open-heart surgery.

In the latter case enormous bolus doses of insulin have been employed.

This presentation reassesses the evidence for the use of insulin in myocardial failure, discusses the logistics of its use, and emphasises the need for the establish/tent of a dose:response relationship.

OF RQ -METHODS AND THERAPEUTIC IMPLICATIONS. J.Eckart,Augsburg,FRG Indirect calorimetry is the method in use today for providing data on total energy expenditure and on the amount and composition of the oxidized fuel. The metabolic rate can he calculated by multiplying oxygen uptake by the oxygen caloric equivalent corresponding to the RQ. The respiratory quotient is the ratio of the volume of carbon dioxide produced to the volume of oxygen consumed over a given interval of time. Whereas fats and carbohydrates are completely oxidized,proteins are not. Each gram of urinary nitrogen signifies a respiratory exchange of 4.7 liters CO 2 and 5.9 liters 0 2 . Handling of water and electrolytes results under physiological circumstances from neural, humoral, cardiovascular and renal mechanisms. Mechanical ventilation interferes with several of these regulatory circles, causing generally water and sodium retention(1). The changes are more marked when positive end-expiratory pressure(PEEP) is applied, and in most studies PEEP was used to amplify the underlying mechanisms. The following factors are involved: l)Increased intrathoracic pressures elevates central venous pressure, decreases venous return and cardiac output. 2) Renal function is impaired by changes in intrarenal perfusion(2) leading to water and salt retention.3) Antidiuretic hormone secretion is increased during both short(5) and long term positive pressure ventilation(4).4) Plasma renin activity and plasma aldosterone are increased(6). ~1ost of these mechanisms are more important during controlled or intermittent mandatory ventilation than during spontaneous breathing with positive airway pressure(8).The application of dopamine or dopamine with dobutamine improves water and sodium excretion in ventilated patients(3,7). It is not clear if water retention during mechanical ventilation is part of an adaptive mechanism to the decreased cardiac filling or simply due to a dysregulation of the mentionned circulatory and neurohumoral factors. Depending upon severity and duration of hypovolemia, hypovolemia-induced activation of the sympathetic nervous system and deterioration of the fluidity of blood, the microvascular perfusion becomes impaired with the final result of tissue hypoxia and depletion of the extravascular fluid compartments. The length of time, a trauma patient has remained hypovolemic is critical for survival due to promotion of multiorgan failure with duration of hypovolemia and shock, respectively. Primary therapy should replace the colloidal fluid actually lost. The aim of initial volume substitution, however, is not only to restore circulating volume and central haemodynamics but to counteract at the same time the secondary impairment of the microcirculation. By virtue of their oncotic power and capacity to retain water within the circulation, red cell free colloid solutions should preferably be used for primary volume replacement in trauma patients. From the artificial colloids, Dextran 70 is considered most suitable due to its long lasting volume and blood fluidity improving effect. Cristalloids have proven satisfactory, however, intravascular refill requires 2.5 -4 times the volume loss and intensive haemodynamic monitoring. Independent upon the initial fluid used, a fall of haemoglobin to 10 -8 g/100 ml calls for red cells. RBC-concentrates/FFP and fresh whole blood are both efficient to control deficits of oxygen carriers and plasmatic clotting factors.

CHEST TRAUHA AND SURGICAL TREATEHEtIT W. Glinz, Surgical C]inlc B, University Hospital Hospital, ZUrich.

Apart of intercostal tube drainage, which is the basic therapeutic procedure in all severe chest injuries, operative intervention is seldom necessary in blunt thoracic trauma. We performed thoracotomy in 8% of such hospitalized patients. There are clearly defined indications for surgery: Hassive and persistent bleeding, acute cardiac tamponade, rupture of trachea, bronchi, aorta, supraaortic branches, and diaphragm. Operative stabilization of the chest wall should be considered in exceptional cases. Hany restricting factors are involved in such an operation, and in many cases it is just not necessary. Conditions are similar in the case of penetrating injuries. Here too, th0racotomy is reserved for precisely defined indications. Many cases can be treated by conservative management~ which does not mean being inactive. Immediate thorac0t0my~ however, is the keyt0 success in all penetrating heart injuries,and this applies particularly to situations that are apparently hopeless. Th0raeotomy was performed in #9 out of 155 penetrating thoracic injuries (#8% of which were gunshot wounds). In this series, 36 heart injuries are included. Mortality was 5.5%. Cardiac failure may be due to abnormal alterations in preload, afterload, contractility and heart rate. Therapy and management of the failing heart therefore has to consider positive inotropic drugs as well as vasodilating and antiarrhythmic management. Acute respiratory failure (ARF) means increase of PaCO 2 and/or low P O 2 ~ed to FIO 2. After trauma a number of pa~h61ogical conditions may cause these symptoms. Hypoxia/hypercapnia are produced by a reduction in ventilation and/or by disturbance of gas exchange, l.Ventilation may be limited mechanically (blocked airways, Pnet~othorax, flail chest), by central nervous depression (coma) or by pain (rib fractures). If causal therapy is not effective in~ediately, mechanical ventilation has to be installed without delay. High frequency ventilation may be adequate in selected cases. 2.Gas Exchange may be disturbed by localined pulmonary lesions preventing ventilation in areas still perfused (pulmonary contusion, aspiration pne~nitis) or in a diffuse change of pulmonary parenchyma beginning as interstitial edema progressing to alveolar edema and later on to interstitial fibrosis (ANDS). Although biochemical mediators may initiate ANDS, the first clinical manifestation is a capillary leak associated with and partially due to hypoxic pulmonary vasoconstriction. Therapy cannot focus on the primary lesions but must (i) increase pulmonary gas voltm~ (mechanical ventilation with PEEP ) to raise P_Or and to reduce hypoxic pulmonary vasoconstrlctlon, pulr~o~ary hypertension and capmllary leakage, and (2) increase cardiac output to reduce pulmonary vascular resistance and right ventricular failure. Intravascular volume must be optimised and vaso-active pharmacotherapy may be necessary. However, the ini0ortant goal is difficult to reach : to get ventilation in match with perfusion. Early prophylactic steps are more effective than late therapeutic attenlos. Cortieostheroids, Heparin, Trasylol, etc. are not proven to have any effect. Head trauma is common but severe brain injury is rare. In a few patients an injury which at first sight seems minor leads to severe brain damage due to intracranial bleeding.

The primary brain injury which occurs at the moment of trauma may be complicated by secondary brain injury if hypoxia, hypocapnia, hypo-and hyper-tension or cerebral venous congestion are allowed to occur.

The main problems in early head injury management are, therefore, I) triage and 2) prevention of secondary brain injury. At the scene of the injury the first requirement is to control the airway.

Regurgitation of vomitus can usually be prevented by correct positioning.

Unskilled attempts at tracheal intubation may make the brain injury worse.

The second requirement is the support of the circulation by controlling haemorrhage and/or infusing colloid.

Once these primary resuscitative measures have been taken the level of coma should he assessed and recorded using the Glasgow Coma Score and the reaction of the pupils examined. The patient should be transported to hospital and a skull X-ray obtained.

If he remains in coma as defined by Jennett as "Not obeying commands, not uttering any recognisable words and not opening the eyes" after correction of respiratory and circulatory inadequacy, then he should be transferred to a hospital able to provide CAT scanning.

This should also occur immediately if there is any sign of worsening of the conscious level. If the CAT scan shows intracranial haemorrhage, craniotomy will usually be necessary and this should be followed by ICP monitoring.

If there is no haematoma but the patient remains in coma, ICP recording should be initiated. Controlled ventilation is indicated for at least 12 hours following any surgical haematoma evacuation.

In the absence of surgery, controlled ventilation is indicated if there is i) no reaction to pain, 2) extensor posturing, 3) convulsions, or 4) ICP over 25 mm Hg.

During controlled ventilation ICP should be kept below 25 mm Hg by mannitol, frusemide and althesin, etomidate or thiopentone. Many severe head injuries also have other major injuries. Necessary surgery should always be covered by adequate anaesthesia, using controlled ventilation and avoiding volatile anaesthetics.

Chest injurSes may require the use of positive airway pressure and this is acceptable providing the patient is nursed head up to reduce cerebral venous pressure.

Limb fractures should be managed by methods which do not require traction in the head down position.

Maxillo-facial injuries can usually wait 24-48 hours before surgery.

It is very necessary to be on guard for visceral intra-abdominal bleeding which can readily be missed during controlled ventilation of the head injured patient.

in trauma victims Peter, K., Franke, N. Institut fGr An~sthesiologie der LM-Univ.

In anesthesia under conditions of trauma and shock due to volume dificiency the pathophysiological alterations caused by these conditions have to be taken into account. On principle, trauma and shock result in an acute disproportion between oxygen supply and oxygen demand of the organism. To the fore of the cardiovascular reaction there is the sympathico-adrenergic stimulation at an reduced cardiac Output which leads to a derangement within the field of microcirculation. In a stage of compensated shock constriction of arterioles, preeapillaries and venules causes an impairment of inflow and outflow. As a result microcirculatory blood flow is redistributed thus leading to a reduction in perfusion of nutritive capillaries as well as to tissue hypoxia and anoxia. The aim of therapy must therefore be an improvement of microcirculatory blood flow. The most important factor in this connection is a rapid and complete volume replacement by dextrane, crystalloids and blood or blood fractions. Adequate oxygenation and reduction of respiratory work must be ensured early and with somnolent patients early intubation is needed in order to protect respiration. With patients in traumatic shock anesthesia should be induced with etomidate, ketamines or opioides because of the low negative inotropy and the modest influence on afterload of these agents. Subsequent to a complete volume replacement narcosis can be maintained by a combination of fentanyl, for instance, and volatile anesthetics (enflurane, halothane). In order to decrease the pressure in situation with circulatory hypertension the administration of nitroelycerine should be preferred, since sodium nitroprusside induces microcirculatory disorders. These necessary measure call for an invasive monitoring including recording of cardiac output and pulmonary pressures. Department of Surgery, Western Infirmary, Glasgow Gll 6NT. The relationship between multiple trauma and infection has been studied in 428 patients admitted to an ITU between 1969 and 1982.

Until 1980 mortality was consistently around 25%. 62% of the non-survivors died in less than 5 days from injury, decreasing from 86% during the first two years to 35% during the last two years.

The majority of the remainder died with septic complications and/or multinle organ failure. During 1981/82 the number of trauma referrals rose by 60%.

Mortality also increased (from 25% to 44%); 33% of the non-survivors died in less than 5 days from injury. There mere no obvious changes in the pattern of referral, severity of injury, nature of bacterial infection or frequency of invasive procedures to explain the increase in mortality.

Other factors influencing host defence both during early resuscitation and subsequent intensive care may adversely affect outcome. We are involved with a prospective study in humans, in hope to clarify the issues. 31 patients with ARDS, septic or non septic shock admitted to ICU have been studied.

All patients were on respirators and PEEP.

Serum for C5a were drawn initially at the time of diagnosis and then daily. 82 samples from 31 patients were collected and analyzed. Complement induced leukocyte aggregation was measured in vitro and graded from i+ to 4+. From 31 patients, 5 were negative, 9 showed i+ to 2+, 17 demonstrated 3+ to 4+ and were considered positive.

SolumedrolR(Methylprednisolone) 30 mg/Kg was given q6hr for 4 or 8 doses in i0 patients.

Classification of each group and results are shown below.

No 

Activation of the alternative pathway of the complement system leads to leukocyte aggregation, entrapment in the pulmonary microvasculature and release of vasotoxie agents which lead to increased physiologic shunting and permeability pulmonary edema. 51 sheep were infused with zymosan activated plasma (ZAP): group I (n=ll) were untreated controls. In group II (n=10) imidazole, a thromboxane synthetase inhibitor, 25mg/kg.h was started 89 h before ZAP infusion. In group III (n=10), PGi 9 in a dose of I00 ng/kg.min was given 30 min before ZAP infusion. Within 5 min, lAP led to a fall in leukocyte count to 2900/mm 5 [p(0.001), rise in plasma thromboxane B 2 (TxB2) from 14 to 246 pg/ml (p(0.001) rise in lymph, TxB 2 from 24 to 609 pg/ml (p(0.001) rise in mes-9, pulmonary artery pressure 0MPA) from 17 to 4SmmHg and Qs/Qt . from 13% to 51%..Both imidazole and PGI 2 prevented the rise in TxB 2 and Qs/Qt and limited the increase in MPA to 25 and S0~ Hg respectively. Imidazole but not PGI 2 prevented the increase in lymph flow which in controls increased from 2.8 to 8.S ml/50min (p40.01) and lymph albumin clearence which increased from 2.2 to 6.0 mi/30 min (p<0.01). The high lymph concentration of TxA 2 suggests a pulmonary site of production and its brqnchoconstrictive action may account for the increased Qs/Qt-TxA2 is only partly responsibl e for the pulmonary hypertension and apparently unrelated to changes in permeability. The protective act/on of infused imidazole against increased permeability appears to be independent of its inhibition of Tx synthetase. 3 groups of 9 patients with comparable aetiology have been constituted according to the duration of the ventilation and the amount of PEEP appreciated with an index as follow: A-Prolonged ventilation (16j.) and low PEEP (index 3) B-Prolonged ventilation (15j.) and high PEEP (index 18) C-Short ventilation because of early death D -Control group of deceased patients without lung pathology Elastic fibers quantitation on randomly distributed prelevment of lung parenchyma stained by orcein has been achieved using i) a point counting method to measure the volume fraction of elastic fibers and the boundary length fraction ; and ii) anautomatic image analyser working on images by successive erosions. The 3 groups A B C show a significant decrease of elastic fibers volume fraction compared with group D. The smallest value is in group C. The 2 groups B and C show a significant decrease of elastic fibers boundary length fraction compared with groups A and D together. The elastic fibers boundary length fraction have been appreciated first by the ratio of the total boundary length onthe volume fraction and secondly by the area of the 4th and 8th erosions given by the automatic device, These histomorphometric results confirm the visual impression of elastic fibers destruction during the early phase of ARDS.

In case of prolonged evolution these results suggest a partial regeneration to occur the quality of which being under the dependance of the type of ventilation. Indeed PEEP seems to maintain the elastic fibers in a fragmentation state close to the state of the first day of the disease. In this group A, where mean venous admixture (40.9 ~ 10.4) was not dif'ferent from group B (41.2 + 8, i) , a similar mean duration of ventilation was obtained despite the lack of consistant PEEP using membrane lung oxygenation in several cases.

Mainz,Abtl.f.Pneumologie u. II.Med.Klinik,D-6500 Mainz ACE is present in high concentration in pinocytotic vesicles on the luminal surface of the pulmonary capillary endothelium.Because damage of the pulmonary vascular endothelium is believed to be a cardinal pathogenetic feature of ARDS it is the question,if serum ACE level can be used as a marker for development and prognosis of ARDS. Until now we have investigated 9 patients with ARDS.ACE activity was measured spectophotometrically in progress of ARDS (range 8-32 days; method described by Cushman and Cheung)and was correlated to data of pulmonary gas exchange (AaD0o,Pa02,PaC0o,O~/~t)and thoraco-pulmonary compllance. 60 h~al~y persons in comparable age were used as contol group. In patients with ARDS ACE activity in serum was found significantly lower(~• =11,9+7,7)than in healthy controls(x~Xs=32,0• -Patients with relatively higher mean ACE levels in progress of ARDS(~=13,0-24,0 U/ml)demonstrated finally before death a marked increase of ACE serum level (range 31,5-43,6 U/ml)corresponding to an extreme rise of pulmonary shunt perfusion. This final increase of serum ACE was not found in patients showing continuously very low ACE values during ARDS(:=3,3-9,3 U/ml).These patients survived a shorter time on average(~=9d) than those with higher ACE levels (~=25d Pulmonary edema following near drowning in fresh water is a frequent and severe early complication. Lung mechanics and pulmonary gas exchange studies were performed 6 and 12 months after the accident in 2 survivors of adult respiratory distress syndrome(ARDS)following prolonged immer sion(lO and 20 minutes)in fresh water. The patients were 25 and 29 years old and non smokers. Assisted ventilation was performed during Ig and 34 days respectively.Measurements included spirometry,lung compliance,distribution of ventilation,transfer factor of carbon monoxyde(DLCO), membrane factor(DM),pulmonary capillary volume, gas exchanges at rest and during exercise. Resulb indicate that lung mechanics returned to normal within 8 months after the onset of ARDS.No alterations of DLCO or changes of the distribution nf ventilation were detected. Conversely,the ratb of dead space to tidal volume was significantly increased at rest and during exercise,and the ra tio of capillary volume to alveolar volume was decreased in both patients.lt is concluded that AROS following pro]onged near drowning in adults causes late abnormalities of gas exchange probably related to decreased pulmonary capillary volumed The aim of this study was to point out the interest of g~ exchanges and compliance monitoring of the patients submitted to controlled ventilation during ARDS. 34 patients supported the study (Mendelson's syndrom 18 -Lung contusion 9 -Viral pneumonitis 4 -Fat embolism 3). The monitoring its based upon the continuous measurement of expired CO2, expired volume and airways pressure by the mean of an individual computerized device 9 The monitoring was reali-9 9 + zed all along the controlled ventilation (4,5 ~ 1,8 days) and allows to obtain the following functional data : VA/V ratio, partial ductances of CO2 (expired/alveolar = DuVCO2 alveolar/arterial = DucC02) -VCO2 -static compliance (Cst). The therapeutic regimen included : CMV with PEEP, isovolemic deshydration, corticotherapy and heparin. 23 patients survived (group S). ii patients (31%) died, (group D). We compared the results (table) of functional data obtained at T6 (6th hour) -T24 -T48 between both groups (Student's t test). Discussion : at the early stage, DuCC02 seems to be the more significant pronostic index. As for recovery the predictive value (PV +) at T6, reaches 85 % when DucC02 0,70. However, when DucC02 < 0,70, the predictive value of death is bad (PV = 50 %). So we tried to define a more accurate index : IEM = DucC02 x Cst which has PV + = 94 % (when IEM 320) and PV -= 74 % (when IEM <20). cm H20 below and repeated at Scs H20 over the cllnically chosen PEEP (13 + 6.5 SD cc H O) at maintenance F 0 (0.57 + 0.18 SD).

was then dlvided into two components, one due to true shunt d 2 an one due to maldistribution, computed as (QVAO2-QsSF6)/g. The study showed:

-OVA/gO 2 and gs/gtSF 6 decreased slgnificantly increasing PEEP (from 0.36 + 0.12 SD to 0.27 + 0.11 and from 0.28 + 0.16 SD to 0.22 + 0.15 SD respectively); -(gVAO2-QsSFd)/O : no s~gsiflcant changes in spite of m~nor unpredlctable variations in both directions (0.08 + 0.05 SD at -5 vs 0.06 + 0.05 SD at + 5); -cardiac output waslnot_ significantly influenced_by~ the PEEP levels (6.67 + 1.98 1 min at -5 vs 6.52 + 1.97 1 mln at + 5); changes were not correlated with changes in QVA/OO or Os/OtSF 2 6' -the fraction of the total OVA/Q0 due to maldistrlbution 2 (QVAO -QsSF )/(gVAO ) (y) showed e significant negative correlation 2 6 2 with FIO 2 (x) (y = 0.75 -0.84 x; r = -0.75). Concluslons:

higher PEEP improved oxygenation mainly through a decreased true shunt. As expected at high F 0 the true shunt was I 2 very similar to OVA/gO . In patients tolerating lower F 0 a 2 I 2 substantial amount of hypoxla was due to maldlstrlbution. I.v. anaesthesia is generally considered to be accompanied by impaired alveolar gasexchange with inpending hypoxemia. TO eliminate this risk use of oxygen enriched inspired air is the role. The aim of the present investigation is to study the role of inspired oxygen fraction For the distribution of $A/~ during i, ~. anaesthesia in patients undergoing lung surgery. Methods. 14 patients (33-72 years) were investigated before lung surgery because of lung tumour with the multiple inert gas elimination technique according to Wagner and West during 4 experimental settings: I/ Before i.lr. ~uaesthesia (FIO 2 = 0.21), 2/ 30 mln after induction of anaesthesia (panicthai, diasepam, fentanyl, pancuronium and controlled ventilation) (FIO 2 0.21), 3/ 30 mi, after increasing FI02 to 0.5, 4/ 40 min after change in body p6sition from supine to lateral with FI02 still = 0.5. Results are presented in table i and 2 as means, x) indicated significance = <0.05 compared with proceeding measurement (paired t-test). ~b I. Ventilation-perfusion relationships (~A/Q), assessed by a multiple inert gas elimination technique, were studied durinq halothane anaesthesia and mechanical ventilation at different inspiratory oxygen fractions (FTOe). Widened %/Q-distributions were observed in all~n~ne patients (mean age 61 years, 4 smokers) during anaesthesia, indicating increased mis-matching between ventilation and blood flow. The ~a/Q-distribution was unaltered when F~O 2 was increased from a mean of 29% to 53%. A further i~crease in FIO 9 to a mean of 85% caused a further widening of the ~A/Q-distribution, and an increase in true shunt (V8/Q = O) from 7% to 10% of cardiac output (p,~ 0.0]). Hdwever, there was no fractional increase in low" VA/Q, except for one patient. On the return to FTO ? of 29%, the VA/O-distribution resumed the same patt@r~ as seen prior to oxygen breathing, and true shunt was reduced to the initial level except for one patient (the one with an increase in "low" 9A/O during oxygen breathing). The findings may fit in wlth release of hypoxic vaso-constriction when FTO 2 is increased to 85%, or the openinq up of a certain population of shunt.vessels. Only one patient displayed a change of the VA/QTdistribution during oxyfen breathing that could be explained by resorption atelectasis. Vasodilators were given to 2 patients with ARDS and who were mechanically ventilated with positive end expiratory pressure (PEEP,cmH20). Hemodynamic, gasometric and lung mechanic parameters and continuous distribution of ventilation-perfusion ratios (VA/Q) determined by the multiple inert gas elimination technique were measured at baseline and after diltiazem (DI 0.5 m~Kg IV) in the first patient and after sodium nitroprusside (N P 7 ~Kg/min IV) in the second patient. In the two eases, pulmonary vasodilation ~ "L,.. The diagnosis of pulmonary embolism (P.E.) is difficult with ventilation-perfusion scanning in patients having a reduced ventilation in the region of the perfusion defects (matched defect) as in.chronic obstructive lung disease (C.O.P.DJ Regional V/Q were therefore computed in each region of a perfumion defect in 44 patients with angiographically confirmed P.E. and 40 patients with C.O.P.D. by dividing ventilation scan by perfusion scan obtained with Krypton 81m in tidal breathing with a gamma-camera linked to a computer. The regional ventilation-perfusion ratios were computed in each region of a perfusion defect. All the patients were studied supine in post6rior view. P.E. was characterized by mismatched defects (reduced perfusion and normal ventilation) with a high V/Q (1.96 • O.1 SD) due to vascular obstruction. A high O/Q (1.37 • 0.2 SD) was also found in the 25 out of 93 defects of perfusion which were matched by a ventilation defect or a radiological opacity in these patients. Low V/Q'S were observed in the non embolic lung due to relative hyperperfusion. C.O.P.D. was characterized by matched defects with low ~/~ in most of the cases (.74 ~ .1 SD) and the two groups (P.E. and C.O.P.D.) with matched defects differed significantly (p < 0.01). This regional quantitative analysis provides a simple mean to discriminate the defects of perfusion due to P.E. from those due to a reduced ventilation. This studydemonstrated that there was no significant difference in CI, P, SI, LVSWI, TPRI, a-vo2, Q2 and P~ between CPPV and HFJV at 100.min and at the same PaCO 2 and PEEP. There was significant difference in Pa02 P < 0.004; QS.OT P < 0.025 and PAP P < 0.001. HFJV should reduce barotrauma because of the reduction in PAP. It is advisable to start with an FIO 2 0.9 and decrease FiO 2 as indicated by PaO 2.

~A/~ <0.005 0.005-0.1-tA/~ log SD 0-10 i0-I00 ~A/~ log ED 1GO

ACCUMULATIVE HIGH-FREQUENCY JET VENTILATION.M.D. Aochi, O Miyano, H.Hizuho-ku,Nagoya City University, Medical School, Nagoya City, Japan. High-frequency jet ventilation (HFJV) in animals and huma -n-be}ngs with normal lungs has been shown to provide adequate ventilation (V) and oxygenation (0) at lower airway pressure than IPPV without causing cardiac depression. However in those clinical cases with pulmonary disorders like ARDS, it is difficult to improve V and 0 by HFJV alone, because of extremely diminished lung compliance and increased dead space. So we have attempted the new method, accumulated highfrequency jet ventilation(Acc.-HFJV), in which the jet flow from HFJV generated by modified Bird Mark 8 has been accumulated by time-cycled regulation of the expiratory valve of CV-2000 ventilator. Aec-HFJV has improved V and 0 without compromising the characteristics of the jet flow, as compared with the superimposed mode. In addition, by using the technique of periodic interruption of the jet flow adjusted by respiratory cycles, it is possible to deliver accumulatiw e jet flow in the inspiratory phase, the expiratory phase or both.of them. We have compared Acc-HFJV with CPPV and conventional HFJV in 40 patients with ARDS at same levels of PEEP.The results have indicated that adquate Vand 0 could be maintained more effectively by using Acc-HFJV than the other two modes of ventilation. No adverse circulatory effects have been observed.

M.Chiaranda, G.Fiore, E.Facco, G.P.Giron, Inst. of Anaesth. and Intensive Care, University of Padua, Italy. The reduction observed in airway pressure when using high frequency jet ventilation (HFJV) instead of intermittent positive pressure ventilation(IPPV),could be advantageous in all conditions in which aminimal variation of intrathoracic pressure markedly modifies cardiovascular functions. This study compares the cardiovascular effects of IPPV and HFJV in experimentally induced haemorrhagic shock. Five healthy Beagle dogs, 12 ~ 15 kg, were anaesthetized, paralized and intubated in the standard fashion. Sistemic, central venous and pulmonary arterial pressures, cardiac output via the thermodilution technique, arterial and mixed venous blood gases, airway pressure (Paw) at the end of the endothracheal tube and oesophageal pressure were monitored, Cardiac and stroke indices (Cl, SI), pulmonary and sistemic vascular resistances (PVR, SVR) and right and left ventricular stroke Work~ indices (RVSWI, LVSWI) were calculated using standard equations Wentilation was performed randomly with a volume-cycled ventilator (3100 Roche) at 15 b/m and with an HFJV apparatus (VJP system) at 60, 120, 240, 480 b/m, regulating tidal volume to maintain PaCO 2 at 38 ~ 2 mmHg. Cardiorespiratory data were obtained for both modes of ventilation before and after lowering CI by inter mittent bleeding to about 50% of the initial value. Mean and peak airway pressures were significantly lower with HFJV at 60 ~ 240 b/m; at 480 b/m, the end expiratory pressure was 3,7 ~ 1,2 mmHg, giving an increased mean Paw. Using HFJV at 60 § 240 b/m during haemorrhagic shock, CI, SI and LVSWI were significantly enhanced, by respectively +20%, +21% and +38%, in comparison to the control values in IPPV; a cardiovascular depression appeared at 480 b/m. These results confirm that when there is an haemodynamic impairment, as in the anaesthetized hypovolemic dog, HFJV reduces the deleterious cardiovascular effects of conven tional positive pressure Ventilation. The influence of different lung volumes on various frequencies of nigh frequency ventilation was studied in 5 anesthetized mongrel dogs (x=28.9kg). After intubation, using a special developed endotracheal tube, dogs were placed in a rigid heat-exchanged pressure box (p.b.) with variable internal pressure (range: -1o to +5 cmH20). Ventilation was performed: A. under conditions of IPPV (f= o.2 Hz), B. as "forced d~ffusion ventilation"(FDV), a ~mdified nigh frequency jet ventilation(f=6Hz,5oHz, cont.=jet stream without any chopper frequency). By variing b.p. (-1o/o/+5cmHgO), different levels of lung volume were achieved. PartiCular combinations of b.p. and respirator settings were selected in random fashion. Blood gases, mixed expiratory CO_ (Fm-CO 9) and minute volume The authors investigated the relationship between physiological dead space (Vd) and tidal volume (Vt) during high frequency jet ventilation (HFJV) at different positive end expiratory pressure (PEEP) levels (0 -2.0 kPa). Four healthy Beagle dogs, IIM7 kg, were anaesthetized, pa ralized and intubated in the standard fashion; ventilatio~ was performed with intermittent (IPPV) or continuous (CPPV) positive pressure at 15 b/m and HFJV at 60, 120, 240, 480 b/m, while minute volume (~)wasregulatedtomaintain PaCO 2 at 38 ~ 2 mmHg. For each mode of ventilation arterial blood samples for gas analysis were taken after 10 min of stabilization. The volume of alveolar ventilation in a single breath (Va) and in a minute (~a) were calculated analyzing the values obtained with a pneumotachograph (Vt) and with Bohr's equation modified by Enghoff (Vd).

The results obtained show that there is a progressive redu ction in Vt,VaandVd and a progressive~increase in Vd/Vtratio with increasing ventilatory frequency. The increase of PEEP values from 0 to 2.0 kPa increases Vd and lowers the respiratory frequency value at which Vt falls below Vd measured at 15 b/m. In conclusion, the authors stress that it is possible to maintain an adequate alveolar ventilation utilizing tidal volumes lower than the anatomical dead space volume. The ha~ to be increased with the increasing ventilatory frequen cy; this is due to an augmented wasted ventilation (~d The enthusiasm about ventilation of patients with higher frequencies has changed a lot from the beginning.Two disadvantages are dominating, one is the impossibility of pressure measurement at alveolar level and the other is the difficulty of application at normal ICU-patients.So a new system has been developed,giving the chance of HFV (high frequency positive pressure ventila= tion),HFO (high frequency oscillation) and CMV (conventional mechanical ventilation) using normal endotracheal tubes with only one small ventilator.This system -nothing more than a small computer -gives the possibility to generate any new ventilatory pattern in future. We compared HFV,HFO and CMV generated with higher or lower frequencies.The research is done using two groups of ICU-patients,one,which does only need mechanically support and another,which has lung disease.ln both groups similar hemody= namic data were found under similar airway pressures.Gas exchange may be better using generated frequencies under some circumstances.

HIGH FREQUENCY JET VENTILATION AND PEAK AIRWAY PRESSURE SYNCHRONIZATION WITH ECG. M.Chiaranda, A.Rubini, M.Trevi san, G.P.Giron, Inst.of Anaesthesiologyand Intensive Care~ University of Padua, Italy. The use of ECG triggered jet ventilation is still controversial.The authors investigatedthe haemodynamic influence of high frequency jet ventilation (HFJV) with inspiratory peak synchronized with different phases of the cardiac cycle. Four;healthy Beagle dogs, II~17 kg, were anaesthetized, paralized and intubated in the standard fashion.Catheters were introduced for the measurement of sistemic, central venous and pulmonary arterial pressures, arterial and mixed venous blood gases; cardiac output was evaluated via the thermodilution technique. Cardiac and stroke indices (Cl,Sl), pulmonary vascular and total peripheral resistan ces (PVR,TPR) and right and left cardiac work indices (RCWI,LCWI) were calculated using standard equations. HFJV was performed at 120 b/m, with minute volume (V) regulated to maintain PaC02 at 38 ~ 2 mmHg and a positive end expiratory pressure (PEEP) from 0 to 2.0 kPa. The haemodynamic influence of HFJV was evaluated when progres sively delaying the inspiratory peak corresponding to th~ onset of the 'R' wave on the ECG trace. The mean pulmonary arterial pressure oscillations decrease as PEEP rises and their value reaches a maximum of 15% at zero PEEP. Lowest PVR and RCWI values were observed when the inspiratory peak coincided with the protomesodiastolic phase of the cardiac cycle (that is at 30-70% of R-Rinter val) and the highest values when the peak drops close tothe R wave (~ 20% R-R). This phenomenon occours independently from the PEEP level applied. The results demonstrate the absence of significant haemodynamic effects due to ECG and inspiratory peak synchroni zation when HFJV is applied to anaesthetized dogs with normal cardiovascular reserve and lung compliance. ~he usefulness of vasodilators in the treahnent of pulmonary hypertension (PHT) following chronic obstructive pulmonary disease (COPD) has not yet been established, especially during acute exacerbation of the disease. phentolamine (Ph), a potent systemic vasodilator drug, is known to act on the pulmcnary circulation as well. The recognized spa~nolytic action of Ph on the bronchial ~noothmuscle could represent a further advantage in the management of COPD.

Seven patients (6 males, I female) with an acute exacerbation of COPD (Pa02 49 + 3 r~n Hg, PaC02 59 + 4 r~n Hg; mean + S~M), aged 49-74,--underwent hemodynamic investigationbeforeandduring Ph infusion (10-15m g/h), without 02 therapy. Baseline hemodynamic data showed a moderate PHT (P~44,5 ~ 7 nm Hg) with raised right atrial pressure and a normal pulmonary wedge pressure (POD 11 + 2 nm Hg; P~ 9 _+ 2 mm Hg) . Ph produced a significant ~eduction in P-~, P~and P~ (p < 0,05; p < 0,01; p 4 0,02). There was no change in cardiac index, Pa02, PaC02 and mean systemic arterial pressure. A further fall in P~ followingan initial reduction with 02 therapy was observed in onepatient.

In conclusion, in patients with an acute exacerbation of COPD, phentolamine has a beneficial effect on pulmonary hemodynamics without detrimental changes in Pa02 , PaC02 and 02 delivery. ~he mechanics responsible for the decreased pressures in the pulmonary circulation induced by phentolamlne are discussed.

M4dicale -HENRI MONDOR HOSPITAL -FRANCE -Serotonin, recognized as a potent vasoactive agent is involved in pulmonary hypertension mechanism in patients with COPD. In order to validate this hypothesis, KETANSERIN (K) a specific 5 HT2 receptor antagonist, was used in COPD. METHODS : 9 patients (ranged 50 to 75 years old) with COPD ~I Research Council Criteria) were studied during acute exacerbation of respiratory failure (4 breathing, spontaneously 5 mecanically ventilated). All were free of treatment except antibiotics. Haemodynamic parameters were systemic arterial pressure (MAP), pulmonary capillary wedge (PWP), pulmonary arterial (PAP), and right atrial pressure (RAP) (Swan-Ganz 7F), cardiac output (CI) (Thermodilutiontechnique), heart rate (HR). Simultaneous sampling or arterial (a) and venous (v) blood permitted measurement of oxygen and carbon dioxide tension, hemoglobin saturation leading to calculate venous admixture (Qs/Qt) (Fi02 : .21). Then 10 mg of (K) were given intravenously followed by a continuous infusion of 4 mg/hour. RESULTS : are summarised in the 2) The constant increase in venous admixture is not always related to CI changed so, when Cl decrease lowers P~02, its variation may explain venous admixture increment 3) On the other hand, the constant decrease in V02 may be related to sympathetic blocking property of this drug. It has been shown, that the ratio between thromboxane and prostacyclin might be an important factor in acute pulmonary hypertension in sheep treated with E.coli endotoxin (i). On this basis we gave ibuprofen to two patients with acute pulmonary hypertension to convert the thromboxane/prostacyclin ratio in advantage of prostacyclin. One patient suffering from severe ARDS (effective compliance below 18 ml/cm H20 , pulmonary artery pressure (PAP) 56/38 mm Hg, P O 2 i0 kPa F.O~ 0 6 PEEP i0 " i z " ' cm H20, balloon occlusion pulmonary angiogram showed pulmonary artery filling defects and thrombi) -and another patient suffering from multiple pulmonary embolism verified by ventilation-perfusion scintigraphy. In the first case a dose of 50 mg/honr of ibuprofen was given i.v., and in the second case a dose of 1200 mg daily was given orally. In both cases PAP fell -in the first case from 56/38 mm Hg to 40/15 mm Hg, and in the second case from 92/37 rm~ Hg to 25/12 mm Hg. PAO2 increased in the patient with ARDS with 4s9 kPa with a constant FiO 2 and constant PEEP. In the second case the P 02 increased from 8,8 kPa to 10,4 kPa. No adverse e~ects were seen. These two cases suggest, that ibuprofen might be beneficial in the treatment of acute pulmonary hypertension. i. W.D.Watkins, P.C.H~ttemeier, D. Kong In order to validate the hypothesis that hypoxemia in P.E. can be related to the shift of perfusion away from the embolie zone we have computed regional VA/Q ratios and measured arterial PO 2 in a group of 44 patients width P.E. proved by angiography. Distribution of VA/Q was obtained with a minicomputer linked to a gammacamera from ventilation and perfusion lung scans with Krypton 81m a short lived radionuclide (TI/2 = 13 sec) delivered by continuous inhalation and infusion in tidal breathing. All 44 patients were studied supine and the ~ercentage of lung field with VA/Q. Iess than 0.75 (VA/Q 0.75) was computer (The overall VA/0 was normslized to i). The mean values of PaO 2 and VA/Q 0.75 were respectively 67 mmHg ~ 2.35 SD~ and 21.6 % ~ 9 SD . There was a significant, inverse relation shi~ between individual PaO 2 and VA/Q 0.75 PaO 2 = -2 VA/Q 0.75 + 112.2, r = ~67 p~/-.O.Ol.

Nine patients with unilateral embolism were also studied on lateral deeubitus with the non embolie lung in dependent position in order to change the distribution of perfusion. In all cases the VA/Q of this non embolic lung decreased due to the increase of perfusion (VA/Q supine 0.83 ~ 0.1 SD, lateral decubitus 0.72 • 0.08 SD p <O.O1). This decrease was accompanied by a decrease of PaO 2 without any change of PaCO 2 (supine 77 • 12 mmHg, lateral decubitus 70 ~ 15 mmHg p~..O.02). These results suggest that over perfusion of non embolic regions is involved in hypoxemia during P.E. The amount of oxygen consumed by the work of breathing can be calculated as the difference in total 0auptake (V02) during controlled mechanical ventilation (CMV) and spontaneous breathing (SB). Whereas this difference is well known in normal subjects (I -2%) the oxygen cost of breathing in patients with mitral valve disease is still subject of discussion. In 8 patients V02 was measured continously according to the method of Neuhof (Crlt.Care Med. 6:155,1978) durlng CMV after mitral valve replacement on the operative day and on the first postoperative day after weanlngfrom the respirator during SB. In every patient mean V02 was increased during SB (172,0 + 14, ml 02/m2/min) in comparlslon to CMV (137,5 ~ 12,~ ml 02/m2/min), the mean difference was 25,5~, with a range of.7,4% to 34,9%. Other factors which might influence V02 (physical activity, body temperature, nutrition, and acid-base status) did not change significantly during both periods of investigation. Because the cardiac index did not vary (2,73 + 0,76 1/m2/mln during SB, 2,68 ~ 0,79 i/m2/min during CMV) the grea~er 02-demand during SB could only be met by an increase in av-D02 (6,50 + 1,22 ml 02/100 ml during SB, 5,15 ~ 1,12 ml 02/I~0 ml during CMY). The results confirm that the oxygen cost of breathing is a significant fraction of whole body V02 during the immediate postoperative period in patients after mltral valve replacement. Lethality among status asthmaticus cases ranges between 7 and 16 % at the peak of the attack.This aecessitates a clinical analysis of each lethal case.Studies were conducted of 320 case histories of status asthmaticus with 38 deaths (11.9 %).An analysis showed that 7 patients died of ass aggravated by cardiovascular insufficiencyi7 more patients died after entering hospital in irreversible hypoxia coma,1 patient developed an anaphylaxic shock from medication,yet another suffered from cerebral blood circulation disorder resulting from the removal of a catheter from the central vein,a patients developed irreversible hypoxia coma durinu couwh fits resultin 9 from tracheal manipulation,1~ patients died as a result of incorrect treatment of the nronchial obturation syndrome or failure to administer treatment.This means that 65.~ % of the l~thal cases could have been avoided by timely and correct treatment. Prevention of lethality in status asthmaticus cases requires : urgent hospitalization of patients sufferin 9 from a prolonged attack of hronthial asthma; resuscitation treatment subdivided into 1) intensive therapy when the patient is fully conscious (control of the cout~h reflex, restoration of sputum theology by microtracheostomia with follow-up bronchoscopy,and 2) resuscitation when the patient is unconscious (artificial ventilation of the lungs,washing,microtracheostomia or forced tracheostomia).

In administerin.q treatment the physician must foresee possible clinical state of the patient. Usual pulmcnary function tests are not easily available for bedside evaluation of bronchial obstruction in patients with status asthmaticus. A global evaluation of lung function and degree of airway obstructien is made possible by the measur~nent of the high frequency respiratory impedance (HFRI) with the forced oscillatiens method. %~nepurpose of this study was to assess the interest of HFRI in evaluating bronchial obstruction during status asthmaticus Preliminary studies on asymptcmatic asthmatics in the laboratory have shown a statistically identical percentage of good classificatien with the HFRI (95,8%) as with the body plethysmography (93,8%). Similarly, the method has given significant correlatiens between several derived mathematical parameters and usual functien tests : Paw Sixteen patients with status asthn~ticus, aged 17-68 y., were investigated in the respiratory unit. Mean Pa02 and mean PaC02 were respectively 66,4 + 19 and 41,9 + 9,6 mm Hg at the time of the first evaluation. The HFRI proved to be able to assess bronchial obstruction and quantify its degree in terms of usual function tests. It allowed an adequate lung functicn monitoring, a reliable evaluaticn of drugs effectiveness and short term evolution. This method demonstrated a frequent discordance between the rapid clinical improvement and a persistant functional impairment. The measurements are short (about ene minute) , easily performed at bedside and do not require any active cooperation. We studied pulmonary gas exchange under different inspiratory Flow patterns in 5 lambs (22.4 ~ 2.3 kg), uslng a high performance ventilator prototype, which can be switched from a modified sine wave (MSW) to a square (S) insplratory Flow curve keeping a11 other parameters unchanged. The animals were tracheostomized, anesthetized and curarized; a OF Swan Ganz catheter was inserted in pulmonary artery, and a peripheral artery cannulated. Each animal was ventilated alternatively with MSW and S. At the end of each period (20 min) we recorded mean arterial, pulmonary artery and airways pressures, and we measured arterial and pulmonary artery blood gases, tidal volume, mean expired CO fraction 9 9 2 and end-tidal CO . Venous admixture (Q /Q) and physiological dead space 2 VA (V /V ) were computed, inspired 0 Fraction was kept constant. After 5 D I ' 2 -i periods with both curves we injecTCed in the animals trachea 10 ml kg of HCI O.01M to produce acute respiratory insufficiency, and we repeated IO to 14 times our measurements varying the inspiratory flow pattern.

Tidal volume, mean airways pressure and total ventilation were identical in the two sets. PaO did not appreciably change with the different ~ e 2 curves, nor OVA/Q did, both in physyological and pathological state; PaCO during basal period was Found significantly lower, by analysis of .2 varlance, with MSW (33.7 + 3.8 vs. 32.0 + 3.t torr, P< 0.02). At the contrary, V /V T was not significantly different; however, changing from NSW to S Dit consistently increased, and correspondingly decreased passing

From S to MSW. The variation was statlstlcally slgniflcant (P":: 0.05).

After pulmonary lesions were established, both PaCO 2 and V /V were significantly lower with MSW: 40.0 + 5.2 vs. 41.8 + 5.1

D T -torr (P< 0.05), and 0.64 + 0.08 vs. 0.87 + 0.08 (P<O.OO).Ventilation with MSW, mutuated from spontaneous breathlng, appears then to allow for better gas distribution within the lungs, at constant volume and alrways pressure, particularly if those have altered mechanical properties. The purpose of this study was to demonstrate that in presence of bilateral lung diseases leading to ARF, the finding of a difference in stat• compliance between the two lungs makes ILV preferable in order to differentiate the ventilatory support. Six patients with ARF caused by bilateral pathology were studied. All of them showed marked difference in static compliance between the two lungs. They were first ventilated in a conventional way, using IMV-CPAP, IRV or HFV.Hemodinamic monitoring was instituted together with control of tissue pH and P02. As their conditions did not show any improvement, they were switched to ILV, thus allowing the demonstration of a persisting gradient in static compliance and the subsequent modulation of the ventilatory techniques. The less compliant lung was ventilated with HFV, the other either with IMV-CPAP or with HEY. The results are consistent with our hypotesis. All patients showed significative improvements of gaseous exchanges together with a normalization in peripheral oxygen transport and utilization. It is concluded that in bilateral lung diseases refractory to conventional ventilatory therapy, a significant difference • compliance must be suspected between the two lungs, thus indicating the institution of ILV with selected ventilatory patterns.

DIFFERENTIAL VENTILATION DURING ROUTINE ANAESTHESIA IN THE LATERAL DECUBITAL POSITION. S. Baehrendtz, C. Klingshedt, G. Hedenstierna and 3. 5antesson. Dept of Medicine South Hospital,Stockholm, Sweden. Differential ventilation (DV) in the lateral posture has shown to improve the gas-exchange in patients with severe acute respiratory failure due to bilateral lung disease. The mechanism is most certainly the countering of airway closure (AC) within the dependent lung. Also patients undergoing sus in the lateral position have a preferential spread off AC towards the dependent lung. The presenl study was thus undertaken to evaluate fihe efiFect of DV during routine anaesthesia in the lateral position. Ten patients without history of cardio-pulmonary disease were investigated prior to major surgery. The patients were intubated by a double-lumen catheter. Ventilation was performed with ONE ventilator (i.e conventional ventilation) and by TWO ventilatos with 50 % of the tidal volume to each lung (DV). Anaesthesia was maintained with enflurane and the patients were kept in the lateral positione during the ceus off the study. Ventilation of each lung was determined by Argon dilution technique and a masspeetrsmeher. Central heamodynamies was measured using a Swan-Ganz thes catheter. In some of the patients , regional perFusion through each lung was studied by a gamma-camera. Results: DV as compared to conventional ventilation resulted in a 26 % decrease in venous admixture and a 23 % increment of arterial oxygen tension (p<O.01)~ while the cardiac output was not significantly altered. In the patients by which the regional distribution of pulmonary blood-flow was studied~ it could be established that DV also resulted in a much more even distribution of pulmonary perfusion between the upper and the lower lung. It is concluded that differential ventilation with 50 % of the tidal volume to each lung improves ventilation versus perfiusion and thus oxygenation during routine anaesthesia in the lateral position.

IMPROVEMENT OF GAS EXCHANGE BY PROLONGED AND DIMINISHED EXPIRATORY FLOW DURING IPPV. W. van Rooyen, J.R.C. Jansen, H.A. Bruining, A. Versprille, Dept. of pulmonary diseases, pathsphysiological laboratory and Dept. of surgery, Erasmus University, P.O. Box 1738, 3000 DR Rotterdam, The Netherlands IPPV, especially to patients with chronic obstructive pulmonary disease, may necessitate either a relatively high tidal volume or a positive end expiratory pressure to obtain sufficient gas exchange. The risks are barotrauma during peakinflation and airtrapping during deflation. To avoid these harmful effects we searched for another mode of ventilation, which improves gas exchange. The effects of prolonged and diminished expiratory flow during IPPV was studied on gas exchange in pigs. Under normal pulmonary conditions these animals were ventllated with a Servo-900 A ventilator an a mean tidal volume of 15 ml kg -I , and a rate of i0 min -I. If spontaneous explration was retarded to the total available expiratory time, a significant increase in Pa02(10%) and decrease in PACO2(8.5%) was observed. Physiological dead space was decreased as well (28%). These effects were compared to those of a PEEP which increased mean tracheal pressure to the same extent as the prolonged expiration. PEEP less improved a PaO ~ (+5%), but did not influence PaCe2 with respec~ to normal IPPV. Cardiac output decreased under both circumszan: ces to the same amount (-18%). We conclude that prolonged and diminished expiration flow has a rationale, when IPPV alone causes problems regarding gas exchange and PEEP is contraindicated. Intrapulmonary gas exchange, effective pulmQna~ blood flow,ventilation-perfusion matching (V/Q), respiratory mechanics (in vivo oxyhemometry,mass spectrometry, pneumotacho~raphy ), systemic and pulmonary hemodynamies (SWAN -GANZ catheter, thermodilution)were studied in 300 patients with respiratory insufficiency (RI) who needed re~iratory support. ECMO ( Oxygenator MOST-122, USSR) was used in11 patients with acute R!: in 5 patients as an extreme measure in the irreversible terminals t ate, and in 6 patients along with other measures of intensive care with mechanical ventilation (MV) in an extremely intensive regime ( high PEEP , toxic FvOo)and progressively reducing PaOo/FvO ~. During EChO PaOo in all the patients in~reas~d and it became possible to pass over to a less intensive MV regime. ( One patient survived smc~g 6 patients). Venous-venous ECMO effectively increases blood oxygenation and favcurably influences the pulmonary tissue metabolism and V/Q. Venous-arterial ECMO is advisable in such cases with concomitant cardiac failure. The low frequency of favourable outlets after EOMO may be explained by limitat~n of indications for it, the retarded beginning, hematologic and technical problems. In the conditions of progressive reduction of PaO~/ FTO ~ (FTO~I,0 ; high PEEP,dehydration, cardiac~the2a h PY2 St is advisable to apply ECMO before the development of critical hypoxia.

H.H. Steinhoff, K.J. Falke, L.F. Samodeldor, Institut fiir Anaesthesiologie der Universit~t Diisseldor f, Germany. An increase in intrathoracic pressure as a consequence of positive airway pressure leads to a depression of renal function. In the past this has been contributed to a rise in antidiuretic hormone secretion mediated by the cardiopulmonary stretch receptors. However, experiments interrupting afferent pathways of this reflex by cutting or cooling the vagus nerves, failed to eliminate the typical renal response to shifts in positive airway pressure. The influence of changes in intrathoracic pressure by intermittent mandatory and controlled mechanical ventilation (IMV /CMV) on renal function without and with a specific blockade of cardiac stretch receptors using 2 % procaine instillation in the pericardium (PLA) was studied in anaesthetized dogs. Institution of PLA during spontaneous breathing depressed all measured parameters of global circulatory and renal function such as urine flow, osmolar-, inulin-, and PAH-clearance. During mechanical ventilation (IMV/CMV) circulatory and renal function parameters further decreased and showed the most detrimental effects during CMV. However, IMV and CMV still caused similar changes in the kidney function as without the blockade of cardiac stretch receptors as described previously by us. Hence, although there is a marked influence of the stretch receptor volume regulating reflex on kidney function, this mechanism does not primarily govern the alteration in renal function induced by shifts in airway pressures. Fluid management in the critically ill patient on high levels of PEEP can be a difficult task. PEEP may cause the cardiac index to fall due to a decrease in left ventricular preload.

However, the high interthoracic pressures produced by PEEP negate the usefulness of the PAo as a measure of left ventricular preload.

Thus, one often has to give fluid challenges with the risk of worsening the acute respiratory failure for which the patient is being treated with PEEP. Inflation of the MAST compresses the venous capacitance reservoir and autotransfers 500-1000 cc's to the central circulation.

This autotransfusion is fully reversible by deflating the MAST.

The MAST were used as a reversible predictor of the effect of fluids on 14 critically ill patients being maintained on PEEP greater than 10 cm N20 with decreased cardiac index. Hemodynamic variables were measured before, during, and after the application and inflation of the trousers.

After MAST inflation, fluids were given in a quantity sufficient to maintain the same PAo after MAST deflation as was achieved with the initial MAST inflation.

The (1). Arterial blood pressure and cerebral blood flow (CBF) may be also modified. AIM : to quantitate the effects of PEEP increments on pha-s~ common carotid blood flow as index of CBF (2) and oxygen metabolism. METHODS : 10 patients (age 30 to 52 years) were studied during anesthesia for cerebral angiography. Relevant parameters were measured of calculated : common carotid blood flow (CCBF) (Pulsed Doppler flowmeter), arterial pressure (CAP), systolic (Qs), diastolic (Qd) flow fractions, systolic (Ts) diastolic (Td) times, systolic (Vs) diastolic (Vd) blood volumes, jugular pressure (Pjug) perfusion pressure (Pperf. = CAP-Pjug), vascular resistance (CVR = Pperf./CCBF), heart rate (HR), arterial and jugular blood gazes at ZEEP and PEEP 15 for arterio venous difference (A-VD 0~) and oxygen consumption (VO~ = A-VD 02 x CCBF). RESULTS-: gazometric and metabolic p~rameters remained unc-~. (3). Furthermore QD is maintained in spite of VD reduction linked to CVR (r=0.72 ; p < 0.001) due to cerebral volume increase (1). REFERENCES : (1) Contro~ed ventilation with PEEP increases intrathoracic pressures and decreases Cardiac output. In order to evaluate the influence of lung and thoracic compliances (CI, Ccw) on the hemodynamic response to PEEP, we studied 11 patients following cardiac surgery. 5 patients had mitral valve replacement (MV) and 6 patients aorto coronary graft (ACG). Static lung and total thoracic compliances were measured in patients sedated and curarised with 1.5 1 syringe and Ccw was computed as i/Ccw = 1/Ct + 1/CI. PEEP levels were increased fr6m 0 to 20 cm H~O with 5 cm H20 steps. For each level, oesophagal, pleuralZand pericardial pressures were measured using water filled catheters and hemodynamic data were recorded from a Swan-Ganz catheter.

PEEP (cm H20 ) 0 20 ACG -0.5 + 1.7 8.9 + 2.1 Pper (mm hg) MV -2.8 u 1.6 6,0 + 2.3 Paw -Ppl (mm Hg) ACG 1.2 u 0.9 9.2 u 1.5 MV 7.87 3.1 12.1 u 2.9 IC % ACG -18.1 u 3.5 HV 13.6 T 4.3 Cl and Ccw were both decreased in MV patients (el 32 + 6; Ccw 165 + 2b ml/cm HpO) compared to ACG patients (CI : 65-+ 6 ; Ccw 5BO ~ 95) and pericardial and pleural pressures-increased in the same range in the two groups of patients while increasing PEEP. The decrease in cardiac output was significantly more important in ACG patients. It is suggested that both C and Ccw could impair the hemodynamic response to PEEP by means of opposite effects on airway pressures transmission. Polymorphonuclear leucocytes constitute a major line of defense for the lung against the dangers of acute bacterial infection by virtue of their ability to phagocytize and kill the invading organisms. Since phagocytosis is associated with an increased phospholipid synthesis presumably serving in renewal and modification of membranes required for phagocytic function we asked thequestionis this function impaired in situations where other cell types may compete for the same substrates of phospholipid metabolism as occurs in the lung where type II alveolar cells are known to have a high phospholipid turnover. We therefore studied the rate of phagocytosis during increased phospholipid synthesis in isolated type II alveolar cells. Type II cells isolated from rabbit lungs were incubated in Krebs-Ringer phosphate (KRP) medium (pH 7.4) containing thyrotropin (TSH) for i hr at 37~

The same volume of solvent was added to paired control incubations. Type II cells were transferred into one of two compartments of a dialysis cell the other one harboring rabbit PMN-leucocytes in KRP-medium. After adding paraffin oil emulsion containing Oil Red 0 incubation was continued for i0 min. Metabolic activity of PMN-leucocytes was stopped with i mM N-ethylmaleimide. Oil Red 0 was extracted from washed cell pellets with p-dioxane and the OD determined at 524 m~. TSH stimulated the uptake of 3H-palmitic acid into type II cells by 60%. 57% • 4.98 (SE) of radioactivity was in lecithin. Particle uptake in leucocytes was 32.17% • 1.71 (SE) less during increased phospholipid synthesis in type II cells. We conclude that phagocytic activity of PMN-leucocytes can be depressed during stimulated phospholipid metabolism of type II cells. The occurence of slow (rate 121-180/min) and rapid (rate ~/181/min) ventricular tachycardias (VT) has been studied In 1348 CCU pts with an automated monitoring system developed at our institutions. Acute myocardial infarction (AMI) was diagnosed in 548 pts. The incidence of slow and rapid VT was 26 and ll%,respectively, in AMI, and 9 and 4%,respectively. in non-AMl pts. On admission, 3 pts with AMI, presented with VT (I slow), all electroconverted. In non-AMl pts 15 (I0 slow) had VT on admission (electroconversion in II). Five pts with VT (rapid; n = 3) after admission, were electroconverted. Thus most runs of rapid VT were short (less than 25 beats). About half of pts with rapid VT had their VT within 12 h after admission. Rapid VT usually developed before slow VT in pts with both forms.

In AMI, about I/3 of pts with rapid VT had their VT without prior signs of hypotension or left heart failure.Prognosis in this group was good. In AMI, rapid VT preceded ventricular fibrillation (VF) in 12 pts, in 5 of these first VT progressed to VF; in one pt VF preceded rapid VT.

In AMI, slow VT was associated with a hospital mortality of 14%, not different from figure in pts without VT. I17%); however, mortality in rapid VT was markedly raised (53%) and 3/4 of deaths occurred in the CCU.

In AMI, the incidence of rapid VT was strongly correlated to myocardial infarction size. In non-AMl, a history of myocardial infarction was related to the incidence of rapid VT. In the total material, a slight but statistically significant relationship between rapid VT and a serum potasium on admission below average was observed. Amrinone (A) is a new non-glycoside and noncatechol positive inotropic agent which probably acts directly on the myocardium. In dO male patients (mean age = 53.g ys) with coro, nary artery disease (CAD) hemodynamic and metabolic measurements were done in the arterial (a) and coronary venous blood before and q5 min. after administration of A (1.5 mg/kg body weight i.v.). Results (5) Although the origin of cardiac arrhythmias is no~ yet understood the role of delayed potentials in ventricular arrhythmias in connection with myocardial infarction has been of increasing interest in recent years. In order to get further information about delayed electrical activity, we studied a group of 50 patients in the acute phase of myocardial infarction with ventricular arrhythmias. Intracardial recordings were obtained in these patients using an intracardial bipolar electrode and an amplifier with different filter frequencies. In 30 of these 50 patients we recorded delayed potentials 150-200 ms after the beginning of ORS. 14 patients showed delayed electrical activity in the diastole even after the T-wave.

In several examples we could show that these delayed potentials initiated ventricular premature beats. 10 of the 50 patients suffered from ventricular tachycardia. These all showed delayed potentials after the T-wave.

In 3patients the onset of sustained ventricular tachycardia induced by delayed potentials were recorded. The same potentials obtained by intracardial recording techniques could also be recorded from the body surface by using a signal averaging technique. These delayed potentials can inhibit pacemakers. We found that the delayed potentials could be abolished by antiarrhythmic drugs. Thus there seems to be an important relationship between electrical activity and the occurence of ventricular arrhythmias. Direct quantitative assessment of parameters of tissue oxygenation and of reserve 02 transport utilization in critically ill patients have remained an unobtained goal. A new approach to these goals is the assessment of nonnutrient blood flow (NNBF).

A set of 4 equations interrelating oxygen delivery (DO2) , oxygen consumption (V02) , and NNBF have been derived: %NNBF=(DO2m/D02 n -i) + (vo2n/v02 m -i) x i00 for DO 2 increased and VO 2 decreased; %NNBF=(vo2n/vo2 m -i) -(DO2n/D02 m -i) x i00 for DO2 decreased and VO2decreased; %NNBF=(DO2m/DO2 n -i) -(vo2m/v02 n -1) x 100 for DO2 increased and consumption increased; %NNBF=-(DO2n/D02 m -i) -(vo2m/vo2 n -I) x 1O0 for D02 decreased and V02 increased.

The superscript (n) values are "normal" values for DO 2 and VO 2 estimated from basal metabolic rate for age and sex and measuring or assuming a value for the resFiratory quotient and assuming a value of (.25) for the oxygen utilization coefficient, lhe superscript (m) values are determined from a single set of measurements of C.I., pVO2 PaO2, and Hgb. Positive values imply "non-nutrient blood flow" or wasted 02 delivery. Negative values imply utilization of reserve 02 Lransport capacity. 63 estimations of NNBF in 16 cardiac patients were done. Graphic analysis of this data demonstrates only a loose correlation between CI and NNB~ with a value for r = .42. Determinations of CI are often used to guide therapy in the post MI or cardiac surgery period. This data suggests Lhat in this subset of patients other factors influenced reserve 02 uransport besides CI and that an expression for reserve 02 transport which integrates the factors responsible for 02 delivery and VO 2 into one expresszon may provide additional physiological information over that obtained from determination of cardiac index alone. I-UC, H DOSE GLUCOSE THERAPY OF CARDIOGEI~C SHOCK 3. Schrezenmeir, 3. Epping, M. Pfeiffer, S. Marcin, W.H. HSrl, Med. Univ.-Klinik, 8700 W~'zburg, FR(3 Glucose controlled infusion systems allow the treatment of cardiogenic shock with pharmacological doses of insulin. Methods: 12 critically ill patients in prolonged cardiogenic shock after myocardial infarction with fatal prognosis and exhausted therapy received a 2000 IF-insulin bolus and 2000 IE insulin per hour in the following. Blood glucose was automatically regulated with 70 % glucose sell,ion by a glucose controlled infusion system (Biostator GC[IS'~). Prior to and 30, 60, 120, 180, 240 min. after the beginning of therapy and 2,4 and l0 hours after therapy the hemodynamic changes, insulin, glucagon, pyruvate, lactate, citrate, 2-oxogluturate, 13-hydroxybuturate and aoetoacetate were determined. Results: Insulin therapy resulted in an increase of cardiac output, ~as significant (p < 0,05) at all times of determination and amounted 11,3 % after 30 minutes and 23 % after 4 hours. Heart rate remained unchanged except for the first 30 rain. where a significant increase of 5.6 % was recorded; according to this stroke volume attained values after 1 h which were 20 % higher than before therapy and remained at this elevated level during the four hours period of insulin treatment. Weiler index decreased significantly about 34 % indicating increased myocardial contractility. Beyond this we found a significant increase of central venous oxygen saturation, a significant decrease of lactate and a tendency of decreasing of citrate and 2-oxogluturate levels. Glucose and potassium blood levels remained in the desired range after substitution. After disruption of insulin treatment the hemodynamic starting level was reattained. Discussion: After exhausting conventional measures -including catee"-o'h~'amines -this therapy resulted in an improvement of cardiac index, myocardial contractility and chemical parameters in blood. The two receptor modetl of insulin effects may explain, why high insulin doses can achieve effects, which cannot be found after lower insulin doses. AR-L 115 BS is a new positive inotropic agent, structurally not related to glycosides nor to catecholamines. AR-L 115 BS was infused at 1,4 mg/min for 6 hours, then discontinued, after another 6 hours the drug was readministered for 6 more hours in 11 patients with severe heart failure (NYHA class IV). Hemodynamic measurements and plasma level were obtained before administration (C), hourly during both perfusions, and during the 6 hours following each perfusion period. The cardiac index (CI), the stroke volume index (SVI) and the left ventricular stroke work index (LVSWI) were significantly increased. The pulmonary artery wedge pressure (PWP) and the systemic vascular resistance (SVR) fell significantly. Heart rate and blood pressure remained unchanged (x = p < O.OO5). In the beginning of recompensation of congestive heart failure (CHF) it was possible to show that of 65 patients 9.2% had accompanying peripheral thrombosis of the deep veins of the legs (1-125 fibrinogen technique). Possible causes may be slowed blood flow or changes in hemostasis. a) Swan-Ganz catheterisation, heart volume by X-rays, liver enzymes, liver function (ICG-and BSP-half live), clotting tests (aPTT,Quick,TZ,fibrinogen) and antithrombin III (AT III, chromogene substrates) were determined in 43 patients with heart failure. AT III levels were compared with 2o healthy persons. b) AT III was tested for 14 days during cardiac recompensation of 19 patients with CHF.

AT III was statistically significant reduced in the group of 43 patients with heart failure. In the subgroup (n=17) with right heart failure and peripheral edema AT III Was 9.6 IU/ml, in the contro]groupof healthy persons 13.2 . AT III showed a.statistically significant correlation with right atrial pressure and liver congestion. Group B showed AT III reduced to 8.7 IU/ml. AT III was 11.5 IU/ml after one week of treatment and 11.6 IU/ml after two weeks. The difference was significant.

These results show changes of hemostasis in CHF by changed liver functions. This could be the reason of the higher rate of peripheral thrombosis in CHF, so that the testing of AT III could be a possible early indicator to prevent thrombosis in CHF. Cardiac output, however, was higher in 20/28 patients while in the supine position, We conclude that the change in body position per se has no beneficial influence on respiratory functions and sitting may decrease cardiac output as a consequence of reduced venous return from lower extremities and increased systemic vascular resistance. LV and aortic pressures (micromanometry), cardiac output, coronary sinus blood flow (CSBF), myocardial oxygen uptake (MV02), myocardial lactate balance and their derivatives were studied before (C) and 2, 5, I0 and 15 min after F (15 mcg/kg, IV) in 9 patients with coronary artery disease (Class 2-3 NYHA) undergoing heart catheterization. Results. F administration produced a decrease of mean aortic pressure (MAP) (significant at 2, 5, I0 and 15 min)

and of systemic vascular resistances (SVR) (significant at 5, I0 and 15 min) with maximal changes from C values observed at 15 min (-16 per cent for MAP) and at 5 min (-16 per cent for SVR). No change in cardiac index was observed.

The following changes in variables related to LV performance were observed: I) a transient increase in heart rate at 2, 5, and I0 min and LV contractility (Vmax) at 2 min (+ 16 per cent and + 9 per cent respectively) ; 2) a sustained (2, 5, I0 and 15 min) decrease of LV end diastolic pressure maximum at 15 min (-35 per cent). MVO was decreased at 15 min (-ii per cent), myocardial lactate uptake was increased (+ 37 per cent at 15 min) but CSBF did not change while coronary vascular resistance was lowered at 2, 5, I0 and 15 min with maximal change at 15 min (-18 per cent). Conclusion.

These findings, i.e., no effect on LV contractility with arterial and coronary vasodilation and decreased MVO 2 suggest that F may be useful for anesthesia of patients with coronary artery disease. A preventive accumulation of the myocardial glycogen stores should save the myocardial viability during hypoxia, since an anaerobic availability of energy is maintained. The therapeutic administration of high doses of insulin together with glucose and potassium achieves a shifting of metabolism us glycogenesis and glycolysis, providing an increased amount of energy rich phosphates. One half out of 16 patients undergoing mitral valve replacement received insulin I U/kg/h together with glucose 0.5 g/ kg/h from onset of anaesthesia until aortic cross clamping. After a 26 min average time of ischaemia a cut-out papillary muscle-tip was immediately plunged into liquid nitrogen and the content of ATP, ADP and CP (creatine phosphate) was determined. In contrast to the control group (Ringer's lactate) the insulin group showed a significant (p<o.oi) higher myocardial content of ATP (7.98 amol/g wet weight vs. 4.54 ~mol/g) and CP (6.78 ~mol/g vs. 5.16 ~mol/ g). Also the ATP/ADP-quotient (5.53 vs. 2.61) and the energy charge potential (ECP; ATKINSON) (o.919 vs. 0.857) was significantly (p 9 higher in the insulin group. It is concluded that the preventive application of high doses of insulin leads in fact no an increased myocardial energy provision during hypoxia and represents an enhanced ischaemic tolerance and an improved myocardial protection. There is a lack of controlled studies on the effect of different pump p~imes on glucose homeostasis. We report here the findings of a prospestive~ randomised study of hypothermic CPB with four different priming fluids~ namely non-glucose/ nonlactate, hi~h-gluoose/non-lactate, high-glucose/hi@h-R lactate and non-glucos~non-lacta te utilising Biostator continuous blood glucose monitoring. Such monitoring allows changes in BG concentration which are masked ~y intermittent sampling to be visualised. We have identified phases of the procedure which are associated with rapid and large changes.

A modest rise in BG occurred after skin incision in all patients (1.3 mmol/l +-0.3 SEN). Glucose+cOntaining, primes resulted in a marked rise in BG (25.5 -0.5 stool/ I) on inststution of bypass followed by a sharp fall to 15 mmol/l -2.5 which level was maintained during the hypothermio phase. With non-glucose containing primes there was a rapid, short-lived fall (2 + 0.5 stool/l) in BG due %o dilution on institution of bypass t followed by a steady rise during hypothermia. On re-~amming there was a further rapid rise in BG of c. 3.0 ~mol/l above 30 ~ C with all primes. A marked rise (3 ~0.4 stool/l) in BG was seen with all primes when sympatho-mime%is a~ents were infused.

Our results have clearly shown the magnitude end rapidity of cha~ges in blood glucose during and after CPB. This new knowledge is of crucial importance in the ma~ment of diabetics requiring this type of surgery.

Mitral s~rgery and pulmonary arterial hypertension (mean PAP superior to 40 mmHg) : post-operative hemodynamic study. C.Isetta*, J.J. Arnulf*, F.Mannot*, F.Montiglio, C.Dusaintpere. Service du Pr V.Dor. Institut A.Tzanck 06 700 Saint Laurent du var. * D.A.R. Pr Maestracci, Nice, France. The postoperative hBx~ynamic effects of mitral valve replacBnent are studied in patients with severe pulmonary arterial hypertension. This hemDdynamic study turns on fifty patients, (mean age 53 + 12 years) with a predominant mitral pathology without any assocT-ated coronaropathy. The preoperative status of these patients shows a cardiac index of 2,6 + 0,76 llmlm2 ; an ejection fraction (EF) of 0,57 + O, II ; pulmonary arterial pressures (PAP) : systolic : 79 + 17, dTastolic : 35 + 8, mean 50 + lO ~ ; a pulmonary capillar~ wedge pressure of 30-+ 5 m~Hg ; a Tight atrial pressure (RAP) : 9 + 5 m~Hg. ~ PAP, RAP,-left atrial (LAP) and arterial pressures Tre continuously menitored in postoperative time. Measures of cardiac output are carried out at the first, 12 th, 24 th hour. An hemodynamic control is made at the third postoperative week. For these fifty patients : the pulmonary arterial (PA) pressures register a significant {~crease of 4~ at the first hour, 32 % at the 12 hour, and 46 % at the third week. The LAP register a significant decrease of 50 % from the first hour and after doesn't change any longer. The RAP and the cardiac index don't differ from their preoperative values al] along the first 24 he~s. Six patie~ die after the eight postoperative day : their PA pressures only register a decrease of 29 % at the first hour and 23 % at the 24 th one, their mean PA stays at a rate of 40 mug. In the preoperative time, only a cardiac index which is lower than two I/me/m2, significantly differs. Five out of these patients present a preoperative tricuspid insufficiency (TI). Seventy eight % of patients presenting a TI linked with the mitral valvulopathy requiere the use of intravenous nitroglycerin and/or tonicardiac drug. The PA pressures progress in two phases. The PA pressures values of the first postoperative hour and the third week control are the same. After mitral valve replace, a mean PA pressure of 40 mHg has a poor prognosis. Two criteria of seriousness are noted : a CI lower than two I/me/m2 in the preoperative time and a TI liaV.ed with the mitral valvulopathy. Feasibility of propranolol (P) infusion to maintain chronic P therapy in the postoperative period was studied when patients cannot take drugs orally. Methods : Seven coronary patients (angina grade III NYHA, 56 ~ 4 yrs) undergoing intraabdominal surgery were selected. They received their last P dose (60-120 mg daily) at 8 pm the day before surgery (Po). Postoperatively P infusion started at about 2 pm (B) (normothermia and normal blood gases) with a first iv dose (0.2 mg/kg in i0 min) followed by a constant infusion (I mg/kg/day) until oral P could be taken again (24-66 hrs). Plasma P concentrations (C) were measured at Po, B and 5, i0, 60 min, 4, 6, 12, 18, 24 hrs after B. Haemodynamic data were simultaneously collected from radial artery and Swan Ganz catheters. Holter monitoring was performed before and after P infusion. Results : From 60 min to 24 hrs mean C (58-80 ng/ml) was in the therapeutic range (i). At B there was a statistically significant increase in heart rate (HR), rate pressure product (RPP) and cardiac index (CI), when compared to Po data. Beneficial haemodynamic results (decrease of HR and RPP compared to B, p<O.01) without any adverse effects (PWP and CI unchanged compared to Po) were observed over the period ranging from i0 min to 24 hrs. Ecg traces showed no ischaemic or rhythmic change. Discussion : In this study expected activity of P is achieved by this infusion way within i0 min over 24 hrs. As compared to a previous study (2) our method is able to achieve P therapeutic level more rapidly because of its starter dose and is performed during a longer time. Closing volume" (CV), closing capacity (CC), FRC, extravascular lung water (ELW) and pulmonary hemodynamics have been measured by the methods based on mass-spectrometry, pneumotachography, Swan-Ganz catheterization and thermodilution in 47 pts before and at the early period after valve replacement (VR). Postoperative changes in ELW were expressed by a ELW/FRC ratio because ELW depended on antropometric characteristics and effective alveolar volume even in healthy persons. The ELW/FRC increased almost twice during first postoperative hours compared with preoperational value and remained at the same level for 2-3 p.o. days. Close correlation (r=0,67) has been observed between ELW/FRC and CC/FRC. CC/FRC increased after VR and became more than 1,0 in most cases. A more pronounced CC/FRC and ELW/FRC increase was observed in pts with LV cardiac failure. Hemodynamic improvement (isuprel, dopamine, nitroprussid) and lung ventilation optimization resulted in a ELW/FRC decrease, gas exchange improvement, and allowed adequate infusion therapy. No dependance of ELW elevation on qua2ity and quantity of infused solutions and on pulmonary artery pressu -~ re was observed. An increase in plasma protein (from 45 to 75 g/I) by albumin infusion was found surprisingly low-effective for water recruitment from lung tissue,probably because of increased capillary wall permeability and elevated perlvascular colloid-osmotic pressure. The ELW increase after open-heart surgery resulted in an increase of pulmonary tissue specific weight and gravitational lung gradient; obviously that contributed to a CV increase, V/Q decrease and blood oxygenation disorders.

COMPREHENSIVE PRE-AND POSTOPERATIVE EVALUATION OF AS-CENDING AORTIC DISSECTIONS (TYPE I AND II) BY ECHOAORTO-GRAPHY. P.Bubenheimer, M. Schmuziger, Rehabilitationszentrum Bad Krozingen, West-Germany From 1977 to 1982 patients underwent operative treatment for ascending aortic dissections (type i: n=15, type II: n=7) by dacron graft replacement of the ascending aortaalone (n=7) or combined with aortic valve replacement (eomposite graft) in 15 patients. For preoperative evaluation and postoperative follow-up of the disease, we prefered echographic imaging of the aorta and its major branches, using a system of multiple short and long axis cross-sections obtained from parasternal, suprasternal, paravertebral, subcostal and abdominal echowindows. In 17 patients the dissection was primarily defined by echography, in 5 patients by angiography prior to the echographic examination. In 14 patients the operation was based on echographic findings without angiography. In all cases the echoaortographic findings were confirmed. Preoperative echoaortography demonstrated aseending aortic aneurysms in 21 (diameter 5-9 cm), aortic valve regurgitation in 15, perieardial effusion in 15, distal extension of the dissection to the aortic arch in 15, to arch branches -especially to the brachiocephalic artery -in 8, to the descending aorta in 14, and down to the iliac arteries in 4 cases. Serial postoperative follow-up (from I week to 5 years) of type I dissections (n=12 survivors) which were corrected only partially by the ascending aortic graft, revealed occlusion of the distal dissection in 8, persistence in 4, and progression in I patient. Descending aortic aneurysms developed within 2 years in 4 patients, leading to successful reoperation in I patient. Conclusion: Echoaortography is not only an excellent method for the diagnosis of aortic dissections, but may also improve the understanding of aortic pathology in chronic dissection by serial longtime follow-up. Systemic O2'extraction (EsO2) is impaired during severe sepsis, leading to a state of ischemia in muscles. To test the hypothesis that E. O2 may also be altered such 6 pts were studied G~sometric, metabolic (lactate) and hemodynamic investigations were performed using Swan-Ganz and Wilton-Webster catheters respectively for cardiac output (CO) and coronary si.nus flow (CSF) by thermodilution technics. Systemic and M 02 consumption (~O2, M?O2) and M lactate extraction (EwL) were calculated. The doses of Db were gradu~lly increased until the vanishing of peripheral signs of shock. The pulmonary wedge pressure was maintained to i0 mm Hg using volume expansion and dopamine (Dp) : 3 Ng/kg/ni~ The doses of Dp were increased to iO pg when the doses of Db were higher than 20 pg, in relation to vasodilatator effects of Db (ArP) has been known as a potent vasodilator. Purine nucleotides play a physiological role in regulating coronary blood flow. ATP may be released from intramural purinergic nerves, acting on purinoceptors in the vessel wall and therefore producing vasodilation. This study evaluates the possible role of ATP in producing deliberate hypotension in dogs. In 6 piritramide anaesthetized dogs, ventilated with 0J N20, a femoral artery catheter and a Swan Ganz thermod~lution catheter were introduced. Mean arterial pressure (MAP), mean right atrial pressure (RAP), mean pulmonary artery pressure (PAP), heart rate (HR)~ pulmonary wedge pressure (PCWP), cardiac output (CO), systemic vascular resistance (SVR), pulmonary vascular resistance (PVR) and stroke volume :(SV) were estimated. A continous i.v. infusion of ATP was given, starting with a dose of I mg/ kg/min and stepwise increasing it every five minutes to a maximum dose of 5 mg/kg/min. All data were analyzed by students' t-test, p~ 0.05 was considered significant. A dose dependent decrease in MAP, PAP, RAP, PCWP, SVR, PVR occured in all dogs inraediately after ATP administration. The decrease in MAPwas 55 % +/-8 % with 5 mg/kg/min. During the whole experiment CO and HRincreased, SV remained unchanged. ATP seems to be a possible agent for deliberate hypotension.

MITOCHONDRIAL SYSTEMS IN SHOCK STATES. G.C. Corbucci, A.Candiani,G.Crimi and A.Gasparetto, Istituto di Anestesiologia e Rianimazione, UniversitA degli Studi di Roma. The source of our results is a group of ten patients with a diagnosis of shock based upon hemodynamic and metabolic criteria together with the monitoring of tissular P02. The above shock states, either hypo or hyperdynamic, were studied with regard to the redox mitochondrial activity and the levels of some substrates of Krebs'Cycle in muscular needle biopsies. During the evolution of shock in the examined patients we have observed a significant reduction (one tenth of normal values) of the activity of cyt. oxydase (COx), of succynate cyt.c reductase (SCR) and of~-oxogIUtarate cyt.c reductase(OgCr). Moreover in all the hypodynamic shock states we have observed an irreversible "inhibition" of cyt.c oxydase. The tissue and plasma levels of some substrates of Krebs'Cycle (cytrate, succynate,malate,~-oxoglutarate,pyruvate and lactBte) have shown a close biofunctional relationship with the redox capacity of the electron transport chain (ETC). We could demonstrate a good correlation among tissue perfusion, intracellular 02 metabolization and mitochondrial oxydative capacity. Our results seem to open new and unexplored fields for the research on shock states in the human being. We have built a new display which measures the blood flow in the descending portion of the thoracic aorta with an oesophageal probe. The instrument consists of : -a Doppler directional velocimeter of continuous frequency emission, -an ultrasonic pulse echo system for measurement of aortic section. This work compares measurements of thermodilution cardiac output (TDco) with descending aortic blood flow (Ao DF 2). The cardiac output range was between 1.9 to 13.1 1 x min -I The correlation between theses methods, based on 300 simultaneous determinations involving 21 patients is the following : TDco : 0,473 = (1,110 xAoF) r = 0,977 p < O,001 The measurement of descending aortic blood flow by the ultrasound oesophageal probe here described is of easy realisation, non invasive and non traumatic for the unconscious patient.

The obtained values are well correlated with those of cardiac output obtained by thermodilution technic. The method has the advantage upon thermodilution of providing continuous recording, which is an important element in the prevention of haemodynamic complications occuring during surgery. I. Chavez I: "Introduction g~n~rale" in "Haemodynamie changes in anaesthesia", Pulmonary function alterations induced by left heart failure (LHF) are known to be related to h~nodynamic impairment. The purpose of this study was to assess the adequacy of high frequency acoustical respiratory impedance (HFRI) measurements by a forced oscillations method in evaluating pulmonary functional disturbances induced by an increased pulmonary capillary wedge pressure (PCW) in critically ill patients.

Repeated measurements of the HFRI and PCWhave been performed inpatients with acute LHF. Discriminant analysis on indices derived from the frequency dependence of the HFRI shows 96% of correct classification between PCW< 15 and ~ 20 n~n Hg. Correlations between PCWandHFPI indices have been ccraputed for patients with and without conccn~nitant chrcnic obstructive pulmonary disease (OOPD). Correlations as high as R = 0.9 (p < 0.001) have been obtained for non rapidly transient changes of PCW in both qroups of patients. Repeated measurements of the HFRI yield a suitable monitoring of the trend of PCW.

In ccnclusion, the sensitivity of the HFRI measurements allows a non invasive estimation of the PC~ at the bedside. The specificity of the HFRI is high enough to distinguish between LHF with and without conccn~nitant COPD. The measure~nent is easily performed and does not require any active cooperation of the patient. Pulmonary acid aspiration syn~ has been recognized as a major hazard in the practice of emergenoy anesthesia This r~sk has been tightly correlated with a gastric fluid pH of less than 2.5 and a volume of more than 25 ml. This study was designed to investigate the onset and duration of action of the new H2-receptor antagonist ranitidine in r~9/1.

After an overnight fast a nasogastric tube was placed in the stc~ach of 12 healthy male subjects. The gastric contents were aspirated and the starting pH was determined. 50 mg ranitidine as i.v. bolus or placebo (0.9% placebo) was given in a randomized fashion. Gastric juice was continuously aspirated over a test period of 6 hours, pH of the gastric aspirates and gastric volt~es were determined in l-hour fraction.

50 mg of ranitid/ne induced a rapid increase in gastric pH.45 rain after ranitidine all 12 subjects had a gastric aspirate pH of more than 5 re~aining at this level over the whole test period. By contrast, in the controls the pH-values ranged between 1.30 and 1.90 over the whole 6 hour test period.

Gastric vol~ne secretion in the ranitidine experiments was reduced from 60 ml/br in the first collection period to volumes below 35 ml/hr (i.e. 45-50 % inhibition), whereas in the placebo-experiments vol~ne secretion remained almost unchanged.

Our data reveal that 50 mg of ranitidine as i.v. bolus induces an immediat~ (within 30 min) increase in gastric aspirate pH and a concomitant drastic reduction of volume secretion in all subjects. This rapid onset of action allows a short time interval between administration of this H~-blocker and induction of anesthesia. The prolonged activity of ranitidine (for more than 6 hours) may indicate that at time of extubation the pH of gastric con tents is still high. Premedication with ranitidine may, therefore, be helpful in patients at risks. The liberation of Minins is generally considered essential for the induction of pancreatic shock, -An AHP was induced in 1 5 pigs by intraductal injection of taurocholic acid (TCA), The 6 controls were only sham operated. In another group of 5 pigs bradykinin (BK) was infused together with the kininase II inhibitor Captopril, Besides blood gas analysis lung biopsies were taken in the course and at the end of the experiments, -All pigs with AHP died of hypovolemic shock. The kallikrein levels increased immediately after the onset of AHP (~x: +75~), The total kininogen decreased about 70~, CO and CVP decr@ased whereas PAP and TPR increased (p<o.ool). PCWP remained nearly constant, Lung biopsies taken early in the course of AHP showed interstitial edema and lateron intraalveolar hemorrhagep alveolar eollaps~ microthromboli, and in some animals hyaline membranes, By infusion of BK a marked interstitial pulmonary edema could be generated in healthy anaesthetized pigs. TCA induced AMP in pigs leads to a significant increase of kallikrein and a decrease of kininogen which means a high liberation of vasoacrive kiniso By altering local capillary permeability they induce a high ascites production and contribute %o a histologically proved respiratory distress syndrom, We studied 126 cases of acute poisoning by phosphoorganic substances. This kind of pathology was treated by a specific therapy consisting in combinated use of cholinolitics (atropin) and reactivators of cholinesterase (isonitrosine).

We payed a special attention to symptomatic therapy aimed at decrease of hypoxia and its secondary effects, control of convulsious, prevention of inflammatory changes of the respiratory organs and control of pulmonary and brain edema. We used as obligatory the gastric and intestinal lavage, intensive ureapoiesis, hemoresorption, hemodialysis and blood transfusion. We noted a marked decrease in fatal outcome in this group of patients as a result of the treatment above described. Despite the broad use of hemocarboperfusion method in a reanimation clinic many problems bound up with the reliable criterions of estimation of effectiveness of this method remain unsolved. We studied the effectiveness of a single and multiple hemocarboperfusions on the dynamics of clinical, biochemical, electrophysiological and psychophysiological indices of hemostasis. 200 patients and 65 dogs were observed in the condition of acute intoxication with phosphoroorganic insecticides, alcohol and psychotropic agents. It was shown that we can speak of effectiveness of hemocarboperfusion, that is, of its curable effect in case of intoxication according to complex estimation of different indices. The most informative characteristics as well as clinical ones are bioelectrical brain activity, attention and memory function, indices of functional condition of sympathoadrenal, cholinergic, kallicreinekinin systems and acid-base. Criterions of effectiveness of a single hemocarboperfusion operation as well as determination of a number of necessary operations in cases of exogenetic intoxication include overall the bioelectrical brain activity, and then the indices of attention and memory function. Neurophysiological mechanisms are analysed on the basis of experimental data. 

TREATMENT IN 17 CASES. F. Beaufils, N, Schlegel, G. Pillion, C, Loirat, J.C. Mercier, H8pital BRETONNEAU, PARIS, We report our experience ~-ith 22 cases of RVT in infants (age I to 87 days). Etiologic factors were : hypertonic dehydration (n=11) perinatal anoxia (n=11). Diagnosis criteria were the following : clinical kidney enlargement gross hematuria, associated with non secreting kidney on intra venous pyelogram IVP and or anuria. RVT was unilateral in 10 cases and bilateral in 12 cases all with anuria.9 infants required peritoneal dialysis.Surgical thrombectomy was rejected, most of these RVT being mainly intrarenal. 5 infants were not treated (3 because of early death). 3 were treated by Heparln (H) and Streptokinase (SK) : loading dose 2000-2500 U/kg/20 minutes maintenance dose 750-1200 U/kg/hours 24-48 hours. 14 were treated by H and Urokinase (UK) : loading dose 4000 UCTA/kg/20 minutes maintenance dose 3000 UCTA/kg/hour 24-72 h. 2 of them received plasminogen 6 hours before UK. 4 infants (3 not treated I SK) died at the acute phase. Bilateral RVT was found in the 3 autopsied eases. ] infant (OK) died after 3 months of chronic peritoneal dialysis. 17 infants (2 not treated 2 SK 13 UK) survived. I (SK) kept end stage renal failure (RF) and hypertension and entered on hemodialysis programm at 6 years of age. I (UK) had mild RF (follow up 21 months). The 15 remaining patients (2 not treated I SK 12 UK) have normal glomerular filtration rate and blood pressure (follow up 5 to 48 months). Evolution of IVP was evaluated for surviving patients according to the overall number of kidneys which were altered in acute phase. 9 out of these 26 kidneys are normal. 9 are non secreting. 8 are atrophic. Finally 6 (I non treated 5 UK) out of the 22 patients fully recovered (normal IVP and renal function). Poor factors of prognosis were bilateralitT (5 deaths 2 BF) and dehydration (5 deaths I RF). The benefit of thrombolytic agents cannot be clearly evaluated because of the small number of patients. Further studies can be performed with UKbecause of its small risk of bleeding (0/14) if appropriate biological monitoring is performed. We report our experience about i00 cases of esophageal atresia admitted to our unit from 1968 to 1900. The aim Of this study is to appreciate respiratory and esophageal complications. 52 % of the infants were considered at high risk (in accordance with the French Society of Pediatric Surgery, classification 1969). 82 % survived after the surgical management of the esophageal atresia. The long term survival rate was 74 % : death oceured as a result of cardiac or multiple malformations, 3 to 36 months after surgery. 74 children were followed up prospectively for periods between 3 and 6 months, then every year 9 Data were obtained from physical examination and esophagography (performed at one year) 9 Other investigations, such esophagoscopy, tracheoscopy were performed in case of respiratory complication. Results : The following diagnosis were established : 9 chronic suppuration of the bronchus in 24 % 9 esophageal stricture in 28 % 9 gastroesophageal reflux in 23 % Correlation between respiratory and esophageal sequelae was noted 9 Our results stress the need for an early and systematic detection of esophageal abnoma]ities, particularly when responsible for the recurrent respiratory problems, and for an antiref]ux procedure 9 We have reported on the possibility of continuous non invasive monitoring of ICP in neonates and its clinical diagnostic relevance (I). It is our aim to develop a sensor which measures ICP with sufficient precision and which is inexpensive and easy to handle. Our latest developement uses a commercially available strain gauge pressure transducer with a flat surface which is in direct contact with the skin of the head above the fontanelle. This method requires the pressure transducer to be pressed against the head skin with a certain force. This is accomplished individually in each application by a step motor drive which can alter the depth of the transducer relatively to the sensor case. The action of the step motor is controlled by a microprocessor which also porvides an automatic routine to find the optimum position by analyzing the signal of the transducer in relation to its deflection. In order to investigate the influence of the skin membrane on the accuracy of pressure transmission between the intracranial system and the transducer, we applied membranes with various compliances to a water filled artificial head with a fontanelle opening. By varying the water pressure below these simulated "head skins" and recording the transducer signal, we found a linear function within the physiological range (10-50 cm water) for all kinds of membranes. We averaged the various slpoes and found a maximum deviation of 6%. For this reason, we assume the real head skin to have similar pressure transmission characteristics. Consequently a sensivity calibration is not necessary before in vivo application. 

Y. Bompard, S. Bobin, J.C. Mercier, P. Narcy, F. Beaufils H8pital BRETONNEAU -PARIs -FPANCE Intubation has been proposed as an alternative of tracheostomy for the airway stabilization in LTB. But this procedure for some authors, might cause secondary subglottic stenosis. The purpose of this retrospec~e study is to analyze the immediate and secondary outcome of 40 LTB treated by NTI. PATIENTS : lh girls (32%), 30 boys (68%), age g weeks to 12 years. Etiology : 9 measles, 10 bacterial proved infections, 10 viral infections, 15 unknown. Tube size was : for 21 patients, 1 size (0,5 mm) smalle~ than the recommanded size for age (I), for 16 patients~/ to the recommanded size for age and unknown for 3 patients. RESULTS : Mortality O;good outcome 37;subglottic stenosis 3 -one intubation(average duration 8h hours)24 patients -prolonged intubation (mean duration 180 hours) 13 patients : 9 unique reintubationk<9 days, 3 multiple intubations, I tracheostomy at 24"hours -delayed subglottic stenosis 3 patients needing I calibration which was successful and 2 tracheostomies (I recovered after 6 months. ] underwent a laryngoplasty after 12 months for failure to recover. Finally 37 (92,5%) out of the intubated LTB had a good outcome. Facilitating factors of subglottic stenosis appeared : measles (2/3)~and too large size tube 3/3. CONCLUSION : Intubation is a safe and efficient procedure in the management of sever LTB. Subglottic stenosis does not seemed to occur if the tube is one size smaller than the recommanded size for a~e. The French group of pediatric intensive care has prospectively studied 90 cases of infectious purpuras which were hospitalized in 1981 ; the purpose of this study was to determine prognosis factors.

The statistic study (X 2 test) of all these cases is in agreement with the litterature data and shows that the mortality is significantly higher when there is : a shock (p<0,001), a coma (p<O,05), an ecchymotic or necrotic purpura (p<O,01), a temperature <36C (p<O,05): no clinical meningism (p<O,O01), a white blood cells count<lO 000 (p/~O,05), a thrombocytopenia <i00 000 (p<O,01), a fibrinogen<l,5g (p<O,O0~ a kaliemia>5mEq (p<O,01), a spinal fluid cells count <20 (p <0,01).

Because the shock is one of the main prognosis factors (23 deaths in 55 shocked patients, vs 2 in 35 non shocked) we have done an other statistic study (with the BENZECRI Method ) to determine a pronosis index in shocked patients : for its determination five initial parameters are used :

-Age, kaliemia, white blood cells count , clinical meningism, platelets count.

Its sensibility is 91%, its specificity is 87%. Our department provides preoperative, postoperative, and general cardiologic care for children with heart disease between the ages of one day and 16 years.

We adapted the alarm limits and conditions of an arrhythmia computer (HP-78525) to our intensive care and long-term ECG monitoring problems and to the field of pediatrics.

For quality control, statistical analysis was used to examine all stored QRS complexes of each patient after the control by the cardiologist and to assess sensitivity and specificity results in a 4-field table. Continuous storage makes possible the detection of false negative as well as false positive results. The control by the cardiologist occurs through the EDIT and RECALL system.

The results obtained over a 24-hour period for 10 patient were presented and compared with the interpretations acquired with Holter monitoring.

The accuracy of QRS analysis ranged between 80%-99% for good ECG quality but dropped to 5% for instable ECG qualitiy for the same patient. Renal dysfunction (oliguria, renal insufficiency a.s.o.) after cardiopulmonary by pass for open heart surgery is a frequent and sometimes severe complication, especially in infants and young children. CaCI2 infusion is known to reverse these urinary troubles. The aim of this study is to try and determine the exact nature of these desorders and the mechanism (s) of CaCI2 infusion action. Six child~n (weight range 8 to 15 kg) with congenital heart diseases have beenLstudied before (period I) and after (period II) infusion of 0.5 mEq/kg b.w. of CaCI2. In one of the patients 2 subsequent studies have been performed. The following methods have been used : standard inulin clearance for glomerular filtration rate (G.F.R.)~P.A.H. clearance for renal blood flow (R.B.F.) measurements. Plasma total and ionized calcium, plasma citrate, urinary sodium (UNa) and osmolality (Uosm) were also measured. Plasma filtration fraction (F.F.) and urinary flow (~ were calculated. Results are summarized in Citrate values and GFR were in normal range and unchanged before and after CaCI2 infusion. Initial values for R.B.F. were low with increased F.F. and urinary osmolarity was elevated, CaCl2 infusion produced an increase of V with a decrease of U osm. Ca §247 tially in the normal-low range,iincrease(Jdramaticany. All these changes were statistically significant as est#~lished by a non parametric test (WILCOXON).

Ine aumors suggest that these phenomenons could be explained as follow ; -low PAH clearance with elevated FF suggest an insufficient RBF -the initial oliguria with elevated Uosm suggest a vasopressin hypersecretion -infusion produced hypercalcemia might induce a stimulation of PGE2.The latter is known to increase RBF, U and antagonize vasopressin tubular action -hypercalcaemia might aĩl sao have a direct inhibitory effect on vasopressin release, Hypercapnia is commonly associated with surimposed metabolic alkalosis (SMA), defined as an excessive increase in bicarbonate concentration, which can itself contribute to further increase the PaCe 2. We tested the hypothesis that correction of SMA by HCI infusion could decrease PaCe 2 in patients with acute ventilatory failure associated with SMA. Concentrated HCI (.25 N) was infused in 15 critically ill patients who presented either with a pH of A40-145 and a PaC02 >45 torr (group A, n=9), or with a pH of Z35-%40 and a PaCO 2 9 torr (group B, n=6). Ten patients had been treated by mechanical ventilation (MV). NaCI administration was either contra-indieated or ineffective to correct SMA. Serum potassium was 4.2 1.4 mM/L. Reduction of FiO 2 had not resulted in any decrease in PaCe2. HCI was infused into the superior vena cava at the rate of lO0 ml/hour until the bicarbonate concentration was below 26 mM/L, or the pH below ~35 (group A) or 730 (group B). The decrease in PaCe 2 was still significant 12 hours after the end of the HCI infusion. In our opinion, these effects on PaCe 2 contributed to the successfull weaning in 7 of the ]0 patients treated by MV, and to a brisk clinical improvement in the other 5 patients. No hematological or vascular complication due to HCI was observed.

We conclude that HCI infusion can significantly improve ventilation and oxygenation in patients who present with hypereapnia and SMA. Aggressive correction of SMA should be actively considered in the management of acute ventilatory failure in critically ill patients.

W.Druml,A.Laggner,K.LenztW.Base,G.Kleinberger;I.Med.

Alka~osis may induce hyperlactemia representing a mechanism of metabolic compensation of the alkalotic State. It is however not known whether alkalosis per se or hypocapnia and whether an increased production or a decreased hepatic utilisation of lactic acid causes the rise of plasma lactate concentration (=LAC).We therefore investigated the influence of alkalosis on the elimination of parenterally infused L-lactic acidT Patients and methods:6 patients(3 2,3 ~,mean age 57.5 a) with respira[ory failure were evaluated, l mmol/kg b.w. L-lactic acid was infused in io min.LAC and blood gas status was determined befere,2,3,1o,2o,3o,6o and 90 min after the infusion.The study was performed in a basal condition with noinnal pH (7.44+O.02)and a pCO of 39.2+

--2 --6.33)and during controlled mechanical hyperventilation and respiratory ~alkalosis (pH 7.59+__O.03)and hypocapnia (pCO 2 29.3+_4.1). Results:Basal LAC was 1.87+0.37 at normal pH and 2.6+ 0.37 mmol/l in alkalosis(#~o%).The elimination half~ime was 2.55+0.25 min at normal pH and 11.04+3.34 min in alkalosis(e320%).The clearance rate decreased from 64.23+ 38.4 ml/kg b:w./min to 25.7+5.01 in alkalosis (-6o%). Despite constant pH and pCO 2 LAC decreased below basal values after 9o min 0.37 mmol/l in normal pH(=m2o%)and 0.4 mmol/l in alkalosis(=-15%)~ Conclusion:Hyperlactemia in alkalosis is caused mainly by a decreased utilisation and is mediated by the pH value and not pCO2.The pH optimum for lactate metabolisation seems to be below pH 7.44.As a clinical implication alkalisation therapy in post-shock states should be performed carefully since the induction of alkalosis may result in an impairment of lactate removal.- Klinik, 8700 W{]rzburg, FIRG In acute disease, catabolism, a negative nitrogen balance and the Joss of funtional proteins remain problems facing parenteral nutrition. When sugar substitutes are used instead of glucose, metabolic disorders can only be detected by an increase in secondary metabolites. The problems of high dose glucose therapy such as variations in osmolarity, alterations in hemoglobin-O~-dissociation and impaired substrate utilization, may be avoided t~y using the artificial pancreas (AP).

24 critically ill patients who received high dose parenteraI glucose therapy and who had blood glucose concentrations of more than 500 mg% and daily variations of more than 500 mg% inspite of insulin per infusion, were placed on the AP. The insulin infusion rate was set to maintain blood glucose values between 150 -200 mg%. With the AP mean blood glucose levels of 198 + 79 mg% were obtained, compared to 552 + 149 rag% (p ~ 0,01T with conventional therapy. Although more ~ucose was infused, the maximal serum glucose amplitude per day was significantly reduced (115 + 97 mg% compared to 262 + 131 mg%; p~0,01) and a s|gnlf~ant fall (p< 0,05) of lactate-blood levels from 2,45 + 1,15 to 1,25 _+ 0924 mg/l with unchanged la~tate-pyruvate ratio and a decrease of citrate and oxo-gIutarate concentrations indicated a better fuel consumption. After attaining a dynamic equilibrium, insulin glucose ratio was 13.2 + 21.1 IE insulin/g glucose and could be used for further therapy without the AP. This parameter of insulin resistance correlated with insulin levels at equilibrium but not with proteolyric activity (digestion of azocasein and phosphorylase kinase} titration of protease inhibitor capacity of plasma, 0(-2-macroglobulin, ~c-l-antitrypsln). Thus the marked resistance in sepsis could not be explained by immunologically detectable insulin degradation or by alteration of the proteolytie activities and protease inhibitors we determined in plasma.

The AP enables us to administer high caloric parenteral nutrition to critically ill patients without the hazards of conventional therapy and to study the mechanisms of insulin resistance in these patients.

Baccaglini, M.Calabrb,E.Pizzinari, C.Tremolada, G.P.Giron A. Tiengo. Universit~ di Padova Italy. Hypo-and hyperglycemic events and acute metabolic derangement are common complications in patients undergoing total pancreatectomy. We evaluated the effect of glucose control with Biostator on intermediary metabolism. 5 non diabetic patients affected by pancreatic adenocarcinoma (age 58• ideal body weight 93• were studied throu ghout operation (S) and 24 H post operative period (PS).-3 patients (PXB) were put under Biostator control, while 2 patients (PXC) were treated according to a conventional protocol. Arterial blood samples were frequently taken for hormone and metabolite determination. In PXB glycemia during S and PS was 11~33mg/dl and 290• in ,PXC. C-peptide became undetectable in all PX (3~9• to <O.2ng/ml). Insulin levels were maintained appropiate in PXB (7~12 ~U/ml in respect to PXC I~3). During S lactate and pyruvate increased similarly in both groups (PXB 2612• PXC 253~700 umol/l), but a marked peak was observed in a PXC (lactate 7169; pyruvate 293~moI/I) at the beginning of PS. Alanine was similar in both groups during S (PXB 20~19, PXC 24~25 ~mol/l). During PS alanine was higher in PXC (35~68) than in PXB (24~38). The most effect of insulin replacement was observed on glycerol (PXB 6%14, PXC 218~8umoI/I) and ketone bodies (PXB 8~14, PXC 567 • ~mol/l). In conclusion: Biostator maintained physiologic glycemia without hypo-and hyperglycemia. Normoglycemia was related to metabolic control avoiding ketoacidosis and hyperlactacidemic risk. Futhermore, reduced levels of alanine in PXB may suggest reduced preteolisis.

UNIT. V. Ninane, M. Marchal, J.L. Vanherweghem, J.P. Degaute, R.J. Kahn. Erasme Hospital, University of Brussels, Brussels, Belgium. Among 2319 patients admitted in a medico-surgical intensive care unit in 1980 and 1981, i01 developed acute renal failure (ARF : blood urea > i00 mg/dl, blood creatinine > 2.5 mg/dl). ARF was related to surgery (n = 33), to one of the following medical insults : hypovolemia, cardiac failure, sepsis, rbabdomyolysis, exposure to nephrotoxic agents (n = 36) or to several of these factors (n = 32). Median age was 62 years (range, 19-89). The management of these patients included fluid challenge and/or furosemide (n = 85), mechanical ventilation (n = 73), parenteral nutrition (n = 36), hemodialysis (n = 32) and peritoneal dialysis (n = 2). Oliguria (urine flow < 400 ml/day for 3 days) was present in 38 patients. Sixty-nine of the i01 patients died. Mortality was not significantly affected by the association of several etiological factors and by the medical or the surgical setting of the patient.

Incidence of non oliguric renal failure was not influenced by fluid challenge or diuretic administration:

Dialysed and non dialysed patients have a similar mortality rate. Patients who presented with a blood urea > 200 mg/dl during hospitalisation have a comparable mortality to other patients.

Mechan Increased survival was associated with younger age, non oliguric ARF and the absence of mechanical ventilation. Our data suggest that the mortality of ARF in critically ill patients is more related to the general status of patients than to the severity and management of uremia.

UNCOVERED T-ANTIGEN AS PATHOGENETIC FACTOR OF ACUTE RENAL FAILURE.G.Lenz, H.Junger, U.Goes, D.Baron, and U.Sugg. Institute of Anesthesiology, University of TUbingen. F.R.G. The enzyme neuraminidase produced by some viruses and many bacteriae unmasks a cryptantigen(Thomsen-Friedenreich-Antigen, T-antigen). on the membrane of red cells and other cells.The T-antigen react with T-agglutinins present in all human sera resulting in a damage of these cells.lncreasing attention has been focused on the in vivo action of neuraminidase as pathogenetic agent for the hemolytic uremic syndrome (HUS) in phildren, Therefore, T-activation of red cells was investigated in 60 surgical ICU patients with suspected sep'ticemia and positive culture results. T-activated red cells were detected by apositive agglutination reaction with peanut agglutinin anti-T (lecitinof Arachis hypogea) and by an indirect hemagglutination Ah test using rabbit antiserum against anti-T. In 17 patients a significant uncovering of T-antigens on red Ah cells was detected in vitro. In 5 of these patients an acute renal failure developped,whereas only 2 of 43 T-negative patients had renal problems. There is strong ~vidence that the unmasking of T-antigens is not only limited to red cells but also does take place in the renal capillary endothelium.According to these results the neuraminidase induced activation of T-antigens may be an important factor in the etiology of acute renal failure in adult septic patients. We therefore suggest that impairment of cell immediated immunity associated with ARF is at least partly due to protein malnutrition. TBP, simple to assess, seems to be a useful tool as it is correlated with some immunologic values. A feasibility study of the efficiency of Dipyridamole (DPM) was made to find out how to prevent clotting of the extra-corporeal circuit during hemodialysis (HD) in 13 high-bleeding risk patients (pts). The bleeding risk factors were active gastro-intestinal bleeding (7 pts), aeute head trauma (3 pts), haemoperieardium (2 pts), others (3 pts). There were 2 bleeding faetors in 2 pts. DPM was infused (25 mg/h) in the "arteria~ line of a unipuncture HD cireuit filled with normal saline. Four benign bleeding complications (6 %) and no neurologieal deterioration were noticed among 66 courses of HD. A clotting of the circuit (C.C.) needing to stop HD.(without any dangerous consequence) occured in 12 courses (18.2 %). A platelet count above 100 O00/mm3 was associated with a 23.3 % oecurenee of C.C., whereas there was none if the count was below 100.OOO. However, this differenee is not statistically significant (0.05 P 0iiO) Bleeding time before HD, the prerinsing of the circuit with heparin and infused doses of saline were not linked with frequency of C.C.

CONCLUSION : I) HD with DPM is feasible. 2) We observed a higher occurence of C.C. than with previously reported methods using either regular saline wash-out or prostaeyclin. 3) This increased oeeurence may, in part, be due to our circuit (unipuncture, low blood-flow rates). 4) So DPM deserves a comparative trial against other methods of HD without heparin.

THERAPY RESISTANT BRONCHIAL ASTHMASUCCESSFULLY TREATED WITH PLASMA EXCHANGE. R. Bambauer +, M. Austgen ++, K. Mic,ka ++, P. Schlimmer ++ und F. Trendelenburg ++ +) Department for Nephrology, ++) Department for Pneumology at the Medical University Hospital, D-6650 Homburg/Saar, W-Germany Therapeutic plasma exchange (TPE) experienced rapid world wide dissemination on the introduction of hollow fibre membranes. New areas of application where TPE combined with conservative therapy appear very promising as a symptomatic treatment procedure are increasingly reported.

In 1978 GARTMANN (Lancet 1978) reported for the first time on the successful use of TPE for bronchial asthma. We were able to considerably improve the condition of 2 patients with therapy resistant bronchial asthma using a technically simplified TPE system. A 33 year old patient with medication dependent, therapy resistant bronchial asthma suffered no further attack after only 1 exchange treatment. The condition of the patient improved so much that she was discharged after 7 further treatments. The extremely high level of IgE decreased under plasma exchange and the pulmonary function also considera'bly improved. A 52 year old patient with therapy resistant bronchial asthma also showed a distinct improvement after only 3 treatments.

The technique, the exchange method of plasma filtration and the course of the diseases treated are reported. The role of colloid-osmotic pressure (COP) in determining intravascular forces in health and disease is generally accepted. In plasmapheresis the Dlasma is separated and has to be replaced by volume-substitution. This therapY has to be iso-colloid-osmotic to prevent patients from renal failure or pulmonary edema. Material and Methods: 9 patients underwent therapeutic plasmapheresis for various reasons. 25 plasmaseparations were performed. COP, albumin(ALB), total protein(TP) and hematocrit(HK) were determined before and after Dlasma-Dheresis and calculated as % increase(+) or decrease(-). In group A 10 Dlasmaseparations were replaced by 5% human albumin, group B: 8 nlasmaseoarations received 10 or less packs of fresh-frozen plasma(FFP) and in group C received plasmasenarations more than 10 oacks of FFP Discussion: After volume-substitution in plasmapheresis no significant change in volume infused, volume seDarated COP, ALB, TP, HK could be documented between the 3 groupm Only ALB increased slightly in groups A and B. Measurement of COP seems to be a reliable parameter for guiding volume-substitution therapy in plasmapheresis. Furthermore it is easy performable and quickly obtainable in modern intensive care units.

AND ACUTE RESPIRATORY FAILURE, L.Holzapfel, P.L. Blanc, L.Thomas, A. Mercatello, M. G6rard, D. Robert, A.Bertoye. H6pital de la Croix-ROusse, Lyon, France. Patients with raised intracranial ~ressure (ICP) may require ventilation with positive end expiratory pressure (PEEP)if acute respiratory failure (ARF) is associated. But PEEP can increase ICP and reduce arterial pressure (AP) ; these changes could critically reduce cerebral perfusion pressure (CPP) and initiate neurological deterioration.

10 patients with coma and raised ICP were studied;age was 37* 18 y; 3 had meningitis, 3 encephalitis, 4 head trauma. During 5.6• 3. I days, PEEP was 12.4 ~ 3.2 cmH20, FI02,47~5, Pa02 83 ~ 19 mm H~, body temperature 37.9 +1.3 ~ C and body weight was increased by 5.2 ~ 2.3 kg due to fluid therapy for hemodynamic support . ICP was monitored in 8 patients with an epidural pressure-transducer (Philips) and in 2 With an intraventricular catheter. Technics for reducing ICP were : -hyperventilation (PaC02 30 + 4 mmHg~ adaptation to the respirator with pancuronium ~omide -drugs given on a regular basis or on an+emergency basis when ICP was ~ 25 mm H _' 9 day and Pentobarbital (P) The prognostic value of somatosensory evoked potentials (SEPs) was evaluated soon after onset of coma secondary to brain anoxia, brain ischemia after cardiac arrest or head trauma. The SEPs were correlated with neurological state and SEPs were followed over several weeks in surviving patients. SEPs were elicited by unilateral stimulation of the median nerve at the wrist. SEPs were recorded at the wrist, Erb's point (NIO), the neck (N15) and the contralateral parietal (PI5-N20-P25 complex) and frontal scalp regions using a Pathfinder II (8 channels) from Nicolet Biomedical. The analysis time was 50 msec. Our group of 30 patients with acute anoxic-ischemic brain injury showed that no patient with major abnormality of the N20-P25 complex recovered; they remained in a vegetative state and SEPs remained grossly abnormal. This predictive value of the abnormality of the N20-P25 complex is not found in the group of patients with acute head trauma. Major abnormality of the N20-P25 complex bilaterally hours after trauma did not exclude recovery of higher cortical function after several weeks.

Zuccarello +, H. Chiaranda, G Trincia § G P. Giron. Institute of Anaesthesiology and Intenslve Care, Dept. of Otolaryngo logy-sect, of Audiology ~ and Dept. of Neurosurgery +, -University of Padua, Italy. Auditory brainstem responses (ABRs) were recorded and ana ly~ed in 35 patients with severe head injuries, who we~ admitted to the intensive care unit] of the university of Padua over the last 12 months. The patients were selected as follows: a) age range from 15 to 60 years; b) presence of a supratentorial post-traumatic lesion and absence of primary brainstem damage at the CT scan; c) absence of temporal bone fracture at the X-ray skull films;d) absence of an anamnestic otologic pathology; e) normal canal otomicroscopy with normal tympanogram. The level of consciousness was defined using the Glasgow Coma Score evaluation and the brainstem disfunction with a clinical assessment of brainstem reflexes and posture. The CT scan findings were classified in four grades, rela ting to the onset of descending transtentorlal herniatio~ The ABRs were recorded and analyzed many times according to the clinical evolution from the moment of admission to the final outcome. Patients who died from causes unrelated to the brain were excluded from the analysis. The results show that a severe br~in~em disfunction, as defined by ABRs, is closely correlated with high mortality while a normal brainstem function is indicative of a good prognosis. ABR seems to provide more reliable information about brainstem damage than the neurological signs and is closely correlated with CT scan findings. Information from these techniques may thus supplement that obtained from clinical examination, allowing early prediction of final outcome.

21 patients in acute anoxie coma after cardiac surgery with cardiopulmanary bypass (9 patients), after cardiac infarction (7 patients), accidental anoxia during anesthesia (2 patients), hemorrhagic shock (I patient), drowning (i patient) and poising by fumes (1 patient) were examined with short-latency median nerve somatosensory evoked potentials (SEP) and EEG.

For the SEPs, the amplitude ratio (R) between the first negative response recorded from the contralateral parietal scalp (N20 -P25) and the negative response from the neck at the level of C2 (N14), and the latency difference between the contralateral scalp response and the eervikal response (Central Conduction Time, CCT) were used as parameters. For the EEG, visual inspection was performed only.

A significant difference of the mean value of R was observed in the group of patients with anoxic coma with a measurable response, as compared to an age-and sexmatched group of healthy volunteers (p~: 0.005, T-test of the means). No significant difference of the CCT was found between the two groups. Furthermore a significant relation betw~n the mean value of R of the 2 hemispheres and the clinical outcome was also found (p ~ 0.002, Fisher's exact probability test).

The only EEG pattern with a clearly dismal prognostic significance -other than the pattern of electrocerebral silence -were suppression bursts. Amplitude of the cortical SEP is therefore considered as the most promising prognostic parameter in acute anoxic coma 9 It is well known that rebound Nypertension following discontinuation of Sodiumnitroprusside can be hazardous and possibly outweigh the benefits of deliberate hypotension, mainly at neurosurgical interventions, thus emphasizing the necessity of an adequate posthypotensive therapeutic management. In a controlled clinical study we compared the effectiveness of Propranolol with that of a new antihypertensive drug Urapidil on 36 patients subjected to controlled hypotension in the course of neurosurgical interventions. In 24 patients intracranial interventions were performed, whereas 12 patients (volunteers) with lumbar disc operations served as controls. One half of each group was treated with either Propranolol or Urapidil pre-and intraoperatively, the desired hypotensive level was 66% of the control value. At certain intervals we neasured Plasma Renin activity, Catecholamines, free fatty acids, arterial blood gas tensions, electrolytes, hemoglobin, P.C.V., whereas NABP and endtidal CO 2 were continuously recorded. Results: 1.-Posthypotensive hypertension following discontinuation of SNP may be due to increased levels of P.R.A. and Catecholamines. 2. In the control group Urapidil was more effective than Propranolol in prevention of hypertension. 3-Urapidil considerably reduces the required amount of SNP. 4. Drug intracranial interventions both drugs were equally effective and prevented hypertensive reactions in 65% of all cases.

CEREBRAL TISSUE ACIDOSIS DURING SODIUNNITR0-PRUSSIDF-INDUCED HYPOTENSION AND ITS PREVEN-TION BY THIOSULPHATE. H.Guggenberger, D.Heuser, P.J. Norris*, Departments of Anaesthesia, Universities of TGbingen (FRG) and Leeds*(UK) Sodiumnitroprusside (SNP) may be regarded as drug of choice to electively induce arterial hypotension, especially when very low levels of systemic blood pressure (NABP) are desirable. However recent animal studies have indicated a progressive development of cerebral e.c.f, acidosis during administration of the drug even when normal CBF values are maintained and the critical dosage of SNP is not exceeded. In order to elucidate whether this acidosis is due to cerebral cyanide intoxication, we tried to prevent development of e.c.f. acidosis by simultaneous administration of thiosulphate in 8 cats out of J9 animals subjected to controlled hypotension with SNP. E.C.F. pH was measured continuously with pH microelectrodes implanted in the cerebral cortex of the animals. In both groups of animals MABP was lowered to 30 mmHg for 30 minutes, followed by a period of 15 minutes with 28 mmHg. The results indicate that atSNP dosage of <dmg/kg severe e.c.f, acidosis develops directly a~erthe onset of SNP administration which can be prevented completely by thiosulphate treatment. In the thiosulphate group development of e.c.f, acidosis did not start before NABP was lowered to 50 mmHg. We conclude that during SNP administration to electively induce arterial hypotension the simultaneous administration of thiosulphate is necessary to prevent cerebral lactacidosis, even when the"critical"SNP dosage is not exceeded.

Margaret HEMMER g Robert HEMMER, Departments of Surgical intensive Gate and of Infectious Diseases. Centre Hospitalier Luxembourg -LUXEMBOURG.

We studied retrospectivel V 91 consecutive patients with severe head injur V ( Glasgow coma scale 3 to B, most under G ) to determine the incidence and the outcome of infection and the value of prophv1actic antibiotics. Mean age WaS 32 veers ( range 1-7G ); ell patients had aorticnsteroid treatment end 45 were induced to barbiturate comb for ICP control. 59% received prophvlsctic antibiotics, mostly Penicillin G, sometimes in association with Gentemicin. 45 patients (32 with prophvlectic antibiotics and 13 without) had one or more episodes of infection. Tvpee of infection were: 27 Septicemias, 22 urinar V tract infections, 14 Sinusitis, ll earl V Pneumonias, 3 late Pneumonias, 3 Meningitis, 30titis, 2 Wound infections. Microorganisms responsible for severe infections (i.e. septicemia, pneumonia and meningitis) were: 9 S.epidermidie; G Streptococcus faecalis; G E.coli; 5 S.aureus; 5 Peeudomonea sp. (among them one P.maltophilie); 5 Proteus sp.; 1 Bacteroidee fregills; i Acinetobacter calcoaaeticua ver. anitratus; 1 Flavobecterium ap.; 1 Corvnebacterium "group J M "; 1 Candide albicans. In 12 patients infection was leading to death. All microorganisms responsible for severe infection were resistent to the prophv1actic antlbiotios administered.

We conclude (1) Nosocomial infection oocured in 49% of our patients with severe head injur V. (2) Prophvlactic antibiotics did not prevent infection. (3)The high proportion of opportunistic and often multiresistant microorganisms ( 5.epidermidle, P. meltophi-liB, Acinetobecter sp., Flsvobecterium sp., Corvnebacterium " group J M ", Osndlda albicans ) suggest that severe heed trauma patients are compromised hosts at high risk for nosscomial infections. We studied, between october 1980 and august 1982, 18 cases of pneumonia which required mechanical respiration and in which CMV was involved. Antibody titers were deter mined by complement fixation, immunofluorescence and ELISA (IgG and IgM for the last two methods). In most of the patients, a check-list for pneumonia in immunocompromised patients was realised by means of bronchic endoscopy (bacteriology; Myco-bacterias, Pneumocystis, Mycoplasma, Chlamydia, Legionella, mycology, virus). CHV were is~ lated on human embryonic lung diploid cells. Nearly all the patients had underlying diseases (hematol gic diseases, marrow transplantations, lung or heart transplantation, plasmocytom, systemic lupus erythematosus, great obesities, chronic respiratory failures, carc~ nologic or vascular surgery, intestinal amibiasis). The chest x-ray often showed a bilateral alveolar pattern, an intersticiai pattern was seen in 28 per cent. Pulmonary pathologic findings showed cytomegalic inclusions cells in 2 of 8 cases, and once in myocard. ELISA method had the best sensitivity, but 6 of 11 patients in which CMV was isolated in the lungs presented no seroconversion.

Other infecticas diseases Were associated in i0 patients. 11 patients finally died. The average duration of mechanical respiration was 5 weeks in the survivors. This study emphasizes the role of CMV in severe pneumonias, often associated with other pathogens. We underline the usual absence of specific symptoms and frequent negative serologies in immunocompromised hosts.

PULMONARY ARTERY CATHETER-RELATED SEPSIS. W. Kaye, M. Wheaton, G. Potter-Bynoe. Brown University-The Miriam Hospital, Providence, Rhode Island, U.S.A. Several reports have documented that' the risk of catheter (cath)-related sepsis for intravenons caths increases after 3 or 4 days (d) of cannulation. However, there is a paucity of data examining the relationship between duration of cannulation and oath-sepsis from pulmonary artery (PA) caths. The purpose of this study was to determine the overall incidence of PA cath-related bacteremia and the relationship between length of time in place and incidence of infection.

From March through August, 1982, 239 PA caths were inserted.

All were followed prospectively, but complete data were available for 120 PA caths (in 107 patients) which were inserted percutanecusly into internal jugular or subclavian vein. A rigid protocol was followed to provide aseptic technique for PA cath insertion, inspection and care of insertion site every 48 hours, and removal. After removal, the excised oath tip was rolled across a blood agar plate which was then incubated at 37 ~ All colony types were enumerated and identified. Following the criteria of Maki et al, 15 or more colonies (pos cult) were considered to signify infection. Cath-related bacteremia was determined by isolating the same organism from the blood. The incidence of pos cult was 9% (Ii/120). Non-fatal bacteremia occurred in 2.5% (3/120) overall or in 27% (3/11) pos cult. The incidence of pos cult relative to duration of qannulation was: 1 d, 2.5% (1/40); 2 d, 2.9% (1/35); 3 d, 21% (4/19); 4 d, 0 (0/8); 5 d, 11% (1/9); 6 d, 40% (2/5); more than 6 d, 50% (2/4). The incidence of pos cult for 2 days or less was 2.7% (2/75) vs. 20% (9/45) for three days or longer. We conclude: (i) there is a real risk of infection (9%) and bacteremia (2.5%) for pulmonary artery catheters;

(2) after 2 days in place, the risk of infection increases sevenfold. However the initial localisation of these germs cannot always be determined. Seventeen tracheotomized patients were studied in order to examine the relationship between cro-pharyngeal flora and broncho-pulmonary secondary infections. All the patients were mechanically ventilated for reasons other than pulmonary infection. Buccal, pharyngeal and bronchial samples were collected within 24 hours of tracheotomy.Durlng the study-period 9 patients developed secondary pneumonia as defined by clinical, biological and radiological criteria. The results show: I) a colonisation of the buccopharyngeal cavity with a high prevalence of negative gram bacilli 2) an early tracheo-bronchlal colonisation by at least one the oro-pharyngeal germs in 16 out oE 17 patients. 3) that the germ held responsible for the superinfection in the 9 cases of secondary pneumonia was always present in the initial buccopharyngeal samples (Proteus:3; Klebsiella: 2; E. coli: 1; pseudomonas Aer.: 1; enterobacter: 1, Serratia: I ). Given the above circumstances, it is reasonable to perform an initial clinical exam as well as systematic bucco-denta] and nasopharyngeal care in order to reduce local bacterial growth. Synergistic bacterial infections are involved in most ex tensive cellulitis oceuring in patients treated by routine anLibiotherapy. This study was performed to evaluate the value of early bacteriological diagnosis snd adequate anti biotherapy CAB)on the prognosis. Patients and methods. Eighteen patients with extensive gan grene were studied (12 men 6 women). Mean age was 57 years (24-82). Eleven patients had signs of systemic toxicity ; mecehanieal ventilation was required in l] patients. Anuric renal failure observed in i patient wss treated with hemodislysis. According to the location of gangrene, patients were divided in 4 groups : perineal (8), abdominal wall (2), limb (6), neck (2) eellulitis. Results. Eleven of 18 patients had been treated before ad mission with penieilline G ossoeiated with imidazole deri vate (n=9) and amimosides (n=6). Four patients developped eellulitis despite treatment withampicillin. Bacteriolo gieal examinations on admission was negative in 5 patients; only gram negative stains (GN5) were isolated in patients having sntianaerobic prophylaxis (n=4) ;GNS were associated with different types of streptococci (n=6). Shaphyloco eci were observed in association with different types of streptococci (n=5), GNS in=l) ; in one patient, D.Strept. was associated to El. perfingens. Specific antibiotics according to bacteriological results stopped gangrene ex tension in all patients. Four patients died ; the death was related to the septic shock in two patients and to other cause in two patients. Conclusions. Early identification of microbial sgents involved zn synergistic-bacterial gangrene has likely played a msjor role in the management of these severe infections. Specific antibiotherapy associated with surge ry may improve significantly the prognosis of these infec tioms. One hundred and six traumatized or oost surgical oatients admitted to our intensive care unit were followed we~kly by skin tests with three types of antigens (PPD, candidin, varidase). Anergy is defined as no reaction to any of the three antigens while reactivity is defined as one or more oositive reactions. The factors related to anergy are : older age, renal failure, major septic complications, lenght of hospital stay and death. Hore interesting the outcome of the ~atients according to their skin t~treactivity : of the 56 reactive oatients at the beginning of the study, 61 remained oositive and 4 of them died; among the 5 oatients who became anergic 4 died. Of the 40 anergic patients, 26 remained anergic during all the study and there were 17 deaths; among the 14 oatients who became reactive only 2 died. 66 reactive 61 reactive; 4 deaths (6 %) 5 anergic; 4 deaths (80 %) p40.01 40 anergic 26 anergic; 17 deaths (65 %) 14 reactive; 2 deaths (14 %) D<0.01 Those anergic who became reactive, had a good outcome comnared to those who stayed positive during all their i~ness Among the factors which influence the skin tests reactivity, we retain the caloric intake : the anergic patient had a mean caloric intake of 25.5 kcal/kg day, compared to a mean of 41.2 kcal/kg day for those who re ~ covered a nositive reaction. Other factor will be discussed.

In conclusion, the reoetition of skintests allowed us to evaluate the efficacy of treatment particularly caloric intake and to Dick uD the high risk oatients in the intensive careunit. In developing countries tetanus is still a real problem in respect to both occurence and mortality rate. In all countries with high medical standard it has become an uncommon disease. In our own hospital the number of tetanus cases per year has decreased from iO -12 in the sixties to 2 -3 in the last three years. Because of this relative rarity, very few clinicians see more than an occasional case. The diagnosis is therefore often missed or recognized quite late as it is still based on clinical observation only. Till now no laboratory or other test is accepted, although various suggestions have been made. The problem of diagnosis is further complicated because cases of pseudotetanus are occasionally seen. Since 1978 we have taken EMGs in all patients who were suspect for tetanus but whose case history or clinical symptoms left some doubt. In 6 cases the EMGs were typical for tetanus with a shortening or absence of the silent period after a stretch reflex or after supramaximal stimulation and with a continuous activity of motor units which could not be influenced voluntarily. In two patients we found a normal EMG. In all 8 cases the acuracy of the electromyographic diagnosis of tetanus was confirmed by the further development of the disease. Therefore we think that Electromyography is a very useful and reliable tool either to confirm or to rule out the diagnosis of tetanus. Immunodeficiency has proved to play an important role in the outcome of critical patients. The present communication will investigate, in a prospective manner, the sensitivity and precocity of the different parameters measuring the immunological competence, as well as its changes as an indez of septic risk. Nineteen patients were studied. All of them were seriously traumatized ( multiple trauma or head injury ). Within the first 24 hours, they were submitted to the following: Skin testing (ST), T lymphocytes, B lymphocytes, total lymphocyte count, and total hemolytic complement (CH50). In a sub-group of seven patients, the same study was repeated a week later. The results showed that T-rosettes was an useful method for prediction on septic risk (p 0,05) as ST were, but with more promptness and statistic significance ( p 0,01 versus 0,05 ). CH50, B lymphocytes and total l~aphocyte count were not specific as an index of evolution nor of septic death or complications. As a conslusion, the determination of T-rosettes has proved to be, in this group of patients, a valuable test, higher than ST, as a prognosis of septic complica~i6ns and as a rapid checking of im/nuno-competence state associated to high risk.

BACTERIOLOGIC DIAGNOSIS OF NOSOCOMIAL PNEUMONIA IN VENTILATED PATIENTS. A. Bouchama, A. Akesbi, J. Pierre, P. Brun, F. Viau, J. Chasfre and C. Giberf. R6animation M@dicaLe, H6pifeL Bichat, Paris, France To determine the relative cuLturaL accuracy of protected specimen brush (PSB) through a fLexibLe bronchoscope in nosocomiaL pneumonia of ventilated patients, we compared this method with histoLogicaL and bacterioLogicaL examination of pulmonary specimen. The procedures were made just after death when venfiLatory supply was stiLL in place. SampLes via the PSB were taken into e branch of the anterior basal bronchus of the Left Lower Lobe. Immediate Left fhoracotomy was performed fo obtain 6 peripheral specimens from the corresponding segment. BacterioLogicaL specimens were eventuaLLy crushed, vortexed vigorously in I mL of medium and quanfilative cultures performed. 26 patients were studied. HietoLogic evidence of pneumonia was associated with pulmonary cuLture > 10 ~ cfu per g. A correL~fio 0 between PSB and Lung cultures was showed (r = 0.57 ; p <0,.01). For patients ventilated Less than 10 days, a better'correLation was found (r = 0.71 ; p < 0.001). Wi~h a threshold of 103 cfu per mL for PSB quantitative cuLtures, no false negative results were observed for the 26 patients and for the 55 microorganisms. With a threshold of 10 ~ cfu per mL, faLse positives wereonLy 25 %, but false negat]ves were observed (17 %). We conclude that cuLture results obtained by PSB should be useful in differentiating bacterial coLonization of the airways frbm pneumonia in ventilated patients in the presence of diffuse pulmonary infiLtrates.

PP~EDICTION OF SEPTIC SHOCK UNFAVOURABLE OUTCOI~E V.A. Gologorslcy, B.R. Gelfand, V.E.Bagdat jev, A. j. Schichov, l.J.Lapschina. 2-rid Moscow ~edical Institute, ~oscow, USSR. A comparative analysis of different physiological parameters was made for prediction oZ outcome in 116 patients with septic shock due to peritonitis. Indicator dilution method, intravascular pressure measurements,blood gas analysis and polieardiogram were used to investigate hemodynamic and oxygen transport parameters.Cardiac performance was estimated by volume loading and almawing Starling curves. Blood volume and arterial pressure did not vary in survived and nonsurvived patients during intensive care.There was significant difference in pulmonary arterial pressure and resistance which increased in nonsurvived patients and remained higher than normal up to the lethal outcome.The most considerable difference was marked in cardiac output,left and right cardiac work,and heart performance index.These parameters are considerable lower in nonsurvivers at all stages of the investigation.According to received date unfavourable outcome prediction or death risk indices were as f~llowing: cardiac index less than #L/min./m=; left cardiac work less than 6kgm/min.;right cardiac work less than 1.Ykgm/min.;total ~ascu~ar resistance higher than 2000dynes/ca /secY.lpulmon ya~i vaGcular resistance higher than 200dynes/ cm /sect;slope of ventric~lar performance curve less than 0,15L/min./m /tort;arterial pO 2 less than 70 mm Hg;arterial-v~nous gradient of oxygen content higher than 5ml/10Om~; oxy~oen availability less than #OOml/min./m~; oxygen extraction higher tan 40%. 

SHOCK. L.F.Kosonogov, V.N.Rodionov, ~edical Institute, Voronezh, USSR. The efficiency of hemosorption and subsidiary circulation with extracorporal oxygenation was studied in 32 patieats with septic shock. Signs of the liver and kidneys diaseases,high plasma toxicity, homeostatic disorders were revealed by clinical and laboratory findings. Before hemoscrption sanative operations were performed to all the patients, and they received conventional intensive therapy without positive results. Hemosorption gave a marked clinical and laboratory effect in all observations. In q2 patients with marked hemodynamic disorders hemosorption was carried out by subsidiary circulation with blood oxygenation from the vein into the artery. Stabilization of the general state normalization of the functions of the liver and kidneys sharp decrease of plasma toxicity were clearly observed after the given therapy. 30 patients out of 52 recovered. Hemosorption has a great detoxicologic effect with intensive therapy of septic shock. (cortisol,prolactin,growth hormone),thyroidal hormones (T4,ET~,TSH) and the renin-angiotensin-aldosterone system were studied by RIA in 15 pts with various underlying acute diseases.Patients with abnormal serum-elec trolytes or blood glucose were excluded.

Stress hormones: As anticipated cortisol and p~o~a~n-[~[s were elevated(cortisol 4o,97~ 2o,89 ug/dl((normal 6-19 ug/dl)) prolactin 59,6+ 32,4 ng /ml((normal rangel-25 ng/dl)).Growth hormone levels were not elevated in our pts (1,43• 1,o2ng/ml, normal range I-5 ng/ml). Thyr2!da!_h2~m2nes:T4,ETR,TSH were normal in all pts. (T4 8, 46 ~ 3, 57, normal 4, (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) 4 Ug/dl; ETR o, 98• o, 13 ;TSH 4,57• 1,o3,normal -6uU/ml)Thus at admittance the pts did not show the well known "iowT3,T4-syndroms

All levels of this system were in the normal range (renin 1,7~o,91 ng/ml/h,normal o,1-2,62ng/ml/h; angiotensin II17,23 Z 5,71 pg/ml;normal 8-25 pg/ml ; aldosteron 143~8o,7 pg/ml,normal 3o-15o pg/ml). Thus the prior described state of hyperreninemic hypoaldosteronismus was not present in our pts at admittance to the intensive care unit.

Abnormalities in stress hormones are an early finding in intensive care unit patients,whereas changes in thyroidal hormones and the renin-angi otensin-aldosterone system either develop later in the course of severe illness or are the consequence of therapeutic interventions. p< 0,01) . Alcohol intoxication increased significantly in 1981 (p<0,01) wheras clinically treated hypnotic drug intoxications decreased. Severety of intoxication was judged by 7 parameters: loss of consciousness, loss of pain reaction, circulatory arrest and/or respiratory failure, areflexia, arterial hypotension SAP< 90 mm Hg, hyperglycemia> 200 mg/dl, artificial respiration. 2 or more of theses parameters were present in 27 (1971) and 33 (1981) pat. and the number of pat. with more than 3 parameters increased significantly. Of the most severe intoxications 9/Z7 (1971) resp. 15/33 (1981) were due to barbiturates. In 1971 39% of pat. were treated in the ICU, in 1981 only 15% (p<0,01). There was no significant difference in frequence of unconsciousness, loss of pain reaction, areflexia, light reaction of pupils, arterial hypotension and hypothermia, but pat. with heart rate<80/min, circulatory arrest, respiratory failure, artificial respiration were more frequent in 1981 (p< 0,01). Gastric lavage was performed in 80% in both periods, forced diureses decreased from 48% (1971) to 13% (1981) (p<O,01). Extracorporeal hemodialysis and hemoperfusion was performed in 11 pat. in 1981. Globale mortality was less than I%, mortality in the ICU was 2% (1971) resp. 6% (1981) (n.s.). The mean duration of stay in the ICU increased from 2.2+ 3.1 (1971) to 5.5 + 8.1 days (1981) (p<0,01). Conclusions:The population of clinically treated pat. with acut poisoning varied in resp. to an increased rate of alcohol intoxication and of very severe intoxications with cardio-respiratory-metabolic failure with a high rate of barbiturates, a decrease of forced diureses, mal nde~a st~o gerunc~ange~.el ec ion of pat. for the ICU. Adrenocortical hypofunction is not uncommon in critically ill patients and daily serum cortisol measurements may have prognostic importance. 1 In a previous study of 132 patients 34 who had oortisol levels remaining <350 nmol/l all died. This report refers to a trial of replacement therapy in cortisol deficient patients. Using a radioimmuno assay [Corning], daily 08.00h serum cortisol measurements were made on 70 consecutive patients, who were allocated randomly to one of 2 groups. 18 [25.7%] patients had cortisol values below the critical level of 350 nmol/l. I 12 of these had l-hour adrenal stimulation tests using synthetic aorticotrophin. All tests were negative, All patients received treatment appropriate to their conditions but additionally patients in group 2 received hydrocortisone replacement while group 1 acted ms controls. Tauroline is applied intravenously as a chemotherapeutic in infections and septic cases. A hypothetic interaction of injected tauroline on the hypertensive effect of dopamine or dobutamine was tested in hypovolemic rats and compared with an injection of saline instead of taurolidine. The blood pressure was lowered in WISTARrats by reducing the blood volume by Iml within 1 minute. The blood pressure was continuously measured with a Statham blood pressure transducer connected to a cannula in the carotid artery and recorded on a Beckman Dyonograph. 4 rats received 3 minutes after th~ artificial blood pressure decrease 0,6 x I0 mol dopamine, followed by an infusion of 2 ml Taurolin ~ NaCl or 2 ml saline within 7 minutes. 4 rats received an infusion of 0,6 x I0 ~mol and 2 ml Taurolin 2% or saline simultaneously. These procedures were repeated several times with the same rats. In all experiments, there was no significant difference in the sympathomimetic effect of dopamine. 2 hypovolemic rats reveived 0,7 xlO -6 mol dobutamine, followed by 2 ml Taurolin NaCI or sal!~e.

2 rats received an infusion of 0,7 x I0 mol dobutamine and 2 ml Taurolin ~. 2% or saline simultaneously.

In these experiments, taurolidine even increased the pressure action of dobutamine when injected simultaneously. A.Gasperi, Maputo Central Hospital, M aputo, Peoples' Republic of Mozambique.

Intensive Care Medicine still represents a "noveery" in most developing countries, because ot~er priorities have to be dealt with by local health authorities, and because of the need of very expensive investments. Nevertless, within the frame of projects of technical cooperation between industrialized and developing countries, some attempts have been made to establish and, where already existing,to ameliorate project~ in the field o9 intensive care medicine. The

Author, after a five years service,spent as responsible os projects of technical assistance in intensive care and anaesthesiology, in african developing countries, comments in the present paper the following points: -definition os intensive care: possibilities and limits in developing countries.

-practical aspects linked with the speciality, such as teaching, training, type of technology needed.

-priorities which should be taken in consideration when planning projects os technical aid.

costs versus benefits analysis. Finally, on the basis of previous experience, the Author presents some suggestions to those, (institutions,aid organisms)who mightbe interested in projects of technical cooperation with developing countries in intensive care medicine. Taurolin is a chemotherapeutic agent with antiendotoxin properties which can be administered intravenously (IV) or intraperitoneally (IP). It is specially indicated in the treatment of peritonitis and septic shock. In vitro, it reacts with Dopamine to inactivate it. It has been recommended that the two drugs should be used alternately in two hour cycles. This sequence is not appropriate in the clinical care situation. Experimentally this has been tested by studying nine dogs anaesthetised and ~entilated. Measurement of pressure was by a tl~r~sducer placed in the right femoral artery~ renal blood flow was measured by a Statham guage in the left renal artery and urinary output by a catheter in the bladder.

These were recorded simultaneously on a Polygraph (Gould ESllOO0 ~. Dopamine was infused in a dose of 6.25 Ug.Eg-min-for 8 hours.

Taurolin 2% i~ Na~l 0.45% was infused in a dose of 15.37+ 1.58 mg.Eg -~ h-for 6 hours commencing one hour after--the Dopamine infusion. A control group was given a saline infusion in place of Taurolin. Taurolin did not significamtly alter the presser effects of Dopamine or its effects on the kidney. Clinically five patients with septic shock who were requiring Dopamine infusions were given Taurolin in addition. Two patients received 200 ml of 2% Taurolin b.d. intraperitoneally while three received 250 ml of 2% Taurolin intravenously over 2 hours twice d~ily. In none of these cases was there any loss of the alpha or beta adversgis aotion of the Dopamine.

From these animal and clinical results it is concluded that there is no centre-indication to the simultaneous administration of these two valuable drugs in the intensive care situation.

Antipyretic therapy with Diclofenac Sodium in ICU patients. Effects and side effects of 109 mq suppository versus 0,2 mq/kg i.m. Zandstra D.F., Stoutenbeek Cb., Miranda D.R. ICU University Hospital, Groningen, The Netherlands.

Diclofenac sodium 100 mg suppository proved to be a potent antipyretic drug. However severe side effects: oligo-anuria and protractive shock necessitating increasing amounts of i.v. fluids and dopamine, make this drug in this dosage unsuitable for use in ICU patients. Whether these effects were dose related or not we studied the effect of 0,2 mg/kg bw km. in 8 ICU patients. Antipyretic effect, as were bloodpressure diuresis and therpeutic measures of both groups were compared. In both groups this drug proved to be a potent antipyretic drug, However dose related effects could be observed, Significant decrease ~in bloodpressure was observed up to 6 hrs in the suppository group, as in the 0,2 mg/kg group only in the first 2 hrs this was seen, In both groups we observed a significant decrease in diuresis, Therapeutic measures to maintain adequate circulation and diuresis were the same in both groups on 3 hrs after application; additional i.v. fluids (600 +75 ml) and a 100% increase of the dose dopamine administered. Conclusion= Serious side effects; acute oligo-anuria and severe fall in bloodpressure necessitating the administration of increasing amounts of depamine and i.v. fluids makes diclofenac sodium ~ven in a small dq.se of 0,2 mg/kg i.m. unsuitable for use in ICU patients, as an antipyretic agent.

Lorenz Boehler, Vienna, Austria High frequency ventilation offers some advantages over IPPV (oxygenation, removal of secretions etc.) but it is difficult to handle and adequate CO2 elimination is not always easily achieved. By superimposing HF Jet impulses on conventional ventilation or even spontaneous breathing the following effects could be observed.

I. Clearing of atelectatic:areas that have been refractory to conventional therapy (bronchoscopic suctioning, CPPV, external vibration). 2. Improvement of blood gas values in ARDS (Fi02 reduced from 1.0 to 0.5 within 8 hours~ 3. In patients breathing oxygen by face mask PaO2 could be raised from.less than 60 torr = 8 kPa (in one case 34 torr = 4.5 kPa) to levels between 90 and 100 torr (12 and 13.5 kPa). This was achieved by repeated 5-min treatment sessions with oral HF-Jet, mechanical ventilation or CPAP was not necessary. 4. It has been successfully used to prevent posttraumatic ARDS. Detailed case histories and x-ray-photos will be presented.

Conclusion: With the exception of a few clearcut indications (eg. large bronchopleural fistulae) many clinical effects seen in HF-ventilation could be achieved by superimposed HF-Jet technique. The new method is easier, safer and cheaper than sophisticated HF ventilation techniques. 2 ~ ) a very significant correlation between mortality and age, showing the importance of the latter (Mann-Whitney test, p < 0,01). However, contrary to previously published results the mortality rate is identical in both age groups (below and above age 45) as long as the I.S.S. is below 38 points (1.2) 3 ~ ) a highly significant correlation between the length of hospitalization in I C U and I.S.S. in the discharged group (Spearman test, p < 0,001). But, in opposition to previous data, there is not a negative correlation between length of survival and I.S.S. in the group of patients who died (3).

In conclusion, the above results are different from previous studies. The reasons for the differences are analysed: population characteristics, quality of care, length of time before admission into ICU, and transport conditions. Ill BAKER: J. of Trauma,1976 , 16: 882-$85 BULL: Acc. Anal. Prey., 1975 , 7:249-255 SEMMLOW: Health Serv. Research, 1976 . The administration of exogenous fibronectin ge_a red to avoid some complications in critically ill traumatic patients (sepsis, ARDS) relates to the important role played by this glycoprorein in the inespecific opsonization of the reticuloendothelial system. The change in the PPL in these patients and its dependence on the severity of the trauma were investigated, The PPL was measured by inm~unoelectrophoresis in 29 patients with multiple injuries and they were related to an injury severity score (ISS). The m~ an value for the FPL was %8,6 ~ 18% as compared to the 98,% ~ 2056% of the control group of healthy human beings (p O,OO1). There was no statistical significance in the correlation between the FPL and the ISS in all the patients studied. Patients with only head trauma (with an ISS = 25,% ~ 0,3) however had a higher FPL (68,5 + 20%) than the rest of patients with multiple injuries (ISS = 35,3 J iO and FPL = 54,8 ~ 16%). These results demostrate that the FPL is reduced in patients with multiple injuries. This l~ vel, however, is better correlated with the volume of tissue involved than with the possible injury of vital organs (head trauma). This is to be taken into account when facing a possible th~ rapeutic infusion of this protein. On 570 multitraumatized patients hospitalized in a traumatologic intensive care unit, severity scores have been established, and compared to clinical evolution. Three types of severity scores were performed at the patient's admission:

-the Injury Severity Score ( ISS ), summarizing the victim's injuries ( based on the Abreviated Injury Scale ), -the Respiratory Index ( RI ) integrsling Pa02 and its corresponding Fi02, -the C.H.0.P. ( using Creatinine, Hematocrit, 0smolarity, and first systolic blood pressure ). Added and compared to these, classical cutaneous immunity tests ( tuberculin, candidine, varidase ) were realized , having only a value on patients hospitalized more than 48 hours. The results show out, first, a very significant difference of Severity Scores Values in the two series individualized:" cured " and " deceased " Secondly, comparing severity scores values together and one to another, by including them into differents tables, an idea on the patient's prognosis can be given in chances of survival. Concerning immunity response, prognosis interest is maximum in the non reactive group ( nearly 60 % of the patients which stay in the unit was over 48 hours ). Based on a two years experience, and on a population of 570 patients, correlating severity scores to clinical evolution can allow an early theoritical prognosis approach of the multitramatized patient survival. The "Glasgow Coma Scale", which is widely used can be a valuable indicator of H.I. outcome in patients breathing spontaneously, but is of little practical use if Controlled Ventilation is required as part of the patient's overall management, especially if relaxant and sedative drugs are used. A Ventilation Observation Chart in usa at Pinderfields Hospital appeared to offer a prediction of ultimate prognosis.

It was based on a simple scoring system of 15 parameters routinely monitored in such patients.

When these charts were subjected to a Discriminant Analysis Package 8 of the 15 parameters were identified as being of significance in predicting outcome and were arranged in order of importance by their weighting factor. Prediction of outcome was calculated using the formula of Sayles et al (1982 The objectives of this study were to compare changes in cardiorespiratory profiles, pre, intra and post TBH to determine the appropriate therapeutic maneuvers which are required to maintain hemodynamic and respiratory stability during TBH. Sixteen treatments, each of 4 hours, were given with the following results.

Pre The values were tested for statistical difference using the Kruskal Wallis test. Urinary creatinine and total N did not differ significantly in G, L and K. Muscle protein breakdown measured as urinary 3MH/creatinine was significantly reduced in K as compared with G or L. Thus, I.V. KIC reduced muscle protein breakdown but did not reduce total nitrogen excretion. Leucine administered under the same conditions did not. Oar managing strategy develops as fellows: 11 a network ef communlcatlenm with peripheral hospltals to collect detailed informations abaut clinical condition and evolution ef the patient; 2) cooperation with private and pubblic Institutions in order t 9 obtain the most suitable means of transport (air and ground); 3) continuous training of a medical team; N)the availability of sanitary equipment of smallest size, rationally arranged in standard kits and battery operated. Our criteria for sanitary transfer of these patients can be summarized: use of the most suitable mean of tranmport (sanitary flight, helicopter, ambulance); patient assessement amd preparation before leaving; sedation and ~urarizatlen of the patients, artificial ventilation with positive end expiratory pressure, possible infusion of sympathomimetic agents; continuous monitoring of respiratory and hemodynamio paramethers; travel performed as fast as possible. We transferred 32 patients from as far as 1.200 km using ambulance (8N%), hellcopter (3%) and sanitary flight (12.5%).Mean PaO on leaving was 51.02 mm Hg (ranging from 95 to 23 mm Rg); on arriva~ was 63.83 mm Rg (ranging from 147.5 to 27.5 mm Hg).Mean PaCO on leaving was ~7.12 (ranging from 63 to 27 mm Hg); on arrival was ~3.9 (ranging from 79.3 to 24.5 mm Rg). 17 of our patients were treated with conventional respiratory therapy, 15 of them underwent LFPPV-ECCO2R; none of them died during the transport.Being our Institution, at present, the only center in Italy where LFPPV-ECCO R is carried out the transfer of patients with ARDS may be considered an integral part in this type of life saving procedure. The alterations of blood coagulation were studied in 16 patients with adult respiratory distress syndrome (ARDS). All patients required mechanical ventilation lot7 days, Fi02 0.8t0.2, PEEP max. 10 ~ 5 cm H20 with a P(A-a)O2/PA02 0.74• 0.13 and Qs/Qtot 38t6%. The following parameters were analized in plasma and bronchoalveolar lavage fluid (BLF) at O, 24, 72 hours and I0 days: fibrinogen ([, Van Claus), fibrin degradation products (PDF, Merskey), factor VIII antigen (F, VIII ag, Laurell), platelet factor 4 (PF4 Ria) and beta-tromboglobulin (~TG, Ria). A control group of 10 patients was included. RESULTS: Mean values of the first 3 days are shown in the In all cases all parameters analized were higher than normal values and an increase of PDF, PF4 and BTG in BLF was observed. These data suggest an endothelial vascular lesion together with an increase of fibrin deposits and platelet aggregation in the pulmonary microcirculation.

Neurogenic ARDS -its frequency and relevance ~c outcome in patients with isolated head injury. Though ARDS following sepsis, trauma or coagulation disturbances is well known, the same does not hold true for centally-initiated impaired pulmonary gas-exchange or neurogenic ARDS, respectively. Therefore, in a prospective clinical study in 129 patients with isolated head injury frequency of ARDS, its relation to severity of cerebral trauma and its relevance to clinical outcome were examined. Severity of head injury was graded by meansof the Glasgow-Coma-Scale (GCS)and pulmonary gas-exchange was expressed by the quotient A-aDOp/PAO p (Q). Patients with additional trauma suc~ as Nemorrhagic shock, fractures of the large bones, direct ehesm-mrauma or aspiration were excluded. Our results demonstrate a relationship between the severity of head injury and impaired pulmonary gas-exchange. 22.5 % of all patients suffer from a severe neurogenic ARDS(Q> 0.6) while in additional 26.3 % pulmonary gas-exchange was moderately impaired (Q>O.L<0.6). With increasing severity of neuro-~ enic ARDS survival rate decreased from 58.3% severe head injury but without ARDS) and 55.3% (severe head injury with a moderate degree of ARDS) down to 24.1 % (severe head injury with severe ARDS) despite an increase of ~ontrolled ventilation from 66.6 to 89.1 %. Moreover, neuro logical outcome is significantly worse in the latter group than in the former ones. Alveolo-cap ilIar ductance for carbon monoxyde (defined by a mathematical expression [(Pl -PA)/PI ] identical to efficient alveolar ventilation) represents the pulmonary blood volume close to ventilated alveolar space. With an original informatic method, i t ' s measurment in s ready s tare cohditions can be performed at bed s [de and in mechanicaly ventilated patients. Aprospective study has been conducted in order to assess the diagnostic and pronostic value of DuACCO in pulmonary embolism. This day 12 patients were entered in the study. On the first day following the admission in the intensive care unit, blood gases and DuACCO were measured and a pulmonary angiography was performed. A long term follow up was obtained by repeated measurments of DuACCO. On day i, DuACCO reached a mean of O.43 (i.e. 61% of the theoreti cal value) with a minimum of O. 24 and a maximum of O. 60, PaO 2 reached a mean o f 53 mmHg with a minimum of 35 mmHg and a maximum of 70 mmHg. The pulmonary hypoperfusion degree, as it can be evaluated on pulmonary angzograms according to WALSH index, reached amean of lOwith a minimumof 2 and amaximumof 15. A closed correlation is observed between Walsh Index and DuACCO (r = -O. 78, p <0.O1). In one case of massive pulmonary embolism (Walsh index = 20) with slight hypoxemia (PaO2 = 62 mmEg) DuACCO remained dramatically lowered (DuACCO = 37, i.e. 53 % of the theoretical value). On day 40, DuACCO reached s mean of 0.59 (i.e. 84 % of the theoret ical value) in 6 patients despite the lack of thrombolytic the rapy. We conclude that DuACCO zs a non invasive reliable measurment allowing severity evaluation and longterm follow up of pulmonary embolism. The experiments were performed. ~n 10 pigs (5 for contro~ and 5 animals received 1,5 mg of escherichia coli endoto ~ xin). The animals were tracheotomised and anesthetized with Halothane (0,5%).the cardiac output (C.O) was measured by thermodilution. The distribution of C.O was measured by injection nuclide labeled radioactive microspheres (15#m in diameter)(New England nuclear) into the left ventricule. Particles were injected before (57 Co) and at 5 (113 Sn) and 60 mn (46 Sc) after administration of end~ toxin. After the pigs were killed, 50 samples of the organs were dissected, weighed and counted in a gamma scintillation counter. The 15#m microspheres measured only the ca~illary blood flow (nutrient blood flow). Particles shunted through the systemic circulation (arteriovenous anastomoses) were stopped in the lungs. 3 groups of organ blood flow were individualized (for 100 g of tissue, percentage of C.O, before and at 5 and 60 mn after infusion).

-coronary (2.32,4.18,6.31") pancreas (0.46,1.03,1.49 ~) aorta (0.12,0.21,0533) blood flow were increased. -liver (0.98,0.63 ,0.54 ~) adrenals (2.57,4.49,2.48) muscle (0.11,015,0.11) and total gastro-intestinal tract blood flow were well maintained. -spleen (3.71,1.85~,2.32 *) brain (l.94,2.97,0.81)kidneys (4.16,4.58 ,2.65 ) and ears (0.042,0.025,0.019) blood flow were decreased. During early endotoxemia the ears and spleen blood flow were decreased in contrast the kidney and brain blood flow were maintained. The pancreas (0.46, 1.03,1.49 ~) adrenals (2.57,4.49,2.48) and pituitary (1.95, 1.80,2.00) blood flow were maintained.

The fraction of cardiac output passing to the lung (arteriovenous anastomoses and bronchial arteries) was significantly decreased (5.68,5.5+,2.21m The technic, largely fused in bronchopulmonary surgery may find a more recent indication concerning intensive care patients. This type of ventilation has been performed on 17 patients of a traumatologic and surgical intensive care unit, using: -P.V.C. endotracheal tubes ( Bronchocath TD ), without ergot, 9 -Two ventilators msking possible a totsly asynchronous ventilation. Asynchronism seems to be one of the necessary conditions to the efficiency of the technic; and this is shown by its positive effects on: 9 hematosis : arterial ( 106 +45 versus 351 + 106 mmHg in Fi02 = I ) and venous blood gases. 9 the haemodyrmmic parameters:cardiac index (2.8 + 1.6 v. 3.2 + 1.6 ), arterial pulmonary resistance , arterial pulmonary pressure ( 17 + 6 versus 13 + 1.5 mmHg ) and pulmnary capillary wedge pressure ( 9+ 5v. 4+ 6mmHg). Local tolerance to the double lumen endotracheal tube, has been appreciated on fiheroptic bronchoscopic criteria 9 Instead of waiting for the advent of a severe hypoxemia, a precocious indication of such technic can be discussed in the following situations :

-after an unilateral aspiration, -when different FF~P levels are requested for each lung, -to finish haemodynsmic tolerance during high PEEP levels is largely improved by the separated artificial lung ventilation.

A.J. Mortimer, J-L. Bourgain, Nuffield Department of Anaesthetics, Radcliffe Infi~nary, Oxford~ U.K.

The purpose of this study was to Lnvestigate the occurfence of ambient gas entrairmmm 9 during jet ventilation at different respiratory frequencies and I:E ratios. _.A model lung (Manley B.O.C_~) o_ficcmpliance 50 ml ~ H90 i mqd resistance 5 mn H~0 1 -sec -was ventilated in aclosed plethysmograp~ by a solenoid-operated jet venti-laP.or. ~ The ventilator output %~s directed into the lung through a i. 8 ma nozzle in a T-piece oonnected to a 9 nm ET tube.

At a driving pressure of 45 psi tb~ degree of entrainaent was measured with I:E ratios of 1:2 and 1:4 at frequencies of i, 3, 5 and i0 hertz.

The tidal volume delivered to the lung (VT I//h~) was derived frGm the pressure swings of the plethysmograph, and the tidal volt.he delivered frcml the jet (VT VISIT) was obtained by dividing the minute volllne of the ventilator by the frequency under observation.

The rat%o of VT I//NG to VT V~NT gave a measure of the entrainment. i.

The F_O^ and humidity of the inspired gas are largely de~e~ined by the gas delivered frGm the jet.

The minute volume of the ventilator is not a useful guide in assessing the minute volume of the patient. Continuous positive airway pressure (CPAP) has been shown to increase functional residual capacity (FRC) in healthy volunteers proportionally to the level of positive pressure (1). The purpose of this study was to evaluate the changes in respiratory pattern and the duration of increase in FRC induced by CPAP (I0 cm HpO) in nine healthy volunteers. Before and after CPAP, FRC was measured by multiple breath nitrogen washout. During CPAP, changes in FRC and respiratory pattern were continuously monitored with pre-calibrated bellows pneumographs. Results : The increase in FRC (+ 22% of control level) ~d under CPAP varied with time (range: + II% to + 41% of control level) and disappeared within the first minute after discontinuation of CPAP. Tidal volume increased significantly during CPAP while minute ventilation remained stable. Conclusions : In patients with normal lungs, the effect of CPAP on FRC ends with the removal of the positive pressure. The mean increase in FRC observed with bellows pneumographs during I0 cm HpO CPAP is similar to the one measured with the helium di?ution method (I) but varies greatly breath by breath and thus may give additional information to the results of former studies where only intermittent measurements of FRC were performed.

I. Gherini S. et al., Chest 76: 251-256, 1979 . High frequency jet ventilation (HFJV) has been widely used during bronchoscopy ; the main problem during HFJV is CO 2 elimination. The accuracy of a cutaneous eapnometer (Roche Kontron) was evaluated in this clinical situation. 6 patients were investigated under general anesthesia. Different methods of ventilation were used successively over a ]0 mn period : manual jet ventilation, HFJV at rates of ]50/mn, 300/mn and 500/mn. The jet ventilator MK 800 (Acutronic mediea! system) was used with a driving pressure of 5 bar and a ]6 gauge injector. Measurements of arterial and transcutaneous PCO 2 were made simultaneously before anesthesia and atthe end of each period of ventilation. Satisfactory correlation was found between transcutaneous PCO 2 and arterial PCO 2. Thus it appears transcutaneous PCO 2 measurement can be used for evaluating CO 2 elimination HFJV has been proved to be satisfactory during bronchoscopy at a frequency of 60/mn. We compared JV with HFJV during laser resection of severe tracheo-bronchial stenosis under general anesthesia, JV using the Sanders technique, HFJV the Jet ventilator MK 800 (Acutronic medical system). During both methods FiO 2 of the injected gases was 0.5, the driving pressure was 5 bar. Ventilatory rate was 20/mn for JV and 300/mn for HFdV with an I/E ratio of 0.25. A 16 gauge tube welded on the bronchoscope permitted ventilation. Arterial blood was withdrawn before anesthesia (A) and at the tenth minute of surgery (B The desirability of changin~ the behaviour of a Respirator, according to the differing and changing needs of the critically ill patient, can only have a meaning if precise control and measurement of relevant data can be achieved.

Our way to comply with these clinical specifications has been the construction of a simple proportional valve with extremely few movable parts but designed specifically to interact with the advanced and reliable microprocessor technology, This combination has provided us a highly versatile operational Respirator, which allows the clinician a very broad range of operationality, concerning any ventilatory mode or function. The device is conceived with clear and precise settings; identified patient alarms; patient monitoring with trends, and warning for erroneous settings. The optimal management of critically ill ventilated patients often requires repeted measurements of static thoracopulmonary compliance (Cs~. However the conventional methods which are-used for this measurement are generally laborious and request a possibly dangerous interruption of the ventilation. The new device Spirolog IIR Dragger which takes into account cycle by qycle informations of expired flow and pressure from the~nasotracheal tube calculates continuously C:S T in ventil'ated patients. The aim of the present study was : I) to compare the values of CST calculated by the Spirelog II (CsT-SP) with those measured by the syringe method (HARP 1975) during inspiration (CsT-iusp) and expiration (CsT -exp), 2) to determine the normal values of CsT-SP by measuring this parameter in anesthetized surgical patients with clinically normal respiratory function and normal X ray chest.

3) to test the usefulness of this calculator for the estimation of the "best-PEEP" level (SUTER 1975) . The results were as follows (mean • SE ml/cmH20) : i) CsT-SP (46,9 • 5,6, n=15) was well correlated (p O,OO1) with csT-ins (64,9 ! 6,5, n = 15) and with C. ST -eXp. (49,1 _ + 4,6, n = 15). CsT-SP was lower than CSTins (p <O,OO1), but was the same as CST-ex p 2) Cs$-SP in normal anesthetized patients was found to be 57,5 -1,9, n = 47, a value which is within the normal range of the already published data.

3) The levels of "best PEEP" estimated with the Spirolog were well correlated with those estimated by the syringe method (p <O,O5) We conclude that Spirolog II is suitable for the clinical management of intensive care ventilated patients. Currently available Peep valves work unidirectionally,i.e. gas flow through them is only possible in one direction. Peep is produced by preloading these valves, which permits gas flow only from a set opening pressure. Once the opening pressure is overcome, ideally gas flow should occur unobstructed. No gas flow is possible in the opposite direction. There are many techniques to produce Peep, they involve water columns, springs, magnens, coun~erpressure, balls, etc..; but one feature common to all currently used valves is their unidirectional function. We present a bidirectional Peep valve which not only exerts Peep in one direction, but also permits unobstructed flow in the opposite direction. The Peep ranges between 3 and 20mbar. Its resistance is 0.8 mbar at a gas flow of 0.5 L/sec and 2.45 mbar at 1 L/sec respectively. Two porns allow suctionning and airway pressure measurements.

The suggested clinical application of the described valve isle produce differential Peep in artificially ventilated patients with assymmetrical lung disease or patients in lateral position during chest surgery. The valve is simply placed on line into the selected branch of a double lumen endotracheal tube. Peep is applied to the less compliant lung in unilateral lung disease Dr to the dependent lung in lateral position during 2hest surgery. The ~echnique of differential Peep is thus much simplified. More complicated techniques like Y-piece sysnemsor the use of two ventilators are avoided. Of course the bidirectional Peep valve may also be used in any other kind of Peep therapy. The valve has then to be placed between the Y-or T-piece of the chosen gas delivery system.

A bidirectional Peep valve is presented. We suggesn ~ts use together with a double lumen endotracheal tube [o produce differential Peep in patients with unilateral lung affections.

COMPLICATIONS ASSOCIATED WITH PROLONGED HIGH FREQUENCY VENTILATION. J. Berr6, A.M. Ros, J.L. Vincent, P.Dufaye, R.J. Kahn. Department of Intensive Care, Erasme University Hospital, Free University of Brussels, Belgium.

The type and the incidence of complications during treatment with high frequency ventilation (HFV) were evaluated in 7 patients (age 17 to 74) who presented with acute respiratory failure. HFV was used for 6 to 34 (median 9) days for management of bronchopleural fistulas (5), associated in one case with %racheo-esophageal fistula, tracheal rupture (i) or voluminous lung abcesses (i). Four of the 7 patients died from multisystem R failure. HFV was delivered using a ventilator ACUTRONIC VS 600 (3) or MK 800(4), at rates 100-250 cycles/min. Inspiratory time was 90-40 % and driving pressure 2-4.5 bar FiO ranged from O.S to 0.9. Jet injector (DESERET ~-142GA-8.3 cm with 2 side-holes) was secured in the endotracheal tube.

Inspiratory gases were carefully humidified.

We identified technical complications associated with insufficient humidification, leading to viscous secretions (3 cases), desiccation of the tracheobronchial tree (2 cases) and total obstruction of the endotracheal tube (one case).

Disconnection or kinking of the catheter were not rare events.

One ease of severe hypcthermia was attributed to humidification with a cold solution.

Ventilator noise was sometimes hardly tolerated and two patients developped acute confusional state. Sleep disturbances were variable.

On the other hand, abrupt disconnection of the noise could give rise to panic. Acceptation of HFV by nursing staff was directly proportional to their knowledge of the device.

Technical problems occured more commonly during nights or week-ends. Inadequate alarms were generally incriminated. Easy, practical solutions can be suggested to improve humidification and monitoring of HFV, and thereby limit the incidence and the severity of complications. Acute colonic dilatation (ACD) were one of the most deleterious side effects of mechanical ventilation. The aim of this work was to assess. i) etiologic factors of ACD 2) ACD treatment. Patients and methods : 9 patients with radiologically proved ACD (ie coecal diameter> 7 cm) were selected. The four following etiologic features were studied : a) ventilation modalities (naso-tracheal tube, tracheotomy), b) hypokalemia, c) cirrhosis, d) patients sedation with phenoperidine. Used treatment were : I) symptomatic medical therapy (enteral nutrition and eventually phenoperidine withdrawa~ nasogastric suction), 2) surgical therapy (decompression and local resection). Results : 6 patients were ventilated by nasotracheal tube, 3 by tracheotomy. As other etiologic factors we found 2 hypokalemia, 3 cirrhosis and 5 patients treated by phenoperidine. None patient were dead during ACD period. High risk of colonic perforation, in spite of initial medical treatment was the reason for surgery in 5 patients. Conclusion : Multiple factors inducing ACD with mechanical ventilation were discussed, but we would emphasize the potential phenoperidine involvement. In fact, for three patients ACD resolution followed phenoperidinewithdrawal.

In one patient a threefold phenoperidine using had induced a threefold ACD. In patients submitted to mechanical ventilation phenoperidine will be used with dramatic carefulness. High risk of colonic perforation due to medical treatment inefficacy allowed surgical therapy. The goals of this study were to assess the feasibility and practicality of HEJV for a heterogeneous group of critically ill patients, with ARF, during the postoperative period, to define optimal initial HFJV settings and to observe the concurrent effects of HPJV with PEEP, endotraeheal suctioning and weaning. One hundred patients admitted to the SICU and NICU were ventilated with HFJV. The age group ranged from 10-92 years with 75 ~ 65 years. The operations included craniotomies and thoracic, abdominal and major vascular procedures. Optimum initial settings were driving pressure 35 psi, frequency 100.min, IE ratio 30:70, FiO2 0.90 and PEEP 3-4 mmHg.

The driving pressure proved to be the pr'imary mode of CO 2 elimination, initial mean PaCO 2 35.8.

FIO2 and PEEP were adjusted to provide the desired PaO 2 and optimum QsP.QT. Cardiorespiratory profiles indicated that Qsp.OT would be reduced a mean of 30% by the addition of PEEP i0 mmHg but was associated with a simultaneous decrease in mean CI from 3.39 to 2.81 L.M 2. Endotraeheal suctioning with HFJV resulted in a A PaO 2 of 15 n~Hg compared to 90 nm~Hg without HFJV.

Weaning was achieved by progressively decreasing the driving pressure.

Patients subsequently reported that they were comfortable spontaneously breathing at a rate of 10-20.min while HFJV continued at 100.min. ICU physicians and nurses found the patients were easyto wean using HEJV. Eightyfive of the i00 patients were extubated within a 24 hour period; fifteen demonstrated criteria consistent with continued ARF and required HFJV for ~ 14 days. HFJV offers a new and improved technique for patients with ARF in the postoperative period. We investigated whether and under which circumstances suppression of spontaneous breathing occurs during high frequency ventilation. Yorkshire piglets were ventilated with a high frequency jet ventilator under pentobarbital anaesthesia. The highest PaCO^ at which the animals z did not breath against the ventllator (Pacopapnea) was established during different patZerns of ventilation. Where necessary for compensation of hyperventilation this was done by adding CO 2 to the respiratory gases. Arterial oxygen tension was kept above i00 mmHg. Suppression of spontaneous breathing activity could be obtained during ventilation with higher frequencies, irrespective the tidal volume and the insufflatory flow velocity. So, the mode of dynamic activation of the stretch receptors was rejected as a causal mechanism. Positive end expiratory pressure (PEEP), increasing with higher frequencies, suppressed the spontaneous breathing activity, also at a constant ventilatory rate. However, a hysteresis effect in the relation PaCo~-apnea versus PEEP was found, which z suggests that the s~ppression of spontaneous breathing duringhigh frequency ventilation depends on the end expiratory volume rather than the end expiratory pressure.

A.Vuori and P.Jaakkola, Department of Anaesthesiology, University of Turku, Turku, Finland. Alveolar carbon dioxide tension is in balance with the pulmonary capillary and arterial blood carbon dioxide tension. Thus, end-tidal carbon dioxide tension (PETCO~) can offer useful information concerning the adequacy ofventilation. In order to find out the correlation between PETC02 and the arterial carbon dioxide tension (PaC02), linear reqression analysis was performed between 250 values of PETCO 2 and simultaneously obtained P.CO 2 values. The PETC02 was determined usine an infrared carbon dioxide analyser (Datex Normocap). The correlation was found good in patients on controlled mechanical ventilation (CMV) after open-heart surgery. The correlation coefficient (r) was 0.858 for a volumecycled ventilator (Servo 900B; n=176) and the r was 0.722 for a ventilator using a continuous fresh gas flow via the T-piece (IMV Bird; n=29). The gas samples were obtained from a sampling adapter interposed between the intubation tube and the T-piece. Thus, with the sampling method used, fresh gas flow Via the T-piece may mix with the respiratory gases and affect the reliability of PETCO 2 monitoring. In another series of samples, obtained from patients on CMV using the Servo 900B ventilators, the r was 0.898 for samples drawn directly from the intubation tube at the mouth level (n=26), while the r was 0.792 for samples drawn from the aforementioned adapter (n=19). Both samples correlate well with the PaC02, but it may be concluded that for best precision it is advantageous to draw the expiratory gas sample directly from the intubation tube.

EVALUATION OF THE ADVANCED EARLY ARTIFICIAL RESPIRATION IN CASE OF A POLYTRAUMA. dePay,A.W., Hohlbach,G. , Kllnik fur Chirurgie der Medizlnischen Hochschule L~beck

The development of post-traumatic respiratory insufficiency has been known for a long time, but the question of when r@spiratory therapy should be initiated is still debated. 21.6% of 236 emergency-treated polytraumatized (PT) patients of the degree II and III showed respiratory insufficiency already at the place of accident and during transport. Only 12.7% were intubated and 11.8% were treated with respiratory therapy prior to hospitalization. 27.9% had to undergo a respiratory therapy after hospitalization Necause of acute respiratory insufficiency. A p]rospective study in 56 polytraumatized patients showed that already 15 minutes after the trauma a clear reduction of the respiratory function could be observed, which could not be adequately treated with the shock therapy. 50% oT the PT II and 75% of the PT Ill patients developed acute respiratory insufficiency within 24 hours, making artificial respiration imperative. Depending on the degree of severity, the patients who were not given artificial respiration at the place of accident were liable to a 20 to 40% higher letality rate. The period of artificial respiration after hospitalization had to be longer to the same extent as the artificial respiration had been delayed after the accident. The conclusion is that all PT II and PT III patients, in particular highrisk patients, should be intubated and given respiratory therapy already at the nlace of accident even if their consciousness has been maintained.

(PEMC02), C. Chopin, J. Mangalaboyi, F. Fourrier, A. Durocher , D. Dubois, F. Wattel , Service de Rdanimation , HBpital Calmette, C.E.U. Lille, FRANCE. In the patient, ~ithout any previous lung disease, the measurement of the difference between PaCO 2 and end ezpiratory PC02. ~ P(a-ET)CO 2) may be an aid for diagnosis of pulmonary embolism (PE). On the contrary~P(a-ET)CO 2 in COLD patients, i8 increased whatever be the co~se of the ARE. Then the difference i8 linked with the ventilation/perfusion ratio mismatching and i8 due to both alveolar mixing defect (ventilatory mechanism) and dead space effect (circulatory mechanism). We speculate that the ventilatory part of the difference could be cancelled by forced expiration. 18 COLD patients with suspicious PE support our study. Pulmonary angiography and PEMCO 2 are performed the 8~e day. PEMCO 2 is calculated by continuous recording of forced ezpired CO 2 and simultaneous determination of PaCO 2. AP(a-PEM)CO 2 is then computed, and also the ratio D = ~P(a-PEM)CO2/PEMCO 2. Diazepam is among the most widely used drugs forthesedation of ventilated patients. However, it does not appear to be generally recognised that this kind of treatment might be responsible for a prolonged depression of the central nervous system and therefore a delay in extubationfor many days. Analysing two cases with massive accumulation of Diazepam and persistent coma the problem of Benzodiazepine-intoxication during sedation of ventilated patients is demonstrated on the basis of measured plasmaconcentrations and the diagnostic and therapeutic use of a new Benzodiazepine-antagonist (Ro 15-1788): i. a previously healthy 70 year old female with respiratory insufficiency following pulmonary embolism remained comatous during 12 days following administration of a total dose of 260 mg Diazep~m withinthefirst 3 days. Plaemaconcentrations of Diazepam and Desmethyldiazepam were 437 and 483 ng/ml, 150 hours after the last medication; the calculated elimination-halflifes of the two substances were 109 and 403 hours. 2. a 63 year old man with preexisting liver disease had to be ventilated because of severe bacterial pneumonia; following sedation during 16 days with a total dose of 245 mg Diazepam he remained cc~atous for additional 6 days. The semiquantitative urinary test for Diazepam was still positive i0 days after discontinuation of the drug. In both cases the Diazepam-induced depression of the central nervous system could be demonstrated and reversed under EEG-controlled conditions by the new Benzodiazepine-antagonist Ro 15-1788. We conclude that Diazepam is not an appropriate drug for long-term sedation because of its pharmacokinetic characteristics; the new Benzodiazepine-antagonist Ro 15-1788 seems to be a valuable tool in the diagnostic and therapeutic management of Benzodiazepine-intoxications. An increase of C5a measured by in vitro aggregation of polymorphonuclear granulocytes (PMN) has been detected in plasma of adult respiratory distress syndrome (ARDS) by Hammerschmidt et al (Lancet, 1980, I, 947-949) . To assess this phenomenon C5a activity was measured by chemiluminescence (CL) in control AB serum before (CS) and after zymosan activation (CSA) and in serum of patients with acute respiratory failure : ARDS, cardiogenic pulmonary edema (CPE), chronic obstructive pulmonary disease (COPD). CSa triggered the production by PMN of free radicals (O2~ H202/02, OH) which were assessed by CL (at the Ist mn). The results were first measured in relative light units (RLU), then a CL index was calculated :

RLU (test serum) -RLU (CS The present study shows that low molecular weight fractions of ARDS serums are able to induce a CL response, suggesting the presence of C5a. The non activation of PMN by whole serum is probably due to CSa inhibitors. By detecting the production of oxygen radicals, only with ARDS serums, our results contribute to the pathogenesis of th~ syndrome. To evaluate the relationship between different colloid osmotic-pulmonary artery wedge pressure gradients(COP-PAPW) and gas exchange,COP-PAPW values (Group A: > 9 torr; B:9-5; C:5-3; D<3)were correlated with shunt(Qs/Qt) and alveolar arterial oxygen difference(A-aDO 2) in 15 septic patients (12 died)in Acute Respiratory Failure(ARF).ARP was characterized by:A-aDOo=313*llS;Qs/Qt=22*9 %;Cardiac Index (CI)= 4,1.3 (mean*sd) ~ and radiological findings of oedema.

The mean~sd values of pressures were:COP=16*3.6 torr,PAPW =13"4 torr,COP-PAPW=3.3• torr. Data were obtained by ontometer and Swan-Ganz thermodilution catheters.

Results. Different COP-PAPW did not correlate with Qs/Qt and A-aDO 2 (see Table) . We compared several demand flow systems with a continuous flow system in respe~ of the inspiratory work of breathing using a mechanical lung model. The continuous flow system consisted of a reservoir of 15 lald a gas Z~ of 7ol/min. Endotracheal tubes of 4.5, 6.5, and 8.5 mmID, various tidal volumes, and different respiratory frequencies were used. 2ressurevolume-, pressure-flow-, and flow-volume-curves were recorded.

A strong dependency between work of breathing and tube diameters became evident. A typical example is demonstrated below, simulating conditions in an infant: tube diameter 4.5 mm, V T 8o ml, RF 24/min, PEEP o. The total inspiratory work of breathing increased with the demand flow system by 2o% and the inspiratory flow resistive work by Ioo%. During the first third of the breathing cycle those values were 24o% and 4oo% respectively.

That enlights the strong efforts which were necessary just for opening the demand valve. We Information regarding 09-uptake, C02-elimination, respiratory quotient an~ metabolic Fate is important for the determination of both certain haemodynamic parameters and relevant supply of parenteral nutrition. Until recently, the most commonly used method of getting these pieces of information involved using a Douglas bag and mass spectrometry. Drawbacks with this method are: -the equipment is huge and expensive.

-the operation of the equipment is time consuming and requires special education. In order to avoid those drawbacks we have designed a device based on indirect calorimetry and intended for measurements on patients connected to an Engstr~m Erica respirator. Inspired and expired gas samples are analysed by an electrochemical 02-sensor and an external CO 2analyser. Tidal volume ~nformation is retrieved from the ventilator. From values measured every 30 seconds, the 02-uptake, C02-elimination, respiratory quotient and metabolic rate are calculated. The results are available both as one-minute mean and as trend values. The difficulties associated with all the methods used consist in that the gas samples are taken at different humidity, temperature and pressure levels.

-inspired and expired volumes differ.

the small concentration differences to be measured. The influence of these factors has been minimized with the aid of automatic 02-sensor calibration, microprocessor-based numerical compensations and gas sample adaptation in a special tubing. Comparison tests with a mass spectrometer using a test model show a difference of less than 8 percent between the two methods. Clinical evaluation on patients is in progress. The aim of this research was to determine the effects of Nitroglycerine (NG) infusion on pulmonary circulation in patients with chronic obstructive lung disease (C.O.L.D.).

In 26 patients with severe C.O.L.D. under controlled ventilation, we have studied, after insertion of right heart and arterial radial catheters, haemodynamic effects and blo~d gas changes for an infusion rate of NG ol 10, 20, 40 pg.mn . In addition, 10 of the 26 patients had a volume loading with NG, and without NG after return to control values.

Before NG infusion, we observed, in all cases, an increase ol mean pulmonary pressures (PAP), an increase of pulmonary resistances, and an increase of right systolic work index. None oI the patients had features of left heart failure. After NG infusion, we noted a significant decrease in the systolic cardiac index producing a reduction of PAP, without any changes in pulmonary or peripheral resistances. Left and right systolic work decreased. All these haemodynamic changes were in direct proportion with NG rate. There were no significant changes in PaO2, PaCO2 except an increase in 2, 3 Diphosphoglycerate with NG. In addition, volume loading, resulted in an increase of right and left pre-loads and after loads, with and without NG. Cardiac out put and the systolic index increased with volume loading with NG, while they did not change with volume loading without NG.

These results suggest that~ during the acute course of respiratory failure in patients with C.O.L.D. under controlled ventilation, NG, may be useful particularly on right overlaod allowing the patient to leave the artificial respirator more quickly. Patients with COLD undergoing acute exacerbations are often malnourished and may require TPN during the initial phase of artifical ventilation. In this study energy expenditure was examined in 6 patients. Energy expenditure (EE)was measured by indirect calorimetry using a new device connected to a respirator. A 0 -microfuelcell takes alternate sampies from t~e inspirator} and exspiratory gases and the difference in oxygen concentrations is multiplied by the ventilatory flow-rate. Measurements were started 24 hours or earlier after initation of controlled mechanical ventilation. EE was continuosly monitored for 2 to 5 days. Measured values were compared to predicted values which were achieved by using Harris-Benedicts formula for BMR and adding 13~ of BMR For each centigrade elevation of body temperature above normal. Furthermore 10% was added because of the so called specific dynamic action of foodstuffs. The patients received TPN including fatemulsion and energy intake was between 19 and 30 keal/kg/24 hours and never exceeded measured EE. Measured EE was between 21 and 37 kca1/Kg/24 hours (mean: 29, 8, SD4, 8) . Measured values of EE were -5~ -+55~ above the predicted values (mean 25,3% SD 18,5}. The reason for this hypermetabolism observed in all patients except one could be the energy cost of filling up energy stores which had been depleted prior to parenteral nutrition. It is difficult to predict the great variations in EE so continual measurements should be made in order to adjust energy intake to requirements.

PNEUMONIA COMPLICATING SELF POISONING WITH PSYCHOACTIVE DRUGS.Offenstadt G,Gabillet G,H~ricord P,Pinta P,Tembely A, Amstutz P -H6pital Saint-Antoine -75012 Paris France. 474 patients admitted in ICU between 1976 and 1981 were retrospectively analysed. Pneumonia (P) was assessed by condensation on chest X ray.

-P developped in 67 patients (group l).In group I mean + . . . .

age (40,4-18 years) was hlgher than zn patlents wzthout P (group II) p~O,O01. In group I,P was present at admission in 53,7 p.eent of the cases. 24 hours after admission 77,6 p.cent of P had appeared.

-Initial location was unilateral in 79,4 p.cent with predilection to the inferior half of the right lung (72p.cent). Fever was more frequent in group I (89,5 p.cent) than in group If (37,7 p.eent) p<O,O01.

-Promoting factors were observed : -Delay between ingestion and admission was much longer in group I 27,6• h than in group II 7,7!9,4 (p<O,O01).

-Vomiting and tracheobronchial embarassment were more frequent in group I,respectively 17,7 p.cent and 63,2 p.cent than in group II 3,7 p.cent and 20,7 p.cent (p<O,OO1).

-Coma was deeper in group I (2,3~0,8) than in group II (1,4~0,9) p~O,OO1.Quantifications of depth was obtained by giving to comas values from 0 to 4.

recovery was simple in 83,8 p.cent of cases in less than 4 days (50 p.cent) and with antibiotics in B3,4 p.cent.

-Abcedation appeared in 3 cases.

The overall mortality was 2,8 p.cent (14 patients); 6 died because only of P,in 4 others,P was aggravating.

-At last duration in ICU was much longer in group I (9~8,1 days) than in II (2,5~2,1) p~O,O01. -P is major complication of self poisoning with psychoactive drugs. Prevention must be a main objectif.

PROCOLLAGEN-III-PEPTIDE IN PROGRESSIVE RESPIRA-TORY FAILURE AFTER PARAQUAT INTOXICATION. J.Lohmsnn, G.Pott, H.P.Bertram, U.Gsrlach, Medizinische Universit~tsklinik,MUnster,W-Germany Patients with severe psraqust poisoning oftsn died because of respiratory failure. Changes in pulmonary function are due to a progressive pulmonary fibrosis 8 to 10 days after intoxication. Procollsgen-III-peptide (P-III-P) was measured enhanced by rsdioimmuno assay in patients with fibrosis of the liver.We determined P-III-P in serum of 4 patients who died after ingestion of paraquat and in 2 petients who survived.ln case of the non-survivors P-III-P increased to 4 fold of normal values. In case of the survivors P-III-P slightly increased then no further progrsssion was seen. Our data suggest that estimation of P-III-P in serum of patients with pera qust intoxications may represent paraQuat-induced pulmonary complications. Dionigi, L.Carnevale, A.Pernigotti, M.Maurelli, G.Lorenzon Vos, A.Mapelli, Rianimazionell, Policlinico, Pavia, Italia Invasive procedures that are potential vehicles ofinfection were scored according to seat, technique and during of application, using ascoring system(ISS)~,in order to value the relationamong nosocomialinfections, extent ofinvasion and other risk factors in IntensiveCare(ICU) patients(pts). I00 pts (9-82 years,mean 50.7), admitted to ICUfor 3days at least because of different ailments, were sorted in 4 groups (A-D) according to the aggregate score of ISS (IS),growing from A roD, during their stay in ICU (15.5 ~9.3 days). The mean infection rate was 60%: 18.5% in group A, 100% in D. Significative correlation was found between IS and number of infectious episodes per patient (ni/p)(p<O.O01! and between ni/p and during of stay (p<O.O02). Infectionrateresulted more significantly related to IS (p<O.Ol) than to during of stay(p<O.O5).No significative correlation was found between infection rate and patient!s age and between infection rate and immunodeficiencyfrom corticosteroid therapy. A moderate invasiveness have not a predominant roll for infection's rise in comparison to the during of stay, the immunodeficiency and the age; it is dangerous when it gets over a medium level; then, it seems to increase any other risk factor. The ISS proved to be usefull for studying comparatively heterogeneous pts and it could have a clinical application, if calculated at the beginning of the stay in ICU, in order to identificate pts with high risk of procedure-related infection. ~Geraci P. etal., Atti 34~ 1982; in press. Before reconstruction and air-conditioning of two neurosurgioal operating-theatres and an adjacent, completely open intensive care unit, traeheobronehial secretions and urine samples from Ioo patients attended in this unit were daily examined for nosocomial and other microorganisms. After the reconstruction project including installation of an air-conditioning system with highperformance filters and strict structural and organizational separation of the unit from other parts of the hospital had beed completed, the daily microbiological surveillance was continued for two years and provided data from 50o patients. A comparison of frequency, species distribution and clinical significance of the microbes obtained from the unit which was run all the time with the same personnel, before and after reconstruction revealed the specific problems associated with a strictly isolated ward. In addition, the changes observed in the spectrum of the nosocomial bacteria, their antibiotic susceptibility patterns and their clinical significance could be used to evaluate the therapeutic measures employed and to establish more appropriate therapeutic schemes if necessary.

ICU ACQUIRED PNEUMONIA: A PROSPECTIVE STUDY. M.Langer, S.Vesconi, M.Solca, A.Graziina. letituto di Anestesiologia e Rianima~ione, Ospedale Policlinico, Nilan/Italy Incidence t time of onset,and mortality of ICU acquired pulmonary infection was evaluated in 96 pts from 2 Units (general and neurosurgical) in a period of 6 months. Pts with an ICU stay less than 48h, preexisting pulmonary infection and irreversible brain damage were excluded. In the 96 stud~Ls the mean value of Therapeutic Index Score System (CULLEN) was 33.9-+ 13. 68 pts (71%)~equired continuous mechanical ventilation fo~ longer than 24h (median 8 days). Pts were considered • have eithe~espi~a tory tract infection if they had persistent fever and consistent clinical data, or pneumonia if a persistend infiltrate on chest X-ray film was also present. Overall mortality was 25/96 pts (264). 46 pts had pulmo-nary~nfection, 25 pneumonia (mortality=60%), 21 respirato-ry~raot infection without infiltrate (mortality=28.5~.~); 50 pts(mortality=8%)did not develop pulmon~.ry infection. Mortality rates observed in the 3 groups are significantly different (p ~ O.OO05). Among pig with pneumonia, the onset of infection was surprisingly early in over 50%of cases.(teble) pneumonia onset(ICU day) 1-3 4-7 8 No pt s/expired 14/5 6/5 6/6 % of TCU acquired pneumonias 5g 24 20

Pneumonia with later onset was associated with a significantly higher mortality (P(0.O25). Beside the well ~own high mortality rate associated with pulmonary infection in ICU pts, this study emphasizes the high incidence of early onset pneumonias. LEGIONEILA INFECTION IN AN INTENSIVE CARE UNIT FROM 1977 TO 1952 Hospital the University of Nottingham, England. Iegionaires disease first appeared in an identifiable form in the Nottingham area in 1977. Initially identification of the disease was difficult and early blood specimens had to be sent to the National Centre for Disease Control in Atlanta, Georgia, USA, for identification of antibodies. This did not help early diagnosis| even when immunological identification became available locally the delay of fourteen days for a significant antibody response precluded the rapid recognition of the disease by that means. Fortunately distinctive radiological, biochemical and haematological changes came to be reeognised. Patients could therefore be treated early on the basis of these findings alone with Erythromycin, which appears to be almost the only available effective antibiotic which is effective in this disease, Almost all the patients referred to the intensive care unit required prolonged lung ventilation with (at times) very high levels of inspired oxygen. Intravenous nutrition was often required for long periods although naso gastric feeding was preferred when ever this was possible. Renal failure, when it occurred required dialysis. A review of the incidence of pneumonia cases since 1973 suggests that there is now a cyclical legionalla factor occurring every two years in those patients requiring mechanical respiratory support. The STAF ~ isolated in blood may be dismissed as contaminants. However, STAF-septicaemlas (SEPT) are well recognized in patients with intravascular fore[ng bodies (cardiac prostheses, ventricular-atrial shunts, vascular grafts) Intravascular catheters (VASCATH) are very commonly used in critically ill patients and they may favour the prese~ ce of STAF-SEPT. in them CRITERIA to diagnose a STAF-SEPT in this study were: I/ clinical septic condition, 2/ isolation of STAF-in at least two blood culture bottles and 3/ presen ce of VASCATH whether or not the STAF -was isolated fro~tem. RESULTS.-Among the 1141 admissions to our ICU, 504 had blood cultures taken and they were positive in 210 patients (15% incidence). In 44 (9% of all blood cultures) STAF-was isolated. However, in 32 (73%) was considered as contaminant and only in 12 (27%) as pathogenic for the above criteria were met. The STAF -was iso]ated in 4 patients in the course of a "multip]e species septicaemia '= and in 8 was the only organism isolated (mean of 3.6 bott]es). One patient with endocarditis in a valvular prostesis was excluded. Hence in 7 patients the STAF-was thought to be the responsible of a " primary STAF-SEPT". Their mean age was 68.3 years (range 51-73 years). All patients had VASCATH (mean 1.9/per patient) with a mean duration of catheterization (n=6) of 7.2 days (range 2-15 days). In 5 STAF-was cultivated from the catheter's tip. One patient developped shock during the SEPT. CONCLUSIONS STAF-appears frequently in blood samples, but usually represents a contamination. However, the repetitive isolation of STAF-in blood samples in patients with VASCATH must be considered as a relevant finding concerning the posibility of a STAF -SEPT. According to literature,mortality of patients suffering from AGP is higher than 50% (in some cases,higher than 90%).In 115 patients (68 primary AGP,47 secondary AGP) oh= served over 20 months,admitted in ICU for a stay over 5 days,were investigated some factors of probable prognostic meaning: pathogenesis (primary, seeondary),source of in= fection (supra-or infra-mesocolic), interval between eli= nical onset and operation, interval between operation and admittance in ICU, stay in ICU, percentage of complica= tione. Total mortality was 40.9% (39.7'% in primary AGP, 42.6% in secondary AGP), while mortality in infra-mesoco= lic AGP was significantly lower. The interval between ope= ration and admittance in ICU was nearly the some in the patients who recovered (1.0+2.6 days) and in those who died (0.6+1.4 days).The interval between clinical onset and operation was remarkably longer for the patients who died. In this last group, the percentage of complications was significantly higher too. According to literature, the investigation confirmed a higher mortality in secondary AGP, a clear connection between operation delay, compliea= tions percentage and mortality. In opposition to literatu= re, it showed a mortality significantly higher in the su= pra-mesocolic versus infra-mesocolic peritonitis. The low total mortality percentage is probably due to the early admission to ICU, immediately after the operation, 3oefore the onset of haemodynamic, respiratory and~or renal com= plications 2

Torr pH I' iP50 Std Nineteen patients (A) admitted in ICU for a severe sepsis were studied:septicemia (7),pneumonia (6),peritonitis (5), extensive eeliulitis (1).There were lO men and 9 women.The mean age wee 57 years. Death rate was 79 %.

On arterial blood we measured:blood gas,hemoglobin (Hb), 2-3 diphesphoglycsrate (2-3 DPG),standard P50 (PSO std)from the oxyhemoglobin dissociation curve (OCD)(dynamio method). 49 samples were performed,ll patients have had daily studies.12 healthy subjects (B) were also studied. Large and unforeseeable variations of PSO Std and 2-3 DPG were observed.There was a etricking right shift of the ODC. We found no definite explanation to this shift.Hypoxemia and anemia are theorical causes,but in our study there was no correlation between PSO Std and Pa 02 or Hb,either between 2-3 DPG and Pa 02 or Hb.

In current practice, blood oxygenation should be assessed by simultaneous measurement of oxygen arterial pressure and oxyhemoglobin saturation. Glucose controlled insulin infusion systems (GCIIS) seem to be a valuable tool in the metabolic treatment of critically ill patients. The incidence of 4320 measurements frcm the GCIIS during a 24-hour control period is the main reascn for developing this method, which permits a ccmfortable handling of a vast number of data. The procedure is based on a BASIC ccrnputer pr6grar~ne,that will be demcnstrate~. /mFatmode: An interface between the GCIIS and the microccmputer PC 800 (MAGIRUS [I~TKNTEC~IK) has been employed to transfer all data directly at one minute intervals. Further, comments concerning actual therapy can be inserted manually. Calculations:The following parameters,characterizing the state of carbohydrate metabolism, may be calculated: -MBG = mean blood glucose -M-value = a measure for the duration and the degree of pathological blood glucose levels (I) -MAGE = mean anplitude of glycemic excursions (2) -an index, indicating the sensitivity to insulin. Record:Blood glucose, insulin rate and dextrose rate are plotted continuously. CcGments will be printed, too. Documentation: All recorded data are stored on a 5 inch floppy disk. Conclusion: The method permits cc~plete processing of GCIIS-data, which is almost inpossible when confined to manual ccmputaticn. Its usefulness in the treatment of critically ill patients, especially during metabolic therapy with glucose-insulin-potassium in acute myocardial infarcticn, will be de/ncnstrated (Supported by the Federico Fo~dation). I. Schlichtkrull, J. et al.:Acta med.Scand.177(1965) Persistance of blood flow through atelectatic areas of the lung cause hypoxia by increased venous admixture.Hypoxie pulmonary vasoconstriction (HPV)should reduce ventilatio~perfusion-mismatching.0ne-lung-anaesthesia is an example of massive atelectasis and a suitable model to study the influence of prostaglandins(PG)on intrapulmonary shunting and HPV. Nine patients were studied during esophageal surgery under two-and one-lung ventilation(TLV,OLV) with 66%02in N20 and neurolept anaesthesia. Systemic and pulmonary arterial pressures and blood gases and cardiac output were measured. During TLV and OLV arterial and mixed venous levels of PGE 2 and PGF~ were determined. Cardiac~output and pCOodid not change significantly,pOo decreased from 219+42mmHg(TLV) to 7~+18 mmHg(@LV)(p~0,01) and Q ~Q~increased from 74i4% (TLV) to 43+10%(0LV)(p~8~O1y.PGE 2 increased f~em 2.83+1,59ng~ml(TLV) to 3,06A2,21ng/ml(OLV)(n.s.) and ~GFo~decreased from 0,35+0,53ng/ml(TLV) to 0,19+0,~ng/ml(OLV)(n.s.).Dur~ng TLV and OLV PGE o was ~ignificantly above normal values (0,76+0,47 ng/ml), while PGFo~ was below normal level(~,87+ 0,2ng/ml). Mixed ~enous PG levels were generall~ below the arterial levels.

These findings indicate a clear prevalence of vasodilating PGE 2 during mechanical ventilation even if massive atelectasis is present. PGE may have partly offset HPV in the presence ~f atelectasis.0ther metabolites of arachidonic acid like prostacyclin could also be inv&ved. 

The purpose of the study was a quantitative determination of functional pulmonary tests in disturbances of cerebral control of breathing and in paralysis of respiratory muscles. Spirometry was made by means of the "POOL" apparatus (USER). Gas composition of theblood and acid-base balance were determined by using the equipment of the "Radiometer" Company (Denmark). Disturbance of cerebral control of breathing occurs in lesions of the corticospinal and corticonuclear connexions: a) Unilateral destruction of these systems in the right hemisphere causes the syndrome of partial loss of cerebral control. Spirometry indicated restrictive forth of respiratory insufficiency, b) Bilateral destruction of the corticonuclear and corticobulbar connexions in the brain stem leads to complete loss of cerebral control and consequently of voluntary breathing and cough. Automatic control provides gas exchange under condition of free airways (tracheostomy, suction from the tracheobronchial tree). Respiratory insufficiency was referred to obstructive type. Diseases of peripheral neuron (paralysis of respiratory muscles and bulbar palsy) are followed by respiratory insufficiency of restrictive-obstructive type. Restoration of pulmonary functions occurs within 8 to 12 months from the development of paralyses in acute polyradiculoneuritis. In severe form of myasthenia gTavis this type of respiratory insufficiency has been observed for many years.

ACIDIC PHOSPHOLIPIDS IN BRONCHOALVEOLAR LAVAGE DURING RESPIRATORY FAILURE OF THE ADULT. J.Tahvanainen, M.Hallman, P.Nikki, Department of Anaesthesia, Helsinki University Central Hospital, and Children's Hospital, University of Helsinki, SF-00290 Helsinki 29, Finland Normal lung surfactant contains prominent phosphatidylglycerol (PG), or phosphatidylinositol (PI), but only trace of phosphatidylserine (PS). These phospholipids were investigated in lung effluent from 14 adult patients with acute respiratory failure. 31 samples were obtained by bronchoalveolar lavage using a 70 cm silicone catheter. PG, PI and PS were measured as percentages of total phospholipids in lavage return. Prior to sampling pulmonary venous admixture (Qs/Qt) and lung thoracic compliance (CLT) were measured. CLT correlated significantly with PG (r=0.61, p<.001)and PS (r=-0.62, p<.O01), but not with PI. No correlation between any of the phospholipids and Qs/Qt was found. Patients who died of respiratory failure (n=8) had higher PS than those who survived (n=8), i.e. 8.4 • 1.2% vs 3.8 • 1.1% (p<.05). CLT was 30 + 3ml/cm H20 in the former group and 46 ! 8 ml/cm H^O in the latter one (p<.O5). PG, PI or ~s/Qt v~lues did not differ significantly between these two groups. Our data suggest that deficient surfactant phospholipids are at least partly responsible of reduced lung elasticity in acute respiratory failure of the adult. Furthermore, a patient with deteriorated lung nechanics and a prominent PS has poor prognosis.

Supported by the Academy of Finland.

NEW SAMPLING TECHNIQUE FOR RESPIRATSRY GAS MEASUREMENTS.

A.Gedeon +, K.Hamilton +, S.H~ggmark, S.Reiz +Engstr~m Medical, Bromma, Regionsjukhuset Ume~, Ume~, Sweden In order to simplify the clinical handling and to improve the response time of medical gas analyzers a new sampling line has been developed that eliminates water fall out from the sampled gas by acting as a selective permeable membrane for water vapor.

Results: As the gas is drawn into a CO 2 or an 02 analyzer it passes along the sampling tube and by diffusion is gradually equilibrated with the ambient air as regards temperature and relative humidity. Continuous measurement of energy expenditure by methods of indirect calorimetry in vsntilatory assisted intensive care patients has therapeutic as well as prognostic implications. The ventilator Erika (Engstr~m Medical AB, Sweden)has been supplemented with a device for oxygen uptake measurement and energy expenditure calculation, the Engstr6m Netabolie Computer (EMC). In the EMC, inspiratory and expiratory 02 eomentrations are determined by an electroDhemical 02 sensor(Catalyst Research). The inspiratery flow (V~ measured by a pneumotachygrsph, is used in the calculation of oxygen consumption. This study examines the influence of different Oe concentrations (F02) , ventilatory frequencies (f), tidal ~ volume~vt) and PEEP on the EMC. Oxygen concentrations (21-i00%) measured by the EMC were eompsredwith values from a masspeetrometsr(Centronie 200). The inspiratory flow signal was compared with values from s Bernstein mechanical spirometer at different ventilator settings.

T~est analysis gave no difference at [02=21% , but significant differences (p:O,O01) for F02~212. The eoeffieient of variation at FO 2 21-100Z was 0,03-0~0~ 02 for the EMCand O,04-O,lOZ O~ for the mssspectrometer. The inspiratory flow meter was ~ependant on F02, f, vt and PEEP. Thus, increasing FO 2 from 21 to lO0~ decreased ~I by 6%;inereasingvt from 0,3 to 1,1 1 increased V I by 6%; increasing ffrom i0 to 20 bpm increased V I by 72 and increasing PEEP from 0 to 20 em H20 decreased 91 by 32. In most clinical situations (Flo ~ 21-60% f 10-20 bpm, vt 0,5-0,9 i t PEEP 0-20cm HgO), the o~ygen and flew meter of the EMC will yield an ~rror in the calculation of oxygen consumption of less than • 5 Z. In order to evaluate the significance of hypoxemia.arising from ventilation to perfusion (VA/Q) abnormalities, PaO2, arterial 02 saturation (Sa02), and transcutaneous 02 tension (PteO 2) were measured on lowering FIO 2 in 47 patients before extubation. All patients had been weaned from mechanical ventilation and CPAP after respirator treatment of at least 36 hours. The patients entered the study on the following extubation criteria: no mechanical breaths,CPAP of 5 cm H20 , FIO 2 of 0.35, PaO 2 60 mmHg and PaCO 2 45 mmHg. In the study, FIO 2 was lowered in the following order: 0.40, 0.35, 0.30, 0.25 and 0.21.

No significant difference was found in PaO 2 or SaO 2 with any FIO 2 between those who were successfully extubated (group S, n=38) and those who required reintQbatlon within 24 hours (group R, n=9). On the other hand, PtcO2/PaO ~ ratios decreased from 0.82 to 0.66 in group K whereas those remained above 0.8 in group S. Hb concentrations ~nd P58 values were lower and septic hypermetabolism more frequent in group R.

These data indicate that hypoxemia arising from VA/Q abnormalities do not solely predict suecesful extubation in patients with eompromized oxygenation. The safe lower limit of PaO 2 depends on extrapulmonary factors, such as Hb, P50 and metabolic state, PteO2/PaO 2 ratio seems to be a valuable eriterla of weanlng, because it probably reflects peripheral tissue oxygenation.

Supported by the Academy of Finland.

Zentrale Anaesthesieabteilung,Bergmannsheil, Ruhruniversit//t Bochum, BRD. The term "preload" usually denotes the force that must be supported by the ventricles of the heart at the end of the diastolic phase: the maximum of diastolic filling pressure, volume an muscle fiber tension ( as given by Laplace relation) is reached; rising ventricular pressure crosses atrial pressure, atrioventricular valve closes. In contrast to this the d~amic point of view leads to quite different conclusions. The prupose of the present investigation is to describe the dynamic relationship between pressure and flow across the mitral valve. Due to inertia of mitral blood flow, valve clnsure is delayed for a certain time after the atrioventricular pres, sure crossover. This delay has been investigated by means of dLmensional analysis. A nondimensional number occurring in this analysis has been "taken from the report of an experiment with dogs and then used to determine the conditions of mitral valve closure ( i.e. the magnitude of dynamic preload) in man. Results: rat/ler high (&p 80 Torr) pressure are obtained during e~ercise( heart rate 172/ rnin. ) and about 10% of the stroke volume enters the ventricle and increases its volt~L~ before the mitral valve closes. The theoretical prediction is 'then verified by echocardiography of the isovolumetric peried(IP).Mitral valve closure and aortic valve opening are picked up simultaneously with two echeoardiographic apparatus. Thus IP can be measured directly.Result: IP decreases at a rate greater than that of the increase in heart rate resulting from exercise and even disappears altogether when the heart rate reaches a sufficiently high level (140/min.). It is possible for the dynamic preload to overlap with the beginning of the afterload period, as indicated by aortic valve opening. Controlled mechanical ventilation (CMV), intermittent mandatory ventilation (IMV) and spontaneous breathing with continuous positive airway pressure (CPAP) were compared in 9 patients with acute myocardial infarction complicated by respiratory failure. Cardiopulmonary function was assessed by measuring thermodilution cardiac output, circulatory pressures, blood gases and the sum of the ST-segment deviations in a 12-lead ECG (AZST).

The ventilator frequency was altered to deliver i00% (CMV), 50% (IMV) and 0% (CPAP) of the patient's minute ventilation with a tidal volume of 12 ml/kg.

The end-expiratory pressure (range 5-10 cmH20) and the inspired oxygen concentration were kept unchanged.

The arterial blood oxygen saturation was > 90% at all times. The respiratory rate during CPAP was 10-24 breaths/min; no CO 2 retention was seen. The essential results were as follows: 2)The withdrawal of ventilatory support is associated with an increase in systemic vascular resistance and may cause myocardial ischemia with reduced cardiac performance even when the blood gases and the respiratory rate are norraal. Within a period of one year, in our intensive care unit five patients developed "torsades de pointes" (TP) with production of syncopes. In all patients E.C.G. revealed QT prolongation longer than 0,55 see and mild bradycardia just before the onset of TP ; at the same time, in three patignts a markedly hypokaliemia (<3 meq/l) was noted, but in two others, kaliemia was normal (4,6 and 4,7 meq/l). The~e two ones, 74 and 79-year-old women, presented no congestive heart failure, cirrhosis, denutrition, diarrhea, rena~ or surrenal dysfunction ; calcemia, magnesemia, protidemla andhlood pH were normal ; no antiarythmic, diuret~6, mineraloeortico~d or laxative drugs were given ; daily urinary potassium was in both eases about 50 meq, balancing imputs ; a measurement of total exchangeable potassium by K 42 revealed a significant depression of 30 % and 28 %. In all our patients ventricular pacing was safely rapidly and temporary used, and at the same time large daily amounts of parenteral potassium were given. In the two normokaliemic patients kaliemia and kaliuresis were without appreciable change, QT returned to normal within three and four days, and under daily potassiu m supplements, none episode of TP occured later without any other treatment. It seems thus that some patients present unexplained potassium deficiency without hypokaliemia and that potassium chronic Supply should be an appropriate measure to prevent dramatic episodes of TP. (~-and B-stimulant) in CPR was compared in this experimental study in dogs.

After five minutes of anoxial cardiac arrest we measured continuosly the following parameters:

hemodynamics, blood gases and electrolytes,binchemic parameters of aerobic and anerobic metabolism.

Left ventricular and central aortic pressure was significantly higher in the E-(n=11) and Ngroup (n=8) than Ln the O-group (n=8) during internal cardiac massage.lncidence of fibrillation was equal in all three groups. The pressures after defibrillation were significantly hLgher with E than with N and 0 (p 0.001).

Consequently the coronary blood flow was better with E as well. This was strongly correlated with survival. ( E: 11/11, N: 5/8, O: 2/8 ) We conclude that epinephrine is superior to N and 0 to restard spontaneous circulation because of its combined 4-and B-stimulating activity.

But epinephrine is not yet the ideal drug in CPR. (3) Germany. Electrical intracardiac stimulation during drug resistant cardiac arrest usually is not available in the ambulance; furthermore correct positioning of the endocardial lead may be difficult and time-consuming during CPR. Therefore we have employed an external mechanical device for cardiac stimulation during the lag period until definitive placement of the endocardial lead. Pressure waves are applied in the 4th intercostal space, using a 2.5 cm diameter thumper with variable rate and adjustable mechanical energy. With proper angulation and sufficient energy, electrical cardiac potentials are induced with subsequent contraction of the heart. In 8 patients with cardiac arrest we used this device during CPR. In 4 of these patients it was possible to obtain a sufficient circulatory function without cardiac massage until stable endocardial electrical pacing could be established. As could be expected, in cases of electromechanical dissociation no effective circulation was obtained. In these 4 patients no mechanical response followed the induced cardiac potentials. This was Confirmed by subsequent electrical stimulation. These preliminary results indicate, that external pacing with a mechanical pressure wave device may be useful in the emergency treatment of patients with cardiac arrest, whileendocardial pacing is accomplished.

BEDSIDE COMPUTER FOR PATIENT SAFETY, DATA COL-LECTION AND DATA PROCESSING. U. Frucht, P. Kunow, Th. Kersting, K. Reinhart and R. Dennhard, Dept. of Anesthesiology and Operative Intensive Care, Free University, Berlin, FRG Introduction: In applying vasoactive substances, false dosages can lead to dangerous cardiovascular alterations. The prevention of blood pressure crises resulting from sudden changes in systemic vascular resistance (SVR) and cardiac output (CO) signifies an advance in therapy with these substances. Methods: We have developed a computer the size of a standard clinical monitor, which , as part of a control system, safely doses vasoactive substances. The advantages of this system (monitor, computer and metering pump) consist in its safety and variability. After input of the applied substance, the maximal dose and a code for the direction of the blood pressure change, the program functions automatically. Fixed limits (dose, blood pressure) are maintained; an alarm is set off when they are reached. In addition, important calculations (hemodynamic profile, pulmonary shunt volume and P50) can be made with a separate computer program (I) and a data logger. Conclusion: The computer stores the dose of the vasoactive substance, the changes in the hemodynamic parameters and the calculated parameters for further statistical evaluation. References: I. Th. Kersting, K. Reinhart, G. Gassch~tz and K.

Eyrich: Easy-to- Hemodynamie parameters obtained by heart chambers and major vessels catheterization are registered automaticall F wi~h the help of computer based monitoring system.Oxygen oalance and acid base status data input to assess the adequacy of circulation is performed with the use of bedside microcomputer key board. The algorFthm is based on clarification of images.Normal type of hemodynamics, hyperkinetic, hFpokinetic type / hypovolemia, compensated and dscompensated heart failure and shock / are defined.Left ventricular and right ventricular pump function are estimated seperately.By the experts' assessment 93% of computer conclusions apNearsd to be correct. A g8 year old patient (pt.) with IHSS having been under medical control for q2 years ran into clinical state IV NYHA. She was treated by means of right ventricular stimulation in VVl-mode only. The preexisting tachyarrhythmia was suppressed by 5o mg Atenolol. Heart rate fell from qq0 to 8o. A stimulation rate of 90 was high enough to provide only pacemaker stimulation. Thus the pt. recovered to such an extend that we decided to implant a permanent pacemaker. The following criteria proved benefit from this procedure at the moment of changing from spontaneous rhythm to pacemaker rhythm: q.Arterial blood pressure raised from 8o to qqo mmHg. 2.Left ventricular high pressure (apex) fell from 23o to qgo mmHg. 3.Echoeardiography showed remarkable reduction of SAM.

The "false" way of excitation of the myocardium causes the apex parts of the heart to contract prior to the septal parts. Therefore the typical trapping phenomenon of IHSS cou~ be diminished. Several and complex problems associated to the hemodynamic conditions and the peculiar pharmacology of dopamine , are frequently met by clinicians tPeatin9 shock syndPom e.

The following report is a preliminary attempt to design a system of physician-computer interaction in the management of shock syndrome and in particular in the adjustment of an optimal pharmacological (dopamine) intervent ion.

The report describes a system of dopamine infusion automatic control as well as the general procedure ado_. pred. Clinical application of the system allows to specify the following advantages: -proper collection and organisation of specific data; -rapid computation of hemodynamic data and estimation of optimal dopamine infusion which is constantly adjusted; -computer-aided measurements and calculations; In order to improve the surgical treatment of patients (pts) with Type II-I aortic dissection we reviewed the records of 36 consecutive pts operated on by our surgical group between march 1971 and november 1982. Sixteen pns required an operation ou emergency basis ~group A) because of signs of impending aortic rupture (12) oliguria (i), development of neurologic defict (2), failure co control hypertension (I). Twenty ~ts had uncomplicated dissection and the operation was performed after stabilization with medical treatment (group B). The overall mortality rate was 33% with i0 hospital deaths in group A (62%) and 2 in group B (10%). We performed 15 operations uszng femoro-femoral by-pass with 7 deaths (46%), 16 using Gott shunt with ~ deaths (25%), 4 using simple cross-cl~aping technique with I death (25%). Only in one patient we used successfully left atrial-femoral artery bypass. The main complications were spinal cord injury in 5 pus, renal insufficiency in 5, left ventricular failure in 4, hemorrhage in 3. We concluded that the cperauive survival race is significanr higher (p(O.05) in stabilized patients; tat spinal cord znjury is unpredictable and probably not related co perfusion techniques; renal failure seems to be related co low perfusion; the Gott procedure zs more satisfactory in providing proximal decompression The optimum treatment of DAA-surgical or solely medicalremains uncertain. Since there is little information about long term results this study persnes prognoeie of operative therapy of DAA over a 7-year period. ~ Since 1974, an unselected cohort of 32 patients had surgical therapy of DAA. All patients were followed up until July 1982, datas were analyzed retrospectively (76.8 pat.yrs,max.surv. 7.2 yrs) . 27 aneurysms were dissected (84%). There were 17 traumatic(S3%) andl3 arteriosclerotic (41%) DAA. In either 16 patients (50%) clinical course was acute or chronic. 30 patients had replacement of DAA by vascular graft, 2 had local correction. All but i patient were operated with ECC or temporary shunt. ~ Overall operative mortality (OM) was 37.5%, dn dissecting DAA 41% (ii/24). OM in traumatic DAA was 18% (3/17), 2/3 patients died in table due to acute rupture. OM in arteriosclerotic DAA was 54% (7/13). Postop. no patient had paraplegia; 4 patients (13%) required hemodialysis because of temporary renal failure, i patient with replacement of DAA died 3 months postop., all other patients were alive on control key date, 61% resuming their professional activities. We conclude that best operative results can be achieved in traumatic DAA . However, the high incidence of intraop. rupture in this group emphasizes the need of rapid diagnosis and surgery. Patients with arteriosclerotic DAA have the highest OM due to poor quality of aortic tissue and to the elevated morbidity of these patients. Although none of our patients suffered paraplegia the incidence of renal dysfunction underlines the importance of ECC or sufficient shunt. Late mortality after surgical therapy of DAA is low, patients generally resume normal private and professional activities.

Comparison of metabolic and endocrine changes produced by balanced anesthesia vs narconeuroleptanalgesia during cardiac surgery.

J.J Arnulf*,C.Isetta*, F.R.Kuntschen**, P.M Galetti *~* Service Pr V.Dor. Institut A.Tzanck 06700 Saint Laurent du Var -F-* D.A.R. Nice Pr P.Maestracci ** Fondation organes articiels. Clinique de Genolier. Pr CH. Hahn -CH-*** Artificiel organ laboratory Brown university RI -USA -To compare the degree of metabolic protection afforded during cardiopulmonary bypass (CPBP) by the two most widely used techniques of anesthesia, 15 patients (14 mons and I women) undergoing elective coronary surgery (If) or cardiac valve replacement (4) had their blood glucose levels contineusly monitored by means of the Biostator glucose controller. The patients received no exogenous dextrose and no insuline from either the anesthesiologist or the priming fluid of the heart lung machine. Blood sa~oles for metabolic and endocrine determinations were obtained before induction of anesthesia, before CPBP, after 30 min of CPBP, at the end of CPBP and 50' after discontinuation of CPBP. Eight patients were operated under balane~ analgesia using enflurane, flunitrazep~n,fentanyl, pancuroniem and nitrous oxide, whereas 7 patients received narceneurolept analgesia using thiopenthal, droperidol, phenoperidine, pancuroniem and nitrous oxide. Starting from the same fasting blood glucose level as the balanced anesthesia group, the patients under narconeuroleptanalgesia exhibited a lower hyperglycemic response to non pulsatile CPBP with the same levels of circulation insulin.This was coupled with lower blood levels of cortisol, and highter levels of glucacon. The narconeureleptanalgesia group also showed significantly lower levels of alanine and 3 -hydroxybutyrate compared to the balanced analgesia group. Lactate, pyruvate, epinephrine, norepinephrine, growth hormone, free fatty acids, glycerol, triglycerides and cholesterol leve|s did not differ significantly between the two groups. NA, K, mg, Ca and P remanied within normal limits. These observations suggest that during narconeuroleptanalgesia, there is an increased glucose utilization by peripheral tissues, coupled with reduced proteolysis and glyconeogenesis as compared with balanced anesthesia. Lipolysis is not affected by the type of anesthesia used for cardiac surgery. The pulmonary lesion produced by ECC causes a decrease in FRC and in arterial oxygenation. In the iranediate postoperative period the addition of PEEP can reverse this effect. 64 patients (42male, 22 female, mean age 55 years) who had undergone CABG (n=4o) or valve replacement (n=24) were included in this study. Group I (n=34) were weaned with CPAP 7, group II (n=3o) were weaned with Q-insufflation. Weaning was started 9.25• hours after ending operation. The patients did not receive any narcotic or sedative, nor were they on any vasopressor agents. Blood gases and respiratory rate (RR) were documented and PAO2-PaO2/PAQ (AaD% quotient) was calculated in each of these 3 conditions: I. CMV, both groups PEEP7; 2. Intubated, spontaneous breathing PEEP 7 (group I), ZEEP (group II); 3. I hour after extubation, both groups ZEEP. Each patient had only I FiQ throughout the study. The data presented in the following Under circumstances of intermittent and continuous positive pressure ventilation (IPPV and CPPV) right ventricular output (Qrv) was measured by means of a flow probe on the pulmonary artery in pigs (closed chest) and analysed beat-to-beat. Calibration was done by means of the Fick's method for oxygen. Qrv decreased during insufflation and recovered during spontaneous expiration via an overshoot to an end expiratory plateau. This plateau was constant when the expiratory pause was lengthened and appeared to be dependent on the level of positive end expiratory pressure (PEEP). The end expiratory flow was considered to be the baseline value for its changes during insuf[lation. The deficit of flow during insufflation was much larger than the expiratory overshoot. Thus, mean cardiac output was lower than end expiratory flow. Decrease of mean cardiac output by PEEP or hypovolemia increased the insufflatory deficit but did not increase the compensatory overshoot. We concluded that -an estimation of cardiac output during an end expiratory plateau phase overestimates mean cardiac output, which was confirmed with the thermodilution technique, -the overestimation is higher at lower values of mean cardiac output, -the end expiratory flow depends on PEEP and mean cardiac output on PEEP and superimposed insufflation. In an attempt to obtain quantitative blood flow informations from Doppler signals recorded in the ascending aorta, the mean Doppler shift frequency must be measured and the extraneous signal from vessel walls represents noise that needs to be recognized and eliminated. If multiple frequencies are present, especially if the signal/ noise ratio is low, zero crossing detectors introduce significant errors (LUNDT 1975) and spectral analysis is necessary to determine the amplitude of each frequency component and to calculate the average frequencies. Several approaches to this problem have been employed, utilizing parallel narrow band filters, or digital fast Fourier transform. We used a laboratory-built Walsh transform analog system to calculate spectral components. The spectrum is displayed in the usual "sonagram" format with frequency as ordinate, the time axis as abscissa, and spectral power as trace brightness. The directional ultrasonic Doppler is also available for digital data acquisition.

The accuracy, reproductibility and ease of use of a pulsed Doppler machine (Alvar 4mHz) associated to the directional spectral analysis system has been investigated in normal subjects and in critically ill patients in artificially or spontaneous ventilation. An ultrasonic probe (diameter lOmm) was positioned in the suprasternal notch, The angle between the ultrasonic beam and the axis of the ascending aorta blood flow is assumed to vary from zero to 3 deg. the angle error is then less than 15 %. From first results, it appears that : 1) spectral analysis provides a valuable velocity signal even though the signal noise ratio is too low for zero crossing detection. 2) individual stroke volumes values calculated from spectral analysis are well correlated with those simultaneously measured by the thermodilution technique. We have studied hemodynamic and renal effects of in creasing dosage of DP in 7 patients with peritonitisand clinical findings of septic shock.

Studies were hegun when each patient had reached asteedy hemodynamic state with diuresis greater than. 7 ml/min while treated with DP.Once the first determinations were recorded~the dose of DP wes increased by 5-~g/Kg/min. and after 45 min the determinations wererepeated.Glomerular filtration rate and renal plasmatic flow were measured by Inulin and paraminohypurate cle~ rances respectively (Cin,Cpah).

We have observed augmentation of stoke index (p<.01), and Blood pressure (p<.01),without significant increase in Heart Rate,total systemic resistances,pulmonary pr~ ssure and pulmonary wedge pressure.Renal response showed augmentation of diuresis (p<.01),Cin (p<.05) and sodium fractional excretion (p=0.2) without significant changes in either Cpah or filtration fraction.

We concluded that incrementing dosage in DP in sep tic shock patients may he useful even with hyh dosageof DP.Increased natriuresis wes not due to changes inplasmatic renal flow. f0r Anaesthesiologie der Universit~t M0nchen +. The assumed prerequisite for blood substitutes with a high solubility for oxygen is that an arterial oxygen tension {PaO 2) of at least 300 mmHg and therefore a high inspiratory oxygen fraction (FiO 2) is necessary for the adequate oxygen supply to the organism. This would seriously limit the application of such substances in clinical practice. In order to investigate the validity of this hypothesis, we measured myocardial oxygen tension (PmO~) using a multiwire surface electrode and myocardial microflow (mMF) by means of hydrogen clearance following hemodilution (HD) with Fluosol DA 20 % (F DA) at FiO 2 of 1.0, 0.5, and 0.3. Eight anesthetised (Fentanyl/Diazepam) and artificially ventilated dogs were thoracotomised and hemodiluted with Hydroxyethylstarch (HES) to hematocrit (hct) 14 % and then with F DA to hct 7 %. Fluorocrit (fct) was 10 %. Results: At FiO 2 1.0 (PaO 2 466 mmHg) , F DA led to a 70 % increase of mean PmO 2 as compared to HES-HD and mMF increased by 7 %. At FiO 2 0.5 (PaOp 212 mmHg) PmO o still exceeded pre-HD control valhes and mMF remained unchanged . At FiO 2 0.3 (PaOp 91 mmHg) there was no further change in mMF an~ PmO 2 de' creased below control values. In those dogs where hypoxic tissue areas were observed acute moycardial failure occured. Conclusion: under our conditions of dilution with Fluosol DA(hct 7,fct 10) myocardial oxygen supply was well maintained at FiO 2 0.5 provided that the coronary reserve was normal. At FiO 2 0.3, however, myocardial hypoxia was observed.

ABSENCE OF EFFECT OF CAPTOPRIL ON PULMONARY HEMODYN~MIC C. RICHARD, A. RI~ILHO, JL. RICOME, Ph. AUZEPY -H0pital de Bic~tre -94270 LE KREMLIN BICETRE (France). We studied effects of oral administrafion of CAPTOPRIL on pulmonary hemodynamic in two groups of six patients, one including subjects with chronic respiratory failure (CRF) and the other subjects with chronic heart failure (CHE) and high plasma aldosterone (PA) level and plasma renin activity (PRA). The reason for this choice of patients was to assess two potential mechanisms of CAPTOPRIL action : hypoxic vasoconstriction inhibition and inhibition of eventual role of angiotensin II on pulmonary circulation. Following Swan Ganz and arterial radial catheter insertion, mean pulmonary arterial pressure (pAP), mean pulmonary capillary wedge pressure (PCWP) and mean systemic arterial pressure (MAP) were measured. Cardiac output was determined in triplicate using thermodilution method. P_RA and PA were measured in the CRF group. All measurements were performed 0 and 60 minutes after oral administration of CAPTOPRIL 50 mg. Statistical analysis was performed within each group using a paired student t test. In the two groups of patients decrease in MAP and in systemic arterial resistance associated with cardiac index improvement was significant (p<O,Ol). In C~m group (PAO2=55 mmHg, PaC02=50 mmHg) there + was no change in PAP (42 i -2 to 41 -i 9 mmHg) and in 

TIC HORMONES DURING SHOCK. J.Lchmann, Th. Hossdoff, N.van Husen, U.Gerlach,Madizinische Uni-versit~tsklinik,M~nster,W-Germany In patients with shock of different etiology activity of alpha-amylase in serum is significantly increased. This increase of total amylase activity represents a relative enhancement of salivary isoamylase while the activity of the pancreatic isoamylase is decreased. Furthermore we measured in serum of 11 patients by radioimmuno assay human pancreatic poiypeptide(Hpp), insulin and C-peptide. ln follow'u p studies during shock the decrease of pancreatic amylase activity correlates with an increase of Hpp. Concentration of insulin and C,peptide are increased in the beginning of the shock later on followed by a decrease. Our results suggest that the exocrine function of the pancreas might be diminished during shock whereas Hpp seems to play a regulative role in the endocrine function of human pancreas.

MODIFICATION OF BLOOD VISCOSITY DURING EXPERIMENTAL HE-MORRHAGIC SHOCK. S.Faenza, S.Taddei, Istituto di Anestesiologia e Rianimazione, Universit~ di Bologna, Italia The purpose of this study was to analyze the changes in blood viscosity during hypovolaemic experimental shock in relationship to the simultaneous variations o2 pulmonary vascular resistance (PVR) and systemic vascular resistance (SVR). Te~ piglets were submitted to hypovolaemic shock for 4 hours, and at the end o2 that period the blood removed was transfused. Arterial and venous blood samples were taken tmder basal conditions (To) , after 4 hours o2 hypovolaemic shock (T~) and 30' a2ter full transfusion (T~). At the same %ime hemodynamic parameters (cardiac output, systemic and pulmonary pressure) were determined using a Swan-Ganz catheter. Viscosity was measuredp on citrated samples~ by means os a cone-plane vlscosimeter under thermostatic_ condition (36.5~ ranging from 646 to 8075 sec . Results showed a decrease in arterial blood viscosity (ABV) from 3.41 • 0.1 centipoise (cp) at T^ to 2,57 • 0.2 cp at T~ followed by an increase to 4.09 ~UO.4 cp at T2; venous ~lood viscosity (VBV) was 3.64 • 0.3 ep at TO, 2,99 ~ 0.1 at T. and 4.91 ~ 0.4 cp at T 2. An increase in ~VR os 27% was seen at T I and a reduction os 2% at T 2 in respect to the basal values. PVR increased o2 67% from T O to Tp. In conclusion the-resul~s showed: ABV was always smaller than VBV (the principal causes seem to be pH, RBC volume,O 2 and CO contents) during hemorrhagic shock we observe~ a reduction in blood viscosity that partially made up for vascular resistance increase (dependln~ from hemodilution mechanism?) at the end of tras2usion (T2) PVR and SVR showed a dlfferent behaviour in relationship to the viscosity modifications: the increase os VBV made worse pulmonary cir-culation while the rise in ABV didn't modify systemic one.

David Belvedere, Dan Thompson, Robert Lumlsh, and John O'Donnell; Merey Hospital, Pittsburgh~ PA, USA A prospective study of all femoral artery catheters inserted in an intensive care unit was undertaken to study the incidence of complications.

A total of 138 catheters were placed and remained in place until they were no longer required, malfunctioned, or were suspeeted as a possible source of infection. Duration ranged from Ii hours to 34 days. There were no instances of mechanical complications.

Of these catheters 105 were available for microbiologic study; 51% showed colonization at the skin site, 20% colonization of the catheter, and 5% resulted in baeteremia. Analysis of risk of baeteremia by days indwelling revealed: 0% at 4'days, 2% at 5-6 days, 5% at 7 days, and 21% at 14 days. This data compares favorably with published data for radial artery catheters, and supports the conclusion that catheters placed by the femoral route are as safe as those placed by the radial approach. Liver perfusi0n scintigraphy in 40 patients with septic shock following neurosurgical operations showed significant diminuition of portal blood flow. Portal blood flow contribution dropped to about 34,5 % ( normal range 50 -75 % ). Following the administration of dopamine ( 3,5 -5,bAng/ks body weight ) a significant increase of portal blood flow (A 39 % ) is observable in all cases except one. Our results indicate that the decrease in portal perfusion is caused mainly by the hepatosplanchnic pooling. Dopamine when administered in doses Which do not affect mean blood pressure elicits selective vasodilation in the renal, superior mesenteric and celiac vascular beds suggesting that this is the underlying mechanism which causes the increase of portal perfusion after dopamine infusion. The method of sequential liver perfusion soin-tigraphy~ the criteria of septic shock and the basic pathophysiological mechanisms are discussed in detail. or the in. or external jugular vein is introduced, and its function is based on a modified Soldinger technique, lfs operational principle is that a normal-lumen catheter is inserted through a small puncture needle and therefore a thin capillary tube. This is achieved by puncturing the pationtTs vessel with a small-lumen needle of only half the diameter of catheter sets used up to now together with a capillary tubs pushed over the needle. After the needle (Braunble syringe system) is removed the normallumen catheter is pushed through the thin, tapered Braun~le syringe. The catheter dilates the capillary tube at pre-formed spots and can thus be pushed further and placed. The capillary tube is then pushed back or is pulled of by splitting it to. tally. The catheter is made of polyurethans. By means of a newly devslopped guide which can be pro.formed it is possible to master even vascular anomalies or severe vessel kinks. This new central ~in catheter has been applied about ?DO timesso far. Vein catheters were inserted in the subclavian vein, the in-or external jugular vein as well as peripherally. Complications occurring during catheterisations are reported, but they are by far less critical than with traditional systems. Advantages of thenew system are as follows:

I Evidently less extensive puncture traumas and thus a deminuation of the dangers of a false puncture (arterioal puncture). 2 A total sealing of the catheter within the vessel: Due to the small puncture opening and the widening of the puncture site by pushing forwand the normal-lumen catheter a firm, tight sealing is created at the vessel/catheter interface which is vital expeoially for anticoagulated patients or for patients suffering from thrombopathy. 3 The catheter is suitable also for juveniles and children due to its small puncture trauma. 4 The new steel guide allows pro-forming in the case of vascular anomalies and shows up much better on opaque material than traditional guides. 5 Guiding the catheter is stabilized substantially by the new guide. 6 Due to the sealing erect at the vessel this system is quite suitable for arterial puncturing (arterial pressure monitoring), too.

INFLUENCE OF EPIDURAI BLOCK. T. Fiore, A. Brienza, A.M. Leone, R. Giuliani, Institute of Anesthesiology, University of Bari, Italy. Spinal block has been proposed for the control of adrene~ gic hyperactivity in tetanus (I) 9 This paper is concerned with haemodynamic changes induced by extradural block. Material and method. Through a spinal catheter introduced into the extradnral space at L3-L4, 6 ml of 0.25% bupivacaine were injected twice with a 6 hrs' interval to 3 patients with severe tetanus, automatically ventilated under neuromuscular block with heavy sedation,in whom signs of adrenergie hyperactivity were evidencedr Values of ha~ modynamic parameters were obtained with a cardiac output thermodilution apparatus using Swan-Ganz catheters;values in the . Only two cases (one with anorexia nervosa and other one with prolonged starvation) showed hypoglycemia and L.A.

2) No statistical correlations were found between blood lactate and alcoholemia in 57 A.A.I. (r = O,O1). 10 cases had lactatemia higher than 3 mmol/l, 4 blood alcohol of which were higher than 40 mmol/l.

3) For the great majority of subjects, glyeemia was normal.

CONCLUSION : Usually severe A.A.I. in adults induces respiratory acidosis but no hypoglycemia. L.A. and hypoglycemia are only reported in cases with starvation or nndernutrition. Care Medecine have reached a high level.However, less atention has been paid to the older patients.One of therea sons for this has been the negative relationship costs/benefits mentioned.We think that the problem must be dealt with. The retrospective analysis of patients admitted in our ICU in the last 22 years has shown that 440 were over 64 years, 273 of them men. The varied pathology that determined admissions is referred to.This is no intention of comparing results as it is difficult to establish groups corresponding either to different previous heath status or to the severity of acute illness on the day of admission. It was found that mortality among the patients over 64 years was 43,4%,being comparable in both sexes.Among the remaining 3.403 patients it was 28,9%,Dividing our sample of aged patients in groups taking into account the mortality we can see an increasing mortality:39,7% in the group 65 to 75 years (n=312),50% in the group 75 to 80 years (n=82) and 56,5% in the group >80 years (n=46).There is a significant difference in these mortalities (xZ=6,328 § p<O,OOl).A more detailed analysis of the data reveals that the statistical significance is mainly due to the age group over 75 years.On the other hand, the longer the stay in the ICU the smaller the rate of mortality (p<O,OOl),as more than half of the patients died during the first 48 hours. The average time of hospitalisation in ICU was 18,1 days against iO,I days among all the other patients. This study does not cover the period after the discharge,because we are just questioning how acceptable and fair intensive care medecine can be in the group of ages over 64 years.~ think that the age, per se, must not be a criterium of exclusion even taking into consideration all sorts of costs. In our opinion the general criteria of investigation,teaching and admission must prevail in these patients. The support of a geriatric ward is assumed. In combination with 52 self designed items we evaluated the answer of 5o patients transferred recently from ICU to regular ward to the question: Are environmental situations as emerging under ICU-conditions comparable in terms of their stress producing power? The patients had to bring the various stress items in a sequence of increasing annoyance just one day after dismission from the intensive care unit. The most annoying factor for all patients, almost sex and diagn3sis independent, was the item lack of information and communication. The item noise and unrest caused by the daily routine and the nursing staff on the intensive care unit was quoted o~ the first place by patients more than 50 y~ars. Helplessness and physical impairment are the most annoying factors by the ventilated patients. The item dependence on technical devices (e.g. ventilators, dialyzer etc.) was surprisingly quoted low by our patients. In conclusion to those facts a human orientated intensive care medicine should be claimed. Further investigations concerning the psychical care and help should be done.

BIOAVAILABILITY OF METRONIDAZOLE ADMINISTERED THROUGH RECTAL AND INTRAVENOUS ROUTE. Fichelle A.~ Farinotti R., Dauphin A., Mourlon M., Desmonts 3.M.

H6pital Bichat -

France. Metronidazole (M) is widely used as prophylactic and curs rive treatment of anaerobic infections in abdominal surge ry. The cost of injectable M led us to use M by rectal route as proposed by Ioannides and el*. This study was designed to compare the bioavailsbility of intravenous (IV) or rectal (R) metronidszole. Patients and methods. The inclusion criteria in the study were the following : no alteration in blood volume, no abnormalities in rectum and in sphincter function, no diarrhea. Eight patients were given successively a single dose of 0.5g via IV and R route. The interval between the successive administrations was 48 hours. The IV M was Flsgyl* (Specie) and the suppositories were made of MO.5g and aerssil 200 O.Olg and suppocire AS2 qsp 2g. The study started by random administration of M either IV or R according to the admission. Plasma levels of M were measured before and 30, 60, 90, 120, 180 min and 4, 5, 6, 8, 12 and 24 hours after administration.

Results. Peak plasma concentration was observed at 3hours after the beginning of administration. The mean value + SD Was 8.2 • 1.7 ug/ml (IV route) and 6.8 + 1.5 ug/ml ~R route). The elimination half life was 6.26~+ 3.04 hours (1V rout~ and 6.29 ~ 2.23 hours (R route) re~pectively.Bio availability expressed as extrapolated area under the cur ve was 102.4 + 41.8 pg/ml/h (IV route)and 91.1 + 40.8~g/ ml/h (R route~in 7 patients having normal hepatTc function ; in one patient with cholestasis bioavailability was increased, 266pg/ml/h (IV route) and 241pg/ml/h (R).

Conclusions. This study demonstrated that M plasma concentrations reached same levels after R administration as after IV route. When dosage and intervals of administration of M is adequately determined, Rroute is s good alternative compared to IV route. The main advantage is the drastic reduction in cost of the treatment : the R form is 35 times cheaper than the IV form. In the present investigation, the responsiveness to mitogens of lymphoeytes prepared from patients after a single injection of qhiopental has been studied. In 8 patients undergoing coronary surgery, anesthesia was induced by • in3ection of T. 5 mg/kg bw. Blood samples were collected before (K) as well as 3 min (3') and 40 min [40') after injection. Lymphoeytes prepared from these samples were kept in culture and stimulated with Phytohemagglutinin (Pli%), Pokeweed Mitogen (PWM) and Concanavalin A ~o~A).

~e stimulation-rates were assayed by 3H-Uhymidine-incorporation.

T-levels were measured by gas-chromatography.

~he results are shown in the cable : Various degrees of impairment of neurohumoral homeostasis according to sympathico-adrenal, kallikrein-kinine, serotonine, histaminreactive system, functional erythrocyte condition data were determined in 310 patients with severe cerebral trauma in an acute period with prolonged loos of consciousness. The following mechanisms of neurotumoral impairment were observed : arterial hypoxemia, respiratory alkalosis, hypercatabolism resulted in activation of the system studied. An impairment of mechanisms of compensation hypoxia was noted : decrease of 2,3 DFG level in erythrocytes, progressive anemia while blood loos was absent. The use of catecholamine and glucocorticoid precursors did nor improve the treatment results. Methods of pharmacological protection of brain from hypoxia were used in a second group of patients with glucocorticoids and an early intensive therapy. The best resluts were obtained with combined administration of barbiturate and dosium hydroxybutyrate. High stability and adaptation of neurohumoral systems to stress and preservation of their reserves, were due to a decreased hypoxia (pa02 + 4,1 kPa), metabolism restoration in erythrocytes, mange of their composition and correlation of hemoglobin fractions. The recovery from coma was shortened by an average of 3 days, fatal outcomes decreased to 15,2 %. In a prospective study of shocked patients we have assessed the prognostic significance of single and serial estimations of whole blood lactate (WBL) and compared their usefulness with simple hae~aodyrmmic measurements.

Thirty patients suffering from shock of varyin~ aetiology were studied. Arterial WBL, mean arterial pressure (MTP), urine voltmm (UV) and core-peripheral temperature gradient (CPTG) were measured after 3 hours and 24 hours resuscitation. There were 15 survivors.

There was no significant difference between survivors and non survivors in either lactate or haemodynamic values at the time of adnission.

In survivors, WBL improved after resuscitation (see table) . However, improvement in haemodynamic values was also significant in survivors.

Using logistic regression analysis, outcc~e in a shock episode can be predicted after three hours with 91.5% accuracy if UV, MAP and CPTG are known.

Adding WBL only increases the accuracy of prediction to 93%.

l~asurements of lactate production in some of the non survivors suggest that in shock, lactate production by skeletal muscle significantly decreases, explaining the relative mreliability of lactate as a predictor of outcome. Increases in intracranial pressure (ICP) occur with the most commonly used antihypertensive drugs. In patients with space-occupying lesions such incrases in ICP may lead to cerebral ischemia from decreased perfusion pressure or even to brain herniation. Ketanserine is a new quinazoline derivate. It is a potent and selective 5 HT-serotonine-blocker. At high concentrations it has alpha I -adrenergic, antihistaminic, antidopaminergic and antiangiotensin properties. This study examines the effects of high dose Ketanserine up to 14 mg/kg on haemodynamic parameters and on ICP in dogs without (Group I) and with (Group II) intracranial hypertension. Haemodynm, ic measurements including mean arterial pressure (b~LP), mean pulmonary arterial pressure (PAP), mean right atrial pressure (RAP), pulmonary wedge pressure (PCWP), heart rate (~), cardiac output (CO), systemic vascular resitance (SVR), and stroke volume (SV) were made. ICP was measured via an intraventricular catheter. Ketanserine was given as a bolus of I mg/kg followed by an infusion up to a total dose of 14 mg/kg. All data were analyzed by student's t-test p~0,05 was considered significant. Uniform blood pressure reductions occured in each group. The maximumdecrease (35 % • 18 in group I, 35 % • 19 in group II) occured at the end of the Ketanserine infusion. An increase in ~ and decrase in SVR could also be observed~ICP did not change. CO, SV, RAP, PCWP remained unchanged. Ketanserine is a safe antihypertensive agent, even in patients with poor intracranial 6ompliance.

Infirmary, Glasgow,Scotland.

The movement of critically ill patients has been described by several authors (Ledingham, 1980; Aitkenhead, 1980) . Special trolleys have been designed or existing dnes adapted for local requirements, depending, for example, on whether or not the trolley has to fit into an ambulance.

The new Mobile Intensive Care Unit (~I.I.C.U.) in Glasgow Royal Infirmary has been designed to use as many standard and familiar components as possible.

It is basically a conversion of a standard hospital trolley and is intended for use only within the hospital complex.

It provides facilities forcardiovascular ~onitoring, continuous infusion of drugs by infusion pumps, I.P.P.V. with P.E.E.P. and I.~1.V. With any nonhypoxic mixture of Nitrous Oxide and oxygen, and spontaneous respiration with any mixture of air/oxygen and nitrous oxide/oxygen.

It is equipped with a defibrillator and suction apparatus and has gas supplies for one hour at normal flow rates.

There is an oxygen failure alarm and also a ventilator failure alarm, and quick change-over from one oxygen cylinder to the other.

A detailed description and photographs in colour will be provided in a poster display.

LEDINGHA~ [, I.McA. and BANKS, J.G. (1980) The prediction differed RSI from the 2 scores 126 times RSI was accurate in 01, AP5 in ~5 cases (X 2 = i0.3, p< O.OO1). Mortality according to RSI is noted in the figure.

We conclude that the very easy to calculate RSI gives an equivalent prediction than APS and could be systematically collected in ICUs. Critically ill surgical patients have a reduced response to recall skin antigens.

The mechanisms underlying this is ~clear but may include altered lymphocyte responsiveness or decreased inflarmnatory cell mobilisation. Peripheral blood lymphocytes from surgical patients in an ITU were tested for their response to the mitogens Concanavallin A (Con A), Phytohaemagglutinin (PHA) and Pokeweed Mitogen (PWM) and peripheral blood mononuclear cells and granulocytes were tested for their chemotaetic response towards casein.

The results (median and range) are shown below:

Control (n=14) Patients (n=17) p Con A (c.p.m.

x 10 ) 111 ( Patients' ly~hocytes showed a reduced response to Con A and phagocyte chemotaxis was increased, This supports the hypothesis that critically ill patients have a defect in T cell function. Many studies have pointed out the interest of GC in trach6 obronchial secretions and of DBA about patients under assis ted ventilation by endotracheal tube. The purpose of this study is to compare the data of GC performed, or estimated on proximal and distal secretions of such patients. The PBA are realised by sterile probe.After dilution the secretions are submitted for culture. Each type of colony is numera~d and identified after 48h of incubation.Germs are consider6d as pathogen if the number of colonies is equal or higher than 106/ml. The DBA are realised on the same day, according to Matthew's technic. The number Of colonies, iA the distal sample is estimated after incubation. According to the data of a previous study of in vitro simulation, we have considered that the DBA were contaminated by the proximal bronchial flora, when the number of colonies/ml is equal or lower than 102, and uncertain between 102 and 104. 83 DBA and PBA have been performed on 48 patients with pneumopathy. 38 PBA (46%) did not have any pathogen germs ("sterile"), 24 (29%) had one~ 21 (25%) more than one. 50 DBA (60%) were "sterile", 12 (14%) had one pathogen germ, 9 (11%) more than one. 12 DBA were uncertain. For 42 (51%) simultaneous PBA and DBA, the bacteriological information was the same. When the results were different, the DBA was able to isolate one germ 8 times, whereas the PBA~terile. On the opposite, on 14 times, the DBA was sterile or containingonly one germ whereas several germs were isolated in the corresponding PBA. The GC in PBA and DBA makes the diagnosis easier in pneumopathies in intensive care, owing to the smaller number of identified microbial species.The GC in DBA may give greater information for the diagnosis when the PBA is out.

ELECTROENCEPHALOGRAPHIC PATTERNS IN ACUTE POST-TRAUMATIC COMA. L.F. Comelli, M.G. Pignatelli, G.F. Testa, M. Avalle, Ospedale San Giovanni, UniversitA di To~ino, Torino, Italia. The aim of this study was to assess the diagno= stic and prognostic value of the electroenceph! logram ~EEG) in acute post-traumatic coma. Five hundred patients were selected in whom the EEG tracings were recorded upon admission and then repeated over successive periods. Five EEG patterns Were identified: borderline (41 cases) diphasic with sleep-like spindles (161 patients) diphasic (139), monophasic (109), and silent (50). Mortality was maximum (100%) in the flat EEG (brain death), very high (86%) in the mono= phasic pattern (deep coma with decerebration), decreased in both the diphasic (60%) and border line (51%) types, and minimal (13%) in the sleep-like pattern. Therefore, the finding os cortical electrical activity resembling that of physiological sleep offered the highest per= centage of remission. On the other hand, the mo nophasic EEG was very unfavourable. Patients with borderline tracings h~d a high mortality rate in spite os the relatively good EEG orga= nization. The present results tend to confirm the utility of EEG assessment in cases os trau matic coma. Moreover, this procedure may be us~ fully employed in evaluating the course os pa: tients with head injury and monitoring the es cacy os the treatment protocol. The study has been carried out to compare blood and cerebro-spinal fluid (C.S.F.) creatine kinase (CK) values, to correlate brain type CK (CK BB) activity with braindy~ function and to assess the prognostic value of CK assay.

-53 severe head injured patients (mean age = 24 + 14 years) have been included in the study. 34 % suffered ~m cranial trauma only and 66 % were multiDle injured pa ~ tients. Blood and ventricular C.S.F. samples have been collected at 9 + 5 hours after the trauma. Total CK activity has been measured spectrophotometrically at 37oC. T~ isoenzymes have been separated electrophoretically and quantified by densitometry. Brain dysfunction has been graded according to the Glasgow coma scale (G.C.S.) and to the LiEge coma scale (L.C.S.) which uses elements of the G.C.S. and the study of 5 brain stem reflexes. The ~atient's outcome after six months has been assesed according to the Glasgow outcome scale.

-There is a good correlation between total CK and CK BB activities in C.S.F. only (r = 0.992). There is no rela-tionshi~ between CK BB activities in the serum and in the C.S.F. (r = 0.139)~ There is an inverse relationshi~ between C.S.F. CK activity and brain dysfunction. The Correlation between these two narameters is better with the LCS (r = 0.8) than with the GCS (r = 0.635). There is a significant difference in C.S.F. CK BB mean values between patient's groups classified according to a bad and a good outcome (F = 5.85 n 0.006), -The CSF CK BB activity is more reliable than serum activity. C.S.F.CK BB activity can be correlated with brain dysfunction. The L.C.S. allows a better assessment of brain dysfunction. High C.S.F. CK BB activity is significantly associated with a bad outcome.

Central somatosensory conduction time in posttraumatic comatose states.

Rumpl,E.,Prugger,M.,Gerstenbrand,F.,Hackl,J.M. and Pallua,A.

Short latency evoked potentials (SEP) were elicted by stimulation of the median nerve at the wrist and recorded simultaneously from the neck and the contralateral scalp in 44 patients with signs of brain stem impairment due to head injury. Brain stem involvement was divided clinically and by CT scan into secondary lesions due to supratentorial mass displacement and primary lesions due to direct violence to the brain stem. The central conduction time (latency between N14 and N20 -CCT) and the amplitude ratio (N20/NI4-AR) were close to normals in patients with supratentorial lesions and good outcome. CCT increased and AR decreased with the worsening of outcome. Asymmetries in SEP indicated moderate or severe disability outcome. Patients with absent SEP over one or both hemispheres died. In patients suffering from primary brain stem dysfunction, confirmed by a normal or slightly abnormal CT scan, prolonged CCT, asymmetric but also absent SEP were also found in patients with good outcome. AR was generally low in these cases. Early appearance of SEP or early recovery of initially distorted SEP and decrease of CCT in control examinations was a favourable prognostic sign. Therefore even absent or severely distorted SEP should be interpreted cautiously, if patients may suffer from primary brain stem involvement. The purpose of the study is to check the interferences of aminophylline with intracranial pressure (ICP).

We have monitored, by intraventricular catheter, transducer, amplifier and recording system, 12 patients of either sex, with mean age of 43,5 years and with hypertensive liquoral pathology. We have monitored also mean arterial pressure (MAP) and PCO . After recording the basic values we have adminzstered aminophylline 2 mg.kg -I in a single bolus intravenously.

In every cases we have achieved rapid decrease whether in ICP or in cerebral pulse. No significant changes in PCO . The values of MAP 2 were steady and consequently we have achieved an improvement of the cerebral perfusion pressure (CPP). BRAIN INJURED PATIENTS -THE OUTCOME ONE WEEK AFTER ACCIDENT. D. Hausmann, K. Rommelsheim, K.U. Josten, Institut fur An~sthesiologie, University of Bonn, K.O. Mosebach, R. Caspari, R. Lippoldt, Physiologisch-Chemisches Institut, University of Bonn, Germany. In order to find out guiding parameters for correction of care, the average levels of about 60 substances in plasma and urine of 18 survivors (S), 13 traumatized (T) and 5 burned (B), as well as 16 nonsurvivors (N), 11 T and 5 B, were compared from day I until about 10 days after accident. Only patients were elected who survived the acute phase (3 days). The combined enteral and parenteral nutrition included 430 g mixed carbohydrates, 70 g fat, and 150 g amino acids per day. The nutritional assessment comprised the aminograms of serum and urine (24hour excretions), the serum concentrations of 5 proteins and 6 hormones as well as the nitrogen balance taking in account the reflux rates following tube feeding. The mortality passed through a maximum between day 6 and day 7 after accident. During this period about 85 % of all N died and the reflux rates reached the highest values (> 80 %!). In the case of N, many parameters showed a corresponding maximum, but I or 2 days earlier, for example nitrogen balance, serum urea, total concentration Of serum amino acids, GOT, GPT, and prolactin. Due to the high reflux rates we propose a TPN at least until 10 days after accident. We have no explanation for the mentioned maxima. 239 387 EXPERIMENTAL ACUTE HYPOXIA. EFFECTS OF PHENOBARBITONE PRE-TREATMENT (HYSTOLOGICAL STUDY) G.di Trapani, S.I. Magalini, A.L. Abbamondi,M.Lazari, A.F.Sabato. Ist. Anestesiologia e Rianimazione,Universita Cattolica S. Cuore, Roma,ltalia. A histological study is carried out on five groups of 50 quinea pigs each (control-after 20min.of hypoxia-2Omin. of hypoxia+2Omin, of oxygen therapy-pretreatment with phenobartitone(lO0 mg kg-l)and 20min. of oxygen therapy). The light microscopy results can be summarized as follows:

The hystological study did not show significant differences between barbiturate-treated and untreated hypoxic brains, in fact, the severity of ischaemic damage as well as its distribution were similar in all the experimental groups. The lesion mainly affects the areas which are known to be most sensitive to hypoxia. It is suggested that in the experimental conditions barbiturates did not act as a protective agents at least as assessedmorphologically.

WATER AND SODIUM BALANCE AND RENAL FUNCTION OVER THREE WEEKS AFTER SEVERE.ACCIDENTAL TRAUMA U. Finsterer, U. Jensen, A. Beyer, U. Schied, W. Kellermann, K. Unertl, K. Peter, Klinikum GroBhadern, University of Munich, Munich, FRG Careful monitoring of water-electrolyte balance and renal function is not yet common practice in intensive care units and might probably help to reduce acute renal failure (ARF). From March 1980 to May 1982 we studied 32 patients after severe trauma :(mainly brain trauma, who could be observed for at least 21 days and did not develop ARF, which had to be treated by dialysis. The most prominent features of this study were: a) positive water balance throughout inspite of a significant weight loss (which points to high fluid losses with perspiration), b) high osmolar excretion, due to high urea excretion with posttraumatic protein catabolism, c) antidiuresis throughout with urine osmolality around 600 mosm/l and high negative free-water-clearance {due to high osmolar clearance), d) cumulated sodium balance was negative and averaged -440 mmol/21 days, e) normal creatinine clearance (occasionally unusual high values), f) normal creatinine excretion. We conclude that patients after severe trauma generally have good renal function and tend to antidiuresis, high osmelar excretion and negative sodium balance.

CORRECTION OF NEUROTRANSMITTER FAILING AS PART OF BRAIN RESUSCITATION. H.Schoeppner, L.Rolf, S.Wagner, J.Zander. University hospita 1 of anesthesiology and crticsl care medicine, Muenster Federals ~epublic of Germany.

Breakdown of neurotransmitter balance represants an essential component o f posttraumatic,ischsmic encephalopsthy. The consequences are seizures by overwhelming of cholinsrgic transmission and cerebral coma by missing of cortical activa tion. 20 patients, suffering from brain lesion ( soma scale 5 -5 7 have been treated with drugs of 6ABAergic and c holinolytic potency ( 5 oases each by infusion of barbiturate, etomidate, gamma -OH -butyrate and procaine.) The effecti~ty control h as been carried out by I. Determination of the cellular ( platelet ) content of serotonin and dopamin by means of spectralfluorometric method. 2. Comparison of the lactate ladles of arterial blood and GSF. 3. Recording of the EEG ( 8 channels ) 4. Determination of cateaholamine and cortisol releasing prae and post the infusion time of 8 hours. The lasting effect, concerning the augmentation of cellularcontent of Serotonin and dopamin has bsenreached ,employing butyrste and procaine. In last case the EEG course changed from generalized subdelta activity %o finally reinte grstion of alpha activity and patients in consciousness.The approach of the lactate levels of arterially blood and CSF showed to be optamally in case of procaine. The depletion of sstecholsmin and sortisol secretion developed step like from barbiturate and atomidate to butyrots and procaine. We report one new case of acute water intoxication (AWl) induced by oxytocin (OXY) infusion during therapeutic abor tion , in a 30 year old primigravida at 20 weeks'gestation After intraamniotic injection ~f Pro~taglandins ~ this patient received 210 units (u) of OXY in 2,5 liters (I) of a 5 % Dextrose in water solution, auring a period of 15 hours (h) 30 minutes (mn)(infusion rate (IR) of about 226 mU/mn), and twice i00 mg IM of Pethidin. Whereas she had vomitings and headache, a grand mal seizure then a deep coma occured. Serum sodium level was 112 millimoles/l and serum osmelarity 259 milliOsmoles/l. With interruption of oxytocin and sodium administation, a marked diuresis occured, with positive free water clearance ; electrolytes nor malized within 48 hours and consciousness cleared within 4 days. She was discharged completely recovered. About this new case and 41 cases previously reported in litterature, circumstances of occurence of this uncommon entirely iatrogenic disease were treatment of incomplete abortion (13 cases), post partum hemorrhage (5), induction of labour (6) or abortion (18). Isolated or associated symptoms were neurologic. In 11 cases, alarm but disregarded signs were headache, nausea, vomiting, or mental confu sion. In 4 cases, no symptom was observed except oliguria which is a constant finding. No correlation was found between severity of clinical signs sodium serum level, total volume perfused, total dose of oxy and duration of treatment (p>0,05). OXY antidiuretic effect depends on its IR. Then concomitant fluid administration may induce an AWl, facilitated by physiologic water and salt retention in pregnancy, individual sensitivity, and morphinics. In 39 patients recovery was complete. In 3 patients, serious neu rologic sequels were reported. 2 maternal deaths occured . 2 newborns presented severe signs of AWl but rapidly recovered. So, such accidents must be prevented by clinical survey, survey of OXY IR, minim fluid and proportional salt intakes. For a sensitive and specific detection of blood we tested the chromium-51-activity in the gastric aspirate after having labelled the patients' erythrocytes with chromium-51. The dosis of the medications were adjusted to a gastric juice pH above 3. From the results with 86 patients we draw the following preliminary conclusions: I. With a sensitive and specific method for detection of blood in gastric juice temporary blood traces could be detected in 30-40 % of patients of group A and B; the detected amounts of blood showed no clinical relevance but were far above the physiological blood loss of the gastrointestinal tract. 2. Visible blood in gastric juice was discovered in 3 patients in group A and in 2 patients in group B so far, an event expected in no treatment patients in 15-50 %. 3. The blood loss in the stomach depends on pH; above a pH of 3 blood could hardly been found, below a pH of 3 the blood losses were much higher and were negatively related to the pH. 4. A guajac paper test is not an appropriate method to detect occult blood in gastric juice contaminated with magnesium-aluminium-hydroxide. Liver is responsible for biosynthesis and degradation of sialoglycoprotein such as~] acid-glycoprotein (o<] AGP) and ~ ] anuitrypsin (~ 1 AT). The plasma of normal subjects contains negligible amounts of asialoglycoprotein because of the ability of the normal liver to clear them almost instantaneously from the circulation. In hepatic diseases, the presence of serum desialylated material has already been evidenced, but in this study, the hepatic insufficiency was often associated with other organ failure (J. Marshall et al, J. Lab. Clin. Med., ]978, 92, 30-37). The actual cause of this desialylation is still not well known : isolated hepatocellular damage (HCD) or multi vi~ ceral failure.

To answer this question we have studied the degree of desialylation ofo< I AGP and o<] AT in serum of 26 patients ; all had multivisceral failure, associated (n = 9) or not (n = 17) with HCD evidenced by usual liver tests. The desialylation of~ ] AGP and ~] AT was determined By compa~ ring radial i~munodiffusion and electroimmunodiffusion ~r each glycoprotein. Results were given in percentage (%) asialoglycoprotein. The presence of a desialylation>5 % was considered as a desialylated serum.

Percentage of desialylated serums Without HCD =< ] AGP% ~I AT% ~j AGP~AT% n = 17 ]7.6 0 17.6 With HCD n = 9 55.5 33.3 77.7 The resulta::were statistically different in the two groins (p<O.O5). There was no correlation between high concentra tion of~ I AGP (x=1,74g/l• and of~! AT (x=4.8Og/l• 1.78) and presence of desialylation. Thus desialylation is not concerned by anabolism ;the pre sence df desialylated~ i AGP ands] AT is in relation with hepato-cellular damage.

TPN WITH SOLUTIONS OF DIFFERENT AA COMPOSITION:EFFECTS ON CRITICALLY ILL PATIENTS NB.-F. Bobbio Pallavicini, V. Emmi, A. Brssehi, C. Reposi, G. Negri, F. Raimondi, M. Fischetti, P. Tosi, G. Iotti,A.Gasperi.Ospedale San Matteo, Pavia, Italia. Authors aim was to investigate how TPN with solutions of different AA composition effects on NB in critically ill patients with a medium-high proteie catabolism.Patients were on TPN for i0 days.The work focused on the identifi= cation of the AA solution,if one,whose composition eoult let in the NB reach positivity sooner,expeeially aceor= ding to significantly different BCAA contents.Three ty= pes of AA solutions(A-B-C)were used.They had the follo= wing contents,expressed as grams i n i000 ml of solution.

A B C The were provided with BCAA:

19.20 28.40 47.77 the equivalent daily AAA :

8.20 6.26 8.20 quantity total calorie 8(130% of their RME).Each of the 32 patients was suffe= ring from acute generalized peritonitis.They were set in one of 3 groups(A-B-C)at random.For survived patients,NB was computed by determinating the urine urea and its evo= lution was daily compared by evaluation of the confiden= tial limits of the means differences,the variances homo= geneity being previously evaluated by F test.No remarks= ble difference in reaching the NB positivy was demonstra= te even if A and C group reached it after 5 days and B group after more than i0 days.Consequently,there is no si gnificant statistical difference inter activity of AA so= lutions used.Altough further studies are necessary,the re suits obtained suggest that "balaneed"AA solutions impro= ve the N utilization in patients with proteic hypercatabo lism.More over BCAA enriching seems not to affect NB fa= vourably. *Preliminary dat a presented to the 4 th ESPEN Congress. The difference between the 2 groups was statistically significant (P<0.025). The analysis of the 2 equations evidenced that: I) NB is always better in M pts; 2) Ni is the major determinant of NB in both groups; 3) Ei effect on protein sparing and utilization is bigger in S pts.

The possible incidence of nutritional disturbances in the oceurence of post-operative infectious complications led us to study nitrogen (N) balance in forty adult patients undergoing surgical procedures with CPB. Patients and methods.

Three groups were definited: group I and group II, 13 patients class NYHA I-II and Ii patients class NYHA III-IV respectively underwent valve replacement 9

The third group, 16 patients, submitted coronary artery bypass (2 + 1 grafts) 9

The study was performed in a time interval starting 3 days before operation (D -3) and ending Ii days after (D + Ii). The scheme of caloric and N intake was similar in the 3 groups : 1950 KCal/day with 13 g N/day before operation and increased from 700 KCal/day with 5,5 g N/day to 2300 KCal/day with 14 g N/day over the postoperative period 9 Nitrogen loss was evaluated from the measurement of daily urinary N output.

The N balance was calculated from cumulative data of three days. Results and conclusion. Our study has shown the following points:

i) postoperative N losses were very high in our patients and were not related to type and severity of cardiac diseases; 2) positive balance could be obtained only 9 days after operation with minimum N intake of 14 g/day and in absence of infectious complications.

If these complications are related to decreased immunologic responses to infection observed in altered nutritional status, some part of prophylaxis could be afforded by a higher caloric and N intake than it is routinely prescribed after cardiac operation. The variations of the phosphate and calcium metabolism after total parathyroidectomy on healthy glands have never been reported. This occurs in patients suffering from carcinoma of cervical esophagus treated by total pharyngolaryngectomy. In two years, 17 patients, 14 men and three women (of mean age 53 years with normal thyroid and parathyroid functions checked) were operated. Serum and urinary calcium and phosphate level measurements have been done after operation. The hydroeleetrolytic adjustements and the parenteral feeding were carried on with glucose and amino-acid solutions avoiding phosphate and containing known and stable calcium rate. From the enterai feeding on, thyroxine, alumine-hydroxyde and eholecalciferol were added.

After total thyroparathyroidectomy, calcium and phosphate serum and phosphate urinary levels significantly decrease (respectively 10 • 9 mg/l p<0,01, 5,6 i 6,3 mg/l p< 0,05; -0,28 • 23 mg/mg urinary cmeatinine p<0,005).

Indicating the lack of parathormone, the urinary calcium excretion increases, depending on the filtrated load (calcium serum level) and as long as the phosphorus intakes remain low, a real low calcium serum level can be noted without clinical signs.

In spite of high frequency surgical complications (93 for 17 patients) the variations of phosphate and calcium metabolism are not affected by them. The work aims to report the results obtained in patients with crush syndrom (C.S.), treated with hyperbaric oxygen therapy (OHB) and peritoneal dialysis (PD) or extra-corporeal dialysis (ED). There were 9 patients aged 17 to 45 years with lesions of the limbs (complicated in 3 cases by gaseous gangrene) and acute renal failure (AR F) (7 oligoaniuresis, 2 aniuresis). 7 to 24 hours passed between the quake and their first aid and a period of 1 to 6 days during which they passed into our care. Besides the contribution of calories and hydroelectrolites, therapy was based on contemporary, daily dialysis (7 ED, 2 PD) ad OHB (2 ATA for 2 hours a day for the first 4 days -2 ATA for 1 hour a day for the following days). OHB was used for 7 to 70 days and when used for more than 23, gangrene was present. Diuresis started again on the 8 th to the 1 lth day with PD and the 6 th to the 15 th with ED 9 Final results: 6 patients cured, 1 amputated, 1 trasfered improved, 1 desceased due to hepatitis. We can therefore confirm the validity of the association of precocious ED with OHB.

BLOOD VOLUME CHANGES DURING SEPTIC SHOCK IN CHILDREN.

D. DEVICTOR*, J.M. ROCCHISANI**, g. HUAULT* * H6pital BICETRE -** H6pitsl COCHIN -PARIS -France Septic shock may induce hypovolemia by increasing fluid leakage from the capillary bed. The purpose of the present study was to test the effects of macromolecular infusions on the red cell mass (RCM), plasma volume (PV), and total blood volume (TBV~. We used Tc 99 labelled erythrocytes to measure RCM in 6 c611d~ren aged from 4 months to 7 years with infections purpura and shock. Haematocrit was measured simultaneously for calculation of PV and TBV For ethical reasona~ the study took place during therapy which included 18 to 72 hours infusions of whole blood and of plasma, albumin, gelatin and dextran. The rate of infusion averaged 180 ml/kg with large interindividusl variations depending on the patient's condition. Results were as follow :

age -watched controls patients RCM ml/kg 18.5 • 3 21 ~ 6 * PV ml/kg 36.6 • 6 45.3 • I0 * TBV ml/kg 57.1T 9 66.5 i 13 *

Since statistical significance was not demonstrated by t-test comparison with Lhe controls, these data indicate that TBV was normalized in the children studied. The fact fhaf the rapid fluid infusion did not increase blood volume beyond its normal limit, suggests that a large fraction of the macromolecular solutions infused disappeared from the vascular bed.

MANAGEMENT OF SEVERE HYPONATREMIA IN CHILDREN: UTILITY OF UREA. G. Siro Brigiani, T. Fiore, R. Giuliani,M. Mieelli, Institute of Anesthesiology, University of Bari, Italy. Urea has been proposed for continued treatment in cases of syndrome of inappropriate secretion of antidiuretic hormone in adults (I) In the present study, rapid correction of hyponatremia has been attempted in children.

Clinical material and method. Two pediatric cases (aged 90 and 8 years respectively) of head injury in deep coma are described, which neurologically exhibited a downfall course, accompanied by I) hyponatremia and hypoosmolality, 2) inappropriately high urinary osmolality and natriu resis, 3) no impairment of renal function,normal volemia, absence of oedema. Fluid restriction having been followed by no favourable result, the acute administration of urea (500 mg/Kg in 50 ml of water) through a naso-gastric tube was carried out twice with an interval of 24 hours.

Results. Values of plasma sodium and osmolality, together with natriuresis, gradually returned to normal within 16 hours after the second administration of urea(see Table) . Conclusions. Urea,which represents the main solute in the countercurrent mechanism, proved to be effective in the correction of decreased plasma sodium content and osmola lity with a sparing effect on sodium loss; a concurrent improvement of the neurological picture was observed. In 22 infants and neonates with respiratory insufficiency we studied total serum protein, colloid osmotic pressure and ventilation parameters.

Positive correlations between high fluid replacement, lowered colloid osmotic pressure and respiratory function problems could be established.

Colloid and crystalloid volume replacement therapy in pediatric intensive care patients with respiratory insufficiency should be controlled by colloid osmotic pressure measurement. Our results were presented and discussed with statistical significance set at p< 0.05 value.

A.Barois, F.Veber, B.Estournet Service de P4diatrie-R6animation Infantile Hapital R.Poincar6, 92 380 Garches -France. 11 children suffering from meningitis associated with a ventricular shunt.were admitted in the department. Mean age : 3 + 3,5 Y (range : 15 days to 12 years~, 8 boys and 3 girls. The ventricular shunt was put for hydrocephalus post neonatal meningitis 5 times, following intracranlal hemorrhage once, complicating a spina bifida 5 times and due to congenital malformations twice. Meningitis occured 26,8 + 40,8 months after the first shunt was put and 9,9 + 8,5 days after a new surgical intervention for obstructed shunt in 9 cases. The germs were meningococcus once, staphylococcus 7 times, streptococcus 4 times and Gram negative bacillus twice. Several germs were found 4 times. The treatment included adjusted antibiotics with good CSF penetration andexternal ventricular derivation, after the contamined material was removed, in all the cases except the meningococcal meningitis. The cerebrospinal fluid (CSF~ became germ-free in 15,8 + 12 days and a new internal derivation (ventriculc-peritoneal shunt 5 times, ventricnlocardiac shunt 4 times) was put after 33,9 + 13,7 days of external derivation. Ten children recovered without new sequelae. One died in two days of staphylococcal sepsis and peritonitis ; she was 17 days old. In conclusion treatment of prolonged and severe shunt infections by external ventricular drainage of long duration and specific antimicrobial therapy may offer improvement in most of the cases. Some data concern the presence inhuman tissues of an atypical creatine kinase (CK) isoenzyme, migrating to the cathod, which is possibly coming from mitochondria (CKm). Sera from 320 infants and children younger than 5 years undergoing open heart surge~/were examined. Samples were done at the arrival in ICU and the morning of the next 3 days. We electrophoretically senarated and fluorometrically detected CK isoenzvmes according to Rosalki (et al) modified by Szasz (et ali. In 38 patients, besides the 3 well kn~nC~l, C~ and CKBB isoenz]anes, we have isolated an atypical CK isoenzyme. We have verified that this abnormal fraction is antigenically "no H, no B" and have the same cathodic mobility t.hat CKm that we have found in organ extracts (hear% lung, liver, ...). From data profile there is 3 groups in patients. Group I (4/38) where CKm is present as early as the arrival in ICU ; Group II (9/38) where CKm appears the day foll~r operation ; Group III (25/38) where CKm is present only the second post operative d~ whereas C~ is nearly vanished. We have verified that CKm does not result from eventual transformation of ~ or BB isoenzymes. CKm disappears the next days. Two of the 38 patients are died, that is ~emean mortality rate. CKm occurrence in infants and children after open heart surge~/ does not seem to have pejorative signification. Significance of this is still uncertain and further studies are necessa~/. Opsonic fibronectin depletion decreases resistance to sepsis (!~.E. Lanser, T.~I. Saba, Ann. Surg., 195, 3 z 1982) with the aim of studying factors predisposing to infections complications, we analyse plasma fibronectin (PF) rate variations following open-heart surgery. This study concerns 10 pediatric patients, aged between 9 months and 13 years, after cardio pulmonary bypass for surgical correction of congenital heart diseases. BLood samples, anticoagulated with Na Citrate, were taken before operation and 6 hours, I day and 7 days after surgical procedure. After centrifugation, plasma samples were immediately frozen. ~leasurements of PF were done by radial immunodiffusion. For statistical analysis non parametric Wilcoxon paired T test were used. Before operation, baseline Level of plasma fibronectin (I) is 303 • 59 mg/l (mean ~ SD). At the sixth postoperative hour (2) PF = 314 • 52 mg/l ; the first postoperative day (3) PF : 270 • 26 mg/l ; the seventh day (4) PF = 330 • 36 mg/l. Statistical comparison gives : for 1 vsa, I VS3, 2 VS3, 3 VS4 : p ~ 0.01 ; but for I VS4 ;IS. Plasma fibronectin levels decrease as early as the 6th hour ; the 7th day PF levels are nearly returned to the baseline rate. Thus, cardiac surgery with cardiopulmonary bypass may induce a drop of PF level. It is possible that fibronectin might lower resistance to sepsis which is a dreaded complication of this surgery. 420 oatients (nremature : 3 %; age~l month : 6 %; 1 to 24 months : 31%; 2 to 6 y.o.: 38 %; 6 y.o.: 22 %) submitted congenital corrective cardiac surgery were studied. Car-dio~ulmonary bYbaSS was used in 82 %. Weights were measured before and after surgery. 125 oatients have lost weight during the procedure, average:l.66 % + 1.58 295 natients have increased their initial BW~ average : 1.84 % + 2.02. Five subgroups were used for statistical analysis : Grou~ 1 : loss> 2 % BW qrouo II: loss 0-2 % BW qroun Ill:gain 1% BW Groun IV: gain I-2 % BW qroun V : gain~ 2 % BW The nostsurgical comnlications : cardiac failure, arrythmias, atelectasis, oleural effusion, and the nostsurgical ventilatory weaning delay were statistically analysed in these 5 grouDs. The number of postsurgical comnlications was increased in correlation with the BW gain during the surgical nrocedure. For instance,, mean incidence of comolications get patients averages 0.51 in group I and 1.15 in grouo V (~<0.001). The nostsurgical ventilato~ weaning orocedure has been more difficult in patients with increased BW;~group I endotracheal intubation duration averages : 9 H; it averages 24 hours in group V (p<O.O01). Conclusion : stability of BW during congenital cardiac surgery is required. BW gain of 1.84 % + 2.02 increases significantly the incidence of nostsuro~cal comolications (cardiac failure, arrythmias, nieural effusion) and delays significantly ventilatory weaning time.

COP was measured ~n duplicate in 316 samples of plasma (217 patients end 99 blood donors)with an IL-WEIL Oncometer,at 24~ measured values were corrected for 37~ the arithmetical mean for paired samples was calculated (C-~'F,37~.Duplicated measurements differed only in 0.22• mmHg (p=O.81).However,paired comparison of mean value of C--0ff

Albumin(A) and total proteins(TP) were measured in each sample, and globulin(G) and A/G ratio calcu]ated. Normal COP in the blood donors group(ambulant)was 31.2• mm llg. CALCULATIONS: COP (in mmllg) was

86 G)'T'O.0735/273 (fc=factor depending of TP

pH)c (c= TP;0.735 = conversion factor of units

68~4.29) was significantly different from mean F-I

Absolute differences between COP,37 and F-I,F-2 and F-3 were respectively 3.26~2.36, 2~ and 2.01• they were correlated with the A/G ratio and G values

IN HIGH-DOSE BARBITURATE THERAPY. C. Krier, K. Wiedemann, H. Polarz, Department of Anesthesia, University of Heidelberg, F R G Despite of the fact that the beneficial effect of barbiturates on the outcome of severe headinjury pat&ents remains controversial,high-dose barbiturate therapy is now commonly used in the management of raised intracranial pressure. The purpose of the study was to investigate the potential side-effects of high-dose thiopentone therapy in a total of 30 patients with focused interest on cardiovascular effects, effects on liver metabolism, qastro-intestinal complications and respiratory complications.The barbituratetreated patients were compared to a group of 27 patients with severe head-injury without highdose barbiturate therapy. The data obtained indicate i. a low incidence of hemodynamic changes when a regimen of slow initial loading is adopted. 2. a low incidence of hepatotoxic effects. 1 case of transient drug-induced jaundice is reported. 3. a high incidence of gastro-intestinal complications, ranging from benign erosions to acute upper gastro-intestinal hemorrhage. The occurence of oesophagitis is related to the esophageal sphincter dilatation and reflux, due to the high-dose barbiturates and demonstrated by endoscopic findings. 4. a h~gh incidence of contamination of the upper respiratory tract and of pneumonia., It is suggested that the suppressive effect of thiopentone on the immune response is 9esponsible for the high incidence of pulmonary complications. It is concluded that the high-dose barbiturate therapy is connected with high risks for the patient due to the multiple side-effects of barbiturates and that this form of therapy requires a extended monitoring and specific preventive measures. At the end of June 1981, 6 workers were intoxicated wlth the above mentioned substances while cleaninq an industrial basin. After a latent period of I to 3 days, symptoms of the neuroloqicpsychic kind occured in all patients with a upward tendency up to a clouding of consciousness and in the mogt severe cases even to a central resoiration depression, A development of pulmonary complications with massive infiltrations was observed. Artificial resniration theraoy was necessary in 3 patients.In 4 oatients epileptic activity was reqistered in the EEG. Iniatially all patients showed a leucocytosis up to 3o.ooo/mm 3, transaminases were slightly increased with an unward tendency. The execretion of tin in the urine corralated with the severity of intoxication as well as a hypokalemia un to 1,9 mval/l. The alasmasenaration done in 4 patients did not lead to the expected decline in the urine tin execretion.The aoplication of metalcaDtase R did not show a tendency to a higher tin execretion in the urine. The main treatment consisted in intensive care procedures. One oatient died because of eulmonary complications, liver cell damage and final kidney insufficiency. The neurohistoloqic findings showed remarkable changes in the central nervous system. I patient still needs intensive care because of hypotone hyperkinetic syndrome. The patient with the third hiqhest tin level still suffers from slightly cerebral deficit.Up to now, there is no report of human toxicological and intensive care experiences in the case of acute intoxications with these substances. This is a review of 30 deep toxic comas studied by BAER to establish its value in diagnos~g and prognosis in association with a physical determination of the level of the coma and toxic analysies.

Importance in diagnosis, a normal response was the most 9ommon finding, but in almost 40 % of the cases, we observed a delayed latency at different level of coma of different toxic etiologies. This result is opposed the commonly accepted idea that BAER are always normal in toxic coma and represent a strong argument to evocate the toxic origin. Therefore, delaye d latency should not exclude the diagnosis. Importance in prognosis : a normal response or a delayed latency is always associated with a good prognosis. But in the latter case the duration of the coma seems longer. So BAER might represent an argument for the use of an epuration method. On the other hand if we note a peak disappearrance as observed in 2 cases, the outcome will be poor~ perhaps due to anoxia. This could be screened for, rapidly, using this method. We studied 135 patients with aconite poisoning fatal dose taken p.o. is 2-4 mg. The main symptom in this state was acute cardiac arrhythmia developing suddenly during the first 30-50 minutes. The changes were characterized by progressive disturbance in conduction, excitability and automatism of the heart. Antidote therapy was not used, as it doesn't exist in this sphere. The patients were treated by intense therapy and resuscitation consisted in use of antiarrhythmic drugs (lidocaSne, novocaTnamide, etal), electric-pulsating therapy, electrocardiostimulating procedure, intensive ureapoiesis, gastric lavage and hemoresorption. We noted the high effectiveness of electro-pulsating therapy. The mortality rate was decreased in this group of patients by the described type of treatment. Stress ulceration is common in patients treated in Intensive Care and when associated with active bleeding carries a high mortality. Antacids effectively increase the stomach pH and decrease the evidence of GI bleeding, whereas cimetidine is not as effective in increasing pH nor in preventing GI bleeding in seriously ill patients. (I) A new histamine H~-receptor antagonist, ranitidine, has not been evaluated in these patients. The object of this study was to evaluate the effectiveness of ranitidine in controlling the gastric pH at above 4. If the pH was ~elow 4 a bolus dose of 50 mgm iv of ranitidine was given and an infusion of Imgm/hr corm~eneed. The infusion was increased up t O 15 msm/hr pr decreased aecording to the pH of the stomach. A further object of the trial was to evaluate a continuous method for measuring gastric pH. This is necessary if the effectiveness of a continuous technique to increase pH is to be assessed. A pH electrode (Novo Indnstri A/S) was inserted together with the NG tube. The pH was continually measured and displayed enabling rapid adjustments to infusion rates of ranitidine to be made. This method was successfully employed in all patients and no adverse reactions attributable to ranitidine were noted. It is hoped that this technique will be the basis for a closed loop servo control system for cnntrolling gastric pH in seriously ill patients. Anaerobic infection is frequently suspected in post surgical acute renal failure (ARF). The aim of this study was to evaluate the pharmacokinetics of parenteral metronidazole in ARF patients treated by hemodialysis. 7 patients with post-surgical ARF, treated every other day by hemodialysis, received parenteral metronidazole in a dose of 500 mg administered in 20 minutes, eight hourl~ In the 7 patients serum values of metronidazole (high pressure liquid chromatography assa~were obtained before hemodlalysis, after the first injection, every 15 minutes for 2 hours and every hour for 8 hours. During hemodialysis serum levels were obtained after a second infusion every 30 mn for 2 hours and every hour to the end of hemodialysis. In 3 patients the therapy was maintained for 3 to 8 days and peak and trough levels were obtained before and after each injection. After the first injection peak levels were 19.03 • 3:47 mg/l (mean • SD). Elimination half life (T I/2 B) was 7.06 • 1.95 h. Trough levels were 5.92 • 2.85 mg/l. These reSults were not different from those obtained in normal subjects. During hemodialysis T i/2 B was 3.45 • 0.5 h corresponding to a mean decrease of 35 % in 4 to 6 h. There was no rebound after hemodialysis. In the three patients receiving repeated infusions the mean peak and trough values for the last five infusions were 18.5 mg/l andP.Omg/l, not different from values obtained after repeated infusion in normal subjects. Hemodialysis prevented significant accumulation of the drug. In conclusion : pharmacokinetics of I.V. metronidazole in ARF are similar to those of normal sujects. Hemodialysis increases the elimination rate. The usual frequency of infusion in normal subjects is suitable for treatment of ARF patients submitted to hemodialysis every other day.

A.Vuori, P.Jaakkola, Department of Anaesthesioiogy, University of Turku, Turku, Finland.The validity and usefulness of capnography in monitoring ventilation in preterm infants was studied comparing the correlation of end tidal pCO^ to arterial pCO 9. A t6tal of 69 sample~ were included in this study.The patients were preterm infants treated with CPAP or IMV for IRDS (hyaline membrane disease) in most cases.The Baby Bird ventilator was used.Because of small tidal volumes, gas samples were taken from the intubation tube near the patient by Cutting the tube and connecting a T-piece between the ends.Gas samples from this T-piece were continuously drawn to a fast-response infrared carbon dioxide analyser connected to a recorder (Datex Normocap DC-102 & CR-I02). The method turned out to be easy to employ and valid.There was a good correlation between ETpCO 2 and arterial pCO 9 in ~ase the plateaux on the capnograms were Droad and/or of equal height.The severity of the IRDS had influence on the correlation, whereas gestational age and birth weight seemed not to affect the correlation.