key: cord-0033832-cnm6vssw
authors: nan
title: PS 0446-0765
date: 2006-08-24
journal: Intensive Care Med
DOI: 10.1007/s00134-006-0320-5
sha: a2dffcab78ea1279969e52a93410e70ef87d3252
doc_id: 33832
cord_uid: cnm6vssw

nan

CONCLUSION. The valved double lumen tracheal tube allowed a higher rate of elimination of carbon dioxide from the test lung than did the standard tube. CONCLUSION. Mortality in ARDSpul did not change, however we demonstrate a better outcome in the extrapulmonary ARDS, related to the early treatment with aCP and to a younger population. The improvement in oxygenation (PaO2/FiO2) was earlier in extrapulmonary ARDS. 2) In our small series, patients with ALI have no changes in EBC compared with a normal population.

3) Inhaled salbutamol significantly increased the pH of EBC in patients with ALI but a direct effect of salbutamol inhalation cannot be discounted. CONCLUSION. We found a direct relation with statistical significance between the higher mortality rate, seriousness of the illness and Protein C values. Though the reduced patients enrolled in the study, this could means that the levels of Protein C disposable to be activated could be important as a prognostic factor in SIRS and septic patients.

Campos L 1 , Porta I 1 , Palomar M 1 , Otal J 2 , Burgueño M 1 , Larrosa N 3 , Vilalta J 4 1 ICU, 2 Preventive M, 3 Microbiology, 4 Neurosurgery, H Vall Hebron, Barcelona, Spain

Ventriculitis is a serious complication of external ventriculostomy drain (EVD) use. Published series about ventriculostomy related infections (VRIs) are highly variable in terms of definition of infection. Criteria proposed by Lozier et al, provides a precise description of VRIs, essential to identify clinically relevant infections.

Prospective observational study (Jan 97-June 05) of all consecutive patients (pt.), admitted to our ICU, with acute non-traumatic intracranial haemorrhage who required an EVD. Criteria proposed by Lozier were used to define VRI: contamination, ventriculostomy colonization, suspected VRI (negative CSF cultures but chemistry and citology strongly suggested a VRI); VRI and ventriculitis. Infection rates were calculated per 100 pt and per 1000 days of EVD.

A total of 250 pt. (114 males), mean age of 54 ±15 y, APACHE II 16.3 ± 9.2 points, GCS 9.6 ± 4.2 required EVD;158 pt had subarachnoidal and 92 intraparenchimal haemorrhage (intraventricular haemorrhage was also present in 189 pt). A total of 359 EVD were inserted. The mean length of catheterization was 20.3 ±16.3 days, and the total days with EVD were 5.075. In 84 (33.6%) pt, 98 VRIs episodes (ep). were identified (70 pt. 1 ep; and 14 pt. 2 ep). The mean time of onset of ventricular infections was 11.9 ± 8.16 days. Rates of VRIs for the 5 groups in table 1. 

Voets A J 1 , Tircoveanu R 1 , Delmee M 2 , Cherifi S 3 , Reper P 1 , Jamart S 1 , Mascart G 4 , Massaut J 1 1 Surgical Intensive Care, CHU-Brugmann, 2 Laboratory of Microbiology, St Luc-UCL University Hospital, 3 Infectious Diseases, 4 Laboratory of Microbiology, CHU-Brugmann, Brussel, Belgium INTRODUCTION. C-reactive protein (CRP), an acute-phase protein, is a sensitive systemic marker of inflammation and tissue damage. Although it has been used for many years as a sepsis marker, its value remains controversial. This is particularly true in surgical patients, who have an inflammatory response not necessarily related to sepsis. The primary aim of this study was to evaluate the value of blood CRP concentrations in surgical patients.

In this prospective, observational study, we measured CRP concentrations daily in all ICU admissions in a 1 month period.

We studied 105 surgical patients, including 66 with uncomplicated surveillance, and 39 with a complicated course, who spent more than 1 day in the ICU. Infection was the complication in 12 patients and problems other than infection complicated the course of the other 27 patients. The septic group presented higher values of CRP throughout the first days in the ICU when compared with the other two groups (Table) . 

All surgical patients have an increase in CRP concentrations, but these are more significant in patients with infection.

In patients with high gastric residual volumes jejunal feeding is recommended. Jejunal feeding tubes can be placed in different ways. The endoscopic technique yields a success rate between 90 and 98% for a correct jejunal placement. However, it requires endoscopic equipement and trained staff. In contrast in small uncontrolled trials different unguided techniques resulted in success rates up to 75%, only. In this prospective randomized trial the success rate of a correct jejunal placement with the endoscopic technique (Freka-Trelumina ® , Fresenius-Kabi) was compared with the unguided frictional technique (Tigertube ® , Cook).

METHODS. 42 patients fulfilling the clinical indication (repeated vomiting, gastric residual volume > 250ml) for jejunal feeding were included in the study. They were randomized in group A (endoscopic technique) or group B (unguided frictional technique). Primary outcome measure was the success rate of a correct jejunal placement. Secondary outcome measures were time from the beginning of placement until successful jejunal placement, number of attempts and necessity of an additional gastric tube.

RESULTS. Data are given as mean ± standard deviation. p<0.05 was regarded to indicate statistical significance. 

Correct jejunal tube placement using the endoscopic technique was successful in all patients. Success rate of the unguided frictional technique was significantly decreased. Additionally, in patients with successful jejunal tube placement using the unguided frictional technique duration of placement and the necessity of an additional gastric tube were significantly increased.

Roderjan C N 1 , Hatum R M 1 , Werneck P 1 , Neto E C 1 , Silva E G O 1 , Vanzan A B 1 , Vianna C C 2 , Cunha C S 2 , Knibel M F 1 1 Intensive Care Unit, 2 Nutritional Support, São Lucas Hospital, Rio de Janeiro, Brazil

The enteral nutrition therapy (ENT) is vital part of the treatment in a majority of the patients in the Intensive Care Unit (ICU). Many patients yet have alterations of the nutritional state in the admission and others have significant risk to develop malnutrition, due to the catabolism of the illness or injury. The current recommendation determines a maximum time of 96 hours (four days) after beginning of ENT to achieve total energy expenditure (TEE). In 2004 the time expense for accomplish of the TEE in a ICU was 4.7 days. Educational measures and training of the multidisciplinary team had been carried through, and new collection of data was made in 2005. The study analyzes the effect of the systematization of control procedures on the average time for attainment of basal energy expenditure in enteral nutritional support of seriously ill patients.

This was a retrospective observational study with 216 patients interned in the ICU before the systematization (in 2004) and 140 patients after the same (2005) .

In the first time 216 patients were studied (56% were women and 44% men), with age average of 74.5 years. One year after we analyzed 140 patients, 62.85% of women and 37.15% of men with average age of 71.02 years. The most prevalent diagnostics found in first and second study steps was respectively: respiratory failure in 28% and 31%; neurological events in 20% and 22%; and postoperative 8% and 18%. The average duration of nutritional support was 30.36 and 23.4 days respectively, and the TEE was reached in 4.7 days before systematization and in 3.13 days after educational measures.

CONCLUSION. In view of the reduction of the mean time for attainment of the TEE (4.7 for 3.13 days) it assure that the systematization of nutritional strategies has relevance in the quality of the offered nutritional support.

Grant acknowledgement. ESHO -AMIL. 19th ESICM Annual Congress -Barcelona, Spain -24-27 September 2006 S133 0502

Roderjan C N 1 , Hatum R M 1 , Neto E C 1 , Werneck P 1 , Silva E G O 1 , Vanzan A B 1 , Vianna C C 2 , Cunha C S 2 , Knibel M F 1 1 Intensive Care Unit, 2 Nutritional Support, São Lucas Hospital, Rio de Janeiro, Brazil

The enteral nutrition therapy (ENT) is important part in the treatment of the patients in the Intensive Care Unit (ICU). Early and aggressive nutritional support can aid in the reduction of complications and reduce the inpatient time. However, some factors limit the full administration of enteral nutrition to these patients. Objective: To identify factors that interferes with average time for attainment of basal energy expenditure (BEE) in enteral nutritional support of seriously ill patients and to raise hypotheses of possible consequences.

This was a retrospective observational study with 140 patients admitted in ICU and remained in nutricional therapy (at least five days) in 2005.

In 140 patients 62.85% were women and 37.15% men. The average age was 71.02 years. We have 55.7% patients with interruptions in the ENT (180 interruptions in the total). Amongst the reasons most prevalent to the interuuptions had been identified: examinations and clinical procedures in 45 cases (27.7% of the interruptions), abdominal distention = 18 (11%), vomits = 22 (13.5%), surgery = 11 (6.79%) and mal adjusted of enteral catheter in 41 cases (25.3%). Analyzing the average time for attainment of the BEE referenced by literature in 4 days, 24 patients (17.14%) not reached this goal, 79% caused by interruptions of diet infusion. Average time to reach BEE was 3.13 ± 2.19 days.

The analysis of this study shows the high incidence of interruptions in the ENT of seriously ill patients, without consequences in the average time to achieve the BEE (average time for attainment BEE was 3.13 days). The adoption and maintenance of nutrition protocols to detect and to minimize the involved factors of inadequate administration of ENT are fundamental activities in a ICU, mainly related to the maladjusted of enteric catheters and accomplishment of clinical procedures.

Grant acknowledgement. ESHO -AMIL.

Liatsi D 1 , Tsapas V 1 , Veroniki F 1 , Tsagourias M 1 , Pampori S 1 , Matamis D 1 1 ICU, Papageorgiou General Hospital, Thessaloniki, Greece

Agitation is a frequent finding in ICU patients. Situations, as opioid or benzodiazepine withdrawal syndrome and brain trauma may induce central sympathetic overactivity and agitation after sedation interruption. The metabolic and respiratory consequences of agitation may have serious therapeutic implications.

Seventeen ICU patients (14 males and 3 females) with an age of 32±12 years (mean ± SD), presenting agitation after sedation interruption were studied by indirect calorimetry. Measurements were performed before and after sedation interruption, during weaning from mechanical ventilation under Pressure Support Mechanical Ventilation. The following metabolic and respiratory parameters were measured: Tidal Volume (Vt), Respiratory Rate (RR), Minute Ventilation (VE), O2 Consumption (VO2), CO2 Production (VCO2), Resting Energy Expenditure (REE), and Resting Energy Expenditure as a percentage of predicted, according to the Harris and Benedict equation. Paired t-test was used for statistical analysis.

The effect of agitation on the respiratory parameters Vt, RR, VE was an increase from 704 ± 160 to 768 ± 157ml, from 12 ± 3 to 24 ± 8 breaths/min and from 9±2 to 16±5 lit/min, respectively (p<0.01). Concerning the metabolic parameters, agitation induced a tremendous increase in VO2, VCO2, and REE [from 270±73 to 404±160 ml O2/min from 231±65 to 382±165 ml CO2/min and 1903±510 to 2904±1177 kcal/day respectively (p<0.01)]. REE, compared to REE predicted was increased from 104% to 160%, (p< 0.01). Respiratory Quotient (RQ) remained constant.

Agitation induces a tremendous increase in the metabolic demands, thus increasing the ventilatory demands in order to balance the O2 consumption and CO2 production. This may be detrimental for patients with limited respiratory reserve.

Dasopoulou M D 1 , Costalos C G 1 , Constantellou E T 2 , Siannis F T 3 1 Neonatal Department, 'Alexandra' Hospital, Athens, 2 Endocrinology Department, 'Agios Panteleimon' Hospital, Nikea, Greece, 3 Biostatistics Unit, Cambridge University, Cambridge, United Kingdom

Leptin is a hormone excreting from adipose tissue and bound to hypothalamus. Leptin is implicated in appetite control, temperature up-regulation and energy expenditure. Our study's aim was to determine the impact of feeding on leptin secretion in a population of preterm neonates.

Leptin study was a prospective one, enrolling 87 preterms born at the hospital. Ethics Committee Approval and parental informed consent were obtained. Eligible neonates were all preterms (<37 weeks of gestation, according to Dubowitz criteria). All neonates were fed within 4-6 hours after birth the same volume according to weight and day. For example on day 1 they received 30 ml/kg with an increment of 30 ml/kg/per day till maximum 250 ml/kg/per day. Neonates were fed the same formula and breast fed infants were excluded as well as those with respiratory distress, birth injury, vomiting, diarrhea, necrotising enterocolitis. Blood sampling during the first 6 hours of life and repeated sampling on day 26 of feeding for serial leptin measurements. 

Ellingsen K 1 , Boganes E 1 , Thorsell E 1 1 Intensive Care Unit, Stavanger University Hospital, Stavanger, Norway

The intensive care unit (ICU) at Stavanger University Hospital in Norway has carried out a quality improvement project on nutrition for the intensive care patient (ICP), involving development of guidelines for escalation (aligorithm) for enteral nutrition, training of nurses and doctors as well as courses and tuition. The purpose was to describe existing nutrition practices in selected registration periods in order to measure the effects of the project in the short as well as the long term.

The project has utilized the prinsiples of the "penetration method". Measurement instrument: Registration of nutrition practices before and after implementation of the new standard (phase 1 and 2), repeated after 21/2 years (phase3). Included:All adult ICP who spent more than two days at the ICU. (2) received a combination of PE and EE. 36% (10) received "free liquids". 4) 52% (76/147) of lay days had documented nutrition targets.

CONCLUSION. The project has improved both the focus on and quality of clinical nutrition at our ICU. Not unexpectantly, the effects were greater in the short term than in the long term. The main effects in the short term were that the calorie intake by the ICP increased by 41%. The use of EE and PE in combination to satisfy the nutritional requirements of the patients also increased. In addition to improved documentation, earlier and increased use of EE is the primary long term effect. The calorie needs of the ICP are still not adequately covered. This show that the ICP still are suffering from malnutrition. This presents both challanges and improvement potential. Main challanges: 1) Adequate nutrition supply. More combination of EE and PE should be concidered in order to meet the needs of the patient. 2) Achiving optimum start of administering (within the first 24-48 hours) and more extensive use of EE. 3) Maintaining focus on and interest for clinical nutrition in the ICU. 4) Regard quality improvement of nutrition for the ICP as a contiuous process.

Gordon D S, Simpson F:"Early enteral nutrition in the critically ill:do we need more evidence or better evidence?" I:Current Opinion in Critical Care 2006,12:126-130.

Dadak L 1 , Stouracova M 1 , Kuklinek P 1 , Stetka P 1 , Sramek V 1 1 Anesthesia and Intensive Care Unit, St. Ann's University Hospital, Brno, Czech Republic

To evaluate impact of synbiotics on enteral feeding tolerance and immune function in long tern ICU patients.

The study was approved by institutional ethical committee. Patients estimated on D1 (D0 = admission) to stay in the ICU > 3 days were randomized. Placebo group (Placebo) received daily tea, treatment group (Synbiotic) Synbiotic 2000 Forte (Medipharm AB, Sweden) until D21. In both groups the enteral nutrition (EN) was administered post pylorus acc. to standard ICU protocol. Monocyte function was monitored on D1 and subsequently every 5 days. Daily amount of enteral nutrition and number of stools were recorded. In this abstract preliminary analysis of early development (D1 -D5 -D10) of CD14+HLADR+ expression and tolerance of enteral feeding is reported. Data are presented as median (Q25; Q75), nonparametric tests were used.

Twelve patients (10 male; age 54 (39; 63)) were randomized (Synbiotic N=7, Placebo N=5). Eleven patients survived ICU stay (S), 1 did not (NS). Mostly surgical patients (N=11) were included, 7 primary admission, 5 secondary admission. APACHE II on admission was 22 (21; 27). There was no significant difference in CD14+HLADR+ on Day 1 (Synbiotic 42 (30; 59) vs. Placebo 54 (33; 66) ). An improvement of CD14+HLADR+ in Synbiotic group was observed during ICU stay (p=0.02) but not in Placebo. On D10 CD14+HLADR+ in both groups was 73 (64; 81) Synbiotic and 52.5 (50; 61) in Placebo. One hundred and two days of enteral nutrition in Synbiotic group vs. 60 days in Placebo group was compared. Daily amount of enteral nutrition was higher in Synbiotic group (905 (560; 1240) ml/day vs. Placebo group 720 (455; 960) ml/day, p=0.05). Number of stools was higher in Synbiotic group (85stools/102 days) than in Placebo group (27stools/60days) during the treatment. Number of days of enteral nutrition intolerance was very low (Synbiotic 2/102 vs. Placebo 1/60 days). 

Berger M M 1 , Chioléro R 1 , Soguel-Alexander L 1 , Revelly J 1 , Eggimann P 1 1 Dpt of Intensive Care Medicine, and Burns Centre, Lausanne, Switzerland

Critically ill patient suffer early glutamine (GLN) depletion. Trauma and burn patients are particularly affected, and suffer addition selenium and zinc deficits. A series of trials have shown clinical benefit from early micronutrient supplementation by the IV route. The goals of the present clinical study were to monitor the clinical introduction of an enteral supplement enriched with GLN and selenium, and to observe its influence on gastrointestinal function.

Case control study of prospectively collected data. Inclusion criteria: critically ill trauma and burn patients requiring enteral feeding after January (INT): 30 g GLN, 300 mcg Se, 30 mg Zn, 500 ml, 250 kcal/day in ® 2005. Intestamin 500 ml -delivery starting within 36 hours of admission, by the naso-enteric tube, for 10 days. All patients received additional intravenous Se and Zn supplements for 5 (trauma) or 15 days (burns). Variables: demographic data, SAPSII, SOFA and ISS scores, daily total, and enteral energy delivery, gut function (gastric residues, stool), length of mechanical ventilation and ICU stay, outcome.

. 18 years, including 46 trauma (ISS 28±8)±Eighty patients were enrolled, aged 42 19% body surface). The patients were critically±and 34 burn patients (burns 30 ill (SAPS 38±13), well matched for age, sex, and severity of injury (trend more ® inhalation injuries in INT group: 13 vs 9, p=0.15). Mean volume of Intestamin 3 days. Enteral energy delivery was higher±delivered was 330±200 ml/day, for 7 in the INT group (1268±827 vs 853±880 kcal/day, plus 415 kcal/day p<0.001) both in burns and trauma patients -which was more than the energy content of the INT solution. GI function under analysis. SOFA score decreased similarly in both groups over the first 5 days. There was no difference in clinical outcome.

The enteral delivery of the solution was well tolerated as shown by the higher energy delivery in the intervention group. No intolerance or toxicity could be detected.

Grant acknowledgement. the study was supported by a grant from Fresenius Kabi, Stans, Switzerland.

Villalobos S J 1 , De La Torre L M 1 , Etulain G J 1 , García Z T 1 , Arcos Z M 1 , Aguirre S J 1 , Martínez S J 1 1 Intensive Care Unit, Center Medical ABC, México, Mexico

: Loose of weight is associated with malnutrition as a consequence of total or partial fasting situations that caused a catabolic state in the patient. The aims of the nutritional therapy in patients with mechanical ventilation (MV) are: to give the necessary energy support, to reduce the ventilatory muscle compromise, to avoid overproduction of CO2 and reverse the nutritional disturbance caused by respiratory insufficiency.

Retrospective study. We analyzed 114 patient's files with the diagnostic of COPD with MV, receiving enteral nutrition for at least 96 hours, divided in two groups: Group 1 with high carbohydrates diet and group 2 with low carbohydrate diet. We compared age, gender, length of stay and ventilation days. Delta of: prealbumin, transferrin, total lymphocytes, urinary nitrogen excretion at the beginning or during the first 24 hours and after 96 hours. We used t-student and X2.

. Fifty one had EPOC and an underlying disease requiring MV. The mean age was 64.5 ± 9.9 years (41-82), 65% (33) were male and 35% (18) female. The mean length of stay was 13.2 ± 4.2 days (7-35), the ventilation were of 7.5 ± 2.5 days (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) . The admission diagnosis were: lung infection 27.5% (14), systemic infection 7.8 (4), immediate postoperative 17.6% (9), stroke 17.6% (9) and acute coronary syndrome 13.7 (7). The mean weight was 67.2 ± 7.6 kg (50-82). The group 1 were 58.8% (30) patients given a high carbohydrates diet, and the group 2 were 41.2% (21) given a low carbohydrates diet. The protein support for both groups was 104 ± 21 (66-142) grams and 1616 ± 280 (1102-2197) kcal. No significant differences were seen in both groups. Delta comparison of urinary nitrogen excretion between the group 1 and 2 was 2.01 ± 1.8 (21) and 0.48 ±2.0 (30) respectively, with a p=0.05. Delta of prealbumin between group 1 and 2 was 0.876 ± 2.5 and 3.78 ± 2.2 respectively with a p=0.04. Delta transferrin between group 1 and 2 was 3.2 ± 17.8 and 29.1 ± 17.2 respectively with a p=0.02 and finally delta of lymphocyte between group 1 and 2 was 186 ± 246.5 and 519 ± 232.5 respectively with a p=0.01.

We found a significant difference in the catabolism degree and in the synthesis of prealbúmina, transferrin and in lymphocyte cell formation in the low carbohydrated treated group.

Grant acknowledgement. Intensive Care unit. "Mario Shapiro". 

We previously observed that critically ill patients requiring prolonged intensive care are vitamin D deficient1. Although 500 IU iv vitamin D supplementation daily did not normalize circulating 25(OH)D nor 1.25(OH)2D levels, an anti-inflammatory effect of exogenous vitamin D was revealed by suppression of serum CRP and IL6-concentrations if circulating 25(OH)D were significantly, albeit moderately increased1. Apart from its classical effects on bone and calcium metabolism 1.25(OH)2D has important immunological effects on cytokine production and monocyte function. Presently, the effect of rapid and full normalization of the vitamin D status on inflammation and calcium metabolism in critically ill patients remains elusive.

In this study, we randomly allocated patients upon ICU admission to receive either 10 days treatment with 15 µg 25(OH)D/d iv (following a 200 µg 25(OH)D iv loading dose, n=11) or placebo (n=13), on top of the currently advised daily 200 IU iv vitamin D supplement. Patients with an anticipated ICU stay of >10 days were eligible for inclusion. Patients younger than 18 years, those suffering from chronic bone or kidney disease and those treated with glucocorticoids before ICU admission were excluded.

. 25(OH)D treatment did not result in a significantly stronger decline in serum CRP over the 10-day observation period compared to placebo. Serum ionized calcium levels remained constant in all patients, while serum phosphorus levels fluctuated similarly in both treatment arms. Nor daily SOFA scores nor length of ventilation were significantly influenced by the 25(OH)D supplementation. Noteworthy, ICU mortality was 1/11 in the 25(OH)D group and 2/13 in the placebo group.

CONCLUSION. These preliminary study results indicate that in contrast to the high 500 IU vitamin D regimen, iv 25(OH)D administration did not significantly effect the indistinct inflammation parameter CRP. The indistinguishable serum calcium and phosphorus levels suggest the absence of adverse effects on mineral homeostasis. Nevertheless, further unraveling of particular cytokine profiles, monocyte function and specific bone formation and resorption indices in these patients are required to deduce the value of intravenous 25(OH)D supplementation in prolonged critically ill patients. INTRODUCTION. Obesity has been described as an independent risk factor for adverse outcomes in critically ill patients, with conflicting results. Recently, lower body mass index (BMI) has been associated with higher odds of death in the intensive care unit (ICU) population. The objective of this study is to determine the influence of BMI on ICU mortality, ICU length of stay (LOS) and nursing resource utilization.

Retrospective analysis of a 19 bed ICU database from 2000 to 2005. The cohort was divided in post-operative and non-operative groups. Coronary patients and readmissions to the ICU were excluded. We created the following five subgroups based on the BMI: <20, 20 to 25, 25.1 to 30, 30.1 to 40, and > 40.0 Kg/m2. Groups were compared by age, APACHE II score, mortality, ICU length of stay (LOS), TISS 28. Association between BMI subgroups and ICU mortality was evaluated using the odds ratio. A multiple logistic regression analysis was used to determine the independent impact of BMI on ICU mortality. P-value < 0.05 was considered significant. The 95% confidence interval (CI) was calculated for the odds ratio (OR).

Of 4104 patient datasets, 3521 were complete. Mean age was 64 years, 50.1% were women and median APACHE II score was 13. BMI < 20 was associated with increased predicted (P<0.001) and observed mortality in the non-operative group (OR =1.56, 95% CI, 1.138 to 2.140) although multiple regression analysis did not identify BMI as an independent factor. We could not find any association between BMI and outcome in the operative patients. There were no differences in ICU LOS or in TISS 28. The severely obese patients were more frequently female (p<0.001).

CONCLUSION. BMI has minimal effects on ICU outcome after patients are admitted to a critical care unit. 

Galdino P 1 , Bortoletto T C 1 , Gurgel A P A 1 , Henrique L M P 1 , Oller A M 1 , Gama R X 1 , Zanini A A 1 , Brandao D F 1 , Faintuch J 1 1 Pharmacy, Hospital Alemao Oswaldo Cruz, Sao Paulo, Brazil

White blood cell count is routinely measured in septic patients and fat intake during parenteral nutrition (TPN) is available in the prescription, but correlation between these findings including clinical outcome has not been often addressed. In a prospective study with seriously ill, malnourished patients, impact of omega-6 lipid administration on clinical course was examined.

Patients submitted to exclusive TPN (n= 96) were investigated on the first day of therapy. Fat intake range (soybean oil emulsion) was 2%-48% of total calories. Variables included initial WBC, WBC change (final -initial), biochemical tests, duration of hospitalization and mortality. Stratification was done according to lipid calories >15% total calories or <15%. Pearson regression analysis was employed to compare the variables. x-rays (CRX) are indicated in patients receiving mechanical ventilation (MV), although the effectiveness of these radiographs is unknown. Thus, the utility of daily routine portable CRX in MV patients (pts) remains controversial. After a survey about CRX indication, we decided that CRX should be obtained only on clinical indication. The aim of this study was to determine 8 and 9 years after its implementation the number of CRX performed, the costs savings and the outcome impact.

To calculate the number of CRX saved with this strategy during 2003 and 2004, CRX volume data were collected retrospectively using two databases and compared to the number of CRX that would have been performed with the daily routine strategy in MV pts. Costs savings were calculated by the product of CRX saved and reimbursement fee and the estimated variable cost of each portable CRX. Pts' characteristics, severity of illness, duration of ventilation, outcome and ratios of observed to expected (O/E) mortality were compared with those of 1995 (before change in practice) and a regional database (Cub-Réa) to evaluate a potential impact on quality of care. 

In a 32-bed intensive care, we started a prospective policy of blood tests ordering reduction to learn if this would trigger a global cut down in physician-induced-costs without affecting clinical outcomes. Simultaneously, we determined over 18 months the financial costs of medical prescriptions (blood tests, radiology investigations, drug therapy). Statistical comparisons were performed with a Mann-Whitney test.

We achieved within 11 months a 34.9% reduction in blood tests prescriptions, and a 21.1% decrease in drugs prescription, a 44.9% reduction in anti-infection drugs, a 19.4% reduction of radiology investigations, and this at constant admission rates (n=127.6 ± 5.7 / month before vs n=121.9 ± 13.1 / month after intervention; p=0.6) and with no increase in mortality. This allowed for savings evaluated at over 900 000 Euros per year of activity with no additional costs. 

A daily-routine chest radiograph (CXR) strategy is recommended by the ACR and practiced in many intensive care units (ICU). Its efficacy is controversial. Diagnostic and therapeutic efficacies, as well as costs, of daily routine CXRs were evaluated and compared with those of clinically indicated CXRs ("on demand").

In this prospective blinded controlled study, daily routine CXRs were obtained from all patients in a mixed surgical-medical ICU for one year. CXRs were evaluated by trained radiologists (to score for predefined items like progressive or new infiltrates, pneumothorax, malposition of tube/lines) and were not accessible for intensivists. In addition to these daily routine CXRs, the intensivist ordered on demand CXRs using a specific form indicating the reason for CXR and suspected abnormalities. Considerable worsening according to predefined criteria on the routine CXR, but not clinically recognized, was communicated with the intensivist. From this data, diagnostic efficacy (the number (#) of CXRs with significant abnormalities/total # CXRs) and therapeutic efficacy (# CXRs leading to an intervention/total # CXRs) were calculated. χ 2 -analysis was used to test differences.

During 12 months, 2816 CXRs in 587 patients were obtained (1890 routine CXRs and 926 "on demand" CXRs). Mean age was 66 ± 16.6. Diagnostic efficacies of daily routine CXRs and "on demand" CXRs were 4.5%, and 17.7% (P<0.0001). Most frequent unsuspected abnormalities observed on routine CXRs were pneumothorax (0.6%) and tubemalposition (0.7%). Most frequent abnormalities of the "on demand" CXRs were infiltrate (5%) and tubemalposition (3.5%). Therapeutic efficacies of daily routine CXRs and "on demand" CXRs were 1.9% and 17.3% (P<0.0001). Most frequent interventions based on routine CXRs and "on demand" CXRs were a change in medication (1.7% and 4.5%) and repositioning of the tube (0.5% and 2.9%). Changing from a routine to an on-demand strategy would have resulted in a reduction of 36% of CXRs (EUR 85.000/yr).

The value of the daily routine CXR is low. Based on these data, daily routine CXRs should be abandoned in the ICU.

Tesinsky P 1 , Svanda J 1 , Cepela P 1 , Dvorakova H 1 1 Dept. of Medicine 2, Charles University Hospital, Prague, Czech Republic

Unnecessary tests, inefficient ordering practices, and collection of more blood than it is required for testing may contribute to iatrogenic anemia in hospitalized patients. Aim of this study was to evaluate the extense of blood losses associated with sampling procedures in a 3 months period in a University Hospital medical ICU setting.

118 consecutive ICU patients (61 males, 57 females, mean age 56.7 years) were enrolled during the period of 3 months. The amount of blood taken for laboratory tests from every patient was monitored during this time. Total hemoglobin levels on admission and at the end of ICU stay were estimated.

Average ICU stay was 5.6 days . Blood sample volume on admission was 36.4 ml on average. Average daily blood withdrawal was 14.6 ml (0-95). Average blood loss during ICU stay was 81.7ml. The most excessive blood loss in a patient due to sampling were 590ml during the period of 39 days of ICU stay. Hemoglobin level on admission was 115.3 g/L, and it was 112.6g/L on discharge, respectively (difference n.s.).

1. Blood sampling during ICU stay did not result into significant blood losses within the study group. 2. The highest blood losses appeared in septic patients and they were associated with frequent repeated blood culture samples. 3. We implicate that patients with prolonged ICU stay requiring frequent monitoring are at risk of iatrogenous anemia due to blood samplings. Adequate ordering practice and appropriate sample sizes should be focused on. 

The Royal Liverpool University Hospital is a 1064 bed hospital based on 2 sites which serves an inner city population and is a tertiary referral centre for specialised services. There are 3 critical care areas: a 13 bed ICU, 4 bed high dependency unit (HDU) and a 4 bed post operative critical care unit (POCCU). On 26 January 2006 intensive care teams performed a census of levels of care, as defined by the Intensive Care Society, [1] for patients in every bed in the hospital.

On the wards 25 patients required level two care, 267 patients required level 1, 20 were deemed inappropriate for escalation of therapy but did not have DNAR orders. In Critical Care, the 13 bed ICU had 4 level 3 patients, and 9 level 2 and 1. The 4 bed HDU had 4 level 2 patients, 4 bed POCCU 2 level 3 patients and 2 level 1 patients. 

Patients suffering from burns are at increased risk of infection, particularly those of the burn wound and urinary tract. These patients are often obligated to stool in the bed, increasing the likelihood that their wounds and urinary catheters will become contaminated with pathogens from the fecal stream. Despite its impact in terms of costs and adverse consequences, there is little published research on techniques of gastrointestinal (GI) waste management techniques. The purpose of this study was to examine the clinical utility and economic impact of introducing a proactive bowel management program using a bowel management system (BMS).

The study was conducted as a Before/After trial in our burn unit. Patients receiving 5±the BMS (n=38) were matched with Control patients (n=38) on gender, age (Years), total body surface area (TBSA) burned, length of hospital stay, length of mechanical ventilation, and the burn location. Urinary tract infections (UTIs) and skin/soft tissue infections (SSTIs) were diagnosed using the National Nosocomial Infection Surveillance (NNIS) criteria by an independent infection control practitioner. Unscheduled dressing changes, and the resources consumed therein, were estimated based on staff interviews and a review of patient charts. Costs associated with the development of infections were estimated using published literature. Assuming each patient had at least one unscheduled dressing change per day for half of their stay, the average costs were $1390.00 and $5497.00 in the BMS and control groups respectively. Sensitivity analyses determined that in order for the BMS to lose this dominance, the infection rates would have to be equal, and the cost of an unscheduled dressing change less than $6.10.

In patients obligated to stool in bed, a proactive bowel management program using the BMS was effective in preventing both UTIs and SSTIs and cost substantially less than the standard reactive practice. 

The antibiotic use rationale influences various aspects of quality of care, and not only impacts microbiologically, but also economically. We believe the interaction of the ICU manager, the intensive care staff and the infectologist on an antibiotic use management program can reduce costs and improve patient safety and quality of the intensive care.

We implemented an antibiotic use program on the third trimester of 2004, based on the best practices regarding empiric therapy, guided by the sensitivity of most prevalent germs on our ICU on 2004, period of treatment, and alteration of antibiotics according to culture results within 48 to 72 hours. The program also included the infectologist working hours at the hospital increase, raising the availability for training the ICU staff on nosocomial infection issues, and monthly presentation of the epidemiological surveillance data to the ICU staff leaderships. We checked the antibiotic use throughout 2004 and 2005, comparing one to each other and to the NNISS publications results.

We observed an average antibiotic use drop in 2005 compared to 2004 of: carbapenems 57 to 12 daily doses/1000 patient-days (79%); piperacilin/tazobactan from 104 to 52 daily doses/1000 patient-days (50%); vancomicin 25 to 8 daily doses/1000 patient-days (68%); fluoroquinolones from 89 to 37 daily doses/1000 patient-days (58.4%); aztreonam from 31 to 12 daily doses/1000 patient-days (61.4%).

A multidisciplinary antibiotic use management program was able to massively impact the amount of daily doses consumed at the ICU, aligning the previous medical culture to current best practices, increasing patient safety by reducing the risk of drug resistance emergency and unwanted side effects, and also, drastically reducing costs. 

Resizing of intensive care services has become increasingly common. The purpose of this work was to examine the utility of discrete event simulation modelling to estimate the need of intensive care resources in a health region that moved from 3 to 2 general ICUs.

Two models representing each of the two restructured ICUs were built using a commercial general-purpose simulation software (ARENA). Controllable parameters of the models included number of ICU beds, number of holding beds, proportion of ICU and HDU patients, proportion of emergency and elective patients as well as number of other parameters. The model itself represents a flow of patients, arriving to ICU and occupying a bed for a random length of stay defined by a probability distribution derived from actual patient records from 2004, as well as the logic for prioritising patients in case all ICU beds are full. The model was validated using expert knowledge and comparisons to historical data.

The two conflicting goals for the organization of the ICUs were to minimize the number of transferred patients (no more than 5%) while maximizing bed occupancy. The simulation model was used to test different combinations of ICU-beds and holding beds (Table) . Additional experiments were done in order to determine required ICU capacity for several special cases. First, if number of patients would increase by 20% the model showed that the current capacities at both ICUs were enough to meet the 5% transfer limit. Second, experiments were done by increasing number of specific patients (postoperative CABG) as well as decreasing the length of stay of the entire cohort. 

Variability in intensive care patient process with respect to resource use or outcome has been shown for specific subgroups. But information is limited whether such variability also exists in general samples of ICU's, and to what extent organisational characteristics or staffing are associated to it. Primary aim of this study is to test if there is wide variability in resource use and outcome in a recently collected dataset [1, 2] . Secondary aims are to test whether outcome and resource use are related to ICU structure and process, and to explore factors associated to efficient resource use.

METHODS. Secondary analysis of SAPS 3 database. Outcome was estimated using standardized mortality rate (SMR). ICU length of stay was used as surrogate for resource use. Standardized resource use (SRU) was calculated with adjustment for severity of acute illness. Each unit was assigned to one of 4 groups: 1) "most efficient": SMR and SRU < median of all units; 2) "least efficient": SMR and SRU >median; 3) "overachieving": SMR < median, SRU > median; 4) "underachieving": SMR > median, SRU < median. Univariate analysis and stepwise logistic regression were used to test for factors of structure and process that might distinguish "most efficient" from "least efficient" ICU's. 

The suicide bombings of July 7th 2005 stretched the resources of 5 hospitals across London. The critically injured casualties were responsible for the bulk of the logistic load in the days that followed the attacks. This abstract discusses the intensive care workload associated with the management of those casualties.

A retrospective review of the 7th July Major Incident case notes and the Intensive Care National Audit and Research Centre (ICNARC) database.

There were 700 casualties in total with 52 immediate fatalities. Following initial assessment at the 5 receiving emergency departments 103 casualties required hospital admission. Of these 20 patients were critically unwell and admitted to intensive care units. There were only 3 subsequent in-hospital deaths with one death occuring before admission to ICU was possible. The 20 critically injured patients accounted for a total of 209.28 days of ICU admission (Mean Length of Stay = 10.46 days, Mean APACHE II = 12.74, Mean SAPS II = 31.68). Injury Severity Score data were sought but were only available for the 7 ICU admissions at the Royal London Hospital. All of the critically injured patients arrived within the same two-hour period and both intra and inter-hospital transfers were required to accommodate them.

The logistic load arising from a major incident is predominantly associated with the identification and treatment of the critically injured casualties and the high rate of arrival of these patients. Small increases in the fraction of patients who are critically injured will have a significant impact on the ability of receiving hospitals to cope. Objectively London coped well during the July 7th suicide bomb attacks, however, to be able to manage higher casualty numbers we must better exploit the spare capacity that exists across the region, reinforce inter-hospital communication and move towards coordinating the response of London hospitals so that they function as well as a network as they did in isolation on 7th July 2005.

Freire M D i C 1 , Knibel M F 1 , Leite C T C 1 , Mota A M 1 , Malvão L 1 1 Intensive Care Unit, Hospital Cardio Trauma Ipanema, Rio de Janeiro, Brazil

Evidence-based medicine has been a tool to improve quality of care of ICU patients and to achieve better outcomes. We believe we could also reduce costs related to drugs use, with best medical practice, evidence-based management of drug choice.

We promoted several meetings with the medical staff from January to April 2005, discussing ten issues related to drugs choice and costs. We established evidence-based protocols that were implemented throughout the first semester. In the year of 2005, we checked the use, cost and profitability of drugs involved in 4 established protocols: enoxaparin, non-fractioned heparin, noradrenalin, dopamine, ranitidine, proton pump inhibitors and midazolan by the Medicine/Day Index (MDI), obtained by the following formula: sum of costs in US dollars of the drugs above, divided by the number of patient-days on the semester. We also obtained the APACHE II Score and the mortality at the ICU.

After the educative meetings and motivating the staff day by day, we reached 97.5% of staff adherence to the protocols and observed a reduction of MDI, on the second semester of 23.37%. There was also an increase on the profitability with the new protocols from 2.7 to 5.0 times the cost. These changes represented a gain in the second semester of US$ 50,540.00. The mean APACHE II Score on the first and second semesters were 11.9 and 8.4, and the mortality rate on the ICU, similar to the risk estimated by the APACHE II, was 6.34 and 5.51%, respectively.

Over the audit period, there were 55 admissions and 56 discharges with a mean length of stay (LOS) of 7.2 days and a median LOS of 3.2days. Occupancy in ICU was 93.7% (9794 hours) during the audit. ICU beds were occupied by level 3 patients for 73.9% of the audit period (7723 hours). In comparison, bed occupancy by patients classified as level 2, 1 or awaiting discharge was 19.8%. 332 bed hours were lost due to critical care beds being occupied by a patients waiting for discharge. This equates to having a bed closed for nearly 2 weeks. In addition, there were two occasions where the unit's capacity was exceeded by 32 hours.

Patient flow analysis is a useful tool not only to measure unmet need but also to highlight inefficient utilisation of limited resources. It has prompted review of discharge planning and level 2 bed capacity. ( 

The uncontrolled structure of antibiotic purchasing and unrestricted usage of antibiotics in the Hospitals in Russia has induced the growth of resistance in the hospital microbial flora in the Intensive Care Unite (ICU).

The methods for solving the problems of quality clinical practice in antibiotic-usage were: 1) system analysis, 2) organizing methods,3) education, 4) monitoring.

1. The proportion of beta-lactam antibiotics in the total antibiotics purchase was reduced from 88% to 63%. 2. The system of control of the antibiotic prescriptions and antibiotic therapy reduced the number of incorrect prescriptions (p<0.01). 3. Implementation of protocols brought about a reduction in the length of antimicrobial-therapy courses (in days) in the ICU of the neurosurgery department (p<0.05). 

The system approach to the organizing of the antibiotic-usage has produced an improvement in the rational consumption of the antibiotics at the City Hospital. The improvement was accompanied by the changes in the antibiotics purchasing structure after 2 years of implementation the microsystem of clinical governance of the rational antibiotics usage. A substantial reduction in incorrect prescriptions was very important for improving the quality treatment and drugs expenditure at the ICU of the neurosurgical department. Databases for the antibiotic prescriptions may be the great breakthrough for prophylaxis of microbial resistance in the ICU. 

Absolute or relative adrenal insufficiencies occur in half of patients with septic shock. Absolute adrenal insufficiency is defined by an unstimulated plasma cortisol level (PCL) < 15 µg/dl. Relative adrenal insufficiency is defined by a maximal increment of PCL after a 250 µg ACTH stimulation < 9 µg/dl. The maximal increment of PCL is the largest difference between PCL 30 and 60 minutes after a 250 µg ACTH stimulation (T30 and T60) and PCL just before ACTH stimulation (T0). The diagnosis of adrenal insufficiency has important therapeutic relevance: a 5-to 7-day period of corticosteroid administration at low dose is associated with earlier shock reversal at 8 days, and a reduction in mortality at day 28. We hypothetized that T30 was of limited interest in the diagnosis of relative adrenal insufficiency in septic shock patients.

During a 3-year period, changes in PCL resulting from a 250 µg ACTH stimulation test were retrospectively reviewed in a series of patients with septic shock. We evaluated relative adrenal insufficiency incidence by taking into account T0, T30 and T60. We then re-evaluated relative adrenal insufficiency incidence by taking into account only T0 and T60. Campaign (SSC) resuscitation and management bundles in university-based teaching hospitals. Concern remains among non-teaching community-based hospitals regarding the resources and organization needed to overcome perceived impediments to achieving the SSC initiative. The purpose of the current study was to 1) Describe the implementation of a sepsis bundle in a non-teaching community-based hospital 2) Identify process issues that hinder compliance with the bundle and 3) Describe the effect of implementation on patient outcomes.

Pre-printed order sets with clinical prompts were developed. "Emergent phase" orders emphasize early recognition, volume resuscitation, antibiotic administration, and intensivist notification. "Critical care" phase orders emphasized early goal-directed therapy (EGDT), adrenal testing, Xigris candidacy, and compliance with ARDSNet and must be evaluated and renewed every 12 hours for the first 48 hours. Presept catheters (Edwards Lifescience, Irvine, Ca.) were placed in all patients by a 24/7 in-house intensivist who was also responsible for care of the patient in the ICU. Nursing and physician staff was educated regarding the initiative. QI data was collected using a computerized database developed for the project. Non-paired variables were compared with the Mann-Whitney test, proportions with Fisher's exact test.

. N, Pre=48. N, Post=50. Baseline age (72±15, 69±15), Apache2 (23±7, 21±8), and # of organ dysfunctions (4±2, 4±2) were similar. Strong trends toward earlier identification and line insertion with less deviation (6.5±9.1, 5.42±6.5h) and number of patients achieving all 3 EGDT endpoints-MAP, CVP, and SCVO2 (77%, 86%) were seen. The amount of volume resucitation at 2h (15±16, 17±17cc/kg) and 24h (5.4±2.9, 5.1±3.4L), pressor days (1.8±2, 1.5±6), ventilator days (4.3±5.5, 3.2±7.7), ICU days (6.1±7.3, 4.7±7.9), and mortality (42%, 48%) were not statistically different.

Despite no statistically significant improvements in promptness of interventions or resource utilization, positive trends resulted. Data collection for QI highlighted multiple areas in need of improvement underscoring its importance to the ultimate success of the sepsis initiative. Development of a sepsis bundle, implementation, and subsequent data collection is feasible in a non-teaching community-based hospital. 

The objective of this study is to determine the mortality of the cirrhosis patients admitted to the intensive care unit and to identify specific predictors of mortality.

Retrospective analysis of patients records (2000 to 2005) . Setting: 4 medical-surgical ICU. Liver cirrhosis was defined histologically or was based on clinical, imaging, and laboratory findings. Measurements: demographics data, disease severity indices and clinical features on admission: Coma (Glasgow coma score <6), hypotension (< 90 mmHg), heart rate, laboratory findings, need of Swan-Ganz, inotropic drugs, dialysis, transfusions and need for mechanical ventilation. The primary analysis compared hospital cirrhotic survivors with nonsurvivors. Student's test was used to compare means, chi-square test for proportions. The outcomes for these cirrhotic patients were compared with those for matched control subjects without cirrhosis, controlling the level of severity (MPM24 +/-5), gender and age (+/-5).

. 184 cirrhotic patients were admitted to the ICU, over 8202 entered patients in the study period, represented 2.2% of the total population, 71.2% men, average age of 59, average stay of 6.8 d, mortality in ICU of 51.6% and the noncirrhotics of 19% (p<0.001). Cirrhotic survivors vs. nonsurvivors: Apache III, MPMO, coma, hypotension, need of inotropic use, arterial line, mechanical ventilation, renal replacement therapy, transfusion of platelets or plasma, parenteral nutrition, bilirrubin and infection were associated with an increase in mortality using univariate analysis. On multivariate analysis the mechanical ventilation OR 3.4 (95% CI 1.1-9.9), APACHEIII OR 1.04 95% CI 1.02-1.05) and the need of inotropic drug OR 8.9 (95% CI 2.7-29) emerged as independent predictors of mortality. Match-control study: There were no significant differences in the mortality observed in the hospital for the cases and the control subjects (56% vs. 49%, p=0.22, 95% CI -4% to 18%) or the ICU stay (7.6 Vs 9.6, p=0.06, 95% CI -0.7 to 4.1).

The cirrhotic patients admitted to the ICU have a poor prognosis, but is not different of the prognosis of patients with the same level of severity. The APACHEIII, the need of inotropics drugs and the mechanical ventilation were independent predicting factors of mortality in cirrhotic patients admitted in the ICU.

Grant acknowledgement. to all my colaborators.

Mosca R 1 , Davies I 1 , Linter S 1 , Lener G 1 1 Sir Humphry Davy Department of Anaesthesia, Bristol Royal Infirmary, Bristol, United Kingdom

Tracheostomy (TS) is the method of choice for long term airway control in critically ill patients. In cardiac surgery the percutaneous dilatational tracheostomy (PDT) has been preferred to surgical TS for reduction in associated risk of mediastinitis.(1) PDT can be performed promptly in ICU once indicated and few personnel are needed. (2) METHODS. We retrospectively analysed 137 consecutive patients (mean age 65.4 years) who received TS in our CICU following cardiac surgery from 1996 to 2005. We divided between early (eTS) and late (lTS) cases respectively performed before and after post-operative day 7. We recorded indication and timing of PDT, in-CICU and in-hospital length of stay (LOS), in-hospital mortality and long term survival (Office of National Statistics ONS).

The majority of cases (53.8%) received a lTS because of failure to wean from ventilatory support for one or more complications including sepsis (55.5%), respiratory failure (49%), cardiogenic shock (26.3%), brain injury (20.4%) or others (16.3%). Of 137 cases, 45 (32.8%) received TS without a period of extubation and 92 (67.2%) had required re-intubation. The majority (54.2%) were reintubated on or by post-operative day 6 and 64% of TS occurred within 4 days of re-intubation. In our records were no sternal wound complications related to PDT. The mean LOS in CICU increased from 18.7 days for not re-intubated eTS to 30.5 days for reintubated lTS. The mean in-hospital LOS was 29.8 days for not re-intubated lTS and 44.9 days for re-intubated lTS. The only statistically significant difference in in-hospital mortality rates was between not re-intubated eTS and lTS (8% vs.37%)(p=0.037; Poisson rate test). Long term survival analysis supplemented with the ONS showed the rate of 40.9% deaths (56/137) in the nine years period.

CONCLUSION. PDT can be safely performed in post cardiac surgical patients without any procedure related mortality or significant morbidity. Our data are supportive of earlier intervention with PDT, particularly in the reintubation group. 

Severe sepsis is associated with high morbidity and high mortality, and represents a substantial health care burden in the United States and European. However, little data on the epidemiology of severe sepis in developing countries is available. The purpose of this study was to investigate the epidemiology of severe sepsis in surgical intensive care unit in chinese multiple academic medical centers. 

The optimal management of severe sepsis and septic shock depends on rapid recognition and aggressive initial management. On this basis, the optimal hospitalization orientation appears to be a crucial element. The aim of this study is to evaluate if the prehospital recognition of the gravity of patients exhibiting a septic pathology influence the place of hospitalization (ICU or emergencies departments).

A prospective, multicenter, observational study included 25320 prehospital medical interventions. In this cohort, we identified 3 groups of patients according to the criteria of the Chest Physicians/SCCM (group 1, G1, sepsis, n=111, group 2, G2, severe sepsis, n=18 and group 3, G3, septic shock, n=34). The main characteristics of the 3 groups (age, sex, workload, length of prehospital medicalisation (LPM) and the Clinical Classification prehospital Emergency Patients (CCMS)) were collected and compared using the Mann-Whitney test and the chi-square test. We also studied the relationship between the place of hospitalization and the 3 groups. Finally, to evaluate the effect of the density of ICU beds on the place of hospitalization, the same relationship was studied distinguishing the urban part of our state (high density of ICU beds) and the rural part of our state (low density of ICU beds).

Mean age was 60±6 years with majority of men (52%), LPM was 63±20 minutes and workload 40 PRN points. If age and sex were not different between the three groups, LPM, workload and CCMS statistically increased with the gravity (p<10e-4) from G1 to G3. No relationship was found between the three groups and the admission in ICU or emergencies departments. However, the density of ICU beds clearly influence the place of hospitalization while the relationship between first the three groups and second the admission in ICU or emergencies departments was significant in the urban part of our state (high density of ICU beds) (p=0.006) but not in the rural part (low density of ICU beds) (p=0.658).

The initial prehospital recognition of the gravity of patients exhibiting a septic pathology is not the guaranty of the optimal hospitalization orientation while this admission place is influenced by the density of ICU beds. Taking into account the weak probability of an increase in ICU beds, those results highlight the need for a better collaboration between emergencies departments and ICU to improve the initial management of severe sepsis and septic shock.

Renal replacement therapy (RRT) carries a significant burden in mortality, cost, and delivery 1 . The aim of this study was to determine overall survival after RRT whilst in intensive care (ICU), to identify the patient sub-groups who survive, and to determine what proportion of these patients will require long term RRT.

We performed a retrospective analysis of all patients admitted to our 16 bedded, mixed (excluding Cardiothoracics) adult ICU, over a 2-year period. Our hospital is unique in being the only centre within a 30-mile radius to offer RRT at an ICU level. The data was collected from Ward Watcher software and renal registry databases. RRT was defined as the need for haemodialysis or haemofiltration whilst on intensive care. 

Delirium in ICU is common and is associated with adverse outcome; it may be potentiated by medication-induced coma. Protocols are believed to result in improved patient care and better outcome. We evaluated the impact of a protocolized approach to sedation and analgesia in the ICU on the incidence of delirium and on the associated outcome.

The study was conducted by reviewing medication use, incidence of delirium and patient outcome in a single tertiary care mixed ICU during two periods: 'pre' (Aug 2003 -Feb 2004 and 'post' (Apr-Nov 2005) . There were no sedation/analgesia protocols in use during the 'pre' period, but there was during the 'post' period. The protocols required nurses to assess and document levels of analgesia, sedation & anxiety, and presence of clinical delirium features. Protocolized interventions included non-pharmacological approaches as well as pharmacological interventions for analgesia, sedation and delirium that were based on validated assessment tools.

. 537 patients were evaluated in the 'pre', and 559 in the 'post', cohort (after exclusion of comatose patients). APACHE scores were higher in the 'post' group (17.4 vs. 16.2, p=0.0057). All patients in the 'post' cohort were assessed as above, but while 57.4% had sedation/analgesic managed by protocol 42.6% did not. The incidence of medication-induced coma, ICU & hospital length of stay (LOS), and percent discharged to dependent care was significantly lower in the 'post' cohort. The 'post' cohort received significantly less opiates, had better analgesia and comparable levels of sedation and anxiety ( Table 1 ). The incidence of delirium was similar (35.2% pre vs. 33.8% post, p=ns), more patients remained cognitively intact (33% pre vs. 42% post; p<0.0001). Although the institution of the protocolized care was associated with significantly improved patient outcome, it was more significant in those patients in the 'post' cohort who were not managed with a pharmacologic protocol.

Protocolized nursing assessments of pain, sedation and delirium were associated with improved short-term and long-term outcome. However, medication administration that was not protocol-driven resulted in even better outcome. These data suggests that it is individualization of care, and not protocolization, which accounts for improved outcomes.

Grant acknowledgement. CICF, FRSQ. 

Patients with chronic liver disease who are admitted to intensive care have a poor prognosis, with a mortality rate between 40 to 92% (1, 2) . There is little data on the outcome of this group of patients in the UK. The aim of this observational study was to determine the ICU, hospital and six-month mortality for patients in a UK hepatology centre and identify any factors that could influence decisions on patient management.

Between 2000-2004, 86 patients with cirrhosis were admitted to the ICU. The aetiology of the cirrhosis, primary reason for admission, APACHE II score and Child Pugh score and their variables at ICU admission were recorded, as well as the level of organ support required during the ICU stay; and statistically analysed against the mortality data.

Overall ICU mortality was 57%, hospital mortality 74% and six-month mortality 78%. There were significant differences in APACHE II scores between survivors and non survivors (p=0.001) and in admission serum creatinine levels (p=0.003). There was no difference in Child Pugh scores between the two groups. The single strongest predictor of mortality during the ICU stay was the use of inotropes (odds ratio 25.7, 95% confidence interval 8. 2 -81) . Of the 49 patients prescribed inotropes, 42 died in ICU (86%) while only 7 survived.

The mortality in this group of patients admitted to ICU remains high. APACHE II score, and not Child Pugh score predicts mortality in this group of patients. Other significant predictors of mortality include admission serum creatinine and the use of inotropes during the ICU stay. 

Thomas M J C 1 , Gould T 1 , Napier S J 1 1 Intensive Care Unit, Bristol Royal Infirmary, Bristol, United Kingdom

Atrial fibrillation is seen commonly on intensive care. After electrolyte correction, amiodarone is the treatment of choice on our unit. The study identified which of our patients are given amiodarone, their outcome and if they remain on amiodarone on discharge from hospital.

We performed a retrospective chart review of the 446 patients admitted to a University Teaching Hospital Intensive Care Unit during a 6 month period and recorded patients who were commenced on amiodarone. We used the ITU database to review diagnosis, apache data and outcome of the 446 patients and the discharge medication of survivors.

. 80/446 (18%) patients required amiodarone. Of these 51 (64%) were surgical patients and 29 (36%) medical. All medical patients were emergency admissions. Of the surgical patients 31 (60%) were general surgical, 6 (12%) vascular and 14 (36%) other, made up of gynaecological, maxillo-facial and thoracic. 17 (33%) had elective surgery and 34 (67%) had emergency surgery. Of the 80 patients requiring amiodarone, 21 died whilst on intensive care and 16 died prior to hospital discharge. This gives an ITU mortality of 26% and an overall hospital mortality of 37/80 (46%) which is significantly different from the hospital mortality of patients admitted to the unit not requiring amiodarone during that period 49/385 (13%) p<0.01 chi square. The patients that required amiodarone did have a higher APACHE II score than the unit average . Of the 43 patients discharged from hospital alive 37 (87%) were on no anti-arrhythmic drugs, 4 remained on amiodarone and 2 had been converted to digoxin.

In this time period the incidence of atrial fibrillation requiring amiodarone on our unit is 18%. The patients most likely require amiodarone were emergency surgical patients. If patients require amiodarone their mortality is significantly increased. The rhythm disturbance is usually short lived with only 6 patients requiring treatment on discharge. This is likely to be because the atrial fibrillation is a consequence of critical illness (eg hypovoalemia, acidosis and sepsis) as opposed to a primary cardiac pathology. This study has allowed us to calculate our incidence of AF and identify patients at risk. It also shows it is a predictor of increased hospital mortality on our unit (46%). However we were reassured to see that drugs commenced on intensive care are modified prior to hospital discharge.

Dolling M 1 , Hall N 1 , Sadler J 1 , Reeseh H 1 , Parker J 1 1 Department of Intensive Care Medicine, Leicester Royal Infirmary, Leicester, United Kingdom

Haematological malignancy patients on ICU have a poor outcome1,2. This is a retrospective observational study of haematological patients admitted to our ICU over seven years, to determine their outcomes and how they differed from that suggested by the literature.

Data were collected for patients with a haematological malignancy admitted to ICU between the years 1999 and 2005. Medical records for these patients were obtained, and demographic, clinical, APACHE II and outcome data were recorded.

RESULTS. 93 of 106 medical records were available. Median age was 55 years and median APACHE II score of 27. 43 (47.3%) survived to ward discharge. Only 27 (32.5%) survived to leave hospital. Admission diagnoses were severe sepsis in 38 (45.8%) patients and respiratory failure in 36 (43.4%). 50% required invasive ventilation with equal numbers remaining spontaneously ventilating or receiving mask CPAP. Survival in the invasive ventilation 'as the only organ supported' group was 5/9 (55%) which was poorer than if self-ventilating with no other organ support ((7/8 (87.5%)). The requirement for inotropic support in addition to invasive ventilation further reduced survival (7/24 (29.1%)). INTRODUCTION. Immunodeficient patients may be at risk of adverse drug event (ADE) because they cumulate toxicity of cytotoxics and of many others drugs. Immunosuppressants have short therapeutic index when few therapeutic alternatives are available (1) . The aim of this study was to assess characteristics and frequency of serious ADEs in immunodeficient patients at the admission to ICU.

We prospectively included all adult patients admitted in a medical intensive care unit during a 6-month period. Immunodeficiency was defined as primary or acquired immunodeficiency. For each drug taken in the previous month and each organ failure, a putative ADE was suspected and a questionnaire systematically filled. Each patient was followed-up to determine if serious ADE contributed to organ (s) failure (s) requiring admission. Follow-up assessed morbidity, mortality, workload for personnel, length of stay, and outcome 28 days after ICU discharge. Then an expert panel estimated the strength of relationship and preventability between drug (s) and potential adverse events by imputation method.

Among the 405 patients admitted from may to october 2003, 79 (19.5%) were immunodeficient. Among them, 35 have presented at least one ADE related organ failure and 23% were hospitalized because of an ADE. In 68% of cases, ADEs were predictable, 37% were potentially preventable, 12.5% were rare, and 21% contributed to death. Mortality, workload for personnel, and length of stay were the same in immunodeficients patients with and without ADE. 71 were still hospitalized 28 days after ICU discharge. Metabolic ADEs linked to glucocorticoids, cyclosporin and immunoglobulins were mostly preventable.

CONCLUSION. ADEs involving critically ill immunodeficient patients were common and potentially life-threatening (2) . Although many of the involved drugs have proven benefit, measures should be put into place to minimize ADE and thereby further improve diagnosis of preventable ADEs. ESS type I or low T3 syndrome, ESS type II or low T3 and T4 syndrome (some authors include low TSH too) and ESS type II with elevated T4. Our aim in this study was to analyse the t.a. in a cohort of patients from our ICU, looking for its prognostic value in relation to mortality.

A descriptive study of 108 patients with medical, surgical and coronary diseases. Thyroid Function (T.F.) (FT3, FT4, rT3 and TSH blood concentrations) were determined at admission, 3th and 7th day of stay. We made an univariate analysis comparing the T.F. between survivors and non survivors. We analysed the Prevalence of ESS in our cohort of patients. We made a logistic regression as multivariate analysis being dependent variable death (exitus) and as independent ones age, gender, APACHE II score and the T.F. above mentioned. For statistical significant hormonal variables in the multivariate, a ROC curve was drawn in respect of mortality, calculating his area below the curve and the optimum cut-off point (o.p.). Then the Sensitity (S) and Specificity (1-S) were determined.

The univariate showed significant differences of T.F. mainly on 3 th day of stay between patients who died and those who survived. 

Recently, increased concentrations of cell-free plasma DNA have been found in various clinical conditions, including trauma, cancer, stroke, and myocardial infarction. It is likely that DNA enters the circulation following cell death. The aim of this study was to investigate the kinetics of the cell-free plasma DNA and its' association with hospital mortality and the degree of organ dysfunction in intensive care unit patients.

We examined the cell-free plasma DNA from 229 ICU pts' sequential blood samples. The first sample was drawn as soon as possible after the patient was admitted to the ICU, the second in the following morning and the third 48 hours after the second sample. Plasma DNA was measured by real-time quantitative PCR assay for the b-globin gene. The pts' clinical data consisted of e.g. age, gender, SOFA and APACHE scores, and mortality data.

The concentrations of the plasma DNA did not differ from each other in sequential samples of a study pt (p=0.874). The first day and the maximum plasma DNA concentrations correlated with APACHE II points significantly (p=.001, p=.01) and the first day plasma DNA concentration correlated also with the first day SOFA (p=.001) and the maximum SOFA during the ICU stay (p=.03). The maximum DNA concentration was significantly lower in hospital survivors than in non-survivors (median 6604 vs. 11346 GE/ml, p=.01). However, the area under the receiver operating curve regarding hospital mortality did not show good discriminative power for plasma DNA concentrations (AUC 0.56-0.62).

Although the maximum plasma DNA concentration was significantly lower in hospital survivors than in non-survivors among the critically ill, we would like to conclude that this laboratory parameter does not improve outcome prediction, at least in this kind of a non-selected patient population. 

There is evidence that provision of more care by registered nurses is associated with decreased complication rates [1] . In the UK, recommended nurse: patient ratio for level 2 patients is 1:2 [3] ; however there is no agreed standard for levels of staffing or skill mix on acute wards. We took part in a census of patient dependency and ward staffing, on one day in our hospital, to help assess whether wards were equipped to care for critically ill patients.

Hospital is a 1064 bed hospital which serves an inner city population and as a tertiary referral centre for specialised services. There are 3 critical care areas with 21 designated levels 2 and 3 beds. A census of levels of care (as defined by the Intensive Care Society [4] ) was performed on 1 day in January 2006. Intensive care teams visited each ward and assessed levels of care of each patient in each bed; in addition the numbers of trained and untrained nurses caring for these patients was recorded.

. 25 patients needed level 2 support on general wards; none of them were staffed to achieve 1 nurse to 2 patients. The overall ratio of qualified (RN) nurses/ patient was 1: 9.74. The ratio of all nurses, qualified and unqualified was 1: 4.81. The range was large, with some wards having 1: 3 ratio, while one ward with 4 level 2, 12 level 1 and 14 other patients had 4 trained staff.

CONCLUSION. It appears that critically ill patients are cared for on wards with less than optimal staff numbers. If these patients are cared for outside ICU, staff need to receive training and there should be adequate numbers. We have found that levels of staffing and skill mix vary considerably between wards of the same hospital. Roles should be redesigned to enable staff to match their skills to the needs of their patients. 

Prolonged length of hospital stay increases the risk of hospital acquired infection and accelerates general debilitation in the elderly. The aim of this study was to identify the causes of delay in treatment and patient discharge, noting in particular the documentation of observations and its impact on treatment decisions.

Care and the Service Improvement Manager examined the notes of 51 patients discharged from Musgrove Park during March 2004 with a length of stay of more than 10 days. For each patient the following was recorded: Date of admission, reason for admission, admitting consultant, date of requests for tests/referrals, date of diagnostics/referrals completed, infection, discharge date, and exceptional events.

The study highlighted omissions in carrying out and recording observations, and enacting doctors' requests. At times we failed to listen to the requests of dying patients, made late referrals to the Palliative Care Team and sometimes admitted patients inappropriately to secondary care. There were significant delays with nutritional support. 

Heintzelmann L M 1 , Poulsen A 1 , Weinhard M 1 1 ICU, Aarhus Sygehus, Aarhus C, Denmark

In 2004 surgeons and onchologists made inquiries about brachy radiation therapy (BRT) right after advanced cancer surgery in the pelvis. Offered in an ICU allows respiraratory and haemodynamic unstabile patients to receive BRT the day after surgery. The idea is to effectuate the BRT as soon as possible to eliminate the remaining cancer cells. Before implemented in the ICU the patient had to wait up to 10 days lying flat on their back before receiving BRT at the oncological ward.

Interstitiel BRT with Pulse Dose Rate during 50 h. is used after advanced cancer surgery in the pelvis. Key-staff from ICU worked out standards for the implementation, after having visited oncological wards with brachytherpy. Also they were involved in the design of the radiation room. ICU special trained nurses were selected from the staff on a non-mandatory basis.14 ICU nurses (50 employed in total) received special training theoretical as well as practical for one day. The rest of the staff received 2 hours about radiation hygiene and must at all times carry a dosemetre to monitor potential radiation.There is one nurse per patient while the patient is admitted in the ICU. Interviews were made by nurses before and after surgery. Standard control was followed according to cancer diagnosis.

During the first 12 months 10 patients were treated with BRT after advanced cancer surgery. Severe complications caused by the BRT were not reported. The technical parts of working with the brachy technology caused only minor delays. The 50 h. therapy were completed with a delay (1-3 h.) The patient comfort has increased with the earlier effectuation of the BRT. Now they spend 7 days less lying flat on their back, which benefits the mental well being of a cancer patient and reduces complications caused by immobilisation. Interviews have reported an increase feeling of more safety and security while cared for and surveillanced by the ICU nurses.

The post-BRT interviews show an increase in comfort due to shorter time lying flat on their back and the 24 h. nursing.

Viner J E 1 1 Critical Care Unit, Torbay Hospital, Torquay, United Kingdom

Respiratory rate has been identified as the most significant indicator for deterioration (1) as has level of consciousness (2) and hypotension (3. This study evaluated compliance with the vital sign standard and identifed the most commonly deranged vital sign in the 12 hours prior to cardiac arrest from January 2004 to December 2005.

METHODS. An audit project was registered with the clinical effectiveness department to evaluate compliance with our vital signs standard and identify premonitory indicators. 412 sets of patient case notes were examined to identify real cardiac arrests, 180 were included. 448 sets of vital signs were scrutinised. All data were collected on specifically designed forms and transcribed to an excel database.

Of the total 448 sets of vital signs there were 272 (60.7%) temperatures; 409 (91.2%) pulses; 410 (91.5%); blood pressures and 374 SpO2 (83.4%). Each of these were examined for frequency and derangement. There were 18 (6.6%) instances of deranged (<35.5) temperature; 63 (15.4%) instances of deranged (<60 ->120) pulse; 157 (60.3%) instances of deranged (<10 -> 26) respirations; 84 (20.4%) instances of deranged (<100 systolic -> 100mmHg diastolic) blood pressure identified and 140 (37.4%) instances of deranged (<95%) SpO2. During the audit period respiratory rate monitoring improved and cardiac arrest rates reduced by 31%.

Respiratory rate is an early and accurate indicator of deterioration and should be measured with every set of vital signs undertaken. Within our Trust, increased teaching and the introduction of a MEWS system have increased vital sign monitoring, improved referral to the outreach team and facilitated early intervention. 

Chest pain is a common cause of seeking medical help. Although its sudden onset strongly suggests that patient has acute coronary syndrome there are many other ethiological causes, that can provoke it. On the other hand there are patients suffering ACS without typical clinical signs. The aim of this study was to establish how often atypical presentation of ACS is present in our population, comparing normoglycemic and patients with diabetes.

The study was performed in prehospital setting in the city of Maribor with population of 120000. Included were patients with afterwards hospitaly confirmed ACS, who were seeking medical help in Centre for Emergency Medicine during one year period. For each patient data were collected including main symptoms and signs, blood sugar level, vital parameters and 12-lead ECG.

198 patients with ACS were included in this study. The main symptom in 144 (73%) patients was acute chest pain, other 54 (27%) patients were having atypical clinical signs including nausea, vomiting, dyspnea, palpitations, malaise, diaphoresis and sudden colaps. When observing subgroup of 55 (27%) patients with elevated blood sugar level it was established, that atypical clinical signs were present in 32 (58%) patients.

As predicted chest pain is not always the main symptom of acute coronary sindrome, especially when diabetic patients are under medical observation. Beside chest pain we should also be cautious when patient is complaining of above mentioned atypical simptoms and signs. In such cases, especially in diabetic patients, 12-lead ECG is mandatory.

Jones C 

Diaries written during the patients stay in the intensive care unit (ICU) are becoming increasingly popular. Initially it was felt that these diaries would help patients understand their illness better and fill gaps in the patients' memory for ICU1. The impact of such diaries on psychological recovery has not been examined before. During a project comparing the psychological recovery of patients we took the opportunity to examine the influence of diaries.

The study was part of a prospective study undertaken in 5 ICUs examining the incidence of post traumatic stress disorder (PTSD). In 3 of the study ICUs some patients received diaries. Starting a diary was not randomised but done when staff had time. This was an opportunistic study of an intervention that was happening at the time of the main study. After ICU discharge the patients recall for ICU was assessed2. At three months post ICU discharge the presence of PTSD-related symptoms was measured using the PTSS-14 screening tool.

. 241 patients were recruited to the main study, with 117 at the three study centres doing diaries. Of the 117 patients, 42 received a diary. The level of PTSD-related symptoms at 3 months post ICU discharge was lower in those receiving a diary (Mann-Whitney U p = 0.04). When just those patients recalling delusional memories for ICU, e.g. nightmares, hallucinations, paranoid delusions, were examined, those receiving diaries had much lower levels of PTSD-related symptoms compared to those who did not (Mann-Whitney U, p = 0.028).

This study suggests that patients receiving an ICU diary have lower levels of PTSDrelated symptoms. The diary may facilitate the working through of traumatic memories, particularly of delusions and be acting like a natural cognitive behavioural therapy. There is a need to perform an RCT of the impact of ICU diaries on psychological recovery. CONCLUSION. DIPEx aims to identify the questions that matter to people when they are ill and is widely used to inform patients, educate healthcare professionals, and provide a patient-centred perspective to researchers and those who manage health services. The DIPEx intensive care study contains around 250 interview clips in video, audio and written formats from patients who talk about their experiences from illness to recovery, including their experiences in ICU, on a general ward and recovery at home. (1). Experiences from intensive care may affect the patients for a long time. The aim of this study was to describe patient empowerment in an intensive care situation. Patient empowerment is usually discussed from the point of view of participation in health care decisions. This study, however, emanated from a more wide perspective were even experiences of increased strength and power were reflected.

The study was based on open-ended interviews with eleven patients in two intensive care units (ICU). The interviews were analysed according to a phenomenological method (2).

All respondents related their experiences of strength and power in ICU situations to their ability to maintain their own inherent joy of life and will to fight, which was said to be of essential importance for their recovery. A positive environment that encouraged feelings of value and motivation, where the patient felt safe, received additional care and participation as he/she wished, had a positive influence. Having a human being, i.e. a next of kin or staff, close by made the patient feel safe as well as information, if it provided answers to questions or wonderings, facilitated comprehension or prepared for coming changes. It was essential for the patient to feel that someone really cared and to receive confirmation that he/she was important as a person. The patient needed to be taken seriously and be listened to, but found it natural to leave medical decisions and treatment principals to the professionals.

CONCLUSION. The importance of strengthening and stimulating the ICU patients' own inherent joy of life and will to fight emphasizes the need of personal care based on a close communication with the patient and his/her next of kin. Time has to be invested in communicating with the patient and family in order to find out what is important for each patient. 

Medical emergency teams (MET) have been established in many hospitals. The establishment is in some cases preceded by an educational effort aimed towards the staff working in the general wards. At our institution we have established a MET and all nurses in the general wards have undergone a one-day course where early detection of organ failure, initial management of organ failure and use of the modified early warning score (MEWS) in assessment of organ failure has been taught. We have then performed a survey among the nurses after the establishment of the MET. The intention was to survey the opinions about the course and about the effects of the establishment of the MET.

A questionnaire was sent to nurses who had taken the course and worked in the medical emergency ward during the first 6 months after the establishment of the MET. Questions asked included the opinion about the course in general, if they felt they had benefited from the course, if they felt a need for a follow-up course, if the MEWS was easy to use and their opinion about the effects of the establishment of the MET on the cooperation between the ICU and the general ward.

RESULTS. 82% of the nurses answered the questionnaire. All nurses who had taken the course felt the course to be useful and made them more confident when assessing patients. 87% felt a need for a follow-up course once or twive a year. 81% found MEWS to be easy to use and a useful tool. 56% used the MEWS regularly. 92% felt that the cooperation between the ICU and the ward had improved after the establishment of the MET. 100% found the establishment to be positive and wanted the project to continue.

In the survey we found that nurses highly appreciated a one-day course prior to establishment of a MET and that the cooperation between the medical emergency ward and the ICU improved after establishment. A primary purpose of establishing METs is earlier detection and treatment of organ failure and the early detection is a key factor. A close cooperation between general wards and the ICU is important in order to achieve this. We believe that providing a course similar to ours and introducing a scoring scale improve the assessment of patients and the cooperation between general wards and the ICU. We also believe the transfer of knowledge between the ICU and the ward to be an important spin-off effect of the establishment of a MET. 

Neurological deterioration often occurs in acute respiratory failure (ARF) due to hypecapnia and/or hypoxemia. Level of consciousness is considered a predictive factor for noninvasive positive pressure ventilation (NPPV) failure, and hypercapnic coma, is an absolute contraindication for this therapy.

We performed a prospective, observational study, with all the patients admitted to ICU with ARF and NPPV between 1997 to 2004, to assess the outcomes of NPPV therapy in patients with ARF, according to consciousness level, tested by Glasgow Coma Score (GCS), at the ventilatory therapy begining. Indication to NIV was presence of dyspnea, respiratory rate > 30 bpm, PaO2/FiO2 < 200, pHa < 7.35 or accessoy respiratory muscle activity. Primary goal of the study was to determinate the success of NPPV (defined as a response to therapy avoiding endotracheal intubation, surviving during the stay in ICU and at least 24 h on a medical ward), depending on GCS. Secondary goal was to identify the variables wich can predict NPPV therapy failure in these patients. Comparison between variables was made by Ji2 and Student T test. Variables related to NIV failure and mortality were analysed by multivariate analysis. 

Noninvasive positive pressure ventilation (NPPV) has been increasingly used over the past decade in the management of acute or chronic respiratory failure and weaning of mechanical ventilation. We performed this clinical study to evaluate the usefulness of NPPV in patients who developed acute respiratory failure or post-extubation respiratory failure.

We analysed thirty four patients (sixteen males and eighteen females, mean ages 58 years) who applied NPPV (BIPAP S/T, Respironics co., USA) for respiratory failure or weaning difficulty at medical intensive care unit (MICU), emergency room and general ward of a tertiary hospital. We evaluated the underlying causes of respiratory failure, duration of treatment, the degree of adaptation, complication and predictive parameters of successful outcome.

The overall success rate of NPPV was seventy-one percent. The duration of NPPV applying time, baseline blood pressure, pulse rate, respiration rate, PaO2, PaCO2, SaO2 were not different between success group and failure group. But, the baseline pH was higher in the success group. Predictors of success were higher baseline pH, patients with underlying disease of COPD, improvement of vital sign and arterial blood gas value after NPPV application. The success rate in patients with post-extubation respiratory failure was eighty percent. There were no serious complication on applying NPPV except minor complications such as facial skin erythema, abdominal distension and dry mouth.

CONCLUSION. NPPV may be effective treatment in patients with acute respiratory failure or postextubation respiratory failure in selected cases. 

Prospective and observational study, with all patients admitted in ICU with pneumonia and treated with NIV between 1997 to 2004. The indication for NIV was dyspnea, respiratory rate >30, PaO2/FiO2 <200, pHa <7.35 or activity of accessory respiratory musculature. The primary goal of the study was to determinate the success of NIV (defined as a response to therapy allowing the patient to avoid endotracheal intubation, and to survive a stay in the ICU and at least 24 hours on a medical ward).

The secondary goal of the study was to identify the variables that can predict a failure of NIV therapy and hospital mortality. The comparison between variables was realized by Ji2 Pearson and Student's T test for independent and matched data. Variables correlated to NIV failure and mortality were analyzed by logistic regression. 

Half of the patients with pneumonia present lack of response to NIV therapy. Respiratory parameters after one hour of NIV and the severity of the patients are factors correlated to worse prognosis. The use of some kind of ventilators is also a determining factor for NIV failure.

Racca F 1 , Ferreyra G 1 , Varese I 1 , Stra E 1 , Berta G 1 , Gregoretti C 1 , Appendini L 2 , Ranieri V M 1 1 Anesthesia, University of Turin, Torino, 2 Pneumology, Fond. S. Maugeri, Veruno, Italy

Although effective in providing non-invasive ventilation with better patient comfort and less side-effect, the use of the helmet to deliver pressure support ventilation (PSV) and continuous positive airway pressure (CPAP) is complicated by carbon dioxide (CO2) re-breathing.

In the present study we tested the hypothesis that turbine-based ventilators with single circuit might improve the carbon dioxide wash-out during CPAP and PSV provided by helmet. 

Our data show that the combination of a turbine-based ventilator with a standard exhalation device at the expiratory access of the helmet is more efficient than an ICU ventilator to decrease CO2 re-breathing during CPAP and PSV delivered by helmet.

Racca F 1 , Berta G 1 , Vittone F 1 , Stra E 1 , Barberis L 2 , Gregoretti C 1 , Appendini L 3 , Ranieri V M 1 1 Anesthesia, University of Turin, 2 Anesthesia, Osp. Maria Vittoria, Torino, 3 Pneumology, Fond. S. Maugeri, Veruno, Italy

In patients with acute respiratory failure (ARF) the use of a helmet to deliver pressure support ventilation (PSV) has proved to be effective in improving gas exchange with better tolerance than mask ventilation. However, use of the helmet to deliver PSV worsened patient-ventilator asynchrony, efficacy of ventilation to unload the respiratory muscles as compared to mask PSV due to the mechanical characteristics of the helmet and the peculiar algorithms governing PSV triggers. In the present study we test the hypothesis that pressure targeted Synchronized Intermittent Mandatory Ventilation (P-SIMV) optimizes mechanical assistance and efficacy of ventilation to unload the respiratory muscles during non-invasive ventilation delivered through the helmet.

Respiratory rate (RR), tidal volume (Vt), pressure-time product for the diaphragm over 1 min (PTPdi/min), the percentage of inspiratory time spent at nominal PSV level (% IA), triggering delay (the time from onset of inspiratory effort to the beginning of detectable pressurization) were measured in 6 healthy volunteers during PSV and P-SIMV delivered through a helmet. Modes of ventilation were randomly applied. All experiments were conducted with 10 cmH2O of pressure assistance and 5 cmH2O of PEEP. RR of P-SIMV were matched with that obtained during a previous mask PSV trial set at the same pressure parameters. To simulate conditions of increased ventilatory requirements a fixed resistor was placed at the airway opening. An Evita 4 ventilator (Draeger, Germany) was used.

Results are expressed in the table: 

Helmet is a recent interface for non invasive ventilation. However, its large dead space may increase the work of breathing and delay response to patient's inspiratory (1) . Little is known of the ability of the recent ventilators to pressurise this interface The aim of the study was to compare, on a lung bench test, new ventilators in terms of pressurisation, triggering and perceived noise during ventilation.

METHODS. Five ventilators were tested in Pressure Support Ventilation (PSV) using a helmet interface (Starmed, Mirandola, Italy) on a manikin: EVITA XL (Dräger, Germany), PB840 (Tyco, USA), ELISEE350 (RESMED, Savigny-le-Temple, France), AVEA (Bird, USA), SERVO I (Siemens, Germany). To simulate spontaneous ventilation we used an active lung bench test (two-chamber Michigan test lung), to simulate 3 effort levels (low, moderate, high). Noise was recorded by a noise dosimeter (Quest Noisepro DLX) with and without Heat and Moisture Exchanger (HME). The pertinent variables were the delay between effort and inspiration (Delta T), the Pressure-Time Product at 0.5 s (which reflects pressurisation), and the noise level. Results are reported as mean (+/-SD). These characteristics are compared to the same protocol via a standard face mask.

PSV evaluation was performed with 2 levels of Pressure Support (10 and 20 cmH2O) at PEEP 8 for 3 levels of effort. Main results of study are reported in the table. (1) . Acute respiratory failure (ARF) due to cardiogenic pulmonary oedema (CPO) is typical indication for NIV.

In our preliminary study we studied 8 pts with ARF due to CPO ventilated noninvasivelly with PAV using total face mask. For PAV we selected following settings: Volume assist=25 cmH2O/L, Flow Assist=4 cmH2O/L/sec,% Set=30% and FIO2=0.6. After 30 min FIO2 was reduced to 0.4. PaO2, PaCO2, pH, saturation were measured and dyspnoea was estimated before and every 30 minutes of ventilation.

PaO2 values before and during PAV are shown in the figure. Mean ventilation time was 4 hours.

In pts with ARF due to CPO noninvasive PAV is: 1) effective and well tolerated, 2) there is no need for sedation and no risk of prolonged weaning from ventilation. 

The present randomized controlled study aims to determine efficacy of NPPV and its effects upon intubation rate and mortality in ARDS.

This study was performed on patients over 18 in the respiratory intensive care unit at University Hospital. 20 patients meeting the American-European diagnostic criteria for ARDS were included in the study. The patients were respectively randomized and then divided into two groups: standard therapy group and NPPV group. Invasive mechanic ventilation indications were determined before the study was begun. Any patient meeting the criteria of intubation was immediately intubated in either group. On the other hand, those not needing immediate intubation underwent a standard medical treatment. In the NPPV group, patients no needing intubation received NPPV in addition to a medical therapy.

RESULTS. 17 patients were male. A total of 18 patients (90%) were followed as pulmonary ARDS. The mean age was 45.2±19.7 year. The mean PaO2/FiO2 was 106.6 mmHg, and the mean APACHE II score was 18.7 in all ARDS patients. The mean survival for all patients was 65%. Eight patients in the standard treatment group and three patients in NPPV group needed urgent intubation. Seven patients in the NPPV group received NPPV via a full-face mask. The mean EPAP was 9.6 cmH2O (range 7-12) and IPAP was 17.4 cmH2O (range 14-22). NPPV was successful in 4 patients (57%), but the other tree patients needed intubation due to failure of NPPV. The duration of NPPV, age, admission APACHE II score and admission PaO2/FiO2 ratio were not dissimilar in the two groups. However, there was a statistical difference between PaO2/FiO2 ratios at 24th hour was regarding the standard therapy and NPPV group (193.0 and 93.3 mmHg, respectively; p=0.003). Immediate intubation was needed in three patients who failed to respond to NPPV. These three patients subsequently died.

While 8 of the patients in the standard group received intubation, the number of patients having received intubation in the NPPV group was 6. Also, while 3 patients died in the NPPV group, this number was 4 in the other group. The fact that all the 3 patients failing to respond to NPPV died afterwards led us to conclude that NPPV could have delayed intubation in ARDS patients. We, therefore, suggest that the first 24 hours should be accepted as reference, and that a patient who fails to benefit from NPPV should be intubated without delay.

Okkonen M 1 , Nunes S 2 , Valta P 1 , Varpula T 1 , Hynynen M 1 , Lindgren L 3 , Tenhunen J 2 1 Anesthesia and Intensive Care, Helsinki University Hospital, Jorvi Hospital, Espoo, 2 Critical Care Medicine Research Group, 3 Clinic of Anesthesiology, Tampere University Hospital, Tampere, Finland

The traditional way to produce CPAP is to use a flow generator together with spring-loaded expiratory pressure valves (S). In the Boussignac-CPAP a virtual valve is created by airflow injected via micro-channels (B). We have compared these two CPAP systems in terms of breathing pattern and respiratory mechanics in eight healthy volunteers.

Subjects were studied in the sitting position in two separate days with different masks. Esophageal and gastric pressures were measured by two separate catheters. Airway pressure was measured by a small tube fitted to the face mask. A respiratory inductive plethysmograph was used to measure respiratory volume. Both sessions included three five-minute periods of CPAP with different positive end expiratory pressure (PEEP) levels in randomized order. Mean airway pressure (Paw), tidal volume (Vt), respiratory rate (RR), dynamic chest wall compliance (Ccw, dyn) and PEEP induced increase in functional residual capacity (∆FRC) were compared between masks.

Vt and RR differed between CPAP systems (Table 1) . Increases in FRC levels were similar with both systems. 

High concentration of S100B has been described in cerebrospinal fluid and serum after several disease of Central Nervous System. S100B, which is released from astrocytes to promote neuronal survival and development, exhibits cytokine-like activity and may play a role in systemic inflammatory response. Aim of this study is to evaluate the relationship between S100B and neurological findings in course of severe sepsis.

METHODS. 16 patients affected by severe sepsis were included. S100B was measured at ICU admission, after 72 hours and after 7 days; normal value is below 0.15 mcg/l. EEG was recorded 72 hours after the diagnosis, if appropriate; Glasgow Coma Scale (GCS) and SOFA score were calculated daily. EEGs finding were classified according the Young score system.

. 8 patients out of 16 showed evidence of severe impairment of consciousness. 5 patients showed milder degrees of coma and 3 patients showed agitation and confusion but a normal GCS. Mean S100B serum level was increased in patients with severe sepsis: S100B above 0.15 mcg/l was found in 9 patients. The GCS score did not correlate with the S100B levels. EEGs findings were abnormal in all cases with a large range of inter-individual variability but we didn't find any correlation between EEG patterns and S100B serum levels.

The main finding of this study is that S100B levels are increased in severe sepsis, without a clear relationship with GCS or EEG. Constantly elevated S100B levels might indicate damage in astroglial cells and could be useful in predicting outcomes of SAE patients but it could be also explained as an inflammatory, not specific, reaction of the brain. 

Markers of mediator and cellular activity have been studied to evaluate the immunological and clinical status of critically ill patients. IL-18 is an inflammatory cytokine, that promotes TNF-alpha, IL-1beta, IL-8, GM-CSF secretion and IFN-gamma production. Moreover IL-18 induces apoptosis and down-regulates the production of anti-inflammatory cytokines [1] . The aim of this study was to evaluate IL-18 serum levels in septic patients.

Ten patients were enrolled in a 4-months study period. Demographic data, SAPS 3, SOFA, C-reactive protein, lactic acid, IL-18, IL-2 and TNF-alpha were measured at the first day of sepsis diagnosis. IL-18, TNF-alpha and IL-2 were quantified by ELISA.

Data are shown in table 1. 

Sepsis is a serious disease with a high mortality rate. A spectrum of changes in immune response appear in the pathogenesis of sepsis, ranging from harmful hyperinflammation to profound immunodeficient immunoparalysis. Neutrophil CD64 expression increases as part of the systemic response to severe infection or sepsis. Other proinflammatory and antiinflammatory cytokines (LBP, IL-6, IL-10) are involved in the pathogenesis of sepsis. The study aimed to determine the importance and feasibility of their evaluation in patients with SIRS, sepsis and septic shock.

METHODS. 67 patients meeting the criteria for the diagnosis of SIRS, sepsis and septic shock according to the Consens conference, were examined. For the determination of granulocyte CD64 expression and HLA-DR on monocytes, flow cytometry was used. Production of TNF-a ex vivo after lipopolysaccharide stimulation was evaluated by Vollblutstimulation Kit (Millenia). The determination of PCT, IL-6, IL-10 were evaluated by use of the chemiluminiscence method.

. CD64 on granulocytes were significantly lower in patients with SIRS compared to patients with sepsis and septic shock (P< 0.0001, resp. P < 0.00001). The determination of CD64/45 and HLA-DR on monocytes using MFI correlated with QuantiBRITE method (rho=0.883, resp. 0.739, P< 0.00001, resp. P < 0.00001). CD64 expression on granulocytes correlates with mediators of systemic inflammation (procalcitonin PCT (rho=0.438, p=0.002), proinflammatory cytokinesinterleukin-6 (rho=0.456, P = 0.001), lipopolysacharid binding protein (rho=0.446, P = 0.002) and also anti-inflammatory cytokines -IL-10 (rho=0.422, P = 0.003). Monocyte HLA-DR expression correlates with C-reactive protein (rho=-0.444, P = 0.038). The values of expression of HLA-DR correlated significantly with values of TNF-alfa in patients with SIRS and sepsis (P = 0.0004, resp. P = 0.0003). The last measured values of HLA-DR on monocytes and production of TNF-alfa were significantly lower in non-survivals compared to survived (P< 0.001, resp. P<0.05).

CONCLUSION. Quantitative CD64 expression on granulocytes, monocyte HLA-DR expression, and production of TNF-alfa ex vivo are useful parameters in long-term monitoring of septic patients and in the context of the clinical status. 

Systemic inflammatory response syndrome (SIRS) is a common syndrome in the intensive care unit that may progress to severe sepsis or septic shock. It is believed that severe sepsis and septic shock are accompanied by the inability to regulate the inflammatory response shown by an imbalance in the pro-and anti-inflammatory response. The aim of this study was to measure potential markers in patients with SIRS, severe sepsis or septic shock that may predict progression of disease and clinical outcome.

We measured the level of soluble mediators osteopontin (OPN), IL4, IL5, IL10, TNF-α and INF-gamma as well as markers of T-cell activation H4/ICOS, CD69, CD25 and HLA-DR in patients with SIRS, severe sepsis, septic shock and healthy controls. Additional blood samples were taken when patients progressed to a different stage and when patients recovered and did not meet the criteria defining SIRS.

Plasma OPN levels (ng/ml) were significantly increased in patients with SIRS as compared to healthy controls. In the more severe stages OPN increased further, while levels decreased when patients recovered (see table) . No differences were observed in the levels of IL10, TNF-α, INFgamma, IL4 or IL5. In patients, the analysis of the markers of T-cell activation showed an expected increase in levels of CD25, CD69 and HLA-DR as compared to the controls. Levels decreased when patients recovered. Of special interest, levels of H4/ICOS (% positive cells) were increased in patients as compared to controls, however levels continued to increase during the recovery of the septic process. 

CONCLUSION. This study suggests that OPN and H4/ICOS may be considered new markers in sepsis. OPN appears to correlate with the severity of sepsis and may play a role in the pro-inflammatory response. H4/ICOS positive T-cells are likely part of the anti-inflammatory response and therefore may play a role in the recovery of sepsis. However, more research is needed to establish predictive roles for these markers.

Grant acknowledgement. Regione Piemonte.

Patients with severe sepsis were characterized by high SAPS II and Apache scores as well as elevated serum levels of procalcitonin as a marker for infection. HLA-DR expression, a marker of monocyte deactivation, was decreased significantly in severe sepsis (9190 receptors per cell versus 23 000 in SIRS). When analyzing monocyte apoptosis it could be shown that the central executioner caspase 3 was activated more in severe sepsis as compared to SIRS or healthy controls (p<0.05). In analogy, PS positive monocytes in severe sepsis were elevated to 20.7% as compared to SIRS (11.7%) and healthy controls (5.2%, p<0.01). The increased PS externalisation correlated with down regulation of monocyte HLA-DR (p<0.05). The protein Bcl-2, which exerts anti-apoptotic effects in the outer mitochondrial membrane, was downregulated to 1010 AU in sepsis verus 1615 AU in SIRS and 2101 AU in healthy volunteers. The downregulation of Bcl-2 strongly correlated with a loss of monocyte surface HLA-DR (r2 0.42, p<0.001).

In conclusion, elevated monocyte apoptosis during severe sepsis correlates with monocyte deactivation. For the first time, there is in-vivo evidence of a link between HLA-DR downregulation on monocytes and apoptosis. It may be speculated that apoptotic cells induce deactivation of monocytes in severe sepsis. 

Infections at the Intensive Care Units (ICU) are a substantial clinical problem due to their high incidence and significant impact on patient mortality, as well as on the duration of their treatment within ICUs. The objective of the present study was to analyse the infections occurring at our department and to identify those micro-organisms responsible for infections, with consideration of their susceptibility to antibiotic treatment. Moreover, an evaluation was performed in respect of correlation between infection incidence, patient mortality and duration of stay at the department.

The study concerned all patients admitted to the ICU between February and July 2004, with a stay duration of >24h. The occurring infections were divided into groups, depending on the first symptoms occurrence, of external infections (<48h) and intra-department infections (>48h). When diagnosing infection symptoms, the CDC definitions were applied.

. 78 patients were qualified to participate in the study, among which external infections were diagnosed in 73%, while infections at the department were diagnosed in 44.9%. The average APACHE II scale scoring in the first day of treatment was 20, higher in the case of external infections. The most commonly occurring infections were: pneumonia (33%), blood infections (17%), lower respiratory tract infections (15%) and urinary tract infections (12%). The most commonly cultured microorganisms responsible for observed infections were: Candida sp., Enterococcus sp., Staphylococcus sp., Pseudomonas sp., Enterobacter sp. and Acinetobacter sp. Among studies patients, the death rate was 37%. High mortality correlated with old age, higher APACHE II scoring and presence of externally acquired infections. The average stay at the department was 10 days and was significantly longer in patients with department infections.

obtained in studies performed in other developed countries. A correlation was evidenced between the occurrence of external, internal infections and, respectively, the increased mortality and length of hospitalisation of patients within ICUs.

Six hundred eleven patients with a median age of 66 (IQR: 53-75) and hospitalized for more than 48 h in the ICU of CHU Liege Belgium were prospectively followed during one year (2004) METHODS. 335 patients who stayed in a mixed ICU in a university teaching hospital for longer than 48 hours were prospectively enrolled during 14-month study period in 2002-2003. Mortality data was obtained and the health-related quality of life was studied by EuroQol-five dimension (EQ-5D) questionnaire in January 2005.

Of the 272 hospital survivors 83 died after discharge, i.e. a total of 146 (43.6%) patients died during the follow-up period. The median follow-up time after ICU discharge was 21.5 months.

In patients without an infection on admission (NIG) (n = 84) the long-term mortality was 45.7% in those who developed an ICU-infection and 26.5% without ICU-infection (p=0.027). In patients with infection on admission the long-term mortality did not differ between those with and without infection (53.3% vs. 46.1%, p= 0.38). EQ-5D response rate was 75% of survivors, without differences in five dimensions between the study groups except that in NIG those patients who developed an ICU-infection had significantly less often autonomy than those without ICU-infection (p= 0.049). The current general health was considered to be at least moderate in 90.6% but when it was compared to the status before ICU admission it was worse in almost half of the responders. Only 36% of those at work returned to their previous jobs.

Our results show higher long-term mortality among patients who did not have an infection on admission, but who developed an ICU infection. Almost half of the patients considered their general health to be worse than before ICU admission and a minority had returned to work.

Nseir S 

Retrospective case-control study conducted in a 30-bed ICU during a 2-year period. All data were prospectively collected. All patients hospitalized >48h in the ICU were eligible. Immunosuppression was defined as active solid or hematological malignancy, organ transplantation, or leucopenia. MDR bacteria were defined as methicillin-resistant S. aureus, ceftazidime or imipenem resistant P. aeruginosa, A. baumannii, S. maltophilia, and extending spectrum ß-lactamase producing Gram negative bacilli. MDR bacteria screening (nasal, anal and axilla swabs; and tracheal aspirate in intubated patients) was performed at ICU admission and weekly. Only MDR bacteria isolated >48h after ICU admission were taken into account. Isolation measures were applied in all patients at ICU admission, in patients with MDR bacteria, and in patients with immunosuppression. Immunosuppressed patients were matched (1:1) with immunocompetent patients according to all the following criteria: age ± 5 yrs, SAPS II ± 5, duration of ICU stay ± 3d, and category of admission. Risk factors for ICU-acquired MDR bacteria were determined using univariate and multivariate analyses. 

Campos L 1 , Palomar M 1 , Porta I 1 , Otal J 2 , Burgueño M 1 , Arikan F 3 , Latorre F 1 1 ICU, 2 Preventive Med, 3 Neurosurgery, H Vall Hebron, Barcelona, Spain

Ventriculitis is a serious complication of EVD use. The presence of the considered risk factors (RF) may influence VRIs incidence. We analyzed RF for VRI (as an attempt to restrain their incidence) and the impact on outcome.

Prospective observational study (Jan 97-June 05) of all consecutive patients (pt), admitted to our ICU with acute non traumatic intracranial haemorrhage who required EVD. Criteria proposed by Lozier et al. were used to define VRIs. Data were analyzed using a descriptive and a Chi-square analysis for categorical variables and a t test for continous variables. 

Risk factors for VRI were SAH, multiple catheters and coinfection. Prophylactic catheter exchange remains a practice option, but our data suggest that this procedure is not justified. VRI was associated with a longer ICU and in-hospital stay but its presence did not influence survival.

Jaimes F A 1 , De La Rosa G 1 , Gomez E 1 , Munera P 1 , Ramirez J M 1 , Castrillon S 1 1 Internal Medicine, Universidad de Antioquia, Medellin, Colombia

Latin America exhibits a wide range of differences, compared to developed nations, in genetic background, health services, and clinical research development (1) . It is valid to hypothesize that the incidence and risk factors for ventilator associated pneumonia (VAP) in our setting may be substantially different of those reported elsewhere (2) . We conducted a study to determine the incidence and risk factors for VAP in a University Hospital from Medellin, Colombia.

Prospective cohort study in three Intensive Care Units (surgical/trauma, medical, cardiovascular) in a 550-bed University Hospital. Critically ill patients (n = 270) who required at least 48 hours of mechanical ventilation (MV) between June 2002 and October 2003 were followed until UCI discharge, VAP diagnosis or death. Independent variables were age, gender, previous use of antibiotics (less versus higher than 48 hours before MV), Glasgow Coma Scale and APACHE II at admission, thoracic or mayor abdominal surgery before admission to ICU, type of enteral nutrition, type of ICU, requirement of paralytic agents, tracheostomy, reintubation, aspiration of gastric content, and comorbid conditions. We performed univariable and multivariable Cox proportional hazards regression analysis to evaluate potential risk factors for VAP.

. 60 patients (22.2%) developed VAP 5.9 ± 3.6 days after admission. The overall incidence of VAP was 29 cases per 1,000 ventilator-days. The daily hazard for developing VAP increased until day 8, and then decreased over the duration of stay in the intensive care unit. The only statistically significant factor after multivariable analysis was gender, with being female reducing 57% the risk of pneumonia (HR: 0.43; 95% CI: 0.19-0.96).

The epidemiologic profile of VAP in terms of incidence, length of stay and clinical course resembles the general pattern described everywhere. Interestingly, however, we could not identify any potentially modifiable risk factor for VAP. A comprehensive multicenter study is warranted. It should provide deep insight about the specific microbiological, genetic and clinic features of VAP in our setting. 

admission. The initial evaluation and treatment is dependent of the focus, comorbilities and place of acquisition being important to consider association with health care environment.

In the context of a prospective, cohort, multi-centred study, on CAS admitted in Portuguese ICUs, from 01/12/2004 until 30/11/2005, we analyse the sub-group of patients with a healthcare related infection (HCRI). HCRI was defined and divided in four groups of patients: a) had wound dressing or iv treatments in the previous 30 days, b) observed at a hospital or haemodialysis centre or had chemotherapy in the previous 30 days, c) admitted to an acute care hospital for two or more days in the previous 90 days, d) living at a nursing home or institution. 

Regarding focus of infection, severity of sepsis and outcome no significant association was seen with HCRI when comparing with non-HCRI. The microbiological profile was different, particularly in the respiratory infections where Pseudomonas and MRSA gain a place among the most prevalent microbes in patients with HCRI.

Muzlovic I 1 , Jereb M 1 , Jurca T 1 , Karner P 1 1 ICU, Department for Infectious Diseases, University Medical Center, Ljubljana, Slovenia

There are many well-known risk factors (RF) that contribute to the high incidence of hospital pneumonia (HP). One of them is previous colonization of the gastrointestinal and upper respiratory tract. Therefore, RF for HP could also influence on previous colonization of respiratory system. The data showing the influence of RF on tracheal colonization (TC) and the influence of these RF on the period between TC and HP are lacking.

In a10 years lasing prospective randomized clinical study in an intensive care unit (ICU) the data were collected last from all consecutive patients who were not on antibiotic therapy at the time of inclusion. 81 patients were included. Most frequent diagnosis was tetanus (63 patients) and encephalitis (15 patients), and 3 patients were treated for other illness. Mechanical ventilation (MV) was performed in 52 (64%) patients. Mean APACHE II score was 19.5 (range, 12 to 24). 26 patients received sucralfate for stress-ulcer prophylaxis, 27 patients received ranitidine, and 28 patients received no stress-ulcer prophylaxis. Statistics: program SPSS.

. pH of gastric juice was significantly higher in patients receiving ranitidine than in other two groups. There were no differences in occurrence of HP between these groups. The most common microorganisms that colonized trachea was Staphylococcus aureus, Pseudomonas aeruginosa, Acinetibacter calcoaceticus, Klebsiella spp. and Enterobacter spp. HP was found in 55 patient and in 66% HP was preceded with TC. Mean latent period between TC and HP (LP) was 2.8 days (range, 1 to 6 days). RFs that were found to have influence on incidence of HP did not always influence the TC. Selected combination of RF (APACHE II > 16, age > 65, MV and sedation) comparing to the single RF could influence on higher incidence of HP and less on incidence of TC. Sedation and previous antibiotic therapy were independent RF that influence on longer LP. Higher values of pH of gastric contents were related to shorter LP. The longest LP was found in Acinetobacter spp.

In most cases HP was preceded with TC. Combination of selected RF could influence on higher risk of TC and HP. The duration of LP was related to the type of RF and microorganisms. 

The complexity of intensive care has been accompanied by increased risk of nosocomial infections (NI). The prevention of NI requires knowledge of rate, sources, pathogens incriminated and antimicrobial resistance patterns. The objective of this study was to analyze the NI in our cardiac intensive care unit (CICU).

All patients (pts) admitted to our CICU during 2005 for more than 24 hours were included in a prospective, observational study. Surveillance infection charts included: demographic data, diagnosis, site of infection, date of onset, pathogens isolated and antimicrobial susceptibility. For the patients with NI who died, the relationship of the infection to death was investigated. Device-associated infection rate (DIR) and device utilization (DU) ratio were calculated. DIR was expressed per 1000 device-days and DU as the ratio of the number of device-days per number of patient-days.

. 588 pts were admitted in CICU during the year and 503 were hospitalized more than 24 h. In 27 pts (5.3%), 55 NI were registered, giving a rate of 10.9%. In 11 pts multi-drug resistant Acinetobacter species was associated with hospital-acquired pneumonia (HAP). Klebsiella pneumonia (ESBL+) was responsible in 5 pts for urinary catheter infections. Gram-positive cocci were responsible for the largest proportion of mediastinites (n=4) and catheter related infections (n= 2). In pts with NI we registered 6 deaths (22% mortality). In pts without NI (n=561) the in-hospital mortality was 8.7%. The ventilator associated pneumonia (VAP)-attributable mortality was 13%. 

In our CICU the most common NI is pneumonia. The rate of VAP, but not the ventilator utilization is higher than those reported by CDC. In our study VAP increased in-hospital mortality. 

Clinical studies, so far, produced conflicting results regarding the role of ventilation-associated pneumonia (VAP) as an independent prognostic factor in terms of mortality, length of hospital stay and mechanical ventilation days. In our study we investigated the contribution of VAP in the prognosis of critically ill patients.

We performed a prospective observational study of 88 patients, aged between 18 and 95 years, who received invasive ventilatory support for more than 48 hours, comparing the group which developed VAP to the control group without pneumonia. The primary endpoints were absolute mortality, length of ventilation support and ICU stay (LOS), while the secondary endpoint was the frequency distribution of microorganisms in the bronchoalveolar lavage. The exclusion criteria were: advanced stage neoplasm or neuromuscular disease and coma Glasgow scale chronically less than 8. Categorical variables were compared by the chi square test and the continuous variables were compared by the T Student test. Potential biases were controlled by multiple linear regression. 

Of the 89 casenotes reviewed there were 69 hospital-acquired pneumonias, 2 patients had both MRSA and pseudomonal pneumonia and were excluded from further analysis. There was incomplete data for 10 patients. 48 patients had pseudomonal pneumonia and 19 had MRSA. The 2 groups of patients were well matched, with similar age, sex distribution, CPIS scores, incidence of bacteraemia and Acute Physiology and Chronic Health Evaluation II scores at diagnosis. There were no outcome differences between the groups with respect to hospital and critical care survival and critical care length of stay (Table 1) . 

Bongo Jr A L 1 , Bautista M S 1 , De Leon N A 1 1 Section of Pediatric Pulmonology, Philippine Heart Center, Quezon City, Philippines

The pattern of nosocomial pneumonia among infants, children and adolescent after cardiac surgery may differ from that in adult ICU.

A retrospective cohort study of patients admitted at the pediatric cardiovacular surgical ICU age 18 years old and below, who developed nosocomial pneumonia after cardiac surgery from January 2003 to June 2004 were reviewed.

A total of 608 patients underwent cardiac surgery, of which 360 patients underwent open heart surgery and 248 underwent closed heart surgery. Among these, 27 patients developed nosocomial pneumonia with the incidence of 4.4%. Independent predictors of pneumonia were: intubated state, feeding through NGT/OGT, previous antibiotic use, use of sedatives and H2 receptor antagonist. A total of 37 pathogenic microbial strains were isolated. Gram negative were the most frequent isolates (33 isolates, 89.18%). The most common etiologic organisms is Stenotrophomonas maltophilia (11 isolates, 29.72%).

Gram negative bacilli remains the most common pathogen isolated in our study. Risk factors in the development of nosocomial pneumonia include intubated state, feeding through the NGT/OGT, previous antibiotic use and use of sedatives and H2 receptor antagonist. It is important to recognize the predominant pathogen associated with nosocomial pneumonia since infections may vary between hospitals as well as between specialized units within same hospital. The findings of the study would recommend to adopt multidisciplinary approach in the implementation and devlopment of strategies to reduce overall infection in the pediatric ICU.

Emergency medicine and acute brain injury 0593-0606 0593

Páramo J M 1 , Ristow A V 2 , Albuquerque A A 1 , Abreu R C 2 , Valim L 1 , Furtado T 1 , FilhoÁ C 1 , Thomaz F 3 1 ICU, 2 Centervasc, Clinica Sorocaba, 3 Radiology, Clínica Radiológica Emílio Amorim, Rio de Janeiro, Brazil

For many years Recombinant Factor VIIa (rFVIIa) has been used with success for treatment and prophylaxis of hemorrhagic events in hemophiliac patients, with inhibitors for VIII and IX factors. Better understanding of its action mechanism, turned it an adjunctive therapy in severe trauma patients or during surgeries in patients without previous coagulopathy. Another problem in clinical scenario are patients using warfarin, that must be submitted to emergency surgical procedures. Recombinant FVIIa, had been used in these patients successfully.

Descriptive report of utilization of rFVIIa to treat a retroperitoneal haemathoma after an endovascular procedure in an anticoagulated patient.

The authors describe rFVIIa utilization in a puerperal patient, who was submitted to an endovascular procedure to correct an iliac arterial-venous fistulae, by part of treatment of Inferior Cava Vein thrombectomy, Left Iliac Vein thrombectomy and venous thrombectomy of left limb, as a result of Phlegmasia coerulea dolens that occurred fifty days before. As a complication of an angioplasty of this vein to correct an extrinsic compression of iliac artery, a large retroperitoneal haematoma was formed with haemorragic shock. 9.6 mg of rFVIIa was used, and bleeding was controlled with prompt haemodynamic stabilization, without surgical intervention. The patient was submitted to abdominal CT at admission, 3rd and 7th day of evolution, which showed complete resolution of haematoma. Laboratory data showed an improved coagulation.

In this kind of vascular access complication, rFVIIa showed to be a safe and effective drug. 

INTRODUCTION. Introduction -Intubation and airway control of injured children is of vital importance, but despite its advantages is associated with meny risks when is performed outside the hospital and by the untrained physicians. Aim of this study was to determine the importance for the survival of the children with craniocerebral injuries, and also for final outcome of treatment.

Methods -This study is a clinical, partly prospective, partly retrospective that includes 60 patients with isolated craniocerebral injures, aged up to 17 yers, and with Glasgow Coma Score under 8, that did not require surgical treatment. Patients were divided in two groups each with 30 patients. The first group included patients that were endotracheal intubated, and the other group included patients that were not intubated.

Results -Theree was no statistically difference between groups regarding the sex, age and Glasgow Coma Score. Regarding the endotracheal intubation there was a statistically significant difference, in the first group 86.7% of the patients were intubated during the prehospital treatment, while 16.7% of the patients from group II were intubated. A greater percentage of patients from group I underwent controlled (66.7%) or assisted (20%) mode of ventilation, and 13.3% of patients were on spontaneus breathing. Among the patients that have survived also a greater percentage underwent controlled (48.6%) and assisted (13.5%) mode of ventilation. Among the patients that did not survive a greater percentage underwent spontaneus breathing (56.5%).

Conclusion -According to the all protocols for the treatment of children with craniocerebral injuries, endotracheal intubation is obligatory for any patient with Glasgow Coma Score under 8. Howevere, contoversies regarding endotracheal intubation of children during prehospital treatment still exist. Endotracheal intubation should be performed by an experienced physician with an adequate equipment, and evrey child with craniocerebral injurie must receave oxygen. 

Repeatedly audit demonstrated that our cardiac arrest trolleys did not provide the minimum level of emergency airway equipment resulting in numerous adverse events. Inaddition there was no cleaning policy for ward-based laryngoscopes [1] nor was equipment for comfirmation of ET-tube placement available.

METHODS. An infection control and airway equipment audit was undertaken of 48 resuscitation trolleys. Equipment levels where assessed against an approved stock list and the laryngoscopes where observed for visible contamination and where it was noted a microbiological sample was obtained.

During the 1st audit no resuscitation trolley was fully stocked and although only 12.5% laryngoscopes where bacterially contaminated 48% were visible dirty. There was a marked improvement in emergency airway equipment provision following the introduction of the airway tray (table 1) and the tray facilitated the introduction of equipment for confirmation of ET-tube placement. CONCLUSION. The lack of emergency airway equipment plus the state of cleanliness is a major cause for concern. The introduction of a pre-packed single-use airway tray containing all emergency airway equipment successfully ensured the provision of clean emergency airway management. 

Papanikolaou P G 1 , Paleologos T S 1 , Moraitou F 2 , Mavrommatis A 2 , Barkas K S 1 , Drosos L 1 , Katsaris G 2 , Kazdaglis K 1 1 Neurosurgical Department, 2 Intensive Care Unit, General Nikaia Piraeus Hospital, Athens, Greece

Postraumatic infection is a serious complication of missile brain wound. Early beginning of appropriate treatment is important for outcome. We present our experience from the use of multimodality neuromonitoring (Intracranial Pressure, Brain Tissue Oxygenation and Microdialysis) in a patient with shotgun penetrating cranial injury. Elevation of lactate / pyruvate (L/P) ratio measured by Microdialysis was an early indication of posttraumatic meningitis, treated intravenously and intrathecally.

Our patient (male, 35yrs) was injured by a small diameter shotgun from short distance. Missile entered cranial cavity via the right zygomatic region and came out via the parietal region homolaterally. Tunnel passed through the right temporal and parietal lobes. There were extradural and subdural haematomas together with multiple right hemisphere contusions. After emergency surgery (for evacuation of clots and debridement) neuromonitoring catheters were placed in ICU ward via a 5.3 mm right frontal burrhole. Catheters' tips were guided inside right frontal lobe white matter, at the penubra of haemorrhagic contusion.

During the first 7 days the levels of monitored neuroparameters were controlled using barbiturates : ICP under 25 mm Hg, PtiO2 under 25 mm Hg, L/P ratio round 25. On day 8 there was an elevation of L/P ratio levels (30-35). A lumbar puncture showed meningitis (acinetobacter sp). There were no indications of brain abscess or cerebritis (no new CT-scan findings, negative cultures from surgical wound). Antibacterial medication (colistin) was admitted intravenously (9 million units per day) and intrathecally (500 000 units every second day). Next two days (9 and 10) ICP levels were also rised (round 30 mm Hg). PtiO2 levels remained normal. From day 11 and on there was improvement of fever and of CSF cell index, together with ICP normalization. Neurological (short-term) outcome was excellent, with no remarkable handicaps.

can be an early indication of posttraumatic CNS infection and can lead to early beginning of effective treatment. 2) Intrathecal colistin administration is valuable in the treatment of meningitis caused by multi -drug resistant acinetobacter. METHODS. 57 years old female had T12 -L1 compression fracture due to traffic accident. After anesthesia induction he was intubated with right bronchial tube and left single lumen subclavian central vein catheter was inserted. During peroperative period the patient had a stable hemodynamic and respiratory situation. 3000 mL fluid infusion and 2 Units of blood transfusion were given during 5.5 hours. Postoperative fluid balance was + 2800 mL. During surgery, left toracotomy was performed and left chest tube was inserted. After surgery she was admitted to ICU for postoperative care. Chest x-ray showed no abnormality within the catheter localization (Picture I). Total fluid infused through the catheter in ICU was 2555 mL and the fluid balance was calculated as + 1100 mL. On the second day she was transferred to ward. According to the files, besides oral intake she received total of 1280 mL IV fluid where total balance was found to be + 1375 mL. The day after she felt shortness of breathing and her physician's early diagnosis was hypervolemia and fluid restriction, diuretic and bronchodilator therapy was started. In the following 8 hours her dyspnea did not resolve and she became tachypneic, started sweating and fever rose to 38.2 • C. Blood gas analysis on spontaneous breathing with face mask (FiO2: 0.5) revealed PaO2: 35mmHg, PaCO2: 38 mmHg, pH: 7.51, HCO3: 31.8 mmol/L, BE:1.3 mmol/L, SaO2: 73.7 and she was admitted to ICU. Chest X-Ray showed enlargement in the mediastinum (Picture II). As this suggested a catheter related complication, thorax CT scan was taken with contrast solution given via the catheter. Catheter tip was found in the mediastinum (Picture III-IV) and the catheter was withdrawn immediately.

Her tachipnea and dyspnea resolved gradually. There was a progressive improvement. Next day control chest x-rays showed that mediastinum reduced to normal sizes (Picture V) and hypoxemia resolved. She was transferred to ward and was discharged on the 8 th day.

One of the central vein catheter complications is malpositioning. In this case the localization site mediastinum which led to clinical worsening and deterioration. We showed with this case that the removal of the catheter and leaving the mediastinal fluid to spontaneous absorption a better and safer approach when compared to an insertion of a drainage tube to the same site.

Paramythiotou E 1 , Ntagiopoulos P 1 , Aggelopoulou H 1 , Koutsodimitropoulos J 1 , Poularas J 1 , Kalogeromitros A 1 , Karabinis A 1 1 ICU, George Genimatas general hospital, Athens, Greece

Road traffic accidents (RTAs) remain one of the most frequent causes of admission in Intensive Care Units in industrialized countries (1) . These accidents are caused either by car or by motorcycle or finally victims are pedestrians struck by a vehicle. Our purpose was to examine, in this preliminary study, the demographic characteristics of these patients as well as their outcome.

During a five -year period (1st January 2000-31 December 2005), all patients admitted in our 10 -bed general ICU unit as victims of road traffic accidents (VRTAs) were included in this retrospective study. Characteristics recorded were age, sex, nationality, length of ICU stay, ISS score and outcome.

During these five years, 1027 patients were hospitalized in total in our ICU. Two hundred eighty eight patients (a percentage of 28%) were VRTAs. 240 p were of Greek origin (84%) and the rest (16%) were foreigners, most of them immigrants. Among these 288 p, 228 were men (79%) and 60 (21%) were women. 16.2 for men, while±Their age range was 14 -80 years. Their mean age was 34.1 38.8 days± 19.5 years. The mean ICU length of stay was 24.6 ±for women 31.0 15.1 (range: 6-75). Finally±(range: 1-378). The mean ISS score was 31.5 sixty-four deaths occurred (22.1%) among the study group.

Our results confirm the fact that road traffic accidents are among one of the main causes of admission in our ICU. Victims are mostly economically productive young adults of male sex and in almost a quarter of the cases accidents lead to patients' death. A prolonged hospitalization is common and many victims suffer from chronic and persistent disabilities. Interventions are needed and drastic measures must be taken in order to diminish the number of victims and the untoward consequences of VRTAs. 

Suberviola B 1 , González-Castro A 1 , Llorca J 2 , González-Mansilla C 1 , Rodriguez-Borregán JC 1 

To evaluate the safety, the complications and effect on neurological outcome of high-dose methylprednisolone sodium succinate (MPSS) versus no medical treatment in patients with acute spinal cord injury admitted to our ICU in the last 11 years.

From January 1994 to December 2005, we retrospectively analyzed 82 patients with acute spinal cord injury admitted to our ICU. Fifty nine patients received MPSS according to NASCIS II protocol and 23 did not received. Mortality and septic complications were recorded during ICU stay. Frankel grade was assessed at admission and at discharge of ICU.

Patients who did not received MPSS had a higher Injury Severity Score (p =0.006). There was no difference between those patients who received steroids and those who did not in mortality (OR: 1.07, 95% CI: 0.90-1.27; p= 0.39) and in functional improvement (OR: 0.96, 95% CI: 0.45-2.04; p= 0.93). There was a trend towards an increase in septic complications in the MPSS group (18% vs 4%; OR: 3.5, 95% CI: 0.47-26.16; p= 0.27) although this did not achieve statistical significance.

The present study confirms the absence of benefit of MPSS therapy in patients with acute spinal cord injury. Not only was there no significant difference in outcome in this study, but also there was a trend to improve septic complications in patients who received MPSS.

Mouriz L 1 , Pastor A 2 , Rocha F 1 , Díaz D 1 , Guede G 1 , Illodo G 1 , López A 1 , Vázquez A 1 1 Anestesiology and reanimation, 2 Neurosurgery, Complexo hospitalario de Ourense, Ourense, Spain

Compartmental abdominal syndrome is due to increase abdominal pressure which produces ischemic bowel and multiple organs failure.

An 82 year old female was admitted to the orthopaedic ward with a hip fracture following a fall. Her past medical history included atrial fibrillation treated with anticoagulant. 24 hours after admission to the hospital, she was transferred to intensive care unit with a hemorrhagic shock. A CT scan of her abdomen revealed a retroperitoneal haematoma. A few hours after the CT scan, she presented with a compartmental abdominal syndrome. Her blood tests showed a low haemoglobin, hematocrit and platelets, renal failure, ATP of 14% and INR of 6.36. An abdominal ecography and CT scan showed a right pararenal retroperitoneal haematoma with major component in the pelvic cavity. Intravenous fluids and inotropic drugs were administrated as well as fresh frozen plasma and blood to treat her clotting disorders and hypotension. To release the high abdominal pressure, the abdominal wall was left opened (contained-open abdomen) and the opening covered with a mesh.

The wound was closed 5 days later due to the improvement of the abdominal pressure and organs function. Her clinical improvement allowed stopping inotropic drugs and sedation. Later on she presented with electrolytic disorder and cardiac arrhythmia which caused myocardial infarction, heart failure and acute pulmonary oedema. She had a cardiac arrest which did not response to resuscitation.

Compartmental abdominal syndrome is a forgotten process in critically ill patients despite its high mortality. It is caused by increase abdominal pressure which leads to multiple organs failure. Abdominal pressure must be measured and decompressive surgery can reduce multiple organ failure which has been associated a bad prognosis.

Anestesiology and reanimation department of complexo hospitalario de Ourense.

Markogiannakis H E 1 , Sanidas E 2 , Michalakis I 2 , Melisas I 2 , Tsiftsis D 2 1 1st Department of Propaedeutic Surgery, Hippocratio Hospital, Athens Medical School, University of Athens, Athens, 2 Department of Surgical Oncology, Herakleion University Hospital, Herakleion Medical School, University of Crete, Herakleion, Greece

Our aim was the identification and analysis of the demographic and injury characteristics along with management and final outcome of trauma patients with blunt liver, splenic and renal injuries.

We reviewed the Trauma Registry data of Herakleion University Hospital of all patients with blunt hepatic, splenic and/or renal trauma during a 4 year period.

One-hundred and eighteen trauma patients with liver, splenic and/or renal injuries were admitted to our hospital. Fifty-five patients (46.6%) had liver trauma, 58 (49.1%) splenic and 42 (35.6%) renal trauma; 17 (14.4%) had concomitant liver and splenic injuries, 14 (11.8%) liver and renal, 10 (8.4%) splenic and renal and 4 (3.4%) liver, splenic and renal injuries. Mean age of the patients was 34.4±2.8 years while 97 (82.2%) were men. Mean Injury Severity Score was 19.3±1.6. Mean grade of hepatic, splenic, and renal injury was 2.3±0.2, 2.6±0.3, 2.5±0.2, respectively. Vehicle accidents constituted the predominant injury mechanism (n=93, 78.8%) while 19 patients were injured by fall (16.1%) and 6 (5.1%) by other causes. Thirty-six patients were in a car (30.5%) and 52 on a motorcycle (44.1%), while 68 were drivers (57.6%), 18 co-drivers (15.3%), 2 passengers (1.7%) and 5 pedestrians (4.2%). Thirty-two had concomitant head trauma (27.1%), 58 thoracic (49.1%), 14 other abdominal injuries (11.8%), 19 pelvic (16.1%), 12 spinal cord (10.1%) and 25 injuries of the upper or lower extremity (21.1%). Eighty-four of them (71.2%) were initially selected for nonoperative management (NOM) while 34 (28.8%) were directly operated. Sixteen NOM-patients failed NOM (19%). Fifty-two patients were admitted in the I.C.U. (44%) and 11 died (9.3%). Mean length of I.C.U. and hospital stay was 9.4±2.2 and 17.8±2.5 days, respectively.

Trauma patients with blunt liver, splenic and renal injuries are usually young male car-or motorcycle-drivers. They frequently sustain concomitant injuries the most common of which are thoracic injuries. Nonoperative management of carefully selected blunt liver, splenic and renal trauma patients is safe and effective. Great caution should be paid, however, so as to identify those patients who are susceptible to fail NOM.

Tomás Marsilla J 1 , Montoiro R 1 , Carcamo A 1 , Martin L 1 , Sanchez N 1 , Montes N 1 , Roche F 1 , Zalba B 1 1 Intensive Care Unit, Clinic University Hospital, Zaragoza, Spain INTRODUCTION. Subarachnoid aneurysm hemorrhage (SAH) is a illness with a high morbidity and mortality. The main complications determinants of outcome after SAH are rebleeding, vasospasm and cerebral infarction and hydrocephalus. The aim of this study was analyzed the association of variables on admission as independient predictors of poor neurologic outcome.

Retrospective study. We examined 69 patients admitted to a critical care unit in a tertiary care university hospital with diagnosis of SAH between January 2003 and December 2005. We evaluated risk factors, clinical scales on admission (World Federation Neurologic Surgeons and Hunt and Hess, and amount of blood seen on initial head computed tomography scanning by Fisher scale. All patients were trated according to standard management protocols with strict hemodinamic control, nimodipine and triple H theraphy once symptomatic vasospasm is evident. In the same way we evaluated the frequency of aneurysm locations as well as arteries reponsibles of cerebral ischemia.

. 60% patients had intracranial aneurysm; 97.6% of them treated with endovascular coils. During the course of hospitalitation 23% of patients had rebleeding, 42% hydrocephalus and the patients with symptomatic cerebral ischemia were 44.9% during the first 14 days.30% patients die or developing to persistent vegetative state. There was a significant association (chi-square) between a higher Fisher grading on admission as well as aneurysm location in anterior communicating artery, both with an elevated incidence of development of symptomatic cerebral ischemia.

Given that higher rate of neurologic consequence, in spite of news therapies in subarachnoid hemorraghes, identifying patients with an increased risk of vasospasm through significants predictors on admission would be useful to decide to start more aggressive treatment. 

Martínez-Barrios M 1 , Callejo-Torre F 1 , López-Pueyo M 1 , Zabalegui A 1 , Cordero J 2 , Gómez-Menéndez A I 3 , Corral E 3 1 Intensive Care Unit, 2 Investigation Unit, 3 INTRODUCTION. Airway trauma is rare but morbid. They are encountered with increasing frequency because of improvements in prehospital care and early initiation of the Advanced Trauma Life Support (ATLS). The study is aimed to give a description of patients with airway trauma, to help providing data for better emergency services & delayed managements of this type of trauma.

METHODS. This retrospective, cross sectional study was conducted in 4university hospitals in Tehran, Iran; from 1997 to 2002 the study was conducted with 28 patients. Age, sex, mechanism of trauma, site of trauma, clinical manifestations, length of hospitalization, complications, mortality & treatment outcome were the main subjects extracted from units. SPSS 10.01 was used for analysis.

. 22 (78.6%) patients were males and 8 (21.4%) were female.. Mean age was 33.1± 2.2 years. 8 (28.5%) cases presented with blunt trauma, 19 (67.8%) with penetrating trauma & 1 had inhaled burning chemical agent. 22 (78.6%) of cases had injuries in cervical airways & the rest 6 (21.4%) in thoracic airways. subcutaneous emphysema 7 (50%), dysphonia & dyspnea each with incidence rate of 6 (42.8%) were the most frequent manifestations. 85.7% of our patients underwent surgical treatment. Mean length of hospitalization was 5±2 days. The total rate of incidence for complications was 14.2%; 2 (7.2%) cases of wound infection, 1 (3.5%) with anastomosis dehiscence & 1 (3.5%) with collapsed lung after thoracotomy. The mortality rate was 10.7% (3 patients 

Martinez-Lucas P 1 , Moreno-Cuesta J 1 , Cuartero del Pozo A 1 , Cortiñas-Saez M 1 , Peyro-Garcia R 1 1 Anesthesia, Hospital General Albacete, Albacete, Spain

To investigate the ICU factors associatted with increased length of stay (LOS), days of mechanical ventilation (DMV), ICU mortality and neurological outcome at six months (GOS) in severe traumatic brain injury (TBI).

A prospective study was carried out in a ten bed surgical intensive care unit of a university hospital. The study included 90 caucasian patients who had experienced a severe TBI within the previous 24 hr. The exclusion criteria was a previous deficit in the central nervous system. The studied variables were: age, sex, type of trauma, GCS, Apache II, size of pupils, simetry of pupils, pupil reactivity to light, present of subarachnoid hemorrhage (SAH), trauma coma data bank (TCDB) CTscale, surgery, early low blood pressure, late low blood pressure, ICP monitoring and cerebral perfusion pressure (CPP). These variables were tested for their association with LOS (days), DMV, mortality and GOS. Data are shown as mean±SD. Confidence intervals were calculated when indicated. A p < 0.05 was considered significant.

LOS is increased in patients without dilated pupils (15±10.2 vs 6.5±11.4), present bilateral pupilar light reflexes (14.3±11.9 vs 9.6±10.9), ICP insertion (17.9±14.7 vs 9.2±7.7) and with a neurosurgery intervention (15.6±11.2 vs 8.4±10.3). DMV are increased in patients with ICP insertion (12.7±10.7 vs 5.7±5.3), and normal pupils light reactivity (9.3±8.6 vs 6.8±8.1). Mortality increased in patients with a low GCS ( [3] [4] 

Interestingly apart from the all previously known score factors over the prognosis of severe TBI we have found that ICP insertion did not improve the outcome, but it got worse the LOS and DMV. Furthermore, the evacuation of focal lesions improved the mortality in ICU and increased the LOS. Finally, the presence of SAH increased ICU mortality in our study.

Leone M 1 , Charvet A 1 , Bregeon F 2 , Antonini F 1 , Martin C 1 , Albanèse J 1 1 Anesthésie et réanimation, CHU Nord, 2 Laboratoire de Physiopathologie respiratoire, EA 2201, Faculté de Médecine, Université de la Méditerranée, Marseille, France

To determine the long-term outcome of patients admitted to intensive care unit (ICU) with chest trauma.

A prospective observational study was conducted between 2004 and 2006. Inclusion criteria were admission to ICU with a thoracic acute injury score >1 and an injury severity score > 15. Exclusion criteria were patients with chronic heart failure or COPD, and pregnant women. A clinical assessment and a chest CT scan were performed on admission. Six months after the ICU discharge, a clinical assessment using a Saint George questionnaire (SGQ), which evaluates clinical and psychological symptoms on a 0 (no symptom) to 100 (maximal symptoms), a chest CT scan, and pulmonary function testing (PFT) were performed. (Table 1) , exclusion consisted on 17 deaths, 15 lost to followup, 10 patients refusing this follow-up, and seven patients unable to undergo assessment. Hence, 54 patients were extensively followed up. At 6 month, 65% of patients had clinical or psychological symptoms, although only 9 (15%) had a SGQ > 15. Chest CT scan was not considered as normal in 59% of cases. During the PFT, both obstructive (20%) and restrictive (11%) deficits were observed. 

Renal failure and replacement techniques 0607-0620 0607 

Regional anticoagulation with citrate is frequently used during continuous hemofiltration. The commonly used concentrated solution of 30% trisodic citrate may result in metabolic alkalosis and hypernatemia. A diluted solution (trisodic citrate 10 mMol/L + citric acid 2 mMol/L) has been recently developed, which can also be used as a replacement fluid. We compared the metabolic and anticoagulant effects of these two solutions.

We randomized 21 patients requiring hemofiltration with citrate anticoagulation to receive either the concentrated solution (n=10) or the diluted solution (Prismocitrate, Hospal) (n=11). In both cases, a Prismaflex device (Gambro-Hospal) was used with the following settings: blood flow 130 mL/min, hemofiltration 2500 mL/h, dialysate 1500 mL/h. The citrate was infused to achieve a target solution of 4 mMol/L, using either 35 mL/h concentrated solution or 2500 mL/h diluted solution. Blood and filter blood samples were obtained every 4 h. The flow rates of citrate and calcium solutions were adapted to maintain a blood ionized calcium concentration between 1. CONCLUSION. This diluted citrate solution can safely be used for regional anticoagulation, and is associated with fewer episodes of hypernatremia and metabolic alkalosis than the concentrated solution.

Grant acknowledgement. Diluted citrate and replacement solutions were provided by Gambro-Hospal. 1 In the UK, CRRT is provided by 90% of the units to 9442 patients annually and 37% units use ultrafiltration rates of 1000ml.h-1. 2 In this prospective audit we set out to determine both the prescribed ultrafiltration rate and mortality in our unit.

The following data was prospectively collected from patients receiving CRRT over 11 months: demographics, weight, RFA, indication, serum UsEs, blood lactate, ABG, anticoagulation, type and flow rates for replacement and dialysate fluids and ICU mortality. We then introduced an educational programme to highlight the benefits of higher ultrafiltration rates and completed the audit cycle over a 7 month period.

A total of 67 patients (48 male) were studied, 36 after the intervention. Median age was 67 yrs, 7% had chronic renal failure, 26% received CRRT post operatively, 51% showed evidence of sepsis and 72% were oliguric. Median serum (range) Na, K, urea, creatinine, pH, base deficit and lactate were 138 (116-167), 4.8 (2.9-8.5), 23 (4.5-61), 301 (59-1020), 7.27 (6.97-7.46), -7.8 (-23-+8), 1.6 (0.6-18) respectively. 

It is possible to achieve higher ultrafiltration rates for CRRT through education and awareness and achieve a clinically significant reduction in mortality in our unit.

The promotion of rapid identification of, and response to a cardiac or respiratory arrest improves survival from cardiac or respiratory arrests. Different countries have varied resources and needs. Simple key recommendations and interventions have been published for resuscitation from cardiopulmonary arrest (1) Unexpected cardiac arrest are usually preceded by signs of clinical deterioration in the hours before cardiac arrest.

Hospital is a 212-bedded acute hospital. A resuscitation team, a resuscitation policy and guidelines, and resuscitation training according to staff members expected ability and roles was implemented in conjunction with published standards with minimal costing and resourcing. A modified Modified Early Warning Scoring (mMEWS) system introduced to relocate the critically ill to the Critical Care Unit. Education about the recognition of the critically ill was undertaken. The number of unexpected cardiac arrests was studied before during and after the implementation of the recommended standards. Data was collected from patients' notes, switchboard and Utstein forms and used to exclude false alarms and non-cardiac arrests.

RESULTS. The incidence of unexpected cardiac arrest was 13.8 per 1000 admissions (13 cases) falling to 1.2 per 1000 admissions (1 case) Odds Ratio 0.08 (CI 0.01-0.65) p<0.01 representing a reduction of 92% in unexpected cardiac arrest rates. CONCLUSION. The implementation of resuscitation guidelines and policies significantly reduces the incidence of unexpected cardiac arrests and has improved quality of care and patient safety, with minimal resourcing and cost. 

Patients readmitted to ICU present a higher mortality rate and a longer hospital stay. A possible solution to assure a safer discharge from ICU and to reduce the rate of ICU readmission could be a follow-up activity by an intensivist. Aim of this activity should be to verify the appropriateness of care on the ward and to support both nurse and medical ward staff. Improving the quality of care directly provided on the ward, this activity should reduce the risk of readmission. In this sudy we have evaluated the effect on the readmission rate and on 28-days mortality of a scheduled follow-up program by an intensivist in patients discharged from an ICU.

We enrolled all patients discharged from an 8 bed ICU of a mixed teaching and notteaching tertiary hospital from 1 January 2004 to 31 December 2004 (n=251). We compared outcome measures of this group with a control group of all patients discharged from the same ICU from 1 January 2002 to 31 December 2002 (n=187). Tracheostomy, gastrostomy and multi-pharmacological prescriptions at ICU discharge were evaluated as factors at risk for readmission. In fact those aspects pose an increased burden to ward staff, and could result in a major risk of complications.

The follow-up activity reduced ICU readmission rate from 14 (7.5%) to 8 (3.2%) (RR=2.34, CI 95% 0.94-6.01, p=0.04). About the considered risk factors only tracheostomy was found to be related to an increased risk of readmission. In the subgroup of tracheostomized patients the follow-up activity resulted extremely effective, reducing the number of readmissions from 19 (79.1%) to 38 (97.4%), (RR= 1.23 95% CI 1.00-1.32 p=0.02). No significant decrease in 28 day mortality was reported.

In patients discharged from ICU, a follow-up activity by an intensivist seems to be able to reduce readmission rate to critical care. In our setting, patients discharged with tracheostomy seem to be at increased risk for ICU readmission. The follow-up activity resulted to be particularly effective in this subgroup of patients.

Endacott R 1 , Chaboyer W 2 , Worrall-Carter L 3 , Ernest D 4 , Edington J 5 1 INTRODUCTION. Physiological bedside assessments obtained on general (non critical care) wards can be evaluated in a systematic way using track and trigger systems. Track and trigger systems allow for early identification of patients at risk of catastrophic deterioration in order to allow for early intervention by critical care teams. We set out to evaluate the inter-rater and intra-rater reliability of the physiological measurements, aggregate scores and triggering events of three physiological track and trigger warning systems.

We performed a prospective cohort study on unselected patients on general medical and surgical wards in one non-university acute hospital. Four investigators (a senior physician, junior physician, senior nurse and junior nurse) examined 114 patients in the inter-rater study and 45 patients in the intra-rater study. Physiological observations obtained at the bedside were evaluated using three tools: the Medical Emergency Team call-out criteria (MET [1] ), the Modified Early Warning Score (MEWS [2] ) and the Assessment Score of Sick patient Identification and Step up in Treatment (ASSIST [3] ).

Inter-and intra-rater reliability were assessed by intraclass correlation coefficients, kappa statistics and levels of agreement.

There was only fair to moderate agreement in most of the physiological parameters and fair agreement on the scores, but better levels of agreement on triggers (Table) . The reliability was partially a function of simplicity: MET achieved higher percentage agreement than ASSIST, and ASSIST higher than MEWS. Intra-observer reliability was better then inter-observer reliability. Using corrected calculations improved the level of inter-rater agreement. 

There was significant variation in the reproducibility of different track and trigger warning systems used by different health care professionals. The systems examined showed better levels of agreement on triggers than on aggregate scores. Simpler systems had better reliability. Inter-rater and intra-rater agreement could potentially be improved by using electronic calculations of scores. 

Hermans B J 1 , Dortmans J A 1 , De Visser A C 2 , Foudraine N A 1 1 Intensive Care, 2 Epidemiology, Viecuri Medical Centre, Venlo, Netherlands

A considerable number of patients who are discharged alive from the ICU will still die in the hospital. Discharge time, TISS and stepdown facilities are reported to be related to post-ICU mortality. As not all patients discharged die unexpectedly (palliation or no-readmission) we studied if TISS could indicate the level of care after ICU discharge for patients who will die unforeseen in the hospital and who will not.

We performed a retrospective, single centre study which included all patients admitted to our ICU between 1 January 2002 and 1 July 2005. Patients < 18 years of age were excluded. Patients who left the hospital alive after ICU discharge (group 1) were compared with patients who died unexpected in the ward or in the ICU after readmission (group 2). Factors related to post-ICU mortality like APII, TISS at discharge, LOS, discharge time (office/night hrs) and discharge facility were studied.

Of the 1358 patients included 328 (24%) died in the ICU, 148 (11%) died after discharge from the ICU of which 97 (7%) died unexpected. 882 patients (65%) left the hospital alive. Risk factors associated with a higher hospital mortality are shown in table 1. Although patients who died foreseen had a comparable TISS with patients who died unexpectedly (20±11 vs 21±10, P=0.59), they were less frequently discharged to the MCU/CCU (35.3% vs 75.3% P<0.000). 56.4% of group 1 patients have been discharged to a MCU/CCU vs 63.2% of all patients who died (P=0.14). 

Associated with in hospital mortality, this study shows that TISS at discharge is not a reliable discriminator to determine the level of post ICU care. Moreover, we found that not discharge time but in weekend discharge is associated with a higher mortality. 

We conducted a chart review of ICU readmissions over a 10-year period. Severity of illness was evaluated by APACHE II and SOFA scores estimation. Duration of mechanical ventilation, tracheostomy, inotropes, readmission interval and ICU LOS were used as surrogate parameters of outcome.

The total number of ICU readmissions was 95 (2.1%) out of 4494 patients (pts). Within 4 days of discharge 53.3%(n:51) of readmissions occured. Readmission was attributed mainly to cardiorespiratory deterioration in 47 pts (49.5%), neurological complications 23 (23.3%), surgical procedures 17 (17.9%) and sepsis 8 (n:9.4%). There was a statistically significant difference between groups (p=0.02). Readmitted individuals had higher APACHE II and SOFA scores compared to initial ICU admission (p<0.01). Overall mortality rate of the readmitted population was 22.1% (n:21). Oxygen mask upon ICU initial discharge, reason of readmission and SOFA 2nd score predicted 72.5% of the variance of poor ICU outcome. 

Intensive Care development has resulted in complex patient care. On general wards, patients can be identified that do not require care on an intensive care unit (ICU) but need more care than can be provided on a general ward 1 . Despite the suggestion that a medium care unit (MCU) promotes efficient and effective care 2 , few hospitals have developed such facilities. Also there is little consensus concerning MCU admission and discharge guidelines 3 .

We investigated the opinion on MCU admission criteria, MCU expansion need, the role of the ICU consultation team and possible advantages of a medical emergency team (MET) by a questionnaire presented to physicians, nursing staff and ward managers in a 852-bed teaching hospital with a 7-bed MCU. On three consecutive days, ten general wards were visited.

Nurses and physicians defined a patient to be MCU eligible in case of medical complexity or when time consuming care was needed. Within the explored time span the need for patient transfer to the MCU was minor (4%). The percentage of MCU patients on the ward throughout the year was estimated to be 6.3% (SD 5.8). Nurses consistently mentioned twice as much MCU eligible patients as physicians and in regard to general wards, the MCU eligibility on surgical wards was assessed twice as high. Nurses expected beneficial patient outcome from an expansion of MCU capacity (53%). The ICU consultation team was valued supplementary (40%) and the added value of a MET team was anticipated (51%).

CONCLUSION. 1) In our hospital, 6% of general ward patients were considered to require complex and lengthy care from both nurses and physicians.

2) The possibility to meet these demands, seems not only to be determined by specific patient needs, but shows to be subject to local organization, population composition, complement of nursing staff and their cumulative skills. 3) Universal criteria and guidelines concerning MCUs are yet to be developed. 

Death on the general floor after discharge from the ICU remains a major concern. We investigated the predictors of post-ICU in-hospital mortality in a large database.

We included patients enrolled in a large, multicenter, observational European study, the SOAP study. Data from all adult patients admitted to the participating ICUs were collected prospectively. Patients were followed up until death, hospital discharge, or for 60 days.

Of the 3147 patients included in the SOAP study, 1729 (54.9%) were discharged to the general floor. The overall in-hospital mortality rate was 24% (n= 747); 4% (n=125) of deaths occurred on the floor. Non-survivors were older, had a higher incidence of hematologic cancer and cirrhosis and greater SAPS II score (and SOFA score on ICU admission, and were more likely to have been admitted for medical reasons than survivors. Residual organ dysfunction/failure on discharge to the general floor was more pronounced in hospital non-survivors than survivors, as reflected by a greater SOFA score on the day of ICU discharge. In a multivariate analysis with hospital mortality as the dependent variable, age (Odds ratio ( CONCLUSION. In addition to age, medical admission, and pre-existing comorbidities on ICU admission, sepsis and/or organ system failure are independently associated with a greater risk of post-ICU death.

Grant acknowledgement. Supported by unrestricted grant from Abbott, Eli Lilly, Baxter, GSK, and Novo Nordisk.

Azevedo M 1 , Silva R 1 , Campello G 1 , Ribeiro L 1 , Dias C 2 , Azevedo L F 2 , Granja C 

Critically ill patients die in the hospital after ICU discharge in a percentage varying within series between 11% and 32%. The associations between post-ICU mortality and several factors inside and outside the ICU have been studied. This study aims to estimate post-ICU mortality and analyse its determinants.

Retrospective analysis of post-ICU mortality in patients admitted between 1 April 2004 and 31 December 2005. Data were collected from an on-going ICU clinical database, including demographic data (gender, age) and ICU data (reason for ICU admission, SAPS II, ICU length of stay, TISS 28 and SOFA score in the last 24 h of ICU stay). A total of 511 patients were admitted, 232 were female gender, with median age of 65 years. 362 patients were admitted for medical reasons, 85 after scheduled surgery, 47 after non-scheduled surgery and 16 for trauma. The median SAPS II was 40 and the median ICU length of stay was 3 days; the median SOFA was 4 and the median TISS-28 was 27. Appropriate descriptive statistics were calculated, comparisons between groups were performed using Chi-square test, non-parametric Mann-Whitney U or Kruskal-Wallis tests as appropriate. For hypothesis testing a value of P<0.05 was considered significant. Statistical analysis was performed using SPSS 14.0 ® software package.

. From a total of 511 patients admitted, 109 (21%) patients died in the ICU. Four patients were discharged directly to their home, and 20 records were lost. From the 378 patients discharged from the ICU, 41 (11%) died in the ward before hospital discharge implying a global hospital mortality of 29%. All patients were discharged from ICU to an intermediate care unit before being discharged to the ward. Patients who died after ICU discharge exhibit a significantly higher age (median 70 years old vs 61 years; p=0.002); a significantly higher SAPS II (median 44 vs 35; p<0.001) and a significantly longer ICU stay (median 7 vs 3; p<0.001) than patients who were discharged alive from hospital. There were no significant differences concerning gender, reason for ICU admission, TISS 28 and SOFA score in the last 24 hours before ICU discharge.

Post-ICU mortality was associated with a higher age, a higher burden of disease at ICU admission and a longer ICU stay. The main factor we think explains the lower post-ICU mortality found in our study, compared with other studies, is the systematic discharge from ICU to an intermediate care unit that we have implemented in our institution. (1), possibly even in sepsis. We hypothesised that increase in LV dyssynchrony is associated with the circulatory dysfunction caused by endotoxin.

In 11 anesthetised juvenile pigs, LV mechanical dyssynchrony was assessed by analysis of long axis segmental volumes measured using a 5 segment conductance catheter (2) . Endotoxin was infused intravenously at 0.25 mcg/kg/hr, together with intravenous fluid resuscitation (Ringers acetate, 20 ml/kg/hr). Systolic ventricular internal flow fraction (IFF) and segmental dyssynchrony fraction (DF) percents were measured before and during endotoxin infusion.

Approximate doubling of MPAP reflects a significant increase in right ventricular (RV) loading during endotoxin infusion. Both systolic DF and IFF increased during peak endotoxin MPAP effects during the first hour, but had even higher mean values after 3 hours of endotoxin exposure. 

Bijuklic K 1 , Kountchev J 1 , Joannidis M 1 1 Internal Medicine, Innsbruck Medical School, Innsbruck, Austria

Levosimendan is a new inodilatator for primary cardiac failure acting by sensitising troponin-C to calcium and opening the ATP-sensitive potassium channels. Levosimendan has also been shown to exhibit immunomodulatory effects which may be an additional pathophysiological mechanism preventing abnormal immune responses in severe heart failure. This study aimed to investigate the effect of Levosimendan on polymorphonuclear leukocytes (PMN).

METHODS. PMN were obtained from venous blood from healthy volunteers. Isolated PMN were incubated with increasing concentrations of Levosimendan ranging from 1ng/ml to 1µg/ml for 2 hours. Afterwards PMN were exposed to fMLP (formly-Met-Leu-Phe) (4µM), PMA (phorbol-12myristate-13-acetate) (4µM) or medium (control), respectively. The production of reactive oxygen species (ROSs) was measured fluorometrically using H2DCF-DA as fluorochrome. Furthermore, to access the effect of Levosimendan on apoptosis and necrosis in PMN annexin-V and propidium-iodide labelled cells were analysed by Flow cytometry.

Levosimendan in a dose dependent manner. The maximal effect of Levosimendan in fMLP-stimulated PMN could be observed at a concentration of 25ng/ml (24% reduction compared to control, p< 0.05). Levosimendan showed an even higher reduction of ROS production in PMA-stimulated PMN (32% reduction at a concentration of 50 ng/ml, p< 0.05). In contrast, Levosimendan did not have an effect on ROS in unstimulated PMN. Furthermore, Levosimendan had no influence on apoptosis in PMN whereas cytotoxic effects could be detected at concentrations higher than 1µg/ml.

Levosimendan effectively suppresses the respiratory burst in PMN stimulated by either fMLP or PMA at concentrations between 10 and 50 ng/ml which correspond to serum levels usually achieved by therapeutic dosage. Levosimendan does not exert an effect on unstimulated PMN. 

Abnormal activation of the ATP-sensitive potassium (KATP) channel contributes to vascular hyporeactivity in sepsis 1 . We have suggested this is due to dysfunctional regulation of the ion channel pore by the sulphonylurea receptor subunit (SUR) 2 . Altered levels of ATP or the ADP:ATP ratio in vascular smooth muscle may be responsible for this dysregulation, though nucleotide levels have not been previously measured in this setting.

At 24 hr and 48 hr timepoints in an in vivo rat model of faecal peritonitis, animals were anaesthetised and the thoracic and abdominal aorta removed. Thoracic aorta was flash frozen in liquid N2 and ATP, ADP and AMP levels later measured using HPLC. Rings of abdominal aorta were placed in an organ bath. Contractility to the alpha-agonist phenylephrine (PE,10 -9 -10 -5 M), and relaxation to the KATP channel opener levcromakalim, (10 -9 -10 -5 M), were used to examine the KATP channel. In separate studies, the pore-blocker PNU 37883A (10 µM), and the SUR inhibitor glibenclamide (10 µM,) were added to pre-contracted rings (using PE, 1µM) which were then relaxed with levcromakalim (3µM). This enabled assessment of the ability to block the channel via either the pore or SUR.

Abdominal aorta from septic rats were hyporeactive to PE at 24 hr and 48 hr (n = 4, P < 0.01), with increased levcromakalim efficacy at both 24 hr (EC50 1.9 vs 3.4 x10 -7 M, n = 4, P < 0.001) and 48 hr (EC50 1.1 vs 2.3 x10 -7 M, n = 4, P = 0.02). Surprisingly, the potency of both PNU 37883A (38% vs 90% reversal, n = 5, P < 0.05) and glibenclamide (6% vs 50% reversal, n = 5, P < 0.01) were reduced at 24 hr, but not 48 hr. ATP, ADP and AMP levels were unchanged at 24 hr.

Glibenclamide did not close the KATP channel in aorta taken from this in vivo model at 24hr. This finding is consistent with a recent study in septic shock patients 3 . The pore-blocker also failed to close the channel at this timepoint. However, the ability of these agents to close the channel was regained at 48 hr. This may be due to either reversal of the mechanism present at 24 hr, or to synthesis of new KATP channels. The data also demonstrate that the reduced potency of the channel closers at 24 hr is independent of nucleotide phosphate levels. In summary, different mechanisms are responsible for vascular hyporeactivity during the septic process. 

Nygren A 1 , Thorén A 1 , Ricksten S 1 1 Cardiothoracic Anaesthesia and Intensive Care, Sahlgrenska University Hospital, Göteborg, Sweden

Clinical septic shock and the vasodilatory shock syndrome after cardiac surgery are characterized by a profound arteriolar vasodilation. Data on intestinal mucosal perfusion in patients with vasodilatory shock are lacking.

Ten patients with vasodilatory shock treated with norepinephrine (NE) after cardiac surgery were compared to ten uncomplicated cardiac surgery patients in a prospective controlled observational study. Mean arterial pressure was 65-75mmHg in both groups. Data on jejunal mucosal perfusion (JMP) (laser Doppler flowmetry), gastric-arterial pCO2 gradient (tonometry) and splanchnic oxygen demand-supply relationship (hepatic vein catheter) and systemic hemodynamics (PA catheter) were obtained during 30 min.

Infusion rate of NE was 0.38±0.23 microg/kg/min. Cardiac index (CI), heart rate (40%, p<0.001), and arterial lactate (60%, p<0.05), systemic oxygen consumption (26%, p<0.001) and delivery (17%, p<0.05) were higher while systemic vascular resistance index (-14%, p<0.05) was lower in the vasodilatory shock group. JMP (61%), gastric-arterial pCO2 gradient (150%) and splanchnic O2 extraction were higher in the vasodilatory shock group while splanchnic lactate extraction did not differ between the groups. 

Changes of blood reology occurring in the acute period of injury and blood loss are the closely related to different thrombotic and thrombembolic complications. The aim of this study was to investigate the blood reological effects of perfluorane in patients with severe injury and blood loss.

We examined and treated 64 patients with severe injury. All of them received a colloid and cristsalloid solutions. Perfluoran (6-10 ml/kg) was added to 33 patients (study group). A control group (31 pts) received isotonic saline at the same dose. We investigated erythrocytic rigidity index, blood and plasma viscosity, intravascular aggregation and ristomycin platelet aggregation.

It was istablished that within the first 48 hours after injury and blood loss there were worse microrheological parameters, which appeared as the increased erythrocitic rigidity and aggregation. At the same time blood and plasma viscosity were decreased as a result of fluid therapy. Perfluoran caused significant decreases ristomycin platelet aggregation during 2 hours after administration (p<0.05; ttest). Erythrocitic rigidity, intravascular aggregation and plasma viscosity also decreased compared to control (p<0.05). These changes were observed for 24 hours after perfluoran administration. Blood viscosity did not differ between groups. Analisis of the treatment results revealed significant reducing the incidence of thrombotic and thrombembolic complications in patients treated by perfluoran.

CONCLUSION. 1) Perfluoran used in a dose 6-10 ml/kg in patients with severe injury and blood loss improved blood rheological properties by stabilizing the microrheology indices-erythrocytic rigidity and aggregation and by decreasing plasma viscosity.

2) The administration of perfluorane in the acute period of injury made it possible to reduce the incidence of thrombotic and thrombembolic complications.

Fund of Assistance to the Russian Medicine. METHODS. 50 patients admitted to the ICU were prospectively recruited in the post-resuscitation phase (>12-24h after admission). StO2 was measured in every patient in the thenar eminency by NIRS for 10 minutes. At the same time, clinical data (BP, HR, RR, T • and SpO2) as well as analytical data (lactate, BD, Hb and SvcO2) were registered. Correlation and study of means for independent variables by t-Student were analysed. Correlations were also analysed for the subgroup of patients with anaemia (defined as Hb< 9g/dL).

Correlation of thenar StO2 and other measured variables (BP, HR, lactate, BD, Hb and SvcO2) could not be shown for the group of patients as a whole. A tendency for lower StO2 in patients with higher BD was observed, but without reaching statistical significance. A relationship between thenar StO2 and the rest of the studied variables could not be found in the subgroup of anaemic patients (n=35), although there was again a tendency for lower StO2 in patients with higher lactate and BD, without statistical signification.

According to the results of our study, we cannot validate NIRS as an estimation or equivalence to analytical variables that are commonly used for evaluation of tissular hypoperfusion in patiens in a post-resuscitation stage. Again, further studies are necessary. 

Schwarte L A 1 , Van Bommel J 2 , Siegemund M 2 , Fournell A 2 , Ince C 2 1 Anesthesiology, University Hospital Duesseldorf, Duesseldorf, Germany, 2 Department of Physiology, Academic Medical Centre, Amsterdam, Netherlands

Mechanical ventilation with positive end-expiratory pressure (PEEP) remains the basic concept for respiratory support in critically ill patients. However, PEEP depresses systemic hemodynamics, e.g. cardiac output, but microcirculatory effects of PEEP remain unclear. Since impaired intestinal microcirculatory perfusion and oxygenation may promote sepsis and multiple organ failure, we studied the per se effects of PEEP on intestinal microcirculation. Herein, impairment of the intestinal microcirculation could result from reduced mesenteric blood flow and/or increased trans-intestinal oxygen shunting. n=6) were anesthetized, intubated and mechanically ventilated. The animals were instrumented for determination of cardiac output and superior mesenteric artery blood flow, and the following intestinal microcirculatory measurements: At the ileum mucosa and serosa, we simultaneously measured microvascular PO2 (µPO2, Pd-porphyrin phosphorescence quenching), microvascular HbO2-saturation (µHbO2, reflectance spectrophotometry) and microcirculatory perfusion (laser Doppler). To quantify trans-intestinal oxygen shunting, we determined the gradient between microcirculatory tissue oxygenation (mucosal+serosal µPO2) and corresponding mesenteric venous PO2 (PO2-gap). After baseline measurements, PEEP was increased from 0 to 15, 20 and 25 mbar. Statistics: Steady state data, mean±SEM, Fisher PLSD, p<0.05.

. PEEP pressure-dependently reduced microcirculatory mucosal µPO2 from 18±2 (baseline) to 10±1 (PEEP 25 mbar), and serosal µPO2 from 32±2 to 20±2 mmHg. Correspondingly, PEEP also reduced mucosal µHbO2 from 49±10% to 33±5% and serosal µHbO2 from 62±6 to 49±3%. Furthermore, PEEP impaired microcirculatory perfusion of the mucosa (from 1.8±0.3 to 1.1±0.2 units) and serosa (from 5.0±0.7 to 3.7±0.6 units). The O2-gap started positive at baseline (+8.2 mmHg), zeroed at PEEP 15 mbar (+0.5 mmHg) and turned negative during higher PEEP-levels (-10.8 mmHg at PEEP 25 mbar).

In summary, PEEP depressed all measures of intestinal microcirculatory oxygenation and perfusion, both of the mucosa and the serosa. Furthermore, our data suggest that -beside reduction of mesenteric blood flow-an increase in trans-intestinal oxygen shunting contributes herein. 

Long periods of gut ischemia can cause organ failure. The protein breakdown occurring in the muscular layer and in the mucosa of the intestinal wall may release amino acids into the circulation or into the intestinal lumen. The aim of this study was to determine whether tyrosine may be released from the cells into the lumen during different modes of intestinal injury, and whether this release can be used as a marker of gut ischemia.

In 24 anesthetized New-Zealand rabbits, superior mesenteric artery (SMA) and aortic flows were measured. A segment from the ileum was isolated using two multilumen catheters with inflated balloons to create a closed segment for perfusion. A tonometry catheter was placed in the ileum. After baseline measurements, animals were allocated into four groups: a sham group without intervention (group I, n=6); group II, given lipopolysaccharide (1000 µg/kg E. coli i-055:B5 EV); group III, submitted to SMA ligation only (n = 8) and group IV, submitted to SMA ligation followed by 1 h of reperfusion (n=5). The concentrations of lactate and tyrosine (fluorometry) were determined in serum and in gut luminal perfusate (GLP).

Tyrosine concentrations increased significantly in the GLP in group III (From 10.1 ± 7.6 to 93 ± 63 mM/mL at 2 hours) and were significantly higher than in group I (25.6 ± 24 mM/mL), group II (16 ± 12 mM/mL) and group IV (11 ± 13 mM/mL) (p <0.05 for all).

Tyrosine is released from cells into the intestinal lumen during ischemia. Measurements of tyrosine levels in the gut may provide a marker of gut injury. 

Heyer L 1 , Rezlan E 1 , Madadaki C 1 , Rabuel C 1 , Schourando P 1 , Payen D 1 1 Anesthesia and CCM, Hopital Lariboisiere -APHP -Universite Paris 7, paris, France

We study the micro-oxygenation impairment of peripheral muscle in post-partum haemorrhagic shock.

Tissue oxygenation and perfusion of thenar muscle were assessed using near infrared spectrometry technique (NIRS) at ICU admission in 33 consecutive post-partum haemorrhagic shock women. Tissue O2 saturation was measured at baseline (StO2) and following forearm arterial occlusion. The slope of the decrease in StO2 during occlusion (S-Occ), the slope of the increase in StO2 following occlusion release (S-Rel) and the deficit of post-ischemic oxygenation recruitment (Deficit) in response to transient ischemia were measured. The haemorrhagic shock group (HEMORRHAGE) was divided accordingly to myocardial ischemia defined by a Troponin I value above 0.08 ng/ml (TROPO+ and TROPO-) . Values obtains in a control group of 11 post-partum women with normal delivery (CON-TROL) were used for comparison.

Tissue oxygenation and perfusion were significantly impaired in HAEMORRHGE group but an additional impairment was observed in TROPO+ sub-group only for S-Occ and S-Rel (p<0.004 et p<0.0001 respectively). In multivariable analysis, the increase in lactate acid and the decrease in S-Rel were the two independent factors associated to Troponin I elevation (p=0.04). 

In 24 elective patients, systemic hemodynamics were recorded at 5 min before and at 10 min during CPB. Side-stream dark field imaging was used to continuously visualize the changes in sublingual microcirculatory hemodynamics in the same location (n=12). Changes were quantified using computer analysis software. Reflectance spectrophotometry was used to measure sublingual microvascular hemoglobin saturation (µHbSO2) in a separate group of patients (n=12). Blood gases, acid-base status, glucose and lactate were determined at the same timepoints in all patients.

The microcirculatory blood flow changed from pulsatile to continuous flow and could be quantified by use of space-time diagrams. At 10 min of CPB, the blood volume output increased while MAP decreased to shock values (49.6±6.8 mmHg), arterial blood temperature decreased to 32.8±1.1

• C and Hb concentration decreased due to hemodilution (hematocrit 20-25%). Functional capillary density (<10 µm) and functional venular density (10-50 µm) both decreased significantly (6.3±2.1 to 3.6±1.6 mm/mm 2 ; p<0.01 respectively, 12.1±2.5 to 9.9±2.6 mm/mm 2 ; p<0.05), showing a 43% fall-out of nutritive capillaries and 19% of small venules. In the remaining perfused venules, the blood velocity increased (349±201 to 563±227 µm/s; p<0.05). The µHbSO2 of the RBC in the remaining blood vessels however increased during CPB (47.9±6.1 to 53.7±5.8%; p<0.001). Blood glucose and pH remained unchanged, while HCO3 and BE slightly decreased. Lactate increased significantly from 1.3±0.6 (low-normal) to 1.8±0.8 mmol/l (high-normal) (p<0.001).

CONCLUSION. The circulatory response to CPB, associated with hypotension, inflammation, metabolic distress and microcirculatory distributive alterations, allow it to be classified as distributive shock. The increased diffusion distance to cells caused by fall-out of capillaries and the reduced O2 off-load capacity due to reduction in exchange surface area explain the raised lactate in combination with the raised µHbSO2. Levosimendan (LS) improves survival in patients with congestive heart failure and seems to improve also renal function by a vasorenal effect 1 . As clinical data on fluid balance are scanty the aim of this study is to evaluate in cardiogenic shock -1 the haemodynamic and volumetric response to LS 2 the changes of extravascular lung water (EVLWI) 2-the effect of LS on diuresis.

METHODS. 20 critical care patients with cardiogenic shock were studied. All patients were mechanically ventilated and connected to an integrated monitoring system (PICCO system/Agilent) by a fiberoptic arterial catheter (pv 2014L16) and by a Swan-Ganz catheter. In all patients LS was infused at 20 micr/Kg/ min in 20 minutes and then at 0.1 micr/Kg min for 24 hours. At basal time (T0), 6 hours (T1),12 hours (T2), 24 hours (T3) during LS infusion the main haemodynamic and volumetric data were evaluated. All data are expressed as mean ± SD. ANOVA test for RM was used to compare changes during times study. P< 0.05 was statistically significant. Table 1 presents the main hemodynamic, volumetric data and 24 hours diuresis. (1), there are no data available regarding its effects under early endotoxaemia, which induces pronounced vasoconstriction with mucosal hypoxia of the gut (2). We investigated the effects of clonidin in endotoxaemia.

After ethical approval 16 anaesthetized (ketamine, flunitrazepam, fentanyl, rocuronium) and ventilated pigs with ultrasonic flow probe around the sup. mesenteric artery, catheters in the femoral artery, mesenteric and portal vein and pulmonary artery and a jejunal tonometer were studied. Tissue surface pO2 was measured by a multiwire surface electrode (3). The intrathoracic blood volume was measured by the COLD ® System. Animals were randomly assigned in two groups: group 1 = CE received basic fluid administration of 10-12 ml/kg h cristalloids, group 2 = CEV received basic fluid administration and continuous crystalloid substitution to maintain ITBV at baseline level. 2h after preparation baseline values (t0) were taken, 1h after clonidin infusion (2 µg kg-1 bolus and 2 µg kg-1 h-1) endotoxin (0.5 µg kg-1 h-1) was given via portal vein. Measurements were made 3 (t1) and 6 (t2) h after endotoxin application. Dept. of Anesthesiology, University Hospital, Duesseldorf, Germany

Adequate microvascular oxygenation of the gastrointestinal mucosa is crucial to maintain an intact mucosal barrier function [1] . The effects of vasodilating drugs on the gastric mucosal microvascular hemoglobin oxygention (HbO2muc) is unclear. Thus we studied the effects of the vasodilators dihydralazine and nitrogylcerine on HbO2muc in anesthetized dogs.

Chronically instrumented dogs (∼30 kg bodyweight, n=12 experiments, with permission) were anesthetized (propofol 15 mg/kg/h) and mechanically ventilated (FiO2 0.3, etCO2 35 mmHg). We measured systemic hemodynamics (mean arterial blood pressure, MAP), systemic oxygen delivery (DO2) and regional microvascular hemoglobin oxygenation of the gastric mucosa (HbO2muc, reflection sprectrophotometry, EMPHO-II). After a stabilization period and obtaining baseline values, equihypotensive dosages of dihydralazine and nitroglycerine were infused. Statistics: Fisher's PLSD and t-test, p<0.05.

In equihypotensive dosages (MAP from ∼80 to ∼65 mmHg in both groups, no difference between groups) only dihydralazine increased HbO2muc (from 64+/-2 to 70+/-2%), whereas nitroglycerine had no effect (63+/-2 to 61+/-3%). At the systemic level, only dihydralazine increased DO2 (from 15+/-1 to 26+/-2 ml/kg/min), whereas nitroglycerine had no effect (14+/-1 to 16+/-2 ml/kg/min). Arterial pH (∼7.37) and lactate (∼1.0 mmol/l) remained stable in both groups.

CONCLUSION. If our results may be transferred to the clinical setting, then dihydralazine -in contrast to nitroglycerine-may improve gastrointestinal mucosal oxygenation and thereby support an intact gastrointestinal barrier function in patients at risk. 

Splanchnic hypoxia, especially of the intestinal mucosa may be a major contributing factor to the development of systemic inflammatory response syndrome (SIRS) and sepsis. (1) Increasing the arterial oxygen partial pressure might avoid or at least to attenuate splanchnic hypoxia. However, there are no data available describing the effect of increased arterial oxygen partial pressure on splanchnic tissue oxygenation.

Following approval by the local animal ethics committee 10 anaesthetised, ventilated and acutely instrumented pigs (catheterisation of pulmonary artery, portal vein, hepatic vein and femoral artery catheter, intestinal tonometer and pO2-electrode onto liver, intestinal serosa and mucosa, ultrasonic flow probes around hepatic artery, sup. mesenteric artery and portal vein (2, 3) ) were treated with increasing inspiratory concentrations of oxygen to reach paO2 levels of 15, 25, 45 and 65 kPa. Measurements of haemodynamic and parameters of oxygenation were made at baseline, and 30 min after changing each inspiratory oxygen concentration. 

Pheochromocytoma is classically associated with paroxysmal hypertension but can present with various non-specific clinical manifestations. A 50-year-old woman came to the ED for epigastric-retrosternal discomfort, right lumbar pain, vomiting and impending sense of doom. BP was 90/50, HR 120, Tc 38.9 • C. The patient rapidly developed cardiogenic shock with respiratory failure, DIC, massive rhabdomyolysis with acute renal failure and hepatic cytolysis. CT showed bilateral pleural effusions with basal infiltrates and a right suprarenal mass. Echocardiography showed severe biventricular dysfunction. Despite massive support refractory shock with elevation of ventricular filling pressures ensued. To unload the failing heart, veno-arterial extracorporeal bypass was started.

A 24 F cannula was positioned percutaneously into the right femural vein, and a 13 F arterial catheter was inserted into the right femural artery. Venous blood was drained using a centrifugal pump (Jostra Rotaflow, Maquet), passed through an artificial lung (Jostra Quadrox, Maquet) and then returned into the arterial circulation. Extracorporeal blood flow was set to 5 L/min and gas flow through the artificial lung to 6 L/min of oxygen. Anticoagulation was maintained with continuous infusion of unfractionated heparin. Liver function was monitored with IGClearance (LiMon, Pulsion).

Rapid hemodynamic improvement was observed, with reduction of filling pressures, increase of SvO2 and correction of lactic acidosis. Treatment with vasoactive dugs was discontinued. Only after institution of the bypass wide fluctuations in blood pressure (typical of pheochromocytoma) became evident and were treated with phentolamine. After 4 days the EC support was discontinued and the patient underwent definitive treatment of the tumor (angiographic embolization followed by right adrenalectomy). After removal of the tumor CVVH was discontinued and a liver function progressively improved. Discharge from ICU was 29 days after admission.

CONCLUSION. This is the second report of a pheochromocytoma-induced refractory cardiogenic shock successfully treated with extracorporeal cardiopulmonary support.

Takahashi A 1 , Kawashima T 1 , Ishii N 1 , Nakao H 1 , Okada N 1 , Toyama K 1 , Yoshida T 1 , Okashiro M 1 , Ando Y 1 , Ri C 1 1 Emergency and Disaster Medicine, Kobe University Graduate School of Medicine, Kobe, Japan INTRODUCTION. Swan-Ganz catheter (S-G) is mostly used for monitoring hemodynamics in intensive care stage. In this study, we adopted PC for monitoring, which is said less invasive, then examined its effectiveness compared to S-G. This is a report of usefulness of PC and TFM. Between December 2002 and February 2006, 44 patients with circulation and respiratory failure were performed hemodynamic monitoring with PC. CI, SVRI, ITBVI, and EVLWI were measured in 438 points and the correlation between ELWI and P/F or ITBVI were examined. Among 44 cases, S-G was used at the same time for 7 cases, then CI, SVRI, and PAWP were monitored, in order to compare with the measured value by PC. For 4 cases, TFM was used at the same time, then CI, TPRI, and TFC were compared with PC measurement.

In the comparison with PC and S-G, the correlation of CI show 0.84852 as correlation coefficient and 2.33015x10-10 as risk rate. The correlation with SVRI show 0.88252 as correlation coefficient, and 5.15376x10-12 as risk rate. The correlation with the preload indicators, ITBVI and PAWP, show no significant correlation. The correlation of EVLWI and P/F show -0.40484 as correlation coefficient, and 1.05400x10-18 as risk rate. The correlation of EVLWI and ITBVI, a total of 438, show 0.1351 as correlation coefficient and 0.00457 as risk rate. However, in the cases with ITBVI below 1000that is, the cases under the normal value, correlation coefficient is 0.02956 and risk rate is 0.74547. Besides, the cases with ITBVI 1001 and over, that is, excessive water ingestion case shows 0.11641as correlation coefficient and 0.04501 as risk rate. In the comparison with PC and TFM, correlation coefficient of CI is 0.65445, and risk rate is 0.001743. The correlation coefficient of SVRI and TPRI is 0.99999, and risk rate is 2.75635x10-31. The preload indicators, TFC and ITBVI, show no correlation.

CONCLUSION. PC can be considered as an alternative of S-G for hemodynamic monitoring. EVLWI shows very good negative correlation with P/F, which is useful for the indicator of oxygenation. For the case with high level of EVLWI with poor oxygenation and that of ITBVI with excessive water ingestion, it is suggested that hydration control is also effective. TFM recognize the correlation with PC, we could expect that they can be more effective indicators.

Picard J 1 , Bedague D 1 , Carcey J 1 , Hallé C 1 , Oddoux M 1 , Ollinet C 1 , Jacquot C 1 , Payen J F 1 1 Département Anesthésie Réanimation 1, CHU, Grenoble, France (Intensive Care Med 2005; 31:1195 . Br J Anaesth 2002 88:65) . We evaluated the ability of ED to assess vascular fluid responsiveness in pts undergoing spinal surgery (SS) in prone position (PP).

After approval of the local Ethical Committee and informed consent 26 pts (52±16 yrs, 12 M/14 F, ASA grade 1-2) were admitted for SS with laminectomy. After induction (target-controlled infusion of propofol (1.6 -4 mg/l) and remifentanil (0.1 -0.4 mcg/kg/min)) a TED probe (ODM2, Deltex Medical, UK) was inserted through the nasal route. A vascular fluid challenge (FC) (200 ml Voluven ® over 10 min) was achieved while the pt was in PP, with no administration of vasoactive drugs. We measured and calculated TED parameters. Data (mean ± SD) obtained before and at the end of FC were compared by paired Student's t-test (p<0.05).

All variables increased significantly after FC while Mean Arterial Pressure (MAP) and HR remained unchanged (table) . There was a large interindividual variability among pts. Defining responder pts (RP; n=15) according to their DeltaSV>10%, there was no significant difference between RP and non RP (n=11) before FC. 

Raineri S 1 , Canzio D 1 , Buscemi R 1 , Pirri S 1 , Pugliese G 1 , Giambartino F 1 , Giarratano A 1 1 Anesthesia and Intensive care, University Hospital, Palermo, Italy

Our aims were to validate the role of Narcotrend (NT) for monitoring the sedation and assessment in critically ill patients and to compare it with an international score as MSAT.

METHODS. 32 patients were enrolled, admitted to ICU more than 48 hours, treated with midazolam (M) e.v.bolo (0.15±0.003mg/kg) followed by continued infusion of M (0.03±0.006mg/kg/h) and remifentanil (R)(0.05±0.01mcg/kg/min) according to international guidelines. We monitorized SpO2, NIBP, EtCO2, HR, and esophageal T • C. NT and MSAT values were registered, like coupled data, each 5 min after pain stimulation (PS) (endotracheal aspiration) until 40 min. We considered an adequate stage of analgo-sedation for NT index score beetween D0-E0, and MSAT value of 2(motor activity), 3-4(arousal) (See Tab.1).

Mean drugs dosage was 0.078±0.026 mcg/kg/min for R and 0.066±0.02 mg/kg/h for M. All patients showed an adequate level of analgo-sedation with NT index score between D0-E0 corresponding to MSAT values of 2/3-4, for 85% of observation time. We registered four cases of over and under sedation, pointed out by Narcotrend in 5 min and MSAT after 15 min, and treated with correction of infusion dose. No significant variation of hemodinamic parameters were registered. 

Arterial stiffness (AS) is a major risk factor of cardiovascular diseases. Augmentation of the central aortic pressure wave (PW) provides an assessment of AS and can be expressed numerically as a percentage of pulse pressure (augmentation index [AIx]). AIx depends on ventricular ejection, and its measurement should be corrected for heart rate (HR). Quantitative analysis of peripheral and central PWs is possible using the pressure recording analytical method (PRAM). We investigated the relationship between HR and aortic and radial AIx, assessed by PRAM over a physiological range of HRs.

Five patients with a mean age of 65 underwent coronary surgery were studied. Radial artery cannulation was established as a routine. After the chest had been opened a 20G catheter was placed in the ascending aorta, and a temporary cardiac pacemaker was positioned. Subjects were paced at 80 beats/min. The pacing rate was increased from 80 to 110 beats/min in 5 beat increments. Radial and aortic PW were analysed simultaneously by PRAM, and AIx was calculated during incremental pacing.

Basal mean HR was 73±5 beats/min; radial systolic, diastolic and mean arterial pressures were 138±9, 51±7, and 80±8, respectively; aortic systolic, diastolic, and mean blood pressures were 100±7, 44±6, 67±5, respectively. Resting radial and aortic AIx were 28% and 25%, respectively. Incremental pacing from 80 to 110 beats/min resulted in an increase in peripheral and aortic blood pressures. Aortic AIx declined linearly with increasing HR (mean AIx% = 25, 22, 19, 16, 15, 14, 13 at HR = 80, 85, 90, 95, 100, 105, and 110, respectively) . Radial AIx showed similar results.

AIx could be potentially useful in identifying patients who may benefit from treatment aimed at reducing AS. We demonstrated that changing HR alters the relative timing of the return of the reflected wave to ascending aorta, and, therefore, that AIx is linearly and inversely related to HR. A better understanding of the interaction between HR and AIx may be of use in studies employing PWA to assess systemic AS. This is the first study assessing AS by PRAM. 

Borrat X 1 , Sanchez-Etayo G 2 , Mercadal J 2 , Tercero F 2 , De la Riba N 2 , Israel T 3 , Adalia R 2 , Zabala E 2 1 Anesthesiology, Hospital Clinic, 2 Anesthesiology, 3 Nursery, Hospital Clinic Barcelona, barcelona, Spain

There is evidence that the development of MOF, and ultimately death is closely related to impaired microcirculation. As a preliminary study, we analyzed the StO2 in three different types of patients to asses the differences in the microcirculation pattern between septic, healthy and hemorrhagic shock patients.

We studied twelve patients (four polytraumatized (PLT), four shock septic (SEPTIC) patients and four healthy volunteers (CONTROL)) and analyzed the microcirculation in the following manner: Thenar muscle StO2 was measured non invasively by NIRS (InSpectra, Hutchinson Technology, USA) before and during upper limb ischemia. The basal StO2 and the rate of the StO2 decrease (down slope) were analyzed.

CONCLUSION. As the rate of StO2 down slope has been related to the rate of O2 consumption, we can hypothesize an altered O2 uptake which can result in a tissular hypoxia both in SEPTIC and PLT patients. Moreover basal StO2 may suggest that the intrinsic mechanism of the altered O2 consumption may be: microvascular shunting for sepsis (high basal StO2) and intense vasoconstriction for PLT patients (low basal StO2). However further research is required to confirm these assumptions. StO2 monitoring may be useful as an indicator of a diminished consumption as a marker of hypoxia during septic or hemorrhagic shock. 

Takeuchi A 1 , Hirose M 2 , Shinbo T 2 , Imai M 1 , Mamorita N 1 , Ikeda N 1 1 Medical informatics, AHS, 2 Clinical Engineering, AHS, Kitasato University, Sagamihara, Japan

Although medical devices have generated alarm signals since the 1980s, a proliferation of alarm signals, particularly of the auditory variety, is causing problems [1] . The purpose of this study was to develop an interactive software system of alarm sounds to present, train, recognize and share problems about alarm sounds.

The system is composed of the alarm sound database, recorded alarm sounds and the alarm sound simulator. The alarm conditions were arbitrarily induced by attaching physical models to medical devices. The system that integrated an alarm sound database and simulator was used to assess the ability to identify the monitor that sounded the alarm for the medical staff.

Sixty alarm sound files (40MB in total) were recorded from 41 medical devices made by 28 companies. The alarm sound database was created in an Excel file (ASDB.xls 170kB, 40MB with photos). Nine columns stored information of each alarm sound, type of device, name of manufacturer, machine model, alarm priority, alarm sound profile, fundamental frequency, A-weighted sound pressure level, alarm condition and file name of alarm sound. The spreadsheet consisted of 91 lines. There were a pair of similar alarm sounds that could not easily be distinguished and had a different priority. The alarm sound profiles were classified by musical impression: a continuous simple sound without a pause (Buzzer, 14 cases), a simple sound with a pause (Beep, 22) , a continuous harmonized sound (Siren, 8) , bursts composed of various pulses (Pulses, 35), a damping burst pattern (Chime, 12). An alarm sound simulator (AlmSS)(260kB, VC++) was constructed with two modules for simultaneously playing alarm sound files and for designing new alarm sounds. The AlmSS was used in the assessing procedure to determine whether 19 clinical engineer can identify 14 alarm sounds only by their distinctive sounds. The overall correct identification rate of the alarm sounds was 46%.

We have developed an interactive software package of alarm sounds. The AlmSS was useful for replaying alarm sounds simultaneously and designing new alarm sounds interactively. Medical personnel and manufactures should recognize the confusing situations and share the information of alarm sounds to improve with cacophony of the current noisy environment. 

Small blood vessel size limits the ability to measure cardiac output (CO) in pediatric and neonatal patients. The problem was overcome by: a) using ultrasound velocity (UV) dilution technology to produce dilution curves by injecting isotonic saline (UV=1533 m/sec) into blood (UV=1570-90m/sec); b) using an extracorporeal AV tubing loop connected to extant arterial and venous catheters through a peristaltic pump. The purpose of the study was to validate dilution measurements with direct readings in different size animals with "gold-standard" transit time technology.

Six rats (230g-712g), three pigs (15-21 kg) and a sheep (65 kg) were instrumented on the ascending aorta or pulmonary artery with transit-time perivascular flow probes (2.5mm, 12mm and 24mm, Transonic Systems Inc.). A disposable AV loop, filled with heparinized saline, was connected between the arterial and venous catheters inserted in the animal. A peristaltic pump circulated blood (4-12 ml/min) from the artery to the vein for 3-5 min to perform 2-3 measurements. Two UV sensors were clamped onto the arterial and venous sides of the AV loop. Isotonic saline was injected into the AV loop upstream from the venous sensor. CO was calculated from the dilution curve recorded by arterial sensor and by a HCP101 Monitor (Transonic Systems Inc.).

In all size animals, ultrasound dilution and transit time readings closely agreed: 

CONCLUSION. There is a benefit in the reduction of duration of mechanical ventilation in the early tracheostomised group. Therefore, tracheostomy and the timing of tracheostomy must be individualised to obtain maximum risk benefit ratio.

Bunker N 1 , Kan M 1 , Watters G 2 , Kinnear J 1 , Higgins D 1 1 Critical Care, 2 ENT, Southend Hospital, Essex, United Kingdom

Tracheal ring fracture (TRF) is a recognised complication of percutaneous dilational tracheostomy (PDT). The aim of this study was to assess whether TRF resulted in an increased incidence of tracheal stenosis.

Over a 26 month period 207 PDT's were performed using either the blue rhino or percutwist techniques. Fibreoptic guidance was used in all cases and allowed identification of TRF's. Patients who sustained a fracture were reviewed at 6 months by a consultant ENT surgeon using laryngotracheoscopy to assess the presence and degree of tracheal stenosis. In addition specific symptoms of upper airway narrowing were elicited.

The results are shown in table 1. Sixteen patients with TRF were identified, at six months 6 of these had died and in the remianing 10 there was no clinical or endoscopic evidence of tracheal stenosis. Laryngotracheoscopy revealed an anterior mucosal ridge in one patient, but there was no significant stenosis. CONCLUSION. This is the only follow-up study designed to specifically assess whether TRF during PDT is associated with an increased risk of tracheal stenosis. Previous studies have identified TRF as a common complication of the percutwist technique 1 , as well as tracheal stenosis following poor placement of the tracheostomy 2 . It is therefore suprising that there was no evidence of tracheal stenosis in our study, which may be explained by the low placement of the tracheostomies in our series (below the first tracheal ring). 

For a period of 4 months, after all successful PT performed in our unit, we visualized the trachea proximal to the TT using a bronchoscope inserted through the oral endotracheal tube. When the fenestration was seen in the trachea we counted the number of fenestration holes visible (maximum of 7 for a FTTT). A picture of the posterior part of the TT was taken. We also recorded evidence of tracheal deformation, trauma and immediate complications.

. 21 PT were performed, all with a Cook Blue Rhino percutaneous tracheostomy kit. In 20 cases a FTTT was used and in 1 case a Smiths Adjustable Flange tube was used. A size 8 tube was used in all 8 female patients and a size 9 tube was used for all 13 male patients. In 11 out of 20 cases (55%) the fenestration was not seen in the trachea and for another 6 patients only one hole was seen in the trachea. Only in one patient (5%) were all 7 holes of the fenestration within the tracheal lumen (table 1). The TT cuff was visible in all cases. The posterior part of the TT was visible in 18 of 21 cases (86%). In 3 cases only the cuff was seen within the trachea. In 5 cases, a severe deformation of the trachea was seen, with a significantly decreased anterior-posterior diameter. In 2 cases, the position of the FTTT was considered to be inappropriate and was changed for a Smiths Adjustable Flange Tracheostomy Tube. In 2 cases, air bubbles could be seen emerging from the fenestration holes, despite the presence of a non-fenestrated inner cannula. One patient sustained a fractured tracheal cartilage and another patient's procedure was complicated by surgical emphysema. CONCLUSION. We suggest that when a fenestrated TT is used, its position should be visualized using a bronchoscope. This may lead to a clinical decision to change the type of TT used. Further studies are required to assess the impact of the position of the fenestration of the TT on the weaning of the patient and whether the position of the fenestration alters with time.

Silva P L 1 , Barata J 2 , Delerue F 1 , Lowmann C 1 1 UMD -Medicina II, 2 Medicina, Hospital Garcia de Orta, Almada, Portugal

Patients with tracheostomy have a high morbility and mortality. The management of tracheostomy tube is a complex process that required careful coordination and consistent follow-up.

The authors made a retrospective study of patients with tracheostomy, admited in the years 2004 to 2005 in a Intermediary Unit (UMD of Garcia de Orta Hospital in Almada, Portugal).

. 36 patients constitue our group, with 36.1% females and 63.8% older than 61 years. 86.1% of patients were transfered from a Intensive Care Unit with a SAPS II at admission in the UMD greater then 30 in 55.5% of patients. All patients were ventilated with positive pressure previously to tracheotomy, with more than 50% of our group ventilated less than 5 weeks. The major reason of mecanical ventilation were pneumonia with or without sepsis. The pathologies associated to our patients were mainly related to cardiovascular and to chronic pulmonary disesases. The major indicacion for tracheostomy in our patients were the need of secrecion remove and a long term positive pressure ventilation. The type of tubes used were Shiley cuffed in 35 patients. The tube of tracheostomy was change until 4 -5 weeks in 25 patients, removed in 1 and never occurred in 5 patients who died previously. All the tubes were changed in the UMD, 92% of then on the firsts 2 weeks of admission in this unit. The removal of the tracheostomy occurred until 2 weeks of the change in 93.3% of the patients, but as soon as possible. In all this proccess the physiotherapy had a fundamental work. 27.7% of the patients of our group (10) died in the hospital.

The tracheostomy is not a disease by it self, but the result of a serious of diseases with a high morbidity. This is reflected in the SAPS II at admission and by the mortality of this patients. In all this proccess the physiotherapy had a fundamental work. 

Although tracheostomy is a common procedure in patients under prolonged mechanical ventilation, its indication and timing remain controversial. It has been suggested that the procedure reduces the duration of mechanical ventilation and intensive care unit length of stay (LOS). Nevertheless, no well designed prospective clinical study has proved this hypothesis to date. We propose to analyze the impact of tracheostomy in the prognosis of patients under mechanical ventilation.

We followed a concurrent cohort of 88 patients in invasive ventilation for more than 48 hours, between 18 and 95 years old. Patients tracheostomized before admission in the intensive care unit (ICU), with Coma Glasgow Scale chronically less than 8, advanced stage neoplasm or neuromuscular disease were excluded. We compared the tracheostomy group (n=38) with the control (endotracheal tube) group (n=50). The primary endpoints were ventilation duration and ICU LOS. The secondary endpoints were mortality and incidence of ventilation associated pneumonia (VAP). Categorical variables were compared by the chi square test and the continuous variables were compared by the T Student test. Potential biases were controlled by multiple linear regression.

RESULTS. The incidence of tracheostomy was 43% (IC=33-54%). In the control and tracheostomy groups, the measured variables were respectively as follows: mean age 75 years (SD=16) versus 72 years (SD=20, p=0.37); mean APACHE II 16 (SD=6) versus 17 (SD=6, p=0.52); the mean initial SOFA 6 (SD=4) versus 5 (SD=3, p=0.20); mean ventilation duration 8 days (SD=5) versus 23 (SD=11, p=6.40x10 -13 ); the mean ICU LOS 14 (SD=11) versus 29 (SD=12, p=8.20x10 -9 ); the incidence of VAP 15.75 and 20.67 cases/1000 ventilation days (p=0.0002) and mortality 54% (IC=40-68%) versus 26% (IC=12-40%, p=0.01). In the multiple regression models, including as independent variables age, APACHE II, initial SOFA, VAP and tracheostomy, only the last two had statistically significant impact in ventilation duration (p=0.02 and 1.20x10 -10 , respectively) and mortality (p=0.001 and 0.0002, respectively).

In our study, tracheostomy was associated with statistically significant longer duration of mechanical ventilation and ICU length of stay, as well as it related to statistically significant decrease in mortality. There was also a statistically significant increase in the incidence of VAP in the tracheostomy patients.

Mallick A 1 , Bodenham A 1 , Elliot S 1 , Oram J 1 1 Anaesthesia, Leeds General Infirmary, Leeds, United Kingdom

Standard length tracheostomy tubes are inserted during most percutaneous dilational tracheostomy (PDT) procedures. Length of the stem of standard tracheostomy tubes has been designed within the perceived limits of anatomic proportions. Tracheostomy tubes that are too short may exert excessive pressure on tracheal wall around the stoma and carry potential risk of accidental decannulation. There is little data on optimal length of the stem of tracheostomy tubes i.e. the distance from skin to the tracheal walls and angle of tracheal stoma. We performed such measurements to determine what should be the optimum length and angle of tracheostomy tubes.

A total of 50 patients requiring tracheostomy were enrolled following informed assent. The procedure of percutaneous tracheostomy was performed as normal using a Blue Rhino ® dilator guided by fibreoptic bronchoscope/monitor. Following the dilation of tracheal stoma a depth gauge was inserted into the trachea to measure the distance from skin to anterior and posterior tracheal walls. The angle of the stoma was then estimated using several preformed angled steel rods choosing the one which aligned best to anterior tracheal wall. After the measurements a tracheostomy tube was inserted. The distance from the carina to tip of tracheostomy tube was then measured with bronchoscope.

Data of two patients were excluded from analysis as the all the measurements were not possible in them. Patients were 22-88years of age with a median of 58 yrs. Their weight and heights were 78.5 ±15.5 kg and 167.6 ± 7.8 cms respectively. No trial extubation was considered in 29 patients (61%), one trial in 11 (23%), two trials in 7 (14%) and three trials in 1 patient. Patients were intubated for duration of 7 ±2 days before performing tracheostomy. Tracheostomy was performed below 2nd tracheal ring in 32 patients (66%), below 1st ring in 14 (30%) and below third ring in 2 patients. The distances measured from skin to anterior and posterior tracheal wall were 25.6±7.0mm and 48±8.2mm respectively. Average estimated angle of stoma was 113±9.0 degree and the measured distance between the carina and the tip of the tracheostomy tube was 41±8.4 mm. These measurements were then compared with both the stem and effective intratracheal segment of 13 commercially available tracheostomy tubes.

CONCLUSION. The measurements in this study suggest that conventional tracheostomy tubes are shorter for the population of critically ill patients and should be made at least 2 cms longer leaving an average distance greater than 2 cms to the Carina. 

In septic patients, muscle tissue deoxygenation rate during stagnant ischemia is well correlated with organ failure extent and clinical outcome. Aim of this study was to evaluate the effects of recombinant human activated Protein C (rhAPC) on muscle tissue perfusion and oxygenation during arterial clamping in patients with septic shock (1).

Prospective pilot study in a nine-bed polyvalent surgical intensive care unit of a University hospital. In 9 consecutive septic shocked patients receiving rhAPC, thenar muscle O2 saturation (StO2,%) was measured by near-infrared spectroscopy (InSpectra TM , Hutchinson Technology Inc., USA) before, during and after a pneumatic cuff inflation at 240 mmHg. The cuff was placed above the elbow and it was deflated at an StO2 value of 40%. The rate of StO2 decreasing during ischemia as well as the time to restore baseline StO2 (re-saturation time) were calculated. Sepsis organ failure assessment score (SOFA) and the main respiratory and cardiovascular parameters were also measured. Data were collected before, during (24 hours) and 24 hours after rhAPC infusion.

In all the patients, 96 hours of rhAPC therapy were completed. The infusion was temporarily stopped to allow invasive procedures in 4 patients and for minor bleeding in 1 patient. Baseline StO2 values observed before rhAPC (79.88±8.5) were very similar to those observed during rhAPC infusion (79.93±11.3) and lower (p<0.05) than those observed 24 hours after the end of rhAPC (77.92±29.7). The rate of StO2 decreasing during ischemia increased by about 36% during rhAPC infusion; thereafter it returned to baseline. The re-saturation time was longer before rhAPC infusion (59.5±50.7 s) than during (43.9±27.8 s; p>0.05) and after (40.7±39.6 s; p<0.05) rhAPC therapy. The SOFA score, the respiratory and the hemodynamic parameters improved during rhAPC therapy in all the patients. The patient survival rate at 28 days after septic shock diagnosis was 66%.

The above preliminary data indicate that rhAPC infusion improves muscle tissue oxygenation and perfusion in septic shock patients. These variations are associated with a contemporary improvement in SOFA score. 

To determine the effect of a treatment with a short-course of antibiotics and indwelling urethral catheter replacement on clinically asymptomatic intensive care unit (ICU) patients with a positive urine culture occurring at least 48 hours after catheterization.

A prospective randomized clinical trial was conducted in medico-surgical ICU of a tertiary care center. Patients admitted to ICU with an indwelling urethral catheter for longer than 48 hours developing an asymptomatic positive urine culture were randomized to receive either a threeday course of antibiotics associated with the replacement of the indwelling urethral catheter four hours after inclusion or no antibiotics, no catheter replacement (standard of care). The primary end-point was the presence of positive urine culture on day 7 and 15. The secondary end-points were the occurrence of septic events as well as the emergence of bacterial resistance after study inclusion.

Duration of positive urine culture was reduced in the study group, compared with the standard of care group (P = 0.07). The number of septic events occurring after inclusion was similar between the two groups. The profile of bacterial resistance was unaffected by the randomization.

Treating a positive urine culture in an asymptomatic patient with an indwelling urethral catheter decreases the duration of bacteriuria but is not beneficial in terms of outcome.

Piñeiro L 1 , Sirvent J M 1 , Ortiz P 1 , Lopez de Arbina N 1 , Bonet A 1 1 Intensive Care Unit, Hospital Universitari Doctor Josep Trueta, Girona, Spain

Gram-positive bacteria (GPB) can cause serious infections in critically ill patients and vancomycin therapy is often initiated empirically. Linezolid is a recent antibiotic proposed for these infections. The aim of this study was to compare clinical success and adverse events of linezolid compared with vancomycin in critically ill patients infected by GPB.

We conducted an open and non-controlled study. The eligible patients with proven or suspected infection due to GPB were treated with vancomycin (VAN) or linezolid (LNZ) according to the physician criteria of intensive care unit (ICU). We registered the demographic data, the characteristics of infection, the adverse events and the outcome. We compared the data by Chi-square test or Fisher test for qualitative variables and t-Student for quantitative variables. The statistical significance was considered at P<0.05. 

In this study, no statistical significant differences were found between LNZ and VAN when clinical success and adverse events were analysed. Vancomycin treated patients were associated with a higher percentage of positive culture at seven day of treatment. 

Since its first description in the early 1960s, Methicillin-Resistant Staphylococcus Aureus (MRSA) gradually has spread within the hospital environment to become a major nosocomial pathogen in many hospitals worldwide. In our hospital, after the first MRSA case detected in 1998, the Epidemiologic Surveillance System (ESS) started up a MRSA control programme.

The study was conducted in a public community hospital with 279 beds, with a medicalsurgical ICU with 9 beds. The study (from 1999 to 2005) was applied to patients with risk factors (patients admitted from another ward or hospital, from residential care facility, and information on prior colonization or infection with MRSA during this admission) to develop MRSA colonization or infection. The patients with risk factors were pre-emptively placed under contact precautions until the results of the screening test (swabs from both nasal cavities, both inguinal areas and wounds) were found to be negative.

The ESS detected an increase in the MRSA cases during the study period in the hospital (79 cases), but not in ICU (22 cases). The incidence rate of MRSA (per 1,000 patient days) in ICU was 2.3. MRSA cases were 68% male, mean age of 65.7, mortality rate was 59%, length of stay 36.5 days and 18.6 days from admission to infection. Twenty-two cases of MRSA were detected in ICU (15 acquired and 7 imported). The sites of infection were the respiratory tract (54.8%), operative wounds (22.6%), bacteriemic infections (9.6%), catheter sepsis (3.2%), soft tissue (6.4%) and others (3.4%). During the same period, the ESS detected 2 colonized patients, 2 nurses of ICU colonized and 5 imported colonized patients (4 from general wards and 1 from residential care facility).

CONCLUSION. In our hospital, MRSA infections were found to be endemic. In our ICU, the control programme (screening test and contact precautions) has demonstrated that the spread of MRSA infections among the patients can be prevented. In general wards, MRSA cases are increasing and thus, MRSA control strategies should be improved. The most frequent infections are related to the respiratory system. 

We report 23 consecutive admissions to Level 2 care unit over 18 months with sepsis 

The mean duration of stay on level 2 care was 2.77 days. Nine patients died (42.8%) and 12 (57.14%) survived from the acute episode. All three patients with Candida died [C.albicans (n=1), C.glabrata (n=2)] All with viral infection survived[Varicella (n=1), EBV (n=1), CMV (n=1), Adenovirus (n=1). The survival from ICU and at 6 months were 7 (43.75%) and 4 (25%) respectively. One out of four patients needing concomitant respiratory support survived (25%). We conclude that early level 2 care and timely transfer to Level 3 care is justified for patients with sepsis and undergoing treatment with intensive chemotherapy for hematological malignancies. Overall survival outcomes have improved for the patients with hematological cancers undergoing intensive chemotherapy regimens and transplant procedures with the advent of better antiviral, antifungal, and antibacterial agents and better supportive care. 

The aim of this study was to define the incidence, the risk factors and the microbiology of the ICU-acquired Bloodstream Infections (BSI) in trauma patients.

It is a 17 month observational prospective study. Forty-six (46) consecutive patients with traumatic injury who were admitted to a 7-bed multidisciplinary ICU and stayed > 48 hrs, were enrolled. BSI and lower respiratory system infections (LRSI) were diagnosed according to CDC criteria. Statistical analysis was performed using chi-square, Fisher exact and t-tests, where appropriate. (14) trauma patients developed 21 BSIs: Primary (PBSI):8, Catheter-related (CR-BSI):8, Secondary (SBSI):5. The incidence rate and the incidence density were 30% and 23.8%, respectively. The characteristics of the patients are shown on the table. Twenty-two (22) pathogens were isolated. Ninety-one percent (91%) of the microbes were gram-negative bacteria. 

High mobility group box-1 protein (HMGB1) is a DNA-binding intranuclear protein. It is also released into the extracellular space by damaged and necrotic cells and acts as a potent pro-inflammatory cytokine. Some studies have shown, that it acts as a late cytokine, which stays elevated several days after onset of severe sepsis. The aim of this study was to evaluate the predictive value of HMGB1 as regarding hospital mortality in adult patient population with severe sepsis and septic shock.

Finnsepsis study was a prospective study about incidence and outcome of severe sepsis in Finland. All adult consecutive ICU admission episodes (4500) were screened for severe sepsis in a 4 month period (from 1.11.2004 to 28.2.2005) . Patients were eligible, if they fulfilled the ACCP/SCCM (1992) criteria for severe sepsis and septic shock. Blood samples for HMGB1 analyses were drawn after the consent on day 0 and after 72 hours. The serum HMGB1 levels were measured by Western immunoblotting analysis with rabbit polyclonal anti-HMGB1 antibodies (BD Pharmingen, San Jose, CA). In addition, a pooled serum sample was used as an inter-assay (inter-gel) control for semiquantification HMGB1 levels in serum. The samples were compared with healthy controls (N=10) by Mann-Whitney test. Organ dysfunctions with SOFA-scores and ICU and in-hospital mortalities were recorded.

Laboratory samples were obtained from 247 patient altogether. Mean age was 59 years (SD 15.6) and mean APACHE II and SAPS II scores 24 (SD 9) and 44 (SD17), respectively. 247 samples were obtained on baseline and 210 72 hours later. ICU mortality was 13.2% (n=33) and hospital mortality 26% (n=65) in study patients. Median HMGB1 levels of healthy controls were 98% (25th and 75th percentiles 91% and 107%). Median HMGB1-levels on day 0 were 108% (25th and 75th percentiles 98% and 119%) and after 72 h 107% (25th and 75th percentiles 99% and 120%), which differed statistically from healthy controls (p= 0.03 and 0.02, respectively). The ROC curve for day 0 or 72 h HMGB1-levels revealed the area under the curve 0.51 and 0.56 (95% confidence limits 0.40-0.61 and 0.47-0.65).

CONCLUSION. HMGB1 values were moderately elevated in severe sepsis and septic shock patients, but did not differ between survivors and non-survivors and were not predictive for hospital mortality in severe sepsis patients.

Palos C A 1 , Silva Z C 1 , França C 1 1 Intensive Medicine, H. Santa Maria, Lisboa, Portugal

Via Sepsis (VS) is an ICU's initiative aimed to improve diagnosis and treatment of septic patients admitted to our 1200 beds universitary hospital according to Surviving Sepsis Campaign (SSC). It covers resources evaluation, awareness, general information and education on sepsis; data registry and analysis.

Non-ICU physicians and nurses were asked to answer an anonymous questionnaire immediately after attendance VS local presentations from May through December 2005. Questions focused on knowledge about SSC before attendance VS presentations; self-knowledge about sepsis before and after such attendance; self-perceived diagnosis and treatment of septic patients at our hospital; importance of VS initiative.

One hundred and sixty-one questionnaires were answered (nurses 55%). Before attendance to these presentations only 38.1% of responders had knowledge about SSC (physicians 57.5%; nurses 30.6%). Previous knowledge about sepsis was self-reported as moderate (50.3%) or poor (42.2%), nurses having lesser expertise (51.1% vs 46.5%) and 89% of responders (nurses 95%) considered to have improved their knowledge about sepsis after VS presentation. Fifty-nine percent of responders reported that patients had a "delayed diagnosis" at our hospital. "Lack of physicians awareness of sepsis" was recognised as the main individual cause to inadequate approach (26.7%), far from "insufficient material resources" (9.9%). VS was perceived as very important by 94.4% of responders. Results about compliance with VS and SSC's guidelines are being collected.

CONCLUSION. VS is an example of how ICUs'staff can lead the change of sepsis approach at the hospital level. Questionnaires were important in order to re-define VS strategy, mainly the need to implement courses for physicians and nurses as well as the need to individualize the first 2h within the 6h-bundle, pointing out the importance of early diagnosis and treatment. In conclusion, every local initiative aimed to implement SSC's guidelines should include evaluations in order to prepare the best suitable answers to local needs. 

Serum cholinesterase activity (CHE) was found to be a sensitive indicator of morbidity and mortality in burn patients, where it was significantly lower in non-survivors 24 hours after admission (1) . As there are no data available on the prognostic value of CHE activity in a large mixed surgical intensive care population, we investigated its prognostic usefulness in an according population.

All consecutive adult patients who needed at least 4 days of ICU treatment admitted to two mixed surgical ICU's during 2004/05 were enrolled. CHE was determined immediately after admission and thereafter daily until discharge or death. CHE of survivors and non-survivors at day one to four, first and minimum values were compared using ANOVA for repeated measurements.

. 441 patients were eligible for the study. The first value of cholinesterase activity was already significantly lower in non-survivors. This difference remained significant on days one to four. The decrease in serum cholinesterase activity versus day 1 was significant in both groups. The minimum values reached during the ICU stay were lower in non-survivors (Table) . CONCLUSION. CHE activity is lower in non-survivors compared to survivors from the first measurement on and is thereby a very early indicator of increased mortality.

(1) L.P. Kamolz; Serum cholinesterase activity reflects morbidity in burned patients; Burns; 2002; 147-150.

A number of questions relating to age and intensive care mortality remain incompletely answered. Is there a threshold of age associated with a lower survival? Is 80 year an upper limit to be taken into account to admit patients on ITU? (1) . A number of investigators have recently examined the effect on the outcomes of older patient (2) . Aim: To determine the effect of age as a factor that influence mortality on patients who are admitted to the mix Intensive Therapy Unit.

We did a prospective observational research study and one hundred and sixteen critical ill patients admitted to a ITU. Enrolment criteria: Any patient older than 12 y. Severity at admission was estimated using the APACHE II score. Age was split on 5 groups: older than 30 y, 40 y, 50 y, 60 y, 70 y, and older than 80 y. Free on any other risk factor of mortality and non-other classification according to type of previous medical problem was taken into account. A descriptive analysis was performed, data are presented as mean ± SD, and student t−test were used to compare the difference of age and mortality among each group. For mortality prediction the ROC analysis tested the best threshold level of the age variable. The statistical analysis was carried out with the SPSS 10 package, and p<0.05 was considered statistically significant.

Death was the outcome variable that was studied. Of 516 patients, 39.3% were male and 60.7% were female. Their average age was 52.4±20.5 years. APACHE II score 12.8±7.3. The overall mortality rate was 9.6%. Mortality and significance are showed on 

Results no demonstrated a significant dominance of age factor on the prognosis of patients admitted on our ITU, due to this, no significance was found for the group of patients older than 40 y. "To be young allow be well, but it does not mean that to be older will die easier". As a result of this, we should expand the analysis of risk factors to mortality for admitted ITU patients based on past medical history (co-morbidities) and physician do not take precarious preliminary decision in front of older patients. Finally, not only age but also co-morbidities must be taken into account on the decision to expend effort to get any benefit. 

Massaut J 1 , Wauthy P 2 , Loscheider M 1 , Goldstein J 2 , Demanet H 2 , Reper P 1 , Deuvaert F 2 1 Intensive Care, 2 Cardiac Surgery, Centre Hospitalier Universitaire Brugmann, Bruxelles, Belgium

Compared to first heart valves operations, valvular surgery reoperation (redo) carries a higher mortality. This study reports a single center experience in redo valvular surgery, analyses risks factors of peri-operative death, computes a pre-operative score for mortality prediction and compares the performances of that score with Euroscore (1).

The data of all consecutive patients who underwent redo valvular surgery in our institution between 1997 and 2005 were retrospectively analysed. Logistic regression was carried out to identify predictive risk factors of death. Multiple logistic regression was used to isolate independent risk factors and compute pre-operative score of mortality prediction. Predictive values of Euroscore were also computed using logistic regression and ROC analysis. Predicitive values of our pre-operative score were compared to the performance of Euroscore.

A total of 227 patients were admitted for redo valvular opération. Hospital mortality was 7.4%. Variables linked to mortality (p<0.05 at the single logistic regression) are emergency, NYHA class, number of redo, pre-operative hemoglobin value, urea, creatinine, lactate dehydrogenase (LDH), duration of operation and extra-corporeal circulation, and transfusions. Pre-operative independent predictive factors of mortality are NYHA class, LDH, and urea. Preoperative score = 1.37190 NYHA + 0.01252 urea + 0.00075 LDH (area under the ROC curve = 0.88). This score seems to perform beter than the Euroscore (area under the ROC curve = 0.60) on our small and homogeneous population.

In redo valvular opérations hospital mortality are highly correlated to a more advanced stage of the cardiac dysfunction independent of the number of previous interventions. A simple pre-operative score of mortality prediction can be computed using NYHA class, preoperative LDH and urea values with good sensitivity and specficity. CONCLUSION. These CVS procedures appeared to be associated to a temporary CBF decline. A SBF and CBF coupling occurred after CVS early in the absence of CBP, which might indicate an altered autoregulation of CBF. 

Effects of different doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a prospective randomised trial

The provision and practice of renal replacement therapy on adult intensive care units in the United Kingdom

Cardiac output by PulseCOTM is not interchangeable with thermodilution in patients undergoing OPCAB

REFERENCE(S). 1. Wiel. Embolie graisseuse cérébrale après traumatisme fermé de la jambe

Des emboles de graisse au syndrome d'embolie graisseuse

Early Diagnosis of Cerebral Fat Embolism Syndrome by Diffusion-Weighted MRI

Magnetic resonance imaging findings in cerebral fat embolism: correlation with clinical manifestations

Vasopressor and inotropic support in septic shock: an evidence-based review

Surviving Sepsis Campaign guidelines for management of severe sepsis and septic shock

Long term survival from intensive care: a review

Grant acknowledgement. Emma Children's Hospital

Cardiac biochemical monitoring for the detection of acute myocardial ischemia

High myocardial lactate concentration is associated with poor myocardial function prior to cardiopulmonary bypass

REFERENCE(S). 1. Glass L. Synchronization and rhythmic processes in physiology

Sample entropy analysis of neonatal heart rate variability

Multiscale entropy analysis of complex physiologic time series

Preoperative risk factors are increased mean age (68.36+/-5.2years), increased mean weight (80.12kgr), diabetes mellitus (40%), NYHA III-IV (60%), left ventricular ejection fraction (LVEF)<35% (32%), pulmonary hypertension (32%). The duration of cardiopulmonary bypass is significantly increased (176.43+/-23.3min), the same as aortic cross clamp time (121.57+/-12.2min). In the majority of patients we certify the low cardiac output syndrome postoperatively (CI<2.0l/min/m2, 56%)) with IABP dependency (28%) and need for increased inotropic support (64%)

REFERENCE(S). 1. Endovascular aortic repair: has there been an improvement in device performance?

Endovascular Stent Graft Repair of Abdominal and Thoracic Aortic Aneurysms: A Ten-Year Experience With 817 Patients

The Influence of Asymptomatic Significant Carotid Diseaseon Mortality and Morbidity in Patients Undergoing Coronary Artery Bypass Surgery

Impaired cerebral vasoreactivity and risk of stroke in patients with asymptomatic carotid artery stenosis

Lactic acidosis revisited

The significance of lactic acidemia in the shock syndrome

Perioperative Medicine And Intensive Care Departement, San Gerardo Hospital, Monza, 2 Surgical Sciences And Intensive Care Departement

We retrospectively analysed data of all patients (pts) undergoing cardiac surgery between January 1998 and June 2005, and discharged from our ICU by 24 hours from surgery. We collected: (i) demographics and anamnestic illnesses (ii) intra-operative variables [cardiopulmonary by-pass (CPB) and aortic cross clamp (ACC) times] (iii) duration of mechanical ventilation, use and type of inotropes. One-Way ANOVA was used for continuous variables whereas, differences in proportions were compared using Chi-squared test. A binary Logistic Regression Model was used to estimate the effect of each considered risk factor on discharging from cardiac surgical to rehabilitative ward, considered as a dycotomous outcome (yes = early < or = 7 days / no = late >7 days). Statistic analyses were performed using SPSS Software. P values less than 0.05 were considered significant.

A total of 1488 pts [median (IQR) age 65 (56-72) ys, 71% males] were discharged from ICU by 24 hours from surgery. Most 67% of them underwent coronary artery by-pass grafting (CABG), while, 28% valve procedures (VP), with a median (IQR) HL of 7 (7-8). The b-LRM was performed, considering discharging from surgical to rehabilitative ward as the categorical dycothomous dependent variable (see methods) and as independent dycothomous variables (i) age (> or <= 65 yrs), gender, diabetes, hypertension, arteriopaty, renal failure, COPD (all yes/no) (ii) NYHA and CCS score (> or <= 2), mitral/aortic pathology and/or coronaropaty (all yes/no) (iii) operation duration (> or <= 240 min), CPB-(> or ≤ 90 min), ACC-(> or <=; 60 min) times and transfusion (plasma or red blood cells) needing (yes/no) (iv) mechanical ventilation duration (> or <=; 8 hours). bLRM showed predictors of longer postoperative hospitalization are (i) mechanical ventilation duration > 8 hrs (p = 0.0000, O.R. = 0.615 95% CI 0.488-0.775) (ii) experiencing both a mitral (p = 0.046, O.R. = 0.671 95% CI 0.453-0.993) and an aortic (p = 0.038, O.R. = 0.624 95% CI 0.4-0.975).

Tracheostomy in the ICU 0663-0669 0663

Shankar Hari M 1 , Barrera Groba C 2 1 Department Of Anaesthesia, Maidstone and Tunbridge Wells NHS Trust, Tunbridge Wells, 2 Department Of Critical Care, Brighton and Sussex Universirty Hospital, Brighton, United Kingdom INTRODUCTION. Aims: to determine whether 1] tracheostomy affects hospital outcome; 2]timing of tracheostomy affects duration of mechanical ventilation and hospital outcome.

Retrospective cohort study [N=703]on patients requiring intubation and mechanical ventilation for > 48hrs.

Tracheostomy was performed in 46%[322/703]of patients. There was no difference in hospital outcome between the 'early' [</=7days; n=168] and 'late'[>/=8Days; n=154] tracheostomy groups (p= 0.36). Survivors in the early group had a significantly shorter duration of mechanical ventilation compared to the survivors in the late group[9 days Vs 15.3 days](p<0.001).Poster Sessions Anticoagulation in sepsis and organ failure 0670-0681 0670

Dixon B 1 , Santamaria J D 1 , Campbell D J 2 1 Intensive Care Unit, St. Vincent's Hospital Intensive Care Unit, 2 Institute of Medical Research, St. Vincents Hospital, Melbourne, Australia INTRODUCTION. Inflammation is a potent cause of tissue injury and multi-organ failure. Unfortunately, the underlying mechanisms by which the inflammatory response causes tissue injury are not well understood. Consequently, interventions to prevent or treat it are limited. We assessed the role of microvascular obstruction in tissue injury and investigated if heparin had a therapeutic role.

Twenty cardiac surgical patients were randomised to a pre-operative infusion of heparin or placebo. To assess microvascular obstruction we measured the changes in the alveolar dead space. To assess the potential actions mechanisms by which heparin may limit microvascular obstruction we measured pulmonary coagulation activation, endothelial permeability and white cell and platelet retention.

The alveolar dead space to tidal volume ratio increased markedly in placebo patients. Administration of pre-operative heparin prevented this (p < 0.05). Pulmonary prothrombin fragments 1+2 production and endothelial permeability were reduced by pre-operative heparin (p < 0.05). Preoperative heparin reduced baseline blood levels of prothrombin fragments 1+2 (p = 0.01) and platelets (p < 0.05) and increased tissue plasminogen activator levels (p < 0.05).CONCLUSION. Microvascular obstruction may be an important mechanism by which inflammation mediates tissue injury. Heparin may limit microvascular obstruction through reduced microvascular fibrin deposition. 

Palaiologou A C H 1 , Botsis P 1 , Eforakopoulou M 1 , Stavropoulos N 1 , Litis D 1 , Ioannidou E 1 1 A' ICU, Hospital KAT, Athens, Greece

Systemic inflammatory mediators activate the coagulation system; conversely systemic inflammation is induced by coagulation. Thus, it may be possible to attenuate the generalized host inflammatory response to septic stimuli through careful regulation of the coagulation system. The aim of this study was to investigate this hypothesis.METHODS. 26 consecutive ICU patients were included in the trial (14 male and 12 female mean age 40±20 years old). Apache II score on admission was ≤10. The study protocol was approved by the local ethics committee. Patients fulfilled the following inclusion criteria :1) Presence of a known or suspected infection, 2) Acute dysfunction of 2 or more organ systems induced by the episode of sepsis and 3) the SIRS criteria. We excluded the patients with 1) active internal bleeding, 2) major surgery within the last 12 hours,3) platelet count <30000/mm 3 , 4) recent (within 3 months) hemorrhagic stroke, 5) history of intracerebral arteriovenous malformation, cerebral aneurysm or central nervous system mass lesion. Patients were randomly assigned to be treated either by adding heparine to standard therapy at a dose of 15 IU /Kg/h for ten days (group A, 14 patients) or not (group B, 12 patients). We estimated the process of sepsis by Sepsis Severity Score (SSS) every day, the proces of the MODS by APACHE II score and finally we recorded the all cause mortality. Statistical analysis was made by x 2 test.

As regards the MODS, the improvement of patients at group A was significantly higher than group B, (p<0.01). 2) Recovery from sepsis was greater at group A than group B, (p<0.001). 3) All cause mortality was lower at group A than that of group B, (p<0.001). METHODS. This was a retrospective study. We analyzed all the files of patients older than 65 years that received aPC between September 2002 and August of 2005. We used descriptive stadistics and estimate the p by t-student for continue variables and X2 for the dicotomic variables. We considered p value less or equal than 0.05 statistically significative.

We analyzed files of 36 patients (p). The mean age was 75 ± 6 years, 72% (n=26) were male, 50% (n=18) has Diabetes Mellitus, 35.7% (n=13) hypertension, 57.1% (n=27) cardiopathy, 42.9% (n=15) cancer and 35.7% (n=13) any chronic pulmonary disease. The etiology of the sepsis was pulmonary in 64.3% (n=23), abdominal in 28.6% (n=10) and urinary in 7.1% (n=3). The severity of illnes and hemodinamic and respiratory management are expressed in table 1. The mortality was 38.8% (n=14). The early administration of aPC (in the first 24 hours) is associated with a lower mortality in comparison to the late of the drug (after the first 24 hours) 11.8% vs 63.2%, p=0.002. The late onset of the infusion of the drug predicts mortality with a momio test of 10.8. 

CONCLUSION. The early onset of activated drotrecogin alfa infusion reduce the length of MV, the requirements of vasopresors and the number of organic dysfunctions at 96 hours. We did not found difference in the bleeding. The early onset of the administration of drotrecogin alfa is associated to an important reduction in mortality in ederly patients.

Janssen van Doorn K 1 , Spapen H D M 1 1 ICU Department, Academic Hospital, Vrije Universiteit Brussels, Brussels, Belgium

Acute renal failure associated with severe sepsis has a high mortality 1 . Starting drotrecogin alfa activated (DrotAA) in sepsis requires the presence of at least two organ failures. Renal failure (RF), defined as the persistence of oligo-anuria following adequate resuscitation, is one of those recognized organ failures. DrotAA may act by enhancing cardiorespiratory recovery as shown by reduction of vasopressors and ventilation needs 2 . However, the evolution of patients with RF and receivingrenal replacement therapy (RRT) is not known.

Twenty-two patients with sepsis-induced RF without pre-existing renal dysfunction were studied over a one-year period. All patients received standard treatment for sepsis, including RRT. Six of them received DrotAA as a continuous infusion of 24 µg/kg/h over 96h.

Patients treated with DrotAA did not differ from untreated subjects with respect to age (63.8 vs 68.0 years), organ failure (2.67 vs 2.56), APACHE II (32 vs 30) and APACHE III-score (88 vs 92). Global mortality was 77%. 75% (12/16) of the patients not treated with DrotAA died as compared to 83% (5/6) of the subjects treated. Fifteen (88%) of the non-survivors, both treatment groups taken together, remained oligo-anuric during the first two days of observation.

Sepsis complicated by ARF has a very high mortality, regardless DrotAA is given or not. Oligo-anuria seems to be a predictive factor for outcome. Although no firm conclusions can be drawn from this small descriptive study, any beneficial effect of DrotAA on the evolution of sepsisinduced RF under RRT remains to be demonstrated. Also, the suitability of RF, as it is actually defined, as a criterion to start DrotAA treatment may be challenged. 

Clark S 1 , Fairburn A 1 , Dyson M 1 , Maddock H 2 , Holland M 3 1 ICU Outreach, 2 ICU, 3 Medicine, Wythenshawe Hospital, Manchester, United Kingdom

Many patients presenting to our medical assessment unit (MAU) have an acute infective illness. Prior to formally implementing The Surviving Sepsis Campaign guidelines on MAU we wanted to assess current practice, and also quantify the severity and numbers of infections presenting to MAU.METHODS. 11 cohorts of MAU patients were audited. Patients were only included once. An audit tool was designed and piloted. Patients with a documented admission diagnosis of infection were included as well as patients with a white blood count <4000/mm3 or >12000/mm3, temperature >38• C or <36• C, raised CRP or antibiotics prescribed since admission. Systemic inflammatory response syndrome (SIRS) criteria and blood pressure (BP) were markers of physiological status. The targets in the Surviving Sepsis resuscitation 6-hour care bundle were the benchmarks for good practice.RESULTS. 202 patients (median age 71yrs, interquartile range 53-82yrs, 114 female) were audited. 102 (51%, median age 75yrs, interquartile range 60-84yrs, 58 female) patients had evidence of an acute infection. Of the infected patients, 5 had a systolic BP</=90mmHg, 8 a MAP</=65mmHg and 69 had 2 or more SIRS criteria. In all patients (n=72 data available), 32% received antibiotics within 3hrs of arriving at hospital (median 3hrs 52 mins, interquartile range 2hrs 22mins-6hrs 58mins). For patients with 2 or more SIRS criteria (n=50 data available), 42% received antibiotics within 3hrs of arriving at hospital (median 3hrs 14mins, interquartile range 2hrs 2mins-3hrs 14mins). In all patients (n=45 data available), 36% had a blood culture taken within an hour of arriving at hospital (median 1hr 42mins, interquartile range 50mins-1hr 42 mins). In 3 patients, blood cultures were taken after antibiotics were given. In patients with MAP</=65mmHg, fluid was initially run at 20ml/kg in only one patient and a lactate was measured in only one patient. Only one patient had a CVP line inserted.CONCLUSION. 51% of patients on our MAU have an infection. The majority are elderly with significant physiological derangement when judged by SIRS criteria. There is a significant delay in the administration of antibiotics, even in those patients who are physiologically most unstable. This data support the view that a systematic approach to identify and treat septic patients on MAU is needed. 

Karlsson S 1 , Pettilä V 2 , Tenhunen J 1 , Lund V 3 , Hovilehto S 4 , Ruokonen E 5 1 ICU, Tampere University Hospital, Tampere, 2 ICU, Helsinki University Hospital, Helsinki, 3 ICU, Satakunta Central Hospital, Pori, 4 ICU, South Karelian Central Hospital, Lappeenranta, 5 ICU, Kuopio University Hospital, Kuopio, Finland INTRODUCTION. Vascular endothelial growth factor (VEGF) is a well known mediator of angiogenesis and a potent vascular permeability factor. Few studies have shown increased levels of VEGF in severe sepsis patients. The aim of this study was to evaluate the predictive value of VEGF regarding hospital mortality in adult patient population with severe sepsis and septic shock.

The Finnsepsis study was a prospective study about incidence and outcome of severe sepsis in Finland. All adult consecutive ICU admission episodes (4500) were screened for severe sepsis in a 4 month period (from 1.11.2004 to 28.2.2005) . Patients were eligible, if they fulfilled the ACCP/SCCM (1992) criteria for severe sepsis and septic shock. Blood samples for VEGF analyses were drawn after the consent on day 0 and after 72 hours. The serum VEGF concentrations were measured in duplicate for each sample using a commercial enzyme-linked immunosorbent assay kit (R&D Systems; Minneapolis, MN) Sciences, UK). The samples were compared with healthy controls (N=10) by Mann-Whitney test. ICU and hospital mortalities were recorded.

Laboratory samples were obtained from 250 patients. Mean age was 59 years (SD 15.6) and mean APACHE II and SAPS II scores 24 (SD 9) and 44 (SD 17), respectively. 250 samples were obtained on baseline and 215 samples 72 hours later. ICU mortality was 13.2% (n=33) and hospital mortality 26% (n=65) in these study patients. Median VEGF-levels on day 0 were 423 pg/ml (25th and 75th percentiles 159 and 858 pg/ml) and after 72h 521 pg/ml (25th and 75th percentiles 182 and 1092 pg/ml). The 72h values differed statistically significantly from those of healthy controls (p=0.03). VEGF levels in day 0 and after 72h were statistically significantly higher in non-survivors (p=0.01 and <0.01, respectively) than in survivors, but the ROC curve for day 0 or 72 h VEGF levels revealed only a moderate area under the curve (AUC) of 0.58 and 0.63 (95% confidence limits 0.48 to 0.68 and 0.54 to 0.72).CONCLUSION. VEGF values are frequently elevated in severe sepsis and septic shock. Values were significantly higher in non-survivors than in survivors, but were not predictive for hospital mortality in severe sepsis patients.Grant acknowledgement. EVO grant from Helsinki University Hospital.

Vandijck D M 1 , Blot S I 1 , Benoit D D 1 , Decruyenaere J M 1 1 Intensive Care Medicine, Ghent University Hospital, Ghent, Belgium

Patients with nosocomial BSI secondary to pneumonia or intra-abdominal infections have been associated with worse outcome compared to those with other sources of infection. The aim of the study was to investigate whether patterns of CRP levels in ICU patients differ according to the source of bacteremia.

A historical cohort was studied with inclusion of 155 adult ICU patients with nosocomial BSI over a two year period from 1 January 2003 until 31 December 2004. Daily CRP levels were recorded from two days prior to BSI until five days after onset of the BSI. For the purpose of analysis, sources of bacteremia were divided into 2 categories: low risk (sinus, urinary tract, IV-catheter, softtissue, and primary sources of bacteremia) vs high risk (low respiratory tract and abdominal sources), respectively.

Mean age was 52.8±17.9 yrs. Of the 155 patients, 117 (75.5%) were defined with low risk vs 38 (24.5%) with high risk sources of bacteremia (P=0.001). Serum CRP concentrations were significantly different on the day of positive blood culture (13.6±8.4 vs 17.4±10.8, P=0.026), and one day after onset of BSI (15.6±8.7 vs 19.5±10.4, P=0.037) . For all other days (day -2, -1, +2, +3, +4, and +5) no statistically significant differences were found between both groups (see Fig. 1 ).

In ICU patients with nosocomial BSI different patterns of CRP levels on day 0 and day +1 are seen according to the source of infection (low risk vs high risk).

Galstyan G M 1 , Serebriysky I I 1 , Berkovsky A L 1 , Vasiliev S A 1 , Rudakova V E 1 , Sergeeva E V 1 , Gorodetsky V M 1 1 ICU, National Center for Haematology, Moscow, Russian Federation INTRODUCTION. Important part of the pathophysiology of sepsis is caused by inappropriate coagulation in the microcirculation. It is known, that haemodilution causes enhancement of coagulation owing to imbalance between procoagulant and anticoagulant factors. HES are often used for correction of hypovolaemia in pts with sepsis. The aim of this study was investigation of the effects of infusions of HES on the hemostasis in septic pts.METHODS. 20 pts received HES 130/04 (Voluven, Fresenius-Kabi) 13.7 ± 1.5 ml/kg within 2-3 hours. During infusion of HES all other infusions were stopped. Before and after infusions we measured in plasma activated partial thromboplastin time (APTT) prothrombin index (PI), activities of antithrombin III (AT III), factor VIII (FVIII), protein C (PC), plasminogen (PG), Willebrand factor (FW), levels of fibrinogen and soluble fibrin monomer complexes (SFMC).

Septic pts had increased levels of FVIII, FW, SFMC and decreased levels of PC, AT III, PG. Infusion of HES decreased levels of fibrinogen, PG, FVIII, FW, PC, PI. Procoagulant factors have decreased more expressed, than anticoagulant factors. APPT before infusion was 34.5±1.7 s. APTT and SFMC did not change after infusion. We determined the frequency, the sources, microbiological identification and outcome.

Blood culture results were reviewed daily. All patients with positive blood culture 48h after admission were included in this study. The sources of infection have been detected by clinical and/or microbiological criteria. Patients were followed until discharge or death.

During this two years 475 patients werre hispital; ized at PICU and 156 (33%) patients had nosocomial bacteremia. The sources of bacteremia were: intravascular catheters (37%), respiratory tract (12%), urinary tract (10%), surgical sites (6%). We have not detected sources in 35% of these patients. The survival rate was 67%. The results of microbiological analysis were:Staphylococcus aureus (21%), Pseudomonas species (19%), Klebsiella pneumoniae (11%), Acinetobacter (9%), Seratia (8%) and other (32%). In-hospital mortality associated with Pseudomonas species (53%) was significantly higher than for other bacterias (26-37%), and Pseudomonas species also was highly resistant on antibiotic therapy.

All information from this study are very important for more successful and more accurate treatment and prevention of intrahospital infections. 

Brain tissue oxygen monitoring is a part of multimodal monitoring. Prevention of secondary brain injury is crucial. Values of parcial brain oxygen pressure (pbtO2) in first hours after brain trauma should predict final patients outcome. The aim of this study was to analyze relationship of early values of brain oxygen in severe head trauma to outcome.

We analyzed data of 18 consecutive adults patients treated in our ICU during time period of 24 months for severe head trauma with Glasgow coma scale (GCS) 8 and less and with monitoring of intracranial pressure (ICP) and partial brain oxygen pressure (pbtO2). We used Codman Neutrotrend multiparameter sensor for pbtO2 monitoring. We placed Neurotrend sensor at the same time as ICP sensor and we used non injured side of brain for aplication. All patients were treated according standard therapeutical protocol. Target of our treatment was to reach optimal pbtO2 levels. Comprehensive records of biological data were made in all patients. We compared data of first 24 and second 24 hours after start of the treatment in ICU with neurological status in the time of leaving ICU among three groups: non survived patients, patients with neurological disability and patients with normal levels of consciousness.

Mean of pbtO2 in the group of non survivers (4 pts) was during first 24 hours 25.99 mmHg and during second 24 hours 21.28 mmHg. Mean of pbtO2 in the group with neurological disability (9 pts) was during first 24 hours 19.91 mmHg and during second 24 hours 15.07 mmHg. Mean of pbtO2 in the group with normal level of consciousness (5 pts) was during first 24 hour 26.31 mmHg and during second 24 hours 26.61 mmHg. We calculated this mean from all measured values during followed time periods and in the individual groups of patients.

We did not find in our group clear relationship between values of pbtO2 and outcome. Monitored values of pbtO2 are dependent on variety of factors during measuring process (e.g. placement in area of lesion or non lesion, deepness of sensor). We consider pbtO2 monitoring as a method to become additonal informations for proper adjustment of therapy. 

Quinine is obtained from the bark of various species of Cinchona. It is used for the treatment of falciparum malaria as well as for nocturnal leg cramps. Quinine or its salts given in usual therapeutic doses may give rise to a train of symptoms known as cinchonism. This is characterised in its mild form by tinnitus, impaired hearing, headache, nausea and disturbed vision, with in its more severe manifestations, vomiting, abdominal pain, diarrhoea and vertigo. The main symptoms of quinine overdosage, which can be fatal, include gastrointestinal effects, oculotoxicity, CNS disturbances and cardiotoxicity. Specialist information and advice on the treatment of poisoning is available in the United Kingdom from the National Poisons Information Service through local poisons information centres. TOXBASE, the primary clinical toxicology database of the National Poisons Information Service, is also available on the Internet to registered users (www.spib.axl.co.uk) providing information about routine diagnosis, treatment and management of patients exposed to drugs, household products, industrial and agricultural chemicals.

We undertook a case study and literature review.RESULTS. An 18-year old girl ingested 5.7g of quinine sulphate with resulting peak serum concentration greater than 13mcg/ml. Initially she suffered with headache, nausea and tinnitus but these progressed to more severe symptoms, including transient loss of visual acuity together with evidence of cardiac disturbances. Advice from a poisons information centre specialist on the management of quinine intoxication focused on decontamination with repeated doses of activated charcoal both in an attempt to limit absorption from the gastrointestinal tract as well as enhancing quinine elimination. Additionally general supportive measures were advised and included correction of electrolyte disturbance, treatment of hypotension, monitoring for cardiac dysrhythmias and assessment of renal function. Glyceryl trinitrate was infused for the visual disturbances she experienced approximately 12-hours following quinine ingestion. Electrocardiographic abnormalities of conduction delay were managed conservatively. She recovered vision after a period of partial blindness.

Overdose with antimalarial agents is hazardous and can be difficult to treat. Specialist management in our case focused on decontamination together with empirical vasodilator therapy for ocular toxicity. These interventions were associated with reductions in serum quinine levels and clinical improvement. 

Acute mechanical large bowel obstruction is a common surgical emergency. Our aim was to identify and analyze the clinical features of these patients along with their management and outcome.METHODS. This is a prospective study of all patients admitted to our hospital with acute mechanical large bowel obstruction between January 2004 and August 2004. Recorded variables were: age, gender, time between the onset of symptoms and arrival at the emergency department, initial vital signs, presenting symptoms, physical examination findings, white blood cell count, abdominal x-rays findings, type of management, time between admission and operation, complications, admission in the Intensive Care Unit (I.C.U.), mortality, and length of hospital stay.

Twenty-four consecutive patients with mechanical large bowel obstruction were admitted. Mean age was 63.6 years; twelve were men (50%) and 12 (50%) women. CONCLUSION. In our study, absence of passage of flatus and/or stool and abdominal distension were the most common presenting symptom and physical finding, respectively, while plain erect abdominal x-rays were essential for diagnosis. The majority of the patients was operatively treated while a substantial portion required immediate surgical intervention. They had relatively low complication, I.C.U. admission, and mortality rates but prolonged hospital stay. METHODS. Monocentric, retrospective study over 5 years (2000) (2001) (2002) (2003) (2004) , of death in observation unit. Statistical analysis used the Epiinfo software with CHI2.

. 102 pts out of 126 were included:74 women,28 men, with an average age of 78 years (from 15 to 97).22 pts were less than 70 years. In 86 cases, death occured during the first 24 hours following the admission, on average 14.4 hours. Principal causes of death were septic (n=34), breathing distress (n=21) and heart failures (n=20). Other etiology were cancers (n=14), and neurological strokes (n=6).Only 25 pts were dependent and 17 were living in institutions or benefiting from home medical care. Maximal cares were dominated by catecholamines (30%), oral intubation (17.6%) and blood transfusion. Among the limited care, antibiotherapy, oxygenotherapy and the non-invasive ventilation were respectively related to 36.3%, 29.4%, and 16.7% of the patients. Palliative care were largely dominated by morphine with 47 pts. Caring's levels were different between the 27 pts medicalized by the SMUR and the 75 pts who were not. In fact, the latter were given the benefit of maximal care in 36.2% of cases (n=37) compared to 14.7¨% (n=15). The decision of stopping all therapy was multi disciplinary in 100% of cases with 59 opinions from the intensive care unit and 23 opinions from cardiology unit. From 3 to 5 different opinions were given for 38 patients. The mobile palliative care unit was called for opinion only twice.

Level of death caring in the short stay unit is important for a unit like this. The collegial decision save emergency doctors making alone care continuation's decision. Intensivists physician's opinion remains essential in final decision. A protocol for "end of life" as its preconized by scientific societies is necessary, as well as a medical-staff training.

Grant acknowledgement. Medical record information dpt.

Slack A J 1 , Aurora S 1 , Watson D 1 , Chadda K 1 , Medici F 2 1 Intensive Care Medicine, 2 General Medicine, Homerton University Hospital, London, United Kingdom

Clostridium botulinum is an anaerobic spore forming bacterium. Spores contaminate wounds, germinate and produce neurotoxin. Seven neurotoxins A-G have been identified, which bind to the presynaptic membrane at the neuromuscular junction, blocking the release of acetylcholine and resulting in a descending flaccid paralysis. Neurotoxin has a high affinity for the receptors causing slow recovery which begins only when new receptors have formed. Soft tissue infections caused by spore-forming bacteria have emerged as a serious problem in intravenous drug users (IDUs). Higher purity heroin permits IDUs who have lost access to veins to use subcutaneous or intramuscular routes, so-called 'skin-popping' and 'muscle-popping'. Use of large amounts of citric acid as a solvent for the heroin or drug contamination may increase tissue damage at the injection site. An anaerobic environment facilitates the germination of C. botulinum.

We undertook a case study and literature review.

Among the 24 consecutive patients with mechanical large bowel obstruction, 6 (25%) were treated conservatively and 18 (75%) were operated. Intraoperatively, no case of reversible bowel ischemia was observed. In contrast, large bowel necrosis was found in 4 patients (16.7% and 22.2% of the study group and the operated group, respectively), perforation of the necrotized large bowel in 3 (12.5% and 16.7% of the study group and the operated group, respectively) and small bowel perforation in one patient (4.1% and 5.5% of the study group and the operated group, respectively). Sigmoid carcinoma was the most common cause of obstruction accounting for 54.1% and 66.8% of the study group and the operated group, respectively (n=13). Adhesions were the second most common cause accounting for 29.1%, 16.7%, and 66.6% of the study group, the operated group, and the nonoperatively treated group, respectively (n=7). Incarcerated umbiligal hernia was the etiology of obstruction in 1 patient (4.2% and 5.5% of the study group and the operated group, respectively), diverticulitis in 1 (4.2% and 16.7% of the study group and the nonoperatively treated group, respectively), sigmoid volvulus in 1 (4.2% and 5.5% of the study group and the operated group, respectively), and ovarian cystadenocarcinoma in 1 patient (4.2% and 5.5% of the study group and the operated group, respectively). Etiology of large bowel necrosis was sigmoid adenocarcinoma in 1 patient (25%), adhesions in 1 (25%), incarcerated umbiligal hernia in 1 (25%), and sigmoid volvulus in 1 patient (25%). Three of the 4 patients with necrotized large bowel had large bowel perforation (1 with sigmoid adenocarcinoma: 33.3%, 1 with adhesional obstruction: 33.3%, and 1 with incarcerated umbiligal hernia: 33.3%) while 1 patient (adhesional obstruction) presented concomitant small bowel perforation.CONCLUSION. In our study, sigmoid carcinoma and adhesions were the most common causes of acute mechanical large bowel obstruction. Great caution should be paid on treatment of these patients since the incidence of bowel perforation and necrosis is significantly high.

Pujol W 1 , Seguy B 1 , Riahi T 1 , Pillet O 1 , Boyer A 1 , Bui N 1 , Gbikpi-Benissan G 1 1 Réanimation médicale, CHU Pellegrin-Tripode, Bordeaux, France

New treatments are needed to control prolonged status epilepticus given the high failure rate of current therapies.

We report the case of a forty-five years old woman admitted in our critical care unit for a Glasgow-scale 3 coma.

The electoencephalographic monitoring showed a non-convulsive status epilepticus. The morphological explorations with cerebral CT-scan and RNM were strictly normal. Two lumbar punction were negative. No aetiology was founded (infectious, immunological, traumatic, toxicologic and vascular aetiologies). The recommended treatments with benzodiazepins and phenytoin were ineffective. Barbituric coma with penthotal was induced during twenty days but ineffective. Valproic acid and propofol were also ineffective. Then ketamine at the posology of 2mg/kg/24h was introduced and broke this refractory status epilepticus in two hours. Intravenous support was performed during ten days and the posology of ketamine slowly decreased.

Ketamine is a non-competitive N-methyl-D-asparate (NMDA)-receptor antagonist. It has been used in a few cases of prolonged refractory status epilepticus (1, 2) . Pharmacological treatment of seizures, including status epilepticus, is therefore often directed at potentating GABAergic inhibition; but functional modification of GABA-A receptors has been described in status epilepticus and this may account for increasing resistance to GABAergic anti-convulsive agent, such BZDs and barbiturics. The management of refractory status epilepticus could therefore include use of NMDAreceptor antagonists, such as ketamine. We discuss with this case-report about indication and optimal posology of ketamine in refractory status epilepticus. 

Maintaining an optimal level of comfort and safety for critically ill patients is an universal goal for critical practitioners. The assessment of sedation level remains a challenge for the intensivist in order to avoid over-or under-sedation phenomena. The indroduction of the bispectral index (an EEG parameter) could bring potential adventages in monitoring sedation. According to the repports, the Richmond Agitation-Sedation Scale (RASS) has been shown to be highly reliable among multiple types of healt care professionals. The RASS has an expanded set of scores at pivotal levels of sedation that are determinated by patients' response to verbal vs physical stimulation, which will help the clinician in titrating medication.METHODS. This is a prospective, nonramdomized, observational study in a surgical and trauma tertiary intensive care of an universitary hospital. Twenty-six consecutive neurosurgical postoperative patients (age range 17-68 yrs, mean age 44 yrs) were included. They were sedated (with propofol by continuos infusion at an initial dose of 2 mg/kg/h, which could be modulated with steps of 0.5 mg/kg/h), in order to mantein an adecuated RASS score. BIS value was continuosly recorded, and manually calculated on a mean average of a minute during the measuring of RASS score, and every 10 minutes for 6 hours on par with RASS score. ECG, SpO2, invasive arterial pressure, ventilatory module, ETCO2, FiO2, temperature were also recorded. For the statistic analysis, Friedman test and Spearman coefficient were utilized.

Nine hundred and thirty-six observations were carried out. The variation range of RASS score was between 0 and -5. BIS range varied from 27 to 96. Statistics analysis of the data obtained pointed out a significative correlation between RASS score and BIS (p < 0.01).

Acording to the reports, Bispectral index correlates with levels of sedation on the sedation scales. In our personal experience, this study demonstrates the utility of BIS and RASS score to track levels of sedation in neurosurgical postoperative patients. 

Tracheostomy represents a prevalent technique in intensive care unit (ICU), and many advantages are attributed to it thought there were many questions about tracheostomy that still are discussed. The tracheostomy in patients with trauma, neurologic and neurosurgical pathology adds particular points. The purpose of this study was to describe tracheostomies carried out in a neurotrauma ICU, characteristics and outcomes of those.

Observational study about patients who required tracheostomy for 2005. All patients who received tracheostomy in ICU were included except if they were exitus in the unit. We collected demographics, admission reason, APACHE II, extubation failure, tracheostomy reason, days from initiation of ventilation to tracheostomy, from tracheostomy to weaning, from tracheostomy to discharge from ICU, type of tracheostomy procedure (surgical versus percutaneous), complications of the technique, ICU length of stay (LOS), time of tracheostomy necessity, pneumonia and late complications. Continuous variables are expressed as means ± standard error and medians and interquartile ranges are also given. Categorical variables are expressed as absolute frequencies and were compared using χ 2 test. Continuous and categorical variables were compared using t-test.

Of 537 admissions, 10.6% required tracheostomy and 12 of them were exitus, so our sample made up of 44 patients: 77% were men with age of 47.73 ± 17.92 years. Admission reason were distributed in 52% traumatic, 43% neurological/neurosurgical and 4% of other etiology. APACHE II was 19.82 ± 6.467. Extubation failure was observated in 36%. The more frequently reason for tracheostomy was the neurologic situation (60%). Time admission-tracheostomy was 13.25 ± 6.078 days, time ventilation after tracheostomy had a median of 2 days with p25 of 1 day and p75 of 5 days. Surgical technique was the more frequently (70.5%). Not complications were observed in 91%, only 6.8% of hypotension and one case of no mayor haemorrhage. Not difference were observed between surgical and percutaneous technique neither in appearance complications, nor ICU length of stay, nor time of ventilation after tracheostomy. Pneumonia was observed in 68% and these patients had a bigger time from admission to tracheostomy than the rest (14.77 ± 6.328 vs 10 ± 4.010 p= 0.01). ICU LOS was 28.68 ± 14.79 days. Time of need of tracheostomy was 35.16 ± 15.908 days and the 87% had not lates complications.

In our ICU the more frequent reason for tracheostomy was the neurologic condition, predominating surgical procedure without complications during the technique and without lates complications. Tracheostomy timing is relacionated with complications as pneumonia.

Tsarenko S S 1 , Petrikov S S 1 , Davidova L A 1 , Potapova N A 1 1 Neurosurgical ICU, Sklifosovsky Scientific Institute of Emergency Medicine, Moscow, Russian Federation

Administration of glucose solutions in patients with severe traumatic brain injury (TBI) is controversial. From the one hand glucose is the main energy substrate for the injured brain. From the other hand hyperglycemia can lead to the hypoosmolality, lactate acidosis, and increase in incidence of poor neurological outcome and infection complications.METHODS. 107 patients with severe TBI enrolled in the study. Patients were randomized into two groups. In the first group glucose solutions were not used as a component of fluid therapy ("Saline" group). As a basic fluid therapy patients received only normal saline and colloids if necessary. Patients of the second group were administrated the same basic fluid therapy, but 30-40% of all intravenous fluid amount was given by 10% and 20% glucose solutions ("Glucose" group). Hypoosmotic 5% glusose solution was never used. "Nutrison-energy" (Nutricia) 30-40 kcal/kg/day was used for enteral nutrition. Patients of the both groups were of the same age, gender, Glasgow Coma Scale (GCS) on admission to the ICU and the length of the ICU treatment. Serum glucose levels, GCS, pneumonia and meningitis incidence, neurological outcomes according to Glasgow Outcome Scale (GOS) were analyzed and compared between the groups.

The serum glucose level on admission to the ICU was (M±SD) 8.4 ± 3 mmol/l in "Glucose" group and 7.9 ± 2.1 mmol/l in the "Saline" group (p>0.05). In the "Glucose" group neurological status improvement in the acute period of TBI was observed in 23 patients (52.9%), but in the "Saline" group only in 25 patients (39.7%) (p<0.05). Pneumonia incidence was 32.2% in the "Glucose" group and 48% in the "Saline" group (p>0.05). Meningitis occurred in 11.5% of patients in the "Glucose" group and in 8% of patients in the "Saline" group (p>0.05). Poor neurological outcomes and mortality (GOS 1.2) were the same in both groups (48% of patients).

Hypertonic glucose solutions don't worsen the neurological outcome and have no influence on the infection complications incidence in patients with severe traumatic brain injury. 

Basal skull fractures with CSF rhinorrhea and the craniofacial injury that requires plastic surgery are usually from the list of the main indications for the early PDT (less than 72hrs) in patients with severe brain injury. An ICP elevation is of great risk and thus limitations arise. From the position of applied medicine, we modified the PDT technique, taking to the consideration the need for intracranial hypertension avoidance.

The investigation represented a prospective observational trial based upon the analysis of PDT procedure in 24 patients with severe traumatic brain injury. GCS 4-8. Continuous ICP monitoring was placed just as SpO2 and ETCO2. One of the inclusion criteria was ICP elevated (22-35 mmHg). We were striving for ICP below 20 mm Hg as a first step. We also performed a test with an ICP reaction to the horizontal positioning beforehand. The endoscopic PDT, Griggs set was used. A discrete endoscopy never exceeded 60 sec. The modification included the peculiarities of the patient positioning: head of the bed elevation -30• , avoiding neck extension.

According to the level of an ICP increase the collected results were divided for two groups. The first group consists of patients whose ICP was never above 25 mm Hg throughout the procedure. The second group is precisely the same contingent based on the traits of pathology but we use to depict an ICP increase over 25 mm Hg. An elevation during procedure never exceeded 5 min. Analyzing the date we came to the conclusion that best results were obtained if favorable ventilation is maintained. The ratio between the diameters of the endotracheal and endoscope supposed to be not less than 3/2.

In patients with severe TBI, complicated by elevated ICP, a timely performed PDT (within 72 hrs) can be safe if modified technique is used.

Mandila C G 1 , Koukoulitsios G V 2 , Stathopoulos G T 1 , Papakonstandinou K 1 , Aggelopoulou E 1 , Karabinis A 2 1 ICU, General Hospital of Athens G Gennimatas, 2 ICU, General Hospital of Athens G Gennimatas, Athens, Greece

Transesophageal echocardiography (TEE) and computed tomography (CT) are routinely used in the emergency and intensive care settings for the evaluation of patients with blunt thoracic trauma and suspected traumatic aortic injury (TAI). We report two cases of traumatic aortic rupture in which the CT scan evaluation was negative for TAI and the suspected diagnosis was confirmed by TEE.

The records of all consecutive trauma patients treated between January 2003 and January 2006 in our intensive care unit were retrospectively reviewed.

. The records of 110 trauma victims were evaluated. 40/110 (36%) had blunt chest trauma among other injuries and/or fractures of the upper ribs, sternum, and/or clavicles on the initial chest X-ray. These patients were evaluated using both CT and TEE. No patient had evidence of TAI on CT. However, two patients with negative CT scans showed evidence of TAI on TEE: one patient had isthmic rupture and the other intramural hematoma.CONCLUSION. Both CT and TEE have high sensitivity and specificity in the diagnostic evaluation of TAI. To minimize the risk of diagnostic errors in this potentially catastrophic situation, CT & TEE should be considered as complementary tests.

Mannitol and hypertonic saline are widely used for intracranial hypertension treatment. The best method of their effectiveness control is invasive intracranial pressure (ICP) monitoring. But it is impossible in all Intensive Care Units. Therefore it's useful to find simple methods of mannitol and hypertonic saline effectiveness assessment.

We investigated 12 patients with intracranial hemorrhage. All patients had invasive ICP monitoring (Codman). ICP elevation above 18 mm Hg was an indication for treatment. In order to decrease ICP we used 15% mannitol -400 ml,1.5 g/kg/h (n=6) and 10% Sodium Chloride, 400 ml/h (n=6). The ICP, plasma osmolality, sodium serum level and neurological signs of brain stem dislocation were studied before and after infusion.RESULTS. 10% NaCl infusion decreased ICP from (M±δ) 29.5±8.3 mm Hg to 15.8±6.8 mm Hg (P<0.05) and was followed by serum sodium level increase from 141±1.9 mmol/l to 146±1.5 mmol/l (P<0.05). Serum osmolality didn't change (280±6 mosm/kg vs 293±12 mosm/kg). Positive changes in neurological status were seen only in 2 patients (33%). Mannitol infusion was followed by ICP decreasing from 30.3±8.8 mm Hg to 19.8±9.6 mm Hg (P<0.05), and plasma osmolality increasing from 296±20.2 mosm/kg to 312±16.9 mosm/kg (0.05). No marked changes in serum sodium level were observed (145.3±2.8 mmol/l vs 145.5±4.7 mmol/l). Positive changes in neurological status were observed in one patient (17%). Changes in ICP and osmolality after 10% NaCl and mannitol infusion differed non-significantly between groups.CONCLUSION. 1)10% NaCl and mannitol effectively decrease ICP 2) Invasive ICP monitoring is the best method of mannitol and hypertonic saline effectiveness assessment.3) Serum sodium level (in case of 10% NaCl) and serum osmolality (in case of mannitol) could be used as alternative methods of hypertonic saline and mannitol effectiveness control. 4) Neurological examination can't reflect ICP changes during hypertonic saline and mannitol infusion. (1). However the precise structure of this channel remains unknown. Under certain conditions, F0F1 ATPase is a channel which can also conduct potassium (2).

By using siRNA techniques on cardiomyocytes, we decreased IF1 expression, a regulator of F0F1 ATPase. With confocal microspcopy, we studied flavoproteins oxidation, a marker of mito K ATP activation, and we measured the time to open the mitochondrial permeability transition pore (Tmpt), a marker of cardioprotection (3). We also measured oxygen consumption (VO2) in presence of venturicidin, a specific inhibitor of F0 subunit. Statistics by t test and ANOVA.

In presence of Dz, flavoproteins fluorescence is increased. However, if IF1 expression is decreased, flavoproteins fluorescence does not increase with Dz. Dz increases Tmpt in control group.* p<0.05 vs control; # p<0.05 vs 100% IF1; $ p<0.01 vs Dz. We hypothesize that the combination of the underlying illness and the side-effects of its treatment will result in permanent damage of the developing heart in children.

SS survivors admitted to our tertiary PICU between 1995 and 2004 and received administration of VA's ≥ 24 hours were included. Cardiac function was evaluated by ECG in rest and during exercise (when > 7 years), 24-hours-ECG-registration and echocardiography. Age at admission and follow-up, severity of illness (PIM II score), length of PICU stay were evaluated.

Ninety of 124 eligible patients were evaluated. In 14 patients abnormalities were detected; such as episodes of ventricular extrasystole during and after exercise (n= 3), rhythm disturbances on 24-hours-ECG registration (n=2), mild left ventricular hypertrophy (n=1) and mild systolic dysfunction of the right ventricle (n=8) on echocardiography. Patient characteristics were not statistically different between patients with and without abnormalities.

A 15% prevalence of ventricular dysrhythmia and mild dysfunction of the right ventricle is found in 90 SS survivors, 1-10 yrs after admission. Since ventricular rhythm disturbances could lead to sudden cardiac death and the significance of dysfunction of the right ventricle in developing children is not known, long term follow-up of survivors of SS in childhood is warranted.

Pulmonary embolism remains an important clinical problem with high mortality rate. Data from this study provide highlights on this problem and may help in raising awareness of the importance of identifying patients at high-risk for developing pulmonary embolism or an unfavorable outcome. Critical care departement, 2 Cardiology departement, 3 Internal medicine departemenyt, Cairo University Hospitals, Cairo, Egypt

Adiponectin is an adipocyte derived plasma protein that shows a number of antiatherogenic properties. It is abundantly present in the circulation and also accumulates in the injured arteries. This study was designed to determine whether a decreased plasma adiponectin level is associated with the extent of atherosclerosis in patients with coronary artery disease (CAD).

Plasma concentrations of adiponectin were measured in 97 patients with CAD and 22 healthy control participants. All patients underwent coronary angiography for clinical purposes. Patients were divided into two groups according to the number 50% diameter stenosis by quantitative angiography ≥ of coronary plaques causing 3 plaques (n= 47) and group 2; ≤(QCA): group 1; >3 plaques (n=50) with a particular focus on the relation between plasma adiponectin concentrations and the presence or absence of cine-angiographic evidence of coronary calcification (a marker of heavy atherosclerotic burden).

. Plasma concentrations of adiponectin in patients with CAD were significantly 2.4 ug/ml, p= 0.03). Patients in± 3.7 vs 7.3 ±lower than in control group (4.6 group 2 had a significantly lower plasma adiponectin level than patients in 4.5 ug/ml, p= 0.05). Patients with calcified lesions± 2.7 vs 5.7 ±group 1 (3.6 (n= 33) had also a significant lower level of plasma adiponectin than patients ± 1.9 vs 5.2 ±with no evidence of cine-angiographic calcification (n= 64) (3.4 4.3 ug/ml, p= 0.05). On the other hand, plasma adiponectin level showed a significant negative correlation with body mass index (BMI) (r= -0.313, p= 0.015). Multiple regression analysis showed that plasma adiponectin concentrations correlated independently with increased number of coronary plaques >3 (p= 0.05).

The present study showed that decreased plasma adiponectin concentrations can predict severity of CAD and may be related to increased atherosclerotic burden. INTRODUCTION. The aim of this study was to assess the effect of surgical coronary revascularisation and associated functional recovery on myocardial metabolism in patients after myocardial infarction. The microdialysis technique was used to estimate changes in interstitial metabolites.

Microdialysis catheters were either implanted into an akinetic area of the left ventricle with transmural extent of hyperenhancement of more than 50% (scar group, n = 10) or into a segment with mild / moderate hypokinesia and less than 50% of transmural hyperenhancement (viable group, n = 10). Myocardial glucose, lactate and pyruvate were analyzed before, during and 24 hours after CABG. The myocardial glucose-lactate-ratio (GLR) and myocardial/systemic glucose ratio (m/s) -as markers of nutritional disorder -were calculated. Myocardial ethanol washout from the microdialysis probe was also measured as a sign recovered of local blood flow.

. Improvement of wall motion was found in all viable compared to the scar segments. After surgical revascularisation recovering of the myocardial flow in these areas resulted in increased glucose delivery to the tissue with a significant better m/s-glucose-ratio. The myocardial GLR and pyruvate levels showed also significant higher values. Restored myocardial blood flow was detected by using the ethanol washout technique.

Our results indicate that revascularisation of chronic, ischemic myocardium with dyskinetic segments resulted partly in an early functional improvement with normalized wall motion. The metabolism of those segments is characterized by a significant increased tissue flow, increased utilization of glucose and a better oxidative nutrition. 

Intraaortic balloon counterpulsation (IABP) is an established treatment for the support of a failing heart. A reduction in complication and mortality rates over recent decades has seen an increased use of IABP with current studies increasingly examining the changing trends in its use. The aims of this study were: to describe indications for IABP use and identify the impact these have on outcomes at an Australian cardiothoracic tertiary referral hospital; and compare these Australian aspects of practice and outcomes with those in a large multinational IABP data registry to benchmark practice. CONCLUSION. Thrombomodulin, an endothelial glycoprotein resulting from the damage to vascular endothelium by the atheromatous pro cess, showed significant correlation with the extent & seve rity of CAD (expressed by Gensini score system). The high er TM level in AMI Gp compared to AP Gp points to the more significant endothelial damage in the former compared to the latter.that stresses the importance of plasma TM as a molecular marker of endothelial dysfunction in acute isch emic syndromes. 

Early, high lactate clearance is associated with improved mortality in severe sepsis and septic shock. We investigated whether early, high lactate clearance was associated with reduced mortality in post-cardiac arrest patients.

We performed a retrospective analysis of post-cardiac arrest patients in an urban emergency department. Inclusion criteria included pre-hospital cardiac arrest patients, and exclusion criteria included traumatic arrest or arrest in the presence of healthcare personnel. The primary endpoint was survival to 24 hours, and the secondary endpoint was survival to hospital discharge. Lactate clearance, defined as the percentage change in lactic acid levels over a given period of time, was tracked at 6 and 12 hours. A multivariable regression model and Student's t-test were used for data analysis.

A total of 79 patients were analyzed with a mean age of 64±17 years and mean APACHE II score of 37.7±5. Of the 79 patients, 27 (34%) died within 24 hours and 66 (84%) died during the hospital course. The mean initial lactate levels were similar among 24-hour and overall survivors and non-survivors. However, lactate clearance at both 6 and 12 hours was significantly higher for both 24hour and overall in-hospital survival ( Table 1 , p < 0.05). A multivariable analysis inclusive of APACHE II score, lactate clearance, initial heart rhythm, and downtime was performed for the primary endpoint of 24-hour mortality. High lactate clearance at 12 hours remained predictive of survival (p < 0.05), whereas all other variables were no longer statistically significant. 

Early, high lactate clearance is associated with early and overall in-hospital mortality in post-cardiac arrest patients. These findings suggest that post-arrest tissue hypo-perfusion plays an important role in mortality.

Stephan M 1 , Schwarte L 1 , Schwartges I 1 , Picker O 1 1 Anaesthesiology, Heinrich-Heine-University, Düsseldorf, Germany INTRODUCTION. Cardiovascular dynamics in health and disease appear to reflect complex regulatory mechanisms [1] , which can be described by analysing haemodynamic entropy. Decreases in entropy have been associated with various critical pathologies, e.g. sepsis [2] . We studied the impact of graded catecholamine infusions on haemodynamic entropy.

Six anaesthetized and mechanically ventilated foxhounds (propofol 20 mg/kg/h, FiO2 0.3, etCO2 35 mmHg) received increasing doses (0, 0.05, 0.1 and 0.2 µg/kg/min) of either epinephrine (EPI) or norepinephrine (NOR). Systolic arterial pressure (SAP) and heart rate (HR) time series (720 cardiac cycles) were analysed using a multiscale entropy algorithm as previously described (vector length 1 to 2, tolerance 0.15, scale 1 to 4) [3] . Statistics: Means ± SEM, ANOVA, p<0.05.

. EPI decreased entropy of HR from 1.6±0.2 to 0.7±0.1 at 0.05 µg/kg/min, whereas NOR did not influence entropy of HR. In contrast, EPI increased SAP entropy dose dependent whereas NOR did not (significant intergroup differences at 0.1 and 0.2 µgkg-1min-1).

Catecholamines have complex influence on haemodynamic entropy. These effects are dose and substance dependent with minor influence of NOR, and marked effects of EPI. Thus, in patients, catecholamine therapy may be one factor differentially influencing haemodynamic entropy. 

Weaning from mechanical ventilation (MV) can be associated with cardiac arrhythmias. Few studies are found comparing their occurrence during weaning with pressure support ventilation (PSV) and T-tube (TT) in patients with and without heart disease. The objective this study is evaluate the occurrence of arrhythmias in these groups of patients during PSV and TT.

Patients without (group 1) and with (group 2) heart disease, under mechanical ventilation for at least 48 hours, were observed during 30 minutes of PSV or TT, in a random order. Variables analyzed were: age, APACHE, length of stay in ICU (LOS), cardiorespiratory variables including respiratory rate, rapid shallow breathing index (f/VT), maximum inspiratory (PImax) and expiratory (PEmax) pressure. Continuous ECG was recorded by Holter method. For statistical analyzes repeated measures ANOVA or ANOVA on ranks were used.

Twenty-two patients were studied, 13 in group 1 and 9 in group 2. Comparisons between groups 1 and 2 showed: no differences were found in APACHE (23 ± 4; 23 ± 8, NS), PImax (32 ± 19; 28 ± 12 cmH2O, NS) and PEmax (24 ± 10; 20 ± 7 cmH2O, NS); f/VT was greater in cardiac patients during TT (PSV: 48 ± 25 versus 41 ± 18; TT: 42 ± 18 versus 57 ± 20, ANOVA: p < 0.05), as well as respiratory rate (PSV: 21 ± 6 versus 20 ± 5; TT: 22 ± 6 versus 25 ± 6, ANOVA: p < 0.05). The occurrence of ventricular arrhythmias (median and interquartile ranges), respectively in PSV and TT were in group 1: 1 (0 -13) versus 1 (0 -5.5) and in group 2: 3 (0.5 -87) versus 21 (4 -61), ANOVA: p < 0.05.

During weaning from MV cardiac patients showed higher respiratory rate and higher f/VT during TT when compared with PSV, as well as a greater occurrence of ventricular arrhythmias in both methods, but principally during TT, when compared with non-cardiac patients. 

Levosimendan is a novel inodilator that has proved to be effective in treating advanced congestive heart failure. The aim of this study was to assess hemodynamic effects of Levosimendan in patients with cardiogenic schok refractory to catecholamines.

Nine patients (53.3±19 years, 5 male/ 4 female) with persisting cardiogenic shock following acute myocardial infarction (5), peripartum cardiomyopathy (2) or dilated cadiomyopathy (2) were candidates to levosimendan infusion. In all patients, high dose of inotropic treatment failed to improve hemodynamic parameters. Levosimendan was introduced at a loading dose of 12 µg/kg followed by a continuous infusion of 0.1 µg/kg/min for 24hours. Hemodynamic measurements were performed using a Swan-Ganz thermodilution catheter at baseline and at 30, 90 minutes, 6, 12, 24 and 48 hours after the starting of levosimendan. Transoesophageal echocardiography was performed at baseline, 12, 24 hours than after 7 and 15 days in survivors.

Levosimendan induced a significant decline of pulmonary capillary wedge pressure and systemic vascular resistances, followed by a significant increase in cardiac index (CI) and mixed venous oxygen saturation. Changes in heart rate and mean arteriel blood pressure were not significant. Left ventricular ejection fraction was improved from 24% to 40% within 48 hours.CONCLUSION. this study showed that the use of levosimendan in cardiogenic shock, improved hemodynamics and left ventricular performance. Additional clinical trials are needed to safely broaden its indications in cardiogenic shock. 

The clinical and radiological signs of cardiogenic pulmonary edema (PE) induced by weaning from the mechanical ventilation are not specific and its diagnosis of certainty is based on the measurement of PAOP. During weaning-induced PE, a low-protein concentrated fluid is transferred toward the alveolar/interstitial space, inducing a relative concentration of the plasma that could result in a significant increase of the plasma protein concentration. We investigated whether the increase in plasma protein concentration during a weaning test could detect reliably weaning-induced PE.

In 29 patients who failed at two consecutive weaning tests on a T tube, a pulmonary artery catheter was inserted. A weaning PE was diagnosed if occurred during a subsequent test (1) tachypnea ≥35/min and/or tachycardia ≥120/min and/or a fall in arterial pulse oximetry ≤90% associated with (2) an increase in PAOP above 18mmHg. The plasma protein concentration was measured before and at the end of the weaning period.

A weaning PE was observed at the end of the weaning test in 23 patients (increase in PAOP from 13±3 to 29±8mmHg, p<0.05). In these patients, the protein concentration increased simultaneously by 9±5% from 58±8g/L. In the patients who did not exhibit PE, protein concentration did not change significantly during the weaning test (3±2% increase from 61±5g/L). More importantly, an increase in protein concentration above 5% during the weaning test predicted the occurrence of weaning PE with a sensitivity of 82% and a specificity of 100%.

Weaning-induced PE can be reliably detected by an increase in plasma protein concentration during a weaning trial.

Prognosis and outcome (II) 0724-0737 0724

Gamsjäger T 1 , Sitzwohl C 1 , Anderle C 1 , Andel H 1 , Kettner S 1 , Krenn C G 1 , Zimpfer M 1 , Weinstabl C 1 1 Anesthesiology and General Intensive Care, Medical University of Vienna, Vienna, Austria

Woodward R 1 , Jansen T C 1 , Mulder P G 2 , Bakker J 1 1 Intensive Care, 2 Epidemiology and biostatistics, Erasmus MC University Medical Center, Rotterdam, Netherlands

Lactate has since long been known as a prognostic marker for outcome. However, SOFA score as an important organ dysfunction score on the ICU has not been related to lactate levels before. The aim of this study was to evaluate the relationship between area under the curve of lactate measurements and serial SOFA scores.

We retrospectively gathered information from consecutive patients admitted to the mixed ICU of a university hospital during 2 months in 2005. Exclusion criteria were liver failure, admission post-liver transplantation and no availability of lactate measurements. Data regarding the first 28 days of stay on the ICU were collected and included serial lactate levels and daily SOFA scores. Total SOFA (sum of daily SOFA scores), mean SOFA (total SOFA divided by number of days) and delta-SOFA (difference in SOFA score between first and second day) were calculated. Area under the curve (AUC) of lactate levels > 2.0 mmol was calculated for the period where lactate levels first exceeded 2.0 mmol/l (within first 24 hours of ICU admission) until normalisation (< 2.0 mmol/l). 

The duration and extent of hyperlactatemia are related to serial SOFA scores, demonstrating more severe organ dysfunction following lactate accumulation over time.

Milic M 1 , Goranovic T 1 , Husedzinovic I 1 1 Clinic of Anaesthesiology and Intensive Care, University Hospital Dubrava, Zagreb, Croatia INTRODUCTION. APACHE II (acute physiology and chronic health evaluation) and SOFA (sequential organ failure assessment) scores could be used as predictors of length of stay (LOS) in various intensive care units (ICUs) (1, 2) . We compared two surgical ICUs (general surgical vs cardiosurgical) to find out usefulness of these scoring systems for predicting LOS in surgical ICUs. We hypothesized that significance of scoring for prediciting LOS is greater in specialized surgical ICUs (i.e. cardiosurgical).

We scored patients in our general surgical ICU (n = 328) and cardiosurgical ICU (n = 158) consecutively on admission (APACHE II-1st day; SOFA-1st day) and on third day of stay (APACHE II-3rd day; SOFA-3rd day) in a 4 month period. There was no significant difference between the study groups according to gender ( 

To study the outcome of elderly patients in an interdisciplinary ICU.

Retrospective 9 year study of all patients staying >48h in the ICU. Patients were compared according to 3 age groups: GroupI <65y, Group II 65-79y, GroupIII >80y. Age, gender, APACHE II, cause of admission, length of stay (LOS), duration of mechanical ventilation (DMV) and patients outcome were studied using multivariate logistic regression. Table I results are shown according to age group. In Table II are shown parameters correlated to outcome. DMV (Odds ratio 0.876, CI 0.864-0.893, p< 0.06) and cause of admission did not influence patient outcome. 

Patients outcome was related directly to age and severity of illness, inversely to length of stay and was not influenced by gender and reason of admission. SOFA score was calculated on admission and every 24 hours until the fifth day in ICU. Delta SOFA score on day 1 was defined as the difference between day 2 SOFA score and admission; delta SOFA score on day 2 was defined as the difference between day 3 SOFA score and admission, and so on. The discriminative power of delta SOFA score for the prediction probability of hospital mortality was tested by the area under the receiver operating characteristic (ROC) curve. The area under the ROC curve summarizes the relationship between 1 minus the sensitivity (number of true positives) and specificity (number of false positives) for all the possible values of the organ dysfunction scores and estimates the ability of the model to assign a higher risk of death to patients who die. P value < .05 was considered significant. Statistical analysis was performed using the SPSS 11.0 statistical package.

Among 305 patients admitted to the ICU (mean age 57.3 years) the severity scores at admission were mean SAPS II score 36.3 points and mean SOFA score 6.7 points. The mean length of stay in the ICU was 14.2 (1-87) days, in the hospital 40 days. The overall mortality in the ICU was 30.8% with a hospital mortality of 41.3%. The outcome measure used was survival status at discharge from the hospital. Compared with the other delta SOFA, delta SOFA on day 5, presented the largest area under the ROC curve (95% CI 0.565-0.701, SE 0.35, p=0.0001), followed by delta SOFA 3 (95% CI 0.515-0.659, SE 0.37, p=0.02) and delta SOFA 1 (95% CI 0.513-0.652, SE 0.35, p=0.02).

Delta SOFA score 4 can identify better than the others delta SOFA score the critical point at which patients exhibits the highest degree of organ dysfunction during the ICU stay. This suggest that delta SOFA score 4 is effective in predicting hospital outcome.

Holtfreter B 1 , Bandt C 1 , Kuhn S O 2 , Schütt C 3 , Gründling M 2 1 Institute for Mathematics and Informatics, 2 Department of Anaesthesiology and Intensive Care Medicine, 3 Department of Immunology, Ernst-Moritz-Arndt-University, Greifswald, Germany

In the current study we validated the predictive value of osmolality, laboratory parameters as well as APACHE II score for hospital outcome. Furthermore, validation of non-linear and logistic regression models for hospital mortality derived from the training data was intended.

Two retrospective observational clinical studies were conducted in a 15-bed mixed ICU in a university hospital each with a length of three years. The training data (TD) involved 931 consecutive patients with an ICU stay >24h, whereas the validation data (VD) involved 1031 patients. Primary outcome parameter was hospital mortality. Receiver Operating Characteristic (ROC) analyses, logistic models as well as were calculated to assess the predictive value of parameters. To test prediction models from the TD goodness-of-fit (GOF) statistics and quality of discrimination were determined.

Hospital mortality was 25.0% (TD) and 22.9% (VD). APACHE II score predicted best in both data sets (AUC 0.784 and 0.749). Serum osmolality revealed best prognosis among clinical parameters (AUC 0.732 and 0.687). However, restricted to long-term patients with ICU stay >5 days, serum osmolality exceeds APACHE II in outcome prediction. Non-linear models with serum osmolality and APACHE II predicting hospital outcome in the training data calibrated well for the validation data (GOF, p=0.97 and p=0.89, respectively) . A multivariate logistic model derived from the training data discriminated well in the validation data (AUC 0.697) with sensitivity of 61% and specificity of 62%.

Here, we report that osmolality is useful in outcome prediction. Though predictive value for osmolality was less than in the training data, the association between osmolality and hospital mortality could be confirmed. Moreover, prognostic regression models predicting hospital outcome derived from the training data could be validated. 

Savi C 1 , Cislaghi F 1 , Corona A 2 , Ferrari C 1 , Condemi A 1 1 Cardioaneasthesia, Luigi Sacco Hospital, Milano, 2 Clinical Epidemiology Laboratory, Mario Negri Insitute, Ranica, Italy

Identifying factors that affect cardiac surgical patient hospitalization length (HL) could improve their healthcare management and costs.

Goswami A 1 , Paddle J J 1 1 Department of Intensive Care, Royal Cornwall Hospital, Truro, United Kingdom INTRODUCTION. The APACHE II system uses the worst pH in the first 24 hours of ICU admission in its calculation of illness severity. Other measures of pH, such as pH on admission to ICU and average pH over the initial 24 hours of ICU admission, may be more valid. We therefore designed a study to compare these three measures of pH and to correlate them with patient mortality.

A retrospective audit was carried out on all patients admitted to the ICU between 5th January 2005 and 1st June 2005. Demographic information and outcome data was obtained from the ICNARC database. The ICU Arterial Blood Gas analyzer (ABL 700 Series, Radiometer Copenhagen) was used to provide data from which admission pH, worst pH and a time-weighted average 24-hour pH were obtained.

Of the 199 patients in the study 159 survived to ICU discharge. The mean age of patients was 56.9 years; the mean APACHE II score on admission was 18.1. The admission pH, worst pH and average 24-hour pH were significantly greater in ICU survivors than non-survivors (p < 0.001). For admission pH, worst pH and average 24-hour pH odds ratios (OR) were calculated examining the risk of ICU mortality within particular pH ranges compared with the risk of ICU mortality within the normal pH range (pH 7.35 -7.45). 

The admission pH, worst pH and average 24-hour pH were significantly higher in ICU survivors than non-survivors. For all three parameters the risk of mortality increased with decreasing pH. For a given pH range the odds ratio for risk of mortality is greatest in the average 24-hour pH group and lowest in the admission pH group, suggesting average 24-hour pH is a stronger prognostic indicator than worst pH or admission pH.

Perioperative intensive care: Long term outcome 0738-0751 0738

Rouge A F 1 , Gomes R V 1 , Nogueira P M M 1 , Fernandes M O 1 , De Souza B V 1 , Spencer L 1 , Xavier S S 2 , Campos L A 1 1 Surgical Intensive Care Unit, Pró-Cardiaco, 2 Cardiology Department, Universidade Federal do Rio de janeiro, Rio de Janeiro, Brazil

Clinical and laboratory parameters that may be useful for estimating risks and monitoring and predicting dysfunctions have been largely studied in the postoperative (PO) period of cardiac surgery (CS). The objective this study is to correlate preoperative BNP levels of patients (pts) undergoing elective CS with pre-and postoperative clinical, hemodynamic and laboratory variables.

Prospective study with a cohort of 83 CS pts selected between August 2003 and September 2005. Their mean age was 67.0±8.55 years, 23 (27.3%) were females, and the mean Euroscore was 4.0±2.60. The BNP was quantitatively measured in the preoperative period (BNPPre) by using immunofluorescence (Biosite Triage BNP Test), and its correlation with 64 previously defined preand PO variables was assessed by use of the Spearman rank correlation.

Mean BNPPre was 181.0 ± 368.95 pg/mL (MED=58.7). The following variables correlated with BNPPre: arterial bicarbonate in the immediate PO period (rho = 0.313; p = 0.004); PaO2/FiO2 in the first hour (rho = -0.275; p = 0.012); base excess in the first hour (rho = 0.325; p = 0.003); pulse pressure variation in the sixth hour (rho = -0.297; p = 0.006); EuroScore (rho = 0.413; p<0.001); Cleveland Pre (rho = 0.348; p<0.001); and age (rho = 0.480; p<0.001).CONCLUSION. The correlation found between BNPPre and the pre-and PO variables supports the hypothesis that that peptide is a good marker of preoperative severity.

Bubenek S 1 , Miclea I 1 , Giurgiuman L 1 , Iliescu V 1 , Moldovan H 1 , Vasile R 1 , Radulescu B 1 1 Anesthesiology and Intensive care unit, Institute of Cardio-vascular Disease, Bucharest, Romania

Our objective was to assess the effectiveness of the augmentation of the mixed venous oxygen saturation (SVO2) over 60% in the patients with low cardiac output syndrome after cardiac surgery necessitating intra-aortic balloon counterpulsation (IABP).

The records of 181 patients with cardiac low-output syndrome who underwent openheart surgery and had an IABP implanted from 1994 to 2005 were reviewed retrospectively. The patients were divided into two groups: group A operated before 2002 (N=121) and group B operated after 2002 (n=60). In the group B the hemodynamic management included a constant interest in aggressively augmenting SVO2 (a value > 60% was aimed) and also a preoperative IABP insertion if necessary. The terms of outcome (i.e. in hospital mortality, weaning of IABP and mortality in IABP-weaned patients) were compared using a StatSoft programe.

Both groups did not differ with respect to their demographic data except for the rate of more complex surgery in the group B (44.4% vs 5.26%). The average in-hospital mortality rate in the group A was 61.9% and 33.3% in the group B. Only 43.8% of the patients in group A could be weaned of IABP as compared to 80% in the group B. The mortality in the IABP-weaned patients was 49.1% in the group A and lowered to 16.6% in the group B.CONCLUSION. the constant and aggressive augmentation of the SVO2 to a value over 60% improves outcome in this high risk patients; it is a mandatory objective in a good hemodynamic management. 

Prolonged ICU and hospital stay is an enough frequent concern (incidence 4.67%) after open heart surgery, which demands to recognize early indicative factors for avoiding the negative consequences in morbidity and mortality rate. Dusfunction of at least 3 organs (MODS) is not a rare cause of prolonged hospital stay after open heart surgery and develops principally in patients with low cardiac output syndrome postoperatively.

Paruk F 1 , Bhagwanjee S 1 , Vanderdonck K 2 , Cronje F 2 , Becker P 3 1 Anaesthesia, 2 Cardiothoracic Surgery, 3 Therapeutic Sciences, University of Witwatersrand, Johannesburg, South Africa

Mortality prediction in patients undergoing cardiac surgery has largely been studied by focusing on preoperative factors. Yet there is no consensus with respect to what constitutes the ideal outcome predictive tool in this subset of ICU patients. Objectives: To: (i) describe the ICU cardiac surgery patient profile, (ii) evaluate the Euroscore (standard and logistic) in coronary artery bypass graft (CABG) surgery and (iii) ascertain factors associated with mortality.

Method: A retrospective study conducted over a 1 year period in the Cardiothoracic ICU, in Johannesburg Hospital, South Africa. Strata 8 software was utilised for statistical analysis.

Two hundred and forty-five patients were admitted during the study period following CABG (n=87), valve (n=127) and miscellaneous (n=31) surgery. The mean age was 49 years and the mean duration of ICU stay was 4.5 days. The overall mortality was 12.2%. The Euroscore (standard and logistic) did not predict outcome and was similar for CABG surgery survivors and nonsurvivors. Preoperative insertion of an intraaortic balloon pump and preoperative cardiac failure were associated with a significant increase in mortality (p<0.05). The intraoperative bypass time was significantly longer in nonsurvivors compared to survivors (p=0.01). Postoperatively, major haemorrhage (>1litre), nosocomial sepsis, extubation failure and organ failure were significantly higher in nonsurvivors compared to survivors (p<0.05). These observations pertain to the entire cohort.CONCLUSION. The Euroscore is a poor predictor of outcome in our patients. The risk factors identified in this analysis need prospective validation.

Riera M 1 , Carrillo A 1 , Amezaga R 1 , Herrero J 1 , Fiol M 1 , Ibañez J 1 , Bonnin O 2 1 Intensive Care and Coronary Unit, 2 Cardiac Surgery, Son Dureta H., Palma, Spain

A tool to assess the potential need of improvements in quality of cardiac surgical (CS) care can be the review of databases. We analyse our database to assess the quality of our initial 3 years experience with postoperative CS care.

Prospective study of 1094 patients undergoing CS admitted in the ICU during 3 years. The collected data and two outcome measures (time to extubation and ICU length of stay) were compared to the European results*.

The average age was 64 years in both groups and there were more women in our group (g.): 34% versus (vs) 29%. The main interventions in our g. compared to the European g. were: Coronary (C.) surgery (s.) with EC-40% vs 57%; Valvular (V.) s.-34% vs 19%; and combined C. and V. s.-14% vs 9%. 

Knoester H 1 , Bronner M B 2 , Bos A P 1 1 Pediatric Intensive Care Unit, 2 Psychosocial Department, Emma Children's Hospital, AMC, Amsterdam, Netherlands

The development of pediatric intensive care (IC) has contributed to improved survival in critically ill children. Little is known about long term physical sequelae in pediatric IC survivors, but they might resemble the findings in adult IC survivors. Awareness of sequelae due to the original illness and its treatment may result in changes in management during the acute phase. The purpose of this study was to evaluate physical sequelae of survivors of Pediatric Intensive Care Units (PICU).

October 2005 all acute admissions to our PICU, were examined in the out-patient follow-up clinic, three months after discharge. A pediatric intensivist evaluated physical sequelae by standardized history taking and physical examination.

116 of 250 eligible patients were evaluated. Patient characteristics, PIM II score, length of stay and length of mechanical ventilation did not differ statistically significant between participants and non-participants. 55 patients suffered a chronic illness after discharge. Ten patients complained of concentration disturbances, 17 of behavioral problems, and 10 of sleeping problems. Nine of 40 school-age children had problems at school. Neurological abnormalities were seen in 24 children, post thrombotic syndrome after a central venous catheter in 5 and hoarseness after extubation in 3 children.CONCLUSION. PICU survivors have various, but substantial physical sequelae three months after discharge. Almost 50% of patients have a chronic illness and almost 20% of patients have neurological problems. Further research is warranted to determine risk factors and methods to prevent sequelae due to intensive care treatment. 

In this follow-up analysis of a large, randomized controlled trial (1), we assessed longterm outcome in 1548 patients admitted after high risk or complicated surgery. Long-term outcome was quantified as (a) 4 years survival, (b) incidence of hospital re-admission, and (c) level of activity and medical care requirements at 4 years assessed by the Karnofsky Score.

Four years after ICU admission, mortality at hospital discharge [5.5% in the high-risk cardiac surgery subgroup (N=970) with a Euroscore predicted hospital mortality of 9.9% and 15.0% for the other patients admitted for complications after other types of high risk surgery (N=578)] had increased to 14.8% and 43.9% respectively. Intensive insulin therapy, previously reported to reduce ICU and hospital mortality (1), was also associated with lower 4-years mortality, which was significant in the cardiac surgery subgroup (2) 

In the weaning period from anaesthesia after major surgery, borderline hypovolemia is frequent. Cardiac output monitoring is not always justified and volume replacement is guided by Central Venous Pressure (CVP) even if its effectiveness is questionable. Analysis of changes of arterial blood pressure has been suggested to be more reliable. Aim of this study was to compare the prediction of volume responsiveness using these two approaches, considering ScvO2 and the response to a volume challenge as independent variables.METHODS. CVP, Pulse Pressure Variation (DPP), Systolic Pressure Variation (SPV) and its component (DDown) were measured in 23 mechanically ventilated patients in VC mode at Recovery-Room admission (T0), then after 500 mL hydroxyethyl starch 6% (T1). Patients were divided up at T0 in fluid responders (R) and non responders (NR), depending on the responsiveness cut-off of each parameter. We established, in each group, the load effect (T0 vs T1) for the ordinating parameter and for ScvO2. We worked out the percentages of concordant classifications (R + NR/total n o patients x 100) for each pair of indicators.

Concordance in responsiveness classification (T0, T1) between CVP and each of the dynamic parameter was poor (range 43.5% -60.9%), while for each couple of dynamic parameters was high (range 73.9 -95.7%).

Dynamic indexes fit better with ScvO2 than CVP. They should be used preferentially to CVP to predict fluid responsiveness after major surgery when invasive hemodynamic monitoring is not available. 

Taking into consideration that elevated blood lactate level is a basic biochemical expression of hypoperfusion, we valuated this increase as a prognostic factor in hemodynamically stable surgical patients who were admitted to our ICU.

We studied 24 high risk surgical patients (aged 3±17 yrs), who were admitted to the ICU postoperatively and who remained hemodynamically stable during their first day of stay. Heart surgery patients or patients who had been treated with inotropic agents during surgery were excluded. During the first day of ICU stay, the patients were on mechanical ventilation and blood lactate levels were evaluated on repeated measurements. Mean arterial pressure, minimum and maximum heart rate, PaO2/ FiO2 ratio and urine output were monitored.

Survivors (Group A, n=20) had statistically significant lower minimum blood lactate levels (1.8±1.2 vs. 3.5±2.8mmol/lt, p<0.05) and statistically significant lower maximum blood lactate levels (2.7±1.6 vs. 5.6±4.5mmol/lt, p<0.05) compared with non-survivors (Group B, n=4). Apart from higher lactate levels, non-survivors had significantly higher APACHE II score on admission (20±1 vs. 14±5, p<0.01) and longer duration of mechanical ventilation (3±4 vs. 10±9days, p<0.05) compared with survivors. There was no statistically significant difference in mean arterial pressure, heart rate, urine output or PaO2/ FiO2 ratio between the two groups.

Increased blood lactate levels are associated with increased mortality in hemodynamically stable surgical patients. In this group of patients, outcome is also affected by severity on admission (as measured by APACHE II score) and length of stay in the ICU. 

Management of patients with concomitant coronary artery disease (CAD) and asymptomatic carotid artery stenosis (ACAS) remains controversial. This study investigates the relationship between the preoperative evaluation of cerebrovascular reactivity (CVR) to assased with with transcranial doppler (TCD) and cerebrovascular events in patients with ACAS undergoing coronary artery bypass grafting (CABG).

Prospective pilot cohort study. Patients with unilateral or bilateral ACAS > 70% were preoperatively evaluated with transcranial doppler to detect CVR to hypercapnia in middle cerebral artery (MCA). Prophylactic carotic endarterectomy was not performed prior to CABG in patients with preserved CVR (Breath Holding Index>0.69) in the territory of middle cerebral artery. Cerebral perfusion was monitored with TCD during CABG. Patients were followed up for the 30 day incidence of type 1 neurological injury. Prospective pilot cohort study. Patients with unilateral or bilateral ACAS more than 70% were preoperatively evaluated with transcranial doppler to detect perfusion characteristic in main cerebral arteries and CVR to hypercapnia in MCA. Prophylactic carotic endarterectomy was not performed prior to CABG in patients with normal symmetric flow velocity in MCA and preserved cerebrovascular reactivity (Breath Holding Index > 0.69) in the territory of MCA. Cerebral perfusion was monitored with TCD during CABG. Patients were followed up for the 30 day incidence of type 1 neurological injury. 

Thaker K 1 , Ismail K 1 1 Anaesthesia, Dewsbury and District Hospital, Dewsbury, United Kingdom

Hypoalbuminaemia increases mortality, length of hospital stay and resource utilisation 1 . We found hypoalbuminaemia detrimentally affected patient outcome.

A retrospective study of 34 patients who underwent major surgery was performed. Data collected included peri-operative serum albumin levels, complications, administration of human albumin solution (HAS) and post-operative hospital stay.

All patients were hypoalbuminaemic post-operatively. Serum albumin fell 43.6% within 24 hours of operation. Mean length of hospital stay increased as post-operative albumin levels fell. Complications rates were 20% and 100% for patients with albumin levels above or below 20 g/l respectively. 23.5% of patients received HAS.

Hypoalbuminaemia increased length of stay and complication rates. Administration of HAS may help counter this. A prompt start to enteral/parenteral nutrition is desirable. 

We conducted this study to assess the impact of prolonged lactate clearance (LacCl) in surgical patients without any confounding factors such as renal or liver failure, severe sepsis or vasoactive drugs.

A prospective observational study in a medical-surgical ICU, was conducted in the Hospital Español de Mendoza, between November 2004 and December 2005. All patients in the immediate postoperative period admitted to the unit who did not show any of the following exclusion criteria were eligible:Acute or chronic renal or liver failure, severe sepsis or septic shock, inotropics or vasoactive drugs on admission and Lac+ < 2 mmol/l. Two blood samples for arterial blood gases, acid-base status and Lac+, electrolytes and hematocrit were obtained from each patient, one on admission (S0) and the other after 12 hours of staying in the ICU (S12). Fluid resuscitation and inotropic or vasoactive agents, when were used, followed our ICU protocol.

During the period of the study 72 elegible patients were admitted to the ICU in the immediate postoperative period. 

Patients in the immediate postoperative period who showed high levels of Lac+ upon admission and did not accomplish a fast clearance within the first 12 hours, had a higher mortality rate. (1), remains questionable before and after volume replacement in CABG patients with open and closed thorax. The aim of this study was to re-evaluate SVV in predicting fluid responsiveness by pulse contour analysis.

After ethical approval and written informed consent, CABG-patients with an EF>35% were included. Patients with intra-cardiac shunts, valvular heart disease or emergency operations were excluded. When asleep, a 5-F thermistor-tipped catheter (PV2015L20A, Pulsiocath ® ; Pulsion Medical systems, Munich, Germany), inserted into the femoral artery, was connected to the FloTrac ® Sensor and the PiCCO pressure transducer. Transpulmonary cardiac output (TPCO) was performed by injecting 15 ml cold isotonic saline. Both PiCCO (Pulsion Medical systems, Munich, Germany) and Vigileo (Edwards Lifesciences, Irvine, CA) enable calculation of SVV described as SVV-PiCCO and SVV-Vigileo. Parameters recorded were: TPCO, mean arterial pressure (MAP), heart rate, central venous pressure (CVP), SVV-PiCCO and SVV-Vigileo before and after volume load with 10 ml kg-1 hetastarch 6%. Patients were ventilated in the supine position with a tidal volume of 8 ml kg-1 during open (t1, t2) and closed chest (t3, t4) conditions. Pearsons' correlation coefficient (SPSS 12.0.1 for Windows XP, SPSS, Chicago, IL) was used to describe the relation between baseline SVV and CI before and after the volume load. P<0.05 was considered significant.

In 17 males and 3 females no vasoactive medication was used. TPCO and MAP increased in response to volume load (p<0.05). No positive correlation between CVP and changes in CI was found. The correlation between SVV-PiCCO and SVV-Vigileo was -0.21 and 0.02 during open thorax conditions, while these correlations were 0.75 and 0.51 during closed thorax conditions.

The present study suggests that SVV is superior to CVP in predicting fluid responsiveness during closed chest conditions, but failed to predict fluid responsiveness during open chest conditions, as was published before (2). SVV-PiCCO was more reliable than SVV-Vigileo. in mechanically ventilated ICU patients. AnaConDa is a filter able to vaporize and trap up to 90% of sevoflurane administered through an infusion pump. The small amount of Sevoflurane escaping in expiratory limb is drained by the standard anaestetic gas scavenging system.

12 consecutive patients requiring sedation and ventilated for no more than 48 hrs. were enrolled in the study (age 68±15) Exclusion criteria were: CNS abnormalities, hemodynamic instability, pregnancy, liver and renal failure. Sedation level was evalueted with Ramsay and RASS score (awake = Ramsay 0, sedated = Ramsay<4 and RASS<3) During Sevoflurane administration (SevAn) concentration (Fins%) was monitored with standard anaesthetic agent analyzer (Drager Scio) Study protocol included 2 hours treatment with standard Propofol+remifentanil ProRe (1) followed by 2 hours of SevAn and a final step of 2 hours with ProRe (2). Between each step drug administration was interrupted in order to evaluate awakening time. Patients were monitored with Bio Spectral Index.

In all patients SevAn sedation was feasible. The equipotent dosage (defined according to the sedation scores) were: ProRe (1) (1) . The aim of this study was to investigate the effects of morphin, sufentanyl, fentanyl and remifentanyl in critically ill non -surgical and pain-free patients needing transient sedation for greater than 1 day.

Ethics Committee as well as familial consent were obtained in all study patients. We defined four groups (G) of 25 randomized patients who were sedated using a blinded narcotic infusion of : 1) G1: morphine (0.001-0.8 mg/kg/h); 2) G2: sufentanyl (0.1-0.5 µ/kg/min); 3) G3: fentanyl (1-5 µ/kg/min); 4) G4: remifentanyl (0.04-0.25 µ/kg/min)) and midazolam (0.1-0.5µ/kg/min) to achieve optimal sedation as defined by a Sedation Agitation Scale (SAS) between 3 to 4 and a Dolor Comportemental Scale (BPS) of 3. After extubation, when the Visual Analogic Scale (VAS) stabilised under 2 (2), pain testing was done using an intravenous infusion of titrated naloxone (1mg) on day 1, 3 and 8. Evaluation of induced pain was estimated by VAS and treated by Paracetamol (1g/6h) and Tramadol (100mg/6h) to keep VAS<4. For statistical analysis a Shapiro-Wilk test, Wilcox and a Student T-test were used.

The different groups were comparable in terms of age (48+/-19 years), gender (35% female), Simplified Acute Physiology Score (SAPS 28+/-11) and mean ventilation time (36+/-18 hours). Mean extubation time was statistically greater in G1 (35 min+/-9) than in G2 (18 min+/-4) compared to G3 (13min+/-5) and G4 (8min+/-5) (p<0.001). Global mean hospital stay (6+/-2days) and mortality (2+/-1%) were comparable. VAS at day 1 were over 6 in each group (G1:6+/-4;G2:6+/-2;G3:6+/-3;G4:6+/-3) (p<0.005). At day 3, VAS stayed over 5 (+/-2) in G1 and showed a similar curve in G2, G3 and G4 (4+/-2). At day 8, VAS diminished under 4 (+/-1) in G1 but was still below 2 (+/-1) in G2, G3 and G4 (p<0.001).

Morphine, sufentanyl, fentanyl and remifentanyl induced hyperalgesia in non-surgical pain-free patients in an Intensive Care Unit setting when these drugs were infused over at least one day. Morphine is the most potent agent and induces the longest hyperalgesic periods. Remifentanyl and fentanyl facilitate a more rapid extubation. 

Critically ill patients may be cared for temporarily by physicians not regularly working in critical care. The effects of drugs used for induction of anaesthesia has been debated (1).

Postal survey of 250 UK anaesthetists examining drug preferences for rapid sequence induction in 4 scenarios: Emergency laparotomy, a critically ill patient, severe traumatic brain injury, and a patient with severe cardiac disease presenting for anaesthesia.

. 134 Surveys were returned. Induction agent preferences are shown in table 1. Adjuvant drug preferences are shown in table 2. Suxamethonium was preferred by 92.5% for laparotomy, 77.5% for the critically ill, 89.5% for head injury, and 78.5% in the cardiac disease scenario. The main alternative agent preferred was rocuronium. 

Our aim was to know the advantages of using remifentanil in the short sedation (less than seventy two hours) of patients undergoing mechanical ventilation in the Intensive Care Unit, according to its peculiar pharmacokinetic and pharmacodynamic characteristics.

A retrospective and descriptive study was carried out between January 2003 and April 2004. Patients undergoing mechanical ventilation, selected according to sedation protocol, were included: hemodynamic instability (Systolic Arterial Pressure < 100mmHg), pain, need for neurological evaluation, and cofactors (hyperlipidaemia, plasma creatinine level > 1.5 mg/ dl). The following parameters were analysed: type of patients, age, gender, APACHE II score, mean ICU stay, mean beginning dose required to reach level 4 of the Ramsay Scale, maintenance dose required to keep Ramsay level 4 or 5, time spent under sedation, need for coadjuvant analgesia or sedation, and possible secondary effects related to sedation.

In total, 50 patients were included (35 males, 15 females), 18 were medical, 26 surgical (from whom 17 were emergency surgery patients) and 6 were traumatological. Mean age was 59 ± 16 years old, APACHE II Score: 17.2 ± 5.8, mean ICU stay: 5.8 ± 3.4 days. The beginning dose was 7.86 ± 1.8 µcg/kg/h and the maintenance dose was 9.9 ± 2.16 µcg/kg/h, time spent under sedation: 45.8 ± 29.2 hours. The mean pre-extubation dose (patient at Ramsay level 2 or 3) was 5.4± 1.92 µcg/kg/h. An 86% of patients did not need additional sedation (midazolam or propofol) and 94% did not need supplementary analgesia. 68% of patients could be extubated within the first twenty minutes, after reaching a Ramsay level 2 or 3. The more common secondary effects seen were nausea and vomiting and the most relevant was bradycardia (5%).

Remifentanil let us reduce sedation and coadjuvant analgesia (up to 90% according to our series), with predictibility of awakening (Ramsay <3:100%) and the possibility of an early extubation (in less than 20 minutes in 68%) with minimum secondary effects. 

Little is known about the population of general surgical patients at high risk of complications and death. Recent research has suggested this population may be much larger than previously thought.

Data was extracted on in-patient general surgical procedures and critical care admissions in two NHS hospitals between April 2002 and March 2005. High-risk surgical procedures were defined prospectively as those for which the mortality rate was 5% or greater. 

A large high-risk surgical population accounts for less than 10% of surgical procedures but 80% of deaths. Fewer than half these patients are admitted to critical care. These figures are consistent with similar estimates for the UK as a whole. 

Hatum R M 1 , Roderjan C N 1 , Silva E G O 1 , Vanzan A B 1 , Werneck P 1 , Do Carmo Neto E 1 , Ferreira D L S 1 , Knibel M F 1 1 ICU, Hospital S, Rio de Janeiro, Brazil

Morbid obesity is one of the most important public health problems in our days and the bariatric surgery became an option to patients with 35 or higher body mass index (BMI) and obesity related complications (hypertension, diabetes, etc). The study objective was to follow-up gastroplasty postoperatorium patients in ICU and find relationship between medical and surgical complications and operation type (open vs laparoscopic procedures).

We analised the early complications and it's correlation with surgical procedure type in 278 gastroplasty postoperatorium patients admited to the ICU in 10 months follow-up. The statistical analysis used Ficher test. We accepted p<0.05 as significance level. (1). We analyzed the impact of these criteria on the incidence of renal replacement therapy (RRT) and on 30 day mortality in postoperative cardiac surgical patients compared with that of acute changes in serum creatinine (DCrea). (2) METHODS. Based on data from 7241 consecutive patients in two University Hospitals we attributed patients to the RIFLE and to DCrea groups, respectively. Incidences of RRT and 30 day mortality were calculated.

Patient distribution according to the RIFLE criteria was different from that according to DCrea: whereas 93% of the patients were in the RIFLE-0 ( Fig.) . Thus, more than 60% of RRTs were necessary in patients of the RIFLE-0 as compared with only 24% RRTs in patients with slightly negative DCrea (Table) .