key: cord-0033161-0y1j99yt
authors: nan
title: 21st ESICM Annual Congress
date: 2008-08-08
journal: Intensive Care Med
DOI: 10.1007/s00134-008-1240-3
sha: 71e7ab6f7761152b1d04828587eddde2063b4195
doc_id: 33161
cord_uid: 0y1j99yt

nan

To present the evolution of multiresistance markers (MMR) in ICUacquired infections according to the site of infection.

Prospective, multicenter, cohort study. Patients admitted to the participating ICUs during the year 2007 were included. Patients were followed until ICU discharge or up to a maximum of 30 days. The following infections were assessed: mechanical ventilationrelated pneumonia (VAP), urinary tract infection related to the urinary catheter (C-UTI), and catheter-related bacteremias/primary bacteremias (PB(VCB). MMR identified were defined according to the CDC (1). Data were collected using a program developed with Acces 97 database. Resistance rates are expressed as % of isolations resistant to the antibiotics selected in respect to each pathogen evaluated. Multiresistance rates for each infection analyzed are presented. Rates of previous years are compared.

A total of 12,453 patients admitted to the participating ICUs were included, 1367 of which (11.0%) developed 1577 infections (12.7%) during their ICU stay. A total of 1625 microorganisms were recovered, 926 (57%) of which were Gram-negative pathogens, 506 (31.4%) Gram-positive, 179 (11%) fungi, and 14 (0.9%) of other bacteriologic families. MMR according to the infection focus are shown in table. baumanni was an independent risk factor for mechanical ventilation (32/43 vs. 23/61,P = 0.04) and mortality (28/43 vs. 19 /61,P = 0.03) as compared to the susceptible A.baumanni BSI. The mean SAPS II(simplified acute physiology score) predicted death rate was significantly higher in patients with MDR A. baumanni BSI as compared with those having susceptible A. baumanni (63% vs. 41%,P = 0.04).

CONCLUSION. MDR Acinetobacter is more likely to be found in patients with solid malignancy, diabetics and grade IV pancreatitis. It was an independent risk factor for mechanical ventilation, mortality and SAPS predicted death rate in patients.

Surveillance of device related infection is one of the mean objectives in Spain. Infection rates (IR) are influenced by both the intrinsic and extrinsic risk factors (RF); the latest vary according to the hospital. Objective: To compare IR in ICU according to the hospital size

Prospective study performed through out 3 months per year during 2 consecutive years (2006 and 2007) , using the methodology and database of the ENVIN-HELICS. Characteristics of the patients, use of devices and IR per 100 patients and per 1,000 days of ICU' stay/ device were analysed. Hospitals were classified as small (SH) \200 beds, medium (MH) 200-500 beds and big (BH) [ 500 beds. udy.

. 112 ICU were evaluated. Their location was: 14 (12,5%) belonged to SH, 45 (40, 1%) to MH, and 53 (47, 3%) CONCLUSION. Major participation of MH and BH. The severity, length of stay and exposition to risk factors are directly proportional to the hospital size, as well as the multiresistant bacterial infection and IR due to more prolonged stay and devices. Stratification is mandatory in the evaluation of IR according to the ICU size.

GRANT ACKNOWLEDGEMENT. Aventis. INTRODUCTION. Fluid resuscitation in the perioperative care of hypotensive trauma patients can exacerbate the injury sustained during hemorrhagic shock. Polymorphonuclear leukocytes (PMN) release human neutrophil peptides (HNP) upon activation and are known to play a pivotal role in reperfusion associated organ injury. We hypothesize that HNP play an important role in mediating proinflammatory and prothrombotic responses during ischemia and reperfusion following hemorrhagic shock.

METHODS. Male Sprague-Dawley rats were subjected to hemorrhagic shock initiated by blood withdrawal to reduce MAP to 40 mm Hg within 15 min. After a 60 min hypotensive period, rats underwent isovolumetric resuscitation using either whole blood (B) or an equal mix of whole blood and lactated Ringer's solution (B/R) to restore values of MAP to 80 -90 mm Hg for 1 hr. Blood was drawn for cytokine analysis. PMN-induced inflammation was assessed in a co-culture of human primary pulmonary artery endothelial cells (HPAEC) with primary human PMN stimulated with HNP. Platelet aggregation and expression of CD62P was examined following stimulation with HNP.

RESULTS. MIP-2 and TNF-lpha levels at baseline were similar between B and B/R rats but following HSR were significantly increased in the B/R group (Table 1) . Direct stimulation of HPAEC with HNP increased PMN adhesion, which was associated with IL-6 and IL-8 production from HPAEC and an increased surface expression of CD11b in PMN ( Table 2) . HNP increased P-selectin expression, induced aggregation, and prolonged the transient aggregation mediated by ADP in both human and murine platelets. 1.00 ± 0.00 1.17 ± 0.14 2.03 ± 0.03* IL-6 (HPAEC) 0.49 ± 0.14 0.68 ± 0.09 11.80 ± 0.69* 13.98 ± 3.45 12.65 ± 1.75 114.2 ± 7.29* Fold-increase in Adhesion and CD11b. Cytokine concentrations are pg/mL. *P \ 0.05 CONCLUSION. Resuscitation using the hemodilution method caused significant inflammation associated with PMN infiltration. HNP released from neutrophils play a crucial role in mediating inflammation through activation of the endothelium, leukocytes and platelets. INTRODUCTION. Aggressive fluid resuscitation may improve outcome in septic patients [1] . For selected patient groups, however, this approach can be detrimental by increasing the risk of lung edema and abdominal compartment syndrome [2, 3] . The aim of this study was to test the effects of a high vs. moderate volume fluid administration strategy on hemodynamics, organ function and survival in two relevant experimental sepsis models.

METHODS. 48 anesthetized pigs were invasively monitored (systemic and regional flows and pressures). They were randomized (full factorial design) to 3 groups (controls, endotoxin infusion and fecal peritonitis) and 2 treatments (high and moderate volume resuscitation), respectively, for 24 hrs or until death occurred. High volume administration consisted of 20 ml/ kg/h, including 5 ml/kg/h HES 6%; moderate volume administration consisted of 10 ml/kg/h). Ventilation was adjusted to keep the animals normoxic and normocapnic, and plateau pressures were kept at \35 mmHg. Differences between groups were assessed by MANOVA using two between-subject factors (model and volume approach) and time as within-subject factor. Preliminary data on control and peritonitis pigs have been presented previously.

RESULTS. The pulmonary artery systolic pressure (PASP) was reduced to 48 ± 27.6 from 94.8 ± 26.6 mmHg in group B, versus 92.8 ± 21.5 from 95.2 ± 22.7 mmHg in group A (p \ 0.05). The mean pulmonary artery pressure decreased to 42.7 ± 12.8 from 55.8 ± 13.7 mmHg in group B, versus 52.6 ± 13.1 from 54.4 ± 12.6 mmHg in group A (p = 0.05). The pulmonary vascular resistance (PVR) decreased to 8.2 ± 3.5 from 10.7 ± 3.8 IU Wood in group B, versus 9.9 ± 3.5 from 10.6 ± 3.2 IU Wood in group A (p \ 0.05). Concerning the increase in the cardiac index, there were no significant differences between the groups. Mortality was 41.2% (10/24) in group A and 15% (3/20) in group B (p = 0.001).

CONCLUSION. Bosentan demonstrates a beneficial hemodynamic and clinical effect and represents a most promising factor with an extension in its use to treat secondary pulmonary hypertension. Congestive heart failure has been reported as a common source of unsuccessful extubation, especially in patients with cardiac disease or chronic obstructive pulmonary disease (COPD). Spontaneous breathing trial (SBT) is recommended prior to a planned extubation. We sought to evaluate the ability of transthoracic echocardiography (TTE) to identify a cardiac source of unsuccessful SBT in high-risk patients.

METHODS. During a 2-month period, all patients with a congestive heart failure (ejection fraction [EF] B 40%) and/or a COPD who were mechanically ventilated [ 24 h were prospectively studied. TTE was performed prior to (T0) and at the end of (T1) the SBT (30 min, unless unsuccessful). The following measurements were performed in triplicate at endexpiration and averaged: left ventricular end-diastolic (LVEDV) and end-systolic (LVESV) volumes with EF using the modified Simpson's rule, maximal Doppler velocity of mitral E and A waves, E/A and E/E' ratios (tissue Doppler at the lateral mitral ring), stroke volume (SV) using the Doppler method at the level of the aortic ring, and the surface of the color Doppler jet normalized by the left atrial surface in the presence of a mitral regurgitation (MR). In each patients, parameters were compared between T0 and T1 using the matchedpairs Wilcoxon test. Results are expressed in median and 95% confidence intervals. Our local Ethics Committee agreed with the protocol.

RESULTS. 16 patients (heart failure, 14; COPD, 3) were studied ; duration of mechanical ventilation: 75 h ). SBT was unsuccessful in 4 patients and another patient who passed the SBT was reintubated within 24 h for progressive accumulation of tracheal secretion due to inefficient cough. The remaining 11 patients underwent both a successful SBT and extubation. Overall, LVEF remained stable between T0 and T1 whereas LV filling pressures significantly increased, as reflected by the higher E, E/A and E/E' values (table; *p \ 0.05). In patients with unsuccessful SBT, E, E/A and E/E' values increased of 19.6 cm/s, 0.67 and 1.97 when compared to 7.9 cm/s, 0.08 and 0.32 in patients who passed the SBT, respectively. When MR was present at T0 (n = 8), its normalized surface increased of 2.1 cm 2 in patient with unsuccessful SBT, and was comparable (+0.14 cm 2 ) in the remaining patients. 6.7/5.8-11.7 T1 126/110-174* 89/67-127 34/27-41 49/42-59 92/70-114* 1.6/0.9-2.2* 9.5/7.0-13.5 CONCLUSION. In this study, SBT significantly increased LV filling pressures and the surface of MR, when present. These variations were more pronounced in the subset of patients with unsuccessful SBT. Even in patients with severe congestive heart failure, the absence of documented variation of LV filling pressure and MR during the SBT appeared as a good predictor of successful extubation.

INTRODUCTION. Induced mild hypothermia therapy (MHyT) improves neurological outcome in post resuscitation cardiac arrest (CA) patients with ventricular fibrillation/ tachykardia (VF/VT). Patients with asystole as initial rhythm were excluded from earlier studies due to poor neurological outcome. The present study enrolled both patients with asystole or VF/VT; one of the objectives was to assess neurological function at ICU discharge and six months after CPR.

METHODS. Prospective multi-center single arm registry in 49 patients with witnessed cardiac arrest and succesful CPR who were selected for MHyT. Patients had to be [ 18 years old and unconscious at ICU/ER admission (GCS \ 8), with a time interval between cardiac arrest and initiation of MHyT treatment of \ 6 hrs. Treatment was performed for 24 hrs with a target temperature of 32,5°C. Neurological status was documented upon ICU/ER arrival, during MHyT treatment, at ICU discharge and 6 months after the event as measured by the Glasgow Outcome Scale (GOS). Temperature measurements were continuously taken throughout the cooling period via bladder catheter. For MHyT treatment the Deltatherm device (KCI, USA) was used.

Of the 49 patients included in the registry 31 (63%) had VF/VT and 18 (36,7%) had asystole/PEA as first rhythm. Good neurological recovery at ICU discharge (GOS 5/4) was seen in 22 (45%) of the patients, 9 (18,4%) were neurologically impaired (GOS 3/2) and 16 (33%) died. In the follow up investigation after 6 months 19 patients (38,7%) showed favorable neurological outcome (GOS 5/4), 2 (4%) were severly neurological impaired (GOS 3/2), 26 (53%) died. 2 patients with GOS 5 at ICU discharge were lost to the follow up, one (GOS 5) died due to another event. Good neurological recovery was seen in 45,16 %(n = 14) of patients with VF/VT and in 27,7 % (n = 5) patients with asystole 6 months after CPR CONCLUSION. MHyT improves neurological outcome in patients with witnessed CA regardless of the initial rhythm. Favorable results in VF/VT patients were similar to preceding studies. Hypothermia also appears to provide neurological protection in patients presenting with asystole at ICU discharge as well as 6 months after the event. Hemorrhage remains a primary cause of death in civilian and military trauma. Volume replacement and surgical bleeding control are the most important issues in the treatment of hemorrhagic shock. Bedside indicators of ventricular preload have been proposed as predictors of fluid responsiveness. This goal could be achieved reducing arterial pulse pressure (DPP) variation induced by mechanical ventilation. The systolic pressure variation (SPV) has been shown to be a predictor of fluid responsiveness. This prospective study was undertaken to determine the DPP, SPV and cardiac output (CO) in a swine model of controlled hemorrhagic shock and volume replacement with lactated Ringers' solution and hypertonic saline + 6% with hydroxyethyl starch.

METHODS. Fifteen pigs were anesthetized, instrumented and randomized into three groups: Sham (S), lactated Ringer solution (LR) and hypertonic saline in 6% with hydroxyethyl starch 130/0,4 (HHES). Hemorrhage was induced by blood withdrawal by a catheter placed through the femoral artery until a mean arterial pressure (MAP) of 40 mmHg was achieved. This pressure was maintained for 30 minutes. At this point, the animals received 32 ml/Kg/ for 20 minutes (LR) or a bolus dose of HHES,4 ml/Kg for 5 minutes. Thirty minutes after the resuscitation, total shed blood was reinfused to the animals. Sham animals were not bled. DPP, SPV and CO were monitored at the following time points: before hemorrhage (Baseline), when a pressure of 40 mmHg was achieved (Tshock), 30 minutes of MAP of 40 mmHg (T0), 30 min, 60 min,120 min and 240 min after treatment (T30, T60, T120,T240). Statistical analyses were performed using one-way analysis of variance followed by Tukey-Kramer test. A p value of 0.05 was considered statistically significant.

RESULTS. During shock, DPP was significant higher in groups LR (51.00 ± 17.59; p \ 0.05) and HHES (53.00 ± 17.32; p \ 0.05) when compared to Sham (16.20 ± 10.96), but no significant differences between them. At time T0, only HHES (49.67 ± 20.21; p \ 0.05) was significantly different from Sham (22.58 ± 11.84) . At time T30, HHES (29.67 ± 17.16; p \ 0.05) differed when compared to Sham (19.18 ± 9.01), but after T60 no significant differences were detected in the three groups. When SPV and CO were analyzed, there were no significant differences at time points.

CONCLUSION. Small volume hypertonic saline + 6% with hydroxyethyl starch promotes similar volume expansion as large volume LR, reducing the large dPP induced by hemorrhage. INTRODUCTION. Nowadays, left ventricular assist devices (LVAD) are increasingly used as a bridge to transplant in heart failure patients. These patients are generally anticoagulated with phenprocoumon and aspirin. During heart transplant severe bleeding frequently poses a serious problem. Despite lack of evidence, desmopressin is often administered successfully when other means fail. We hypothesised that apart from aspirin-induced cyclooxygenase inhibition an additional platelet dysfunction is present in these patients. In order to evaluate this aspect we employed multiple monitoring devices for platelet function.

METHODS. Blood from 12 out-patients after elective LVAD implantation and 12 healthy matched volunteers was screened with the help of thrombelastography (TEG Ò ), platelet function analyzer (PFA-100 Ò ) and a new whole blood aggregometer (Multiplate Ò ) [1] . Additionally, we measured von Willebrand Factor (vWF) antigen levels and vWF:RCo activity (vWF ristocetin co-factor). The anticoagulation regimen for LVAD patients consisted of 100 mg of aspirin PO daily combined with phenprocoumon (INR: 2.5-3.5) [2] .

Oral anticoagulation produced an average INR of 3.5. Aspirin reduced MA-AA (Maximum Amplitude Arachidonic Acid; TEG Ò ) by [ 60% in patients (p = 0.001 vs. controls). The effectiveness of aspirin to suppress arachidonic acid-induced aggregation was also reflected by an almost 75% decrease in the ASPItest (Aspirin Test; Multiplate Ò ) (p \ 0.001 vs. controls) while ADP-induced aggregation was not altered in patients (p = 0.805). Platelet function under high shear was invariably and severely compromised: CADP-CT (collagen adenosine diphosphate closure times; PFA-100 Ò ) were 2.5-fold longer in patients than in controls (p \ 0.001), and 50% of patients had maximal CADP-CT values of 300 s. While vWF antigen levels were 80% higher in patients (p \ 0.001), vWF:RiCO was subnormal (\60%) in 5/12 patients. Ristocetin-induced aggregation was also 3-fold higher in controls than in patients (p \ 0.001).

CONCLUSION. This indicates that LVAD out-patients have an additional functional platelet defect, which affects the glycoprotein Ib/vWF axis. Therefore, desmopressin use in the management of bleeding in these patients seems justified. Near Infra-Red Spectroscopy is a novel method for rapid and noninvasive assessment of tissue oxygen saturation (StO2). An association between StO2 (%) and oxygen delivery has been demonstrated during shock, trauma and resuscitation. We sought to explore StO2 during cardiac surgery and its association with post-operative outcome.

METHODS. 74 adult patients undergoing first-time, elective, single procedure cardiac surgery requiring cardiopulmonary bypass (CPB) were studied. StO2 was measured from the thenar eminence (Inspectra Tissue Spectrometer Model 325, Hutchinson Technology Inc, USA) through anaesthesia and surgery, and arrival on the Intensive Care Unit (ICU). Outcome was defined as length of ventilation and ICU stay, post-operative length of stay and morbidity outcome (POMS) on post-operative days 1, 3, 5, 8 and 15 (the presence or absence of a infectious, pulmonary, cardiovascular, wound, haematological, pain, renal or gastrointestinal complication). All clinical information was prospectively collected.

RESULTS. StO2 rose from baseline during induction of anaesthesia (81.7 to 88.5, p \ 0.001) then fell during surgery (mean 78.9) with a significant change in minimum StO2 during CPB (75.9 to 68.2 p \ 0.0001). The mean StO2 during the 1st 5 minutes of anaesthesia was lower in those patients with than without POMS on D15 (81.1 ± 7.0 v 87.6 ± 7.7, p = 0.04) with the mean StO2 during those first 5 minutes being predictive of the presence of D15 POMS (ROC 0.762, p = 0.01). Lower mean StO2 during the first 20 minutes of ICU monitoring was observed in those with than without POMS on D3 (72.9 ± 12.9 v 85.5 ± 8.4, p = 0.009) and was predictive of the presence of D3 POMS (ROC 0.793, p \ 0.0001).

CONCLUSION. StO2 during anaesthesia and ICU is predictive of post-operative complications on D15 and D3 respectively, with lower StO2 being associated with poorer outcome. Such data suggest that reduced tissue oxygen delivery, as measured by StO2, may be associated with poorer outcome -an association which may in theory be causal. Were this the case, interventions targeted at maintaining StO2 may beneficially impact on such outcome. This hypothesis seems worthy of further exploration.

GRANT ACKNOWLEDGEMENT. This work was supported by an unrestricted educational grant from Hutchison Technology (Inc). We are indebited to the patients and staff at the Heart Hospital, London for their support and cooperation.

T. Günther 1 , A. Liebold 2 , G. Noeldge-Schomburg 1 , D. A. Vagts* 1 1 Dep. of Anaesthesiology and Intensive Care Medicine, 2 Dep. of Cardiac Surgery, University of Rostock, Rostock, Germany INTRODUCTION. Increased arterial blood lactate levels have been identified as a good predictor of postoperative morbidity and mortality. However, little is known about factors that might influence this increase. Therefore we examined the changes in blood lactate levels during and after cardiopulmonary bypass (CPB) as well as blood gas levels, pressure and flow parameters, metabolic values, body temperatures and substitution of catecholamines.

We conducted a retrospective clinical study with consecutive patients (n = 403) undergoing cardiopulmonary bypass for cardiac surgery at a university hospital. The difference between pre-and postoperative lactate levels was obtained and its relationship to the perioperative factors was analysed. Patients were assigned to three groups: slight increase in lactate levels (SIL, n = 154, increase \ 0.8 mmol/L), medium increase in lactate levels (MIL, n = 105, 0.8 mmol/L B increase \ 1.4 mmol/L) and high increase in lactate levels (HIL, n = 144, 1.4 mmol/L B increase). Differences between the three groups in the potential risk factors were identified by one-way ANOVA and Bonferroni correction. Correlations were assessed with either Spearmans rank correlation test or Pearsons correlation test, as appropriate.

RESULTS. Analysis of variance indicated that the three groups differed significantly in maximum glucose concentration during CPB (p \ 0.001) and during ICU stay (p \ 0.001), application of insulin during CPB (p = 0.001) and during ICU stay (p \ 0.001), total perfusion duration (p \ 0.001) and aortic clamp duration (p = 0.002). Correlation analysis identified all of these potential risk factors to be positively correlated to the increase of arterial blood lactate levels (p \ 0.001). Correlation of increase in lactate to the administration of catecholamines was not verified.

CONCLUSION. Hyperglycemia and long CPB duration are associated with increased lactate levels. Attempts to reduce glucose concentration and CPB duration may result in lower lactate concentrations after cardiac surgery. INTRODUCTION. ECG abnormalities after cardiac valve surgery may be similar to those seen during spontaneous myocardial infarction, and few reports describe them. In patients who have undergone valve surgery, new ST-T abnormalities are common but not necessarily diagnostic of myocardial ischemia. However, when new Q waves appear in territories other than those identified before surgery, myocardial infarction should be considered, particularly if associated with elevated biomarkers, new wall motion abnormalities or haemodynamic instability (1) (2) . The purpose of this study is to asses the incidence of ST-T abnormalities and perioperative myocardial infarction, clinical manifestations (electrocardiographic, echocardiographic, biomarkers elevation) and prognosis, in patients with cardiac valve surgery without coronary artery disease.

METHODS. Retrospective analysis of patients who underwent cardiac valve replacement in a Tertiary University Hospital over a ten year period. After surgery, the patients were routinely admitted to the Intensive Care Unit (ICU). We excluded patients with angiographically significant coronary lesions before surgery and those with aortic dissection. Statistical analysis was performed with SPSS 12.0. (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) ; Mechanical ventilation (hours) 40 ± 13 (0-360); Intraaortic balloon pump 23.3% (n = 7); Low cardiac output 37%;(n = 11); ST changes Inferior leads:46% (n = 14); Inferior + right:20 % (n = 6); Anterior:16.7 % (n = 5); New onset Q waves 60 %;n = 18; Cardiogenic shock 13.3%;n = 4; Emergency coronary artery bypass 10 %;n = 3; Mortality causes: Cardiogenic shock n = 4; Multiple organ failure n = 2; Mortality: 20%; n = 6.

CONCLUSION. perioperative myocardial infarction following valve surgery is a fairly uncommon complication, that must be suspected in patients with ST abnormalities; in our patients, it is associated with increased mortality. Inferior leads were the most frequently affected, with high proportion of right ventricular dysfunction. Strategies must be directed towards early diagnosis, including angiography in patients hemodynamically unstable. 

I. Sekkat* 1 , P. Lequeux 2 , J. J. Massaut 3 1 Anesthesiology, ULB, Bruxelles, 2 Anesthesiology, CHU Tivoli, La Louvière, 3 Intensive Care, CHU Brugmann, Bruxelles, Belgium

INTRODUCTION. The monitoring of mixed venous saturation(SVO2) necessitates the use of a Pulmonary Artery Catheter. SVO2 measurement is now frequently replaced by the measure of venous saturation in the superior vena cava (ScVO2). The aim of our study is to analyze the correlation between ScVO2 and SVO2 and their predictive value of low output state following cardiac surgery.

METHODS. SVO2, ScVO2 and cardiac output were monitored in 15 patients during 24 h following cardiac surgery. SVO2 and continuous cardiac output were monitored using optical pulmonary Swan Ganz catheter and Vigileo monitoring from Edwards LifeScience. ScVO2 was monitored using optical central venous catheter and monitoring from Pulsion. Data were collected every two seconds. Paired data for ScVo2 and SVO2 were analyzed with T test and linear regression. Logistic regression was used to predict low output state from ScvO2 and SVO2. The statistical analysis was performed using Stata 8 for UNIX.

RESULTS. ScVO2 underestimates SVO2 (SVO2 = 62.88 +-6.20, ScVO2 = 57.35 +-9.20, p \ 0.0001 at paired T test) and poorly correlates with SVO2 at linear regression (R 2 = 0.57, ScVO2 coefficient = 0.446, const = 37.45). Predictive value of ScVO2 for low output state as defined by a CI \ 2.4 l/min was poor (*LROC = 0.6) at logistic regression as compared with SVO2 (LROC = 0.69).

CONCLUSION. ScVO2, on the average, underestimates SVO2 and is only of low value to detect low output state following cardiac surgery. INTRODUCTION. ASV maintains an operator preset alveolar minute ventilation by providing automatic selection of ventilatory settings and continuous breath by breath adaptation (1) . We studied the influence of ASV on need for sedatives (propofol) and tes (morphine) in patients after CABG.

METHODS. Randomized controlled trial comparing ASV with pressure control/pressure support (PC/PS)-mechanical ventilation (MV) in 128 patients after Coronary Artery Bypass Grafting (CABG). Data on prescribed propofol and morphine were retrieved from the patient data management system (in the ICU) and from the patient's charts (in the OR). Data are means ± SD, or medians [IQR] where appropriate. Statistics: T-tests were used to analyze differences between groups; non parametric tests were used when applicable RESULTS. There were no differences between ASV-and PC/PS-MV patients regarding demographics, length of stay in ICU/hospital, and creatinine levels. Sedation use in the OR was similar. Duration of intubation and mechanical ventilation was 16.7 ± 6.1 h versus 17.9 ± 1.0 h with ASV and PC/PS-MV, respectively (P = 0.35). There were no differences in prescribed propofol and morphine (Table 1) . INTRODUCTION. Acute renal failure requiring renal replacement therapy (RRT) is common in critically ill patients. Mortality in this group of patients is high, ranging from 28 to 90%. The purpose of the study was to define the outcome of patients requiring haemofiltration (HF) during critical illness and outcome following discharge from our adult general ICU, over a 3 year period, of patients still requiring RRT at time of hospital discharge.

Between January 2004 and December 2006, 436 patients (total admission were 3890) received RRT in our ICU as noted by our database. We looked into the ICU and hospital mortality and their relation to renal plus respiratory or renal plus cardiovascular or both. All patients requiring long term renal support were available from our computer based registry.

Of the 436 patients, 47% of patients were admitted with a respiratory cause and 35% with cardiovascular cause. The average number of ventilator days was 9.3 and average haemofiltration 5.49 days. Our ICU mortality was 37%. 11.9% subsequently died in hospital. 17% were transferred to other hospitals in the region following critical care discharge. 54 patients were identified as having pre-existing renal impairment and hospital mortality in this group was 46.29%. Of the 382 patients admitted to ICU with normal renal function pre critical illness, 10 still required RRT at time of hospital discharge. At the end of 1 year of follow up, 4 had died, 3 were still on haemodialysis (Abdominal aortic aneurysm repair, perforated duodenal ulcer and diabetic ketoacidosis) and 2 patients were being monitored. By 2 years post follow up no patients required RRT. The monitored patient's diagnoses were TTP and SLE. The mortality of patients admitted with specific organ failures were CONCLUSION. Of the survivors of critical illness requiring RRT, 2.61% required RRT for up to one year following ICU discharge. 2 years after discharge no patients with normal renal function preadmission required RRT. The mortality of patients discharged from hospital still requiring RRT was 40%. These results have implications for what patients, or the patient's relatives, are told when RRT is deemed necessary during critical illness. It also has implications for those commissioning RRT services in survivors of critical illness. Our mortality for patients requiring RRT is low. The lower mortality for 3 organ failure versus 2 probably represents non escalation treatment limitations and needs further investigation. INTRODUCTION. Patients suffering critical illnes are at risk of developing post-traumatic stress disorder (PTSD). We assessed prevalence and risk factors for PTSD symptoms in patients who required mechanical ventilation one year after ICU stay.

METHODS. Two-hundred and eighty-seven patients who required mechanical ventilation [ 48 h were enrolled in a prospective, nonrandomized, multicenter study in Chile (ClinicalTrials.gov NCT00403208). Main exclusion criteria were neurologic impairment and end-stage organ disease. One-year mortality was 48%, and survivors were screened telephonically for memories of traumatic experiences during ICU stay (nightmares, panic, pain, and suffocation) and PTSD symptoms by means of the post-traumatic stress syndrome-10 (PTSS-10) scale. PTSS-10 score [ 35 was defined as PTSD. Statistical analysis included MannWhitney and Fisher's exact tests for univariate analysis, and multiple logistic regression to test risk factors for PTSD.

RESULTS. Fifty percent (75/149) of patients answered the questionnaire (52 could not be reached, 14 refused consent, and 8 were not able to respond). Average PTSS-10 score was 29 ± 14, and 20 (37%) patients had a score [ 35. Gender (M:24 ± 11 vs. F:35 ± 14, p \ 0.001), medical condition (p \ 0.015) and oxygenation index at 24 h (p \ 0.009) were correlated to PTSS-10 score. No relationship was found between hypnotics, analgesic and neuromuscular blockade use and doses, or level of sedation during mechanical ventilation and PTSD symptoms. Patients with traumatic memories had a greater prevalence of PTSS-10 [ 35 (p \ 0.005). Multivariate analysis showed only gender to be predictive of PTSD symptoms.

CONCLUSION. Prevalence of PTSD symptoms in patients surviving critical illnes and mechanical ventilation was more frequent in women, and similar to other studies. None of sedation variables studied was related to PTSD symptoms. INTRODUCTION. Intensive Care Unit (ICU) survivors are at risk for physical or psychological sequelae after critical illness. Medical practioners outside the ICU may not recognize or understand these sequelae. While there are no consensus recommendations of how to organize ICU follow-up, experiences of others suggest that repeated follow-up may be valuable for evaluation after intensive care.

In 2007 a multidisciplinary follow-up team was formed, consisting of three physicians, two nurses and a physiotherapist from the General ICU. Hospital psychiatrists were involved to dsicuss consensus for patient referrals, as well as the Pain Clinic.

Patients treated for 4 days or longer in the General ICU are seen in the ward by a nurse from the follow-up team 2-5 days after ICU discharge. Patients are briefed about their ICU stay, and the ICU follow-up is described shortly. Visits to the ICU Follow-up Clinic are 3, 6 and 12 months after ICU discharge. Prior to each visit patients fill in ICU Memory Tool, Short Form-36, Hospital Anxiety and Depression Scale (HAD) and Impact of Event Scale (IES). At each visit the patient meets a nurse, a physician and a physiotherapist from the ICU. Memories from the ICU are ventilated and the course of illness and treatments during the ICU stay are described. Patients with [25 points in IES or [10 in any of the two subscales of HAD are referred to a collaborating psychiatrist. Patients with mean daily pain score above 3 on the Visual Analog Scale or with untreated neuropathic pain are referred to the Pain Clinic. Questionnaire results and potential further referral are discussed with the patient. Muscle strength, mobility and restrictions in physical functioning are evaluated and the patient performs a 6 minute walking test. If necessary patients are referred to their local out-clinic physiotherapist.

RESULTS. During the first year of follow-up, 95 patients who spent [96 h in the ICU and survived to follow-up were invited and 57 came for follow-up. Eight patients were referred for more extensive psychiatric evaluation, of which four were found to be in need for specific treatment (cognitive behavioural therapy, antidepressive treatment) or repeated assessment. Two patients were referred to the Pain Clinic. Thirty-eight patients were found to be in need of physiotherapy recommendations or referral. Evaluation of the follow-up indicates a high degree of patient satisfaction.

CONCLUSION. Follow-up of ICU survivors by an ICU-based multidisciplinary team is feasible and can help identify unmet physical and psychological needs of ICU survivors during the first year after intensive care. A liaison between ICU clinicians and other specialists for referral of patients with psychological, somatic or pain sequelae is essential. INTRODUCTION. Morbidity after intensive care spans from impaired physical function to mental problems. However, few ICUs have regular follow-up routines. In rural areas it is inconvenient for patients to visit an out-patient clinic. From January 2006 we implemented a two-step follow-up in our ICU, and the experiences from the first two years are presented METHODS. The study was conducted in a 10 bed general ICU. All survivors ([ 16 years) with a LOS [ 48 h were eligible. 3 months after ICU discharge they were contacted by telephone, two questionnaires (EQ 5D and Karnofsky score) were applied, and a semistructured interview performed. This part was performed by one of six trained ICU nurses. If the patient expressed a wish for a follow-up, or the interviewer deemed this necessary, patients were invited to the out-patient clinic at the hospital.

RESULTS. During the first two years, 738 adult patients were treated in the ICU, mean SAPS II score was 41.9, hospital mortality 24%. 214 (75%) of 285 eligible patients were followed up, and 57 patients wanted a second follow-up at the hospital. The figure shows results from EQ-5D (mean sum before 0.78 and after 3.25). The mean Karnofsky score was 66. An unpleasant memory was reported by 57%, most often hallucinations (24%) or anxiety (11%).

In 6% pain was remembered as a problem.

CONCLUSION. Three months after the ICU stay the daily life of a large part of our patients were still reduced, with perceived QOL worse than pre-ICU. However, most patients seem to cope with such problems without the need for a second consultation in our ICU, making this model cost-effective. INTRODUCTION. Acinetobacter baumanii is an important cause of infections in patients in intensive care units. Sulbactam combinations are bactericidal against A.baumanii in in vivo models. Tigecycline, the first clinically available semisynthetic glycylcycline approved for clinical use, has been shown to have potent activity against a wide variety of gram-positive and gram-negative pathogens, including multidrug resistant (MDR) strains. We prospectively studied the efficacy of tigecyclin versus a high dose regimen of ampicillin-sulbactam (A/S) in patients with Ventilator Associated Pneumonia (VAP) caused by A.baumanii.

METHODS. Twenty-three mechanical ventilated patients [ 72 h (mean age: 70 ± 7) that developed VAP and had positive A. baumannii tracheal aspirates were enrolled in the study. The case was considered to be aetiologically confirmed if A. baumannii was isolated and quantitative culture of broncho-alveolar lavage ([ 104 cfu/mL) was achieved. The antimicrobial susceptibility of the isolates was determined using the disk-diffusion (Kirby-Bauer) method, the VITEK II system and the E-test method (AB Biodisk, Solna-Sweden). All isolates exhibited resistance to almost all antibiotics routinely tested, excluding colistin. Susceptibility to tigecyclin was intermediate. Ten of the patients received tigecyclin intravenously (Group A), as monotherapy (50 mg twice daily) while thirteen of them received 9 g A/S (Group B) three times daily (cumulative dose of 9 g sulbactam daily, adjusted for creatinine clearance). Follow up cultures and clinical evaluation of all patients was performed five days after the initiation of therapy. Clinical success was defined by a lessening of the signs and symptoms of VAP, while microbiologic success was defined as eradication of the pathogen in BAL cultures.

RESULTS. Follow up BAL revealed microbiologic success in 8 patients of Group A (80%) and 10 patients of Group B (77%) (p = 0.86). Clinical success was observed in 7 patients of Group A (70%) and 10 patients of Group B (77%) (p = 0.7). There was no significant difference in 14 days and 30 days mortality between the two groups. Adverse reactions occurred in 2 patients (15.3%) of Group B (reversible nephrotoxicity and diarrhea), which did not lead to discontinuation of treatment. Toxicity was not observed with tigecyclin use.

CONCLUSION. In this study tigecyclin and high-dose ampicillin/sulbactam were comparably effective treatment options for MDR A.baumanii VAP, while tigecycline therapy seems safer. The administration of these regimens could prevent the excessive use of polymixins in intensive care units and the subsequent development of pandrug resistant strains. INTRODUCTION. Retrospective studies have reported good clinical and microbiological success rates using intravenous colistin as monotherapy to treat multiresistant A. baumannii ventilator associated pneumonia (VAP), comparable to that obtained with imipenem at the time to treat VAP due to imipenem-susceptible strains. However, colistin as monotherapy led to poor results in mouse pneumonia by multiresistant A baumannii, due to inadequate penetration into the pulmonary parenchyma.

METHODS. Between November 2003 to December 2006, 74 consecutive patients developed microbiologically documented VAP due to A. baumannii susceptible only to colistin. Twenty four from these patients were excluded from the study as they were requiring antibiotics against Gram negative bacteria for infections other than VAP, or they were considered to suffer from multimicrobial VAP. The rest were prospectively, randomized to administer colistin only (3 MU three times daily, adjusted for creatinine clearance) (Group A) or combination therapy with colistin (dose as group A) and ampicillin/sulbactam 24 (16 + 8) gr IV (Group B). Eleven more patients were also excluded in the course of the study due to development of confirmed infections from microorganisms other than A. baumannii. Finally, Group A consisted from 19 patients (age 56.6 ± 14.3) Group B consisted from 20 patients (age 56.9 ± 18.7). The clinical response of VAP (improvement of symptoms, clear decrease of purulent bronchial secretions, constant decrease of vasopressors) was assessed on day 5th of treatment.

RESULTS. Group A did not differ to Group B in APACHE II on admission to ICU (14.5 ± 3.1 vs. 16.5 ± 4.7), SOFA score on diagnosis of VAP (7.6 ± 2 vs. 7.9 ± 3.1), the clinical presentation with septic shock (52.9 vs. 47.1%), or the existence of bacteremia. In univariate logistic regression analysis only the combination treatment (70% vs. 15.8, p = 0.001, OR = 12.44 (2.6-59.3) ) and SOFA B 8 (58.3 vs. 20%, p = 0.024, OR = 0.179 (0.04-0.8)) were associated with clinical response of VAP.

CONCLUSION. Combination therapy with colistin and high dose of ampicillin/sulbactam seem to be more effective in VAP due to multiresistant-only colistin sensitive-A. baumannii VAP.

INTRODUCTION. The emergence of multidrug resistant (MDR) gram-negative bacteria necessitated the revival of older antibiotic drug classes such as polymyxins (colistin) and tetracyclines. Tigecycline, the first clinically available semisynthetic glycylcycline approved for clinical use, has been shown to have potent activity against a wide variety of grampositive and gram-negative pathogens, including MDR strains. We retrospectively evaluated the efficacy of tigecyclin versus colistin in patients with Ventilator Associated Pneumonia (VAP) caused by A.baumanii.

METHODS. Twenty-five mechanical ventilated patients [ 72 h (mean age: 70 ± 7) that developed VAP and had positive A. baumannii tracheal aspirates were enrolled in the study. The case was considered to be aetiologically confirmed if A. baumannii was isolated and quantitative culture of broncho-alveolar lavage ([ 104 cfu/mL) was achieved. The antimicrobial susceptibility of the isolates was determined using the disk-diffusion (Kirby-Bauer) method, the VITEK II system and the E-test method (AB Biodisk, Solna-Sweden).Interpretation of the susceptibility results was in accordance to the Clinical and Laboratory Standards Institute (CLSI). All isolates exhibited resistance to almost all antibiotics routinely tested, excluding colistin. Susceptibility to tigecyclin was intermediate. Ten of the patients received tigecyclin intravenously (Group A), as monotherapy (50 mg twice daily) while fifteen of them received colistin intravenously, as monotherapy or in combination with cefepime (3*106 IU three times daily, adjusted for creatinine clearance) (Group B) . Follow up cultures and clinical evaluation of all patients was performed five days after the initiation of therapy. Clinical success was defined by a lessening of the signs and symptoms of VAP, while microbiologic success was defined as eradication of the pathogen in BAL cultures.

RESULTS. Follow up BAL revealed microbiologic success in 8 patients of Group A (80%) and 10 patients of Group B (66,6%) (p = 0.86). Clinical success was observed in 7 patients of Group A (70%) and 9 patients of Group B (60%) (p = 0.7). There was no significant difference in 14 days and 30 days mortality between the two groups. Adverse reactions occurred in 5 patients (33%) of Group B (reversible nephrotoxicity), which did not lead to discontinuation of treatment. Toxicity was not observed with tigecyclin use.

CONCLUSION. Colistin and tigecycline seem equally effective in eradication of multidrug resistant A. baumanii in BAL cultures of patients with VAP. However, tigecycline administration seems to be safer than colistin in the clinical treatment of these patients. More data are needed, in order to clarify the role of these regiments in the treatment of A.baumanii infections.

INTRODUCTION. Whilst antibiotics have reduced mortality from sepsis, their use is associated with increased healthcare associated infections, emergence of drug-resistant organisms, increased hospital length of stay, morbidity and mortality and increased cost. There is now evidence that strict control over antibiotic prescription, with appropriately shortened durations of regimes reduces cost, does not compromise outcome, and may be beneficial in critically ill patients [1, 2] .

We performed a retrospective analysis of 56 patients admitted consecutively to our ICU with severe pneumonia over the past 12 months and evaluated the suitability of antibiotic chemotherapy based on choice of agent, duration of therapy, modification following assessment of the clinical state of the patient and microbiological data. We excluded patients with chronic immunodeficiency states or those who died within 24 h of admission to the ICU. Data analysis was performed using the InStat v.3 statistical package (GraphPad Software Inc, USA).

Of the 56 patients, 31 (55.4%) received appropriate antibiotic chemotherapy, 8 (14.2%) received an incorrect antibiotic, and 17 (30.4%) received an inappropriately long duration of therapy based on microbiological data, clinical response and current hospital guidelines. Sputum cultures/tracehal aspirates taken on admission were positive in 13 (23.6%) patients despite clinical evidence of pneumonia. The incidence of hospital acquired infection (HAI) was significantly higher in those patients receiving incorrect antibiotic choice or prolonged duration of therapy (12 vs. 6 infections, p = 0.023). When stratified by duration of therapy, those patients receiving B 5 days of antibiotics had significantly reduced ventilator days (5.8 vs. 10.2 days; p = 0.027), duration of ICU stay (6.9 vs. 12.4 days; p = 0.02), and reduced HAI (3 vs. 14 episodes; p = 0.0093) despite being comparable for age, relevant comorbidities, and severity of illness (APACHE II score 19.6 vs. 19.4; p = 0.9).

CONCLUSION. 1. Shortened antibiotic regimes based on clinical assessment, response to empiric therapy and microbiological data are associated with reduced incidence of hospital acquired infections, ventilator days, and duration of intensive care length of stay. 2. Accurate diagnosis of infection may be improved using biochemical markers such as procalcitonin assay or clinical infection scores such as CPIS (Clinical Pulmonary Infection Score) given the low yield from sputum culture/tracheal aspirate. INTRODUCTION. An optimal anticoagulation strategy during CRRT in critically ill still remains unresolved issue. It is well known that cooling of blood decreases its ability to clot. Therefore, we developed and experimentally tested a device designed for regional blood cooling in CRRT extracorporeal circuit as a non-anticoagulant measure to prevent circuit clotting.

We developed the device for selective cooling of extracorporeal circuit (20 centigrade) allowing blood rewarming (38 centigrade) just before returning into the body. 12 anesthetized and ventilated pigs were randomized to receive either 6 hrs of continuous hemofiltration with application of this device (COOL; N = 6) or without it (CONTR; N = 6). In neither of these groups any anticoagulant was used. In 15, 60, 180, 360 min after starting hemofiltration variables related to circuit patency (time to clotting [TC], number of alarmtriggered pump stopping, venous circuit pressure [VP]) were assessed. In addition, thrombinantithrombin complexes [TAT circ], a marker of coagulation activation in circuit were measured.

RESULTS. Data are median and interquartile range. While the patency of all circuits treated with selective cooling was well maintained within the observation period, 5 of 6 sessions were prematurely clotted in untreated group (TC 197 min (180; 240) ). Similarly, the number of alarm-triggered pump stopping was significantly higher in nonintervened group (COOL 1.5 (0; 2) vs. CONTR 6 (5;7)). Vt's of 6-8 ml.kg -1 of ideal weight have been recommended; this weight is calculated using a formula based on patients' height. Our unit advocates use of this guidance in determining ventilator settings for patients with ARDS and ALI (based on oxygenation criteria), but our compliance with this guidance was not known. The Soehnle Professional 3020 system is an industrial use electronic scale which allows patients to be weighed without being moved; its use has not previously been described in a critical care setting.

Prospective audit of tidal volumes achieved per kg of actual and ideal weight in fully ventilated patients with ARDS or ALI. During October and November 2007, 20 patients were assessed. Vt's were recorded for each of the first 24 h of ventilation. Values were excluded if an assisted breathing mode was in use.

Notes and x-rays were reviewed to confirm diagnoses of ARDS/ALI. Patients without X-ray changes were included in the analysis. Patients were measured to calculate ideal weight, and weighed using a Soehnle Professional 3020 system. This uses chocks placed under bed wheels to weigh bed and patient together, allowing deduction of the weight of the bed and equipment to derive patient weight. The weight of all equipment on our ICU was measured prior to introduction of the scales. PaO 2 at the time of highest FiO 2 requirement was recorded, and highest Pinsp used during the first 24 h.

RESULTS. 17 patients met oxygenation criteria for ALI, 12 of these met criteria for ARDS. Actual weight was greater than ideal weight in 12 patients. Average actual weight was 149.9% of ideal weight. Only 5 patients received low Vt ventilation (6-8 ml kg -1 ideal weight). A further 5 had median Vt between 8.0 and 8.5 ml kg -1 ideal weight. 1 patient required peak inspiratory pressure over 30 cm H 2 O. Median Vt was 9.2 ml kg -1 of ideal weight and 6.7 ml kg -1 of actual weight. There were no adverse incidents related to use of the Soehnle scales, and staff required only 10 minutes training in their use.

CONCLUSION. Compliance with low Vt was good in terms of actual weight but relatively poor in terms of ideal weight. Patients were being ventilated with excessive Vt. Compliance with low Pinsp was very good. Ideal weight was lower than actual weight and this seems likely to be a factor in the tendency to excessive Vt. The Soehnle 3020 system is a safe way of weighing unstable critically ill patients. INTRODUCTION. During volume controlled ventilation (VCV), several factors can modify the volume (VT) really delivered to the patient's lung, such as ambient pressure and temperature, water content, and the amount of compressed gas in the tubings. Consequently, the preset VT can differ from the VT delivered to patient's lungs. This bench study aimed at assessing the ability of ICU ventilators to accurately deliver a pre-set VT, by compensating compressed gas and accurately predicting final VT in body temperature and pressure, saturated with water vapour (BTPS) conditions, from initial ambient temperature and pressure dry gas conditions (ATPD).

METHODS. Ten ICU ventilators with compensation of compressed gas (Avea, Elisée350, Engström, Esprit, Evita XL, Extend, PB7200, PB840, Servo I) and three ventilators without compensation (Bird 8400, Avea compensation off, Servo I compensation off) were evaluated using a Michigan lung test model. We compared three pre-set VT (300, 500 and 800 ml), with three simulations of the respiratory system (obstructive, restrictive, normal), with and without an inspiratory pause. With two pneumotachographs, we measured the VT delivered by the ventilator in ATPD conditions and the VT received by the test lung, and used appropriate corrections for simulating BTPS conditions.

Overall, the compensation algorithms for compressed gas allowed to reduce errors on delivered VT in ATPD condition (P \ 0.05). Errors persisted, however, from at ATPD to BTPS condition with frequent overestimations (the figure averages values for all ventilators without versus with compensation). Large errors persisted frequently for all respiratory system conditions, from -10% to +24 % of the preset VT.

CONCLUSION. Errors in delivered volumes were significantly different between both groups of ventilators. Despite technological improvements, performances of ventilators remain heterogeneous regarding the accurate delivery of VT. INTRODUCTION. Injured lungs consist of fast and slow compartments characterized by fast and slow inhomogeneously respiratory time constants (au). So far, regional au has only been determined in animal models of acute lung injury (ALI) by dynamic computed tomography. The aim of our study was to determine regional au and the dependency of au on different positive end-expiratory pressure (PEEP) levels in patients with acute ALI using electrical impedance tomography (EIT).

Five adult patients with ALI caused by pneumonia were included. To determine the inspiratory au, sustained step increase in airway pressure was performed from 0, 8 and 15 cm H 2 O of PEEP to 35 cm H 2 O. EIT scans were obtained at a rate of 25 cycles/s using the Goe-MF II device (Cardinal Health, Hoechberg, Germany). Regional au and the percentage of the corresponding lung tissue were calculated by a bi-exponential fit of the data from the global, ventral and dorsal lung regions in the chest cross-section.

The results of the global regions are shown in Table 1 . Higher au values and percentages of lung tissue with slow au were found in dorsal than in ventral regions. Higher PEEP levels decreased the fast and slow au in global and dorsal regions without significant changes in compartment size. INTRODUCTION. Lung protective ventilation strategies for acute respiratory distress syndrome (ARDS) consider the use of a reduced tidal volume (V T = 6 ml/kg) (1) . Extracorporeal CO 2 removal with the aim of further reducing V T may additionally limit injurious effects of ventilation in severe ARDS. However, ventilation with V T \ 6 ml/kg may cause pulmonary de-recruitment. The aim of this prospective experimental animal study was to evaluate the distribution of regional lung ventilation during ventilation with 5 ml/kg V T combined with extracorporeal CO 2 removal compared to 10 ml/kg V T without extracorporeal CO 2 removal by using electrical impedance tomography (EIT).

METHODS. 10 anesthetized supine pigs (50 ± 5 kg, mean ± sd) were ventilated in a pressure-regulated volume-controlled mode. A pumpless arteriovenous extracorporeal membrane oxygenator (interventional lung assist (ILA), Novalung, Hechingen, Germany) was connected to one femoral artery and vein before acute lung injury was induced by repeated saline lavage. Inspiratory oxygen concentration was increased to 100% and positive endexpiratory pressure (PEEP) was set 2 cmH 2 O above the lower inflection point determined by a constant low flow inflation maneuver. Subsequently, 5 animals were ventilated with 10 ml/kg V T and clamped ILA for 30 min followed by a 30 min period of ventilation with 5 ml/kg V T and open ILA. The other 5 animals were ventilated in reversed order. Regional lung ventilation was determined from EIT measurements (EIT system Goe MF II, Cardinal Health, Höchberg, Germany) during baseline conditions at ventilation with 5 and 10 ml/kg V T in the normal lung, after induction of lung injury as well as 5 and 30 min after ventilation with low V T and open ILA or high V T and clamped ILA, respectively. Changes in ventilation distribution were calculated as anterior-to-posterior shifts in ventilation evaluated from ventilation profiles and so-called centers of ventilation (2).

. A significant shift in ventilation towards the dependent lung regions was observed in injured lungs after having set PEEP using pressure-volume curve. This shift in ventilation was not influenced by the lower V T ventilation combined with ILA and maintained over the whole time period independent of the chronological order. The center of ventilation was located at 50.5 ± 2.5 % of the anterior-to-posterior chest diameter during low V T and open ILA and 50.1 ± 2.2 % during high V T and clamped ILA, respectively.

CONCLUSION. Spatial distribution of ventilation in the lavaged lung is not impaired by the use of lower V T ventilation strategy in combination with ILA under the prerequisite that PEEP is adequately set. INTRODUCTION. Ventilation with high tidal volumes may harm the lung in acute lung injury (ALI) by increasing both lung inflammation and permeability of the alveolar-capillary membrane, resulting in increased mortality. Whether high or low PEEP or prone positioning (PP) improves patient outcome remains controversial. Mathematical analysis of the expiratory pressure-volume (P-V) curve allows determination of specific points (inflection point (c) and point of maximal compliance decrease (Pmcd)) which might help tailoring PEEP level from individual characteristics of respiratory mechanics. The aim of this study is to evaluate regional lung inflammation, perfusion, ventilation and aerated volume assessed with Positron Emission Tomography (PET) in supine position (SP) at 2 PEEP levels, and in PP during experimental ALI.

METHODS. ALI was performed in 21 piglets by tracheal instillation of hydrochloric acid. Animals were then randomly assigned to 3 experimental groups, based on the characteristics of their expiratory P-V curve. 14 animals were studied in SP with PEEP = c + 2 cm H2O (c + 2 group, n = 7), or = Pmcd + 2 cm H2O (Pmcd + 2 group, n = 7). The 7 remaining animals were studied in PP with PEEP = c + 2 cm H2O (PP group). Tidal volume was then adjusted in all groups to keep plateau pressure below 30 cm H 2 O. 2 hours after randomization, ventilation and aerated volume, tissue fraction (FT) and perfusion were assessed with PET using inhaled 13 N, density analysis and 15 O labelled water, respectively. Lung inflammation was assessed with PET as lung uptake of 18 F-FDG (Ki) normalized by tissue fraction (Ki/FT).

RESULTS. PEEP was significantly higher in the Pmcd + 2 group (13 ± 3) than in c + 2 (5 ± 3) and PP (4 ± 2 cm H2O) groups. Tidal volumes were significantly reduced in the Pmcd + 2 group (6.2 ± 1.6 ml/kg) as compared to c + 2 (9.4 ± 0.7 ml/kg) and PP (9.4 ± 1.4 ml/kg) groups. In SP, increasing PEEP level did not significantly modify anterior-to-posterior distribution of both ventilation and perfusion. PP significantly redistributed perfusion and ventilation towards ventral regions. The values of Ki/FT were significantly lower in the PP group, as compared to both SP groups.

CONCLUSION. In this model of ALI, high or low PEEP level did not reduce lung inflammation. PP decreases lung inflammation, as compared to SP. This effect might be related to a protective effect of PP on ventilator-induced lung injury.

GRANT ACKNOWLEDGEMENT. Hospices Civils de Lyon and IBA radio-isotopes France.

INTRODUCTION. Contrast-induced nephropathy (CIN; increase in serum creatinine of [=0.5-mg/dl and/or [=25% within 48 h after contrast-medium) is the third leading cause of hospital-acquired acute renal failure. CIN is frequently found in ICU-patients, and it is associated with increased mortality and prolonged hospitalisation. Hydration as well as theophylline are among the most promising prophylactic approaches. Several studies suggest, that hydration with sodium bicarbonate might be superior to hydration by saline. Therefore, it was the aim of our double-blind study to compare the effects of a hydration with sodium bicarbonate (B) to hydration with saline (S) in addition to prophylaxis with theophylline. 

After major trauma to the muscles, myoglobin is released into the systemic circulation. Myoglobin can precipitate thereby impeding renal function. This can lead to myoglobinuric renal failure, with an incidence of 4-33%. A recommended therapeutic approach to prevent myoglobinuric renal failure is the alkalinization of the urine by administering sodium bicarbonate. Retrospective clinical studies and small trials advocate this therapy. However, two recent larger retrospective trials suggested that alkalinization of the urine does not offer any benefit over hydration and maintaining forced diuresis in these patients. We therefore tested the hypothesis, that forced diuresis with large-volume crystalloid infusion is equally effective in protecting renal function in patients with severe rhabdomyolysis than additional alkalinization of the urine.

After institutional approval 30 consecutive trauma patients were enrolled in this study. To be eligible the patients had to have a serum myoglobin above 2000 mmol/l. Patients were excluded from the study if they had any preexisting renal disease. After enrollment the patients were randomly assigned to two treatments. In the Control Group patients received 5 ml/kg/h Ringer Lactate to achieve a diuresis of 3-4 ml/kg/h. If diuresis was not sufficient 100 ml of Mannitol 20% was administered three times daily. If diuresis still did not achieve targeted volumes a continuous infusion of 0,05 mg/kg/h furosemide was started, and titrated to achieve the targeted diuresis. In the second, the Alkaline Group, patients received the same volume and diuretics management as the Control Group. Additionally active alkalinization of the urine was performed by the administration of sodium bicarbonate. An initial bolus of 100 mmol Sodium Bicarbonate was given followed by a continuous infusion of 0,1 mmol/kg/h. The continuous infusion of Sodium Bicarbonate was then adjusted to keep the urinary pH [ 7. Renal Injury was defined according to the RIFLE Criteria for Acute Kidney Injury. If diuresis decreased \ 0,5 ml/kg/hr despite the above mentioned efforts to increase diuresis, or the patients blood urea nitrogen increased [ 100 mg/ml, or serum creatinine increased C 4 mg/dl, continuous veno-venous hemofiltration was started. The study was stopped after 72 h or when serum myoglobin decreased \250 mg/dl.

. 30 patients were included in the study, 15 in each group. The APACHE scores were similar in both groups. The incidence of Risk, Injury and Failure were comparable in the two groups (table) . One patient in the Alkaline Group and no patient in the Control Group had to undergo hemofiltration. Urinary ph in the Alkaline/Control Group after 24 h was 7,3 ± 0,4/ 6,8 ± 0,5 and after 48 h 7,5 ± 0,3/7,2 ± 0,3. 

Growing interest surrounds the role of microvascular dysfunction and renal hypoxia in the initiation and development of acute renal failure. Regarding ischemiareperfusion(I/R)-induced ARF, this role is poorly understood. In the kidney nitric oxide(NO) is important for adequate endothelial function and regulation of oxygen metabolism. Therefore the aim of this study was to investigate if early I/R-induced renal microvascular hypoxia can be prevented by nitric oxide synthase(NOS) inhibitors.

Prospective animal study, where 28 rats were divided in 4 groups: (1) I/R group underwent a 30 minutes supra-renal aortic clamping (n = 7), (2) I/R treated with a non selective NOS inhibitor L-NAME (n = 7), (3) I/R treated with a selective iNOS inhibitor L-NIL(n = 7) and (4) time-control group(n = 7). Cortical (C\muPO 2 ), and medullary(M\muPO 2 ) microvascular oxygen pressure, renal oxygen delivery(DO 2ren ), renal oxygen consumption(VO2ren) and renal oxygen extraction(O 2 ER) were measured using oxygenquenched phosphorescent method throughout 2 h of reperfusion (1) . Renal function was assessed with creatinine clearance and excretion fraction of Na + .

RESULTS. While renal arterial resistance increased (+192% vs. baseline, p \ 0.05), renal blood flow, C\muPO 2 and M\muPO 2 dropped 2 h after the reperfusion (-70, -42 and -42% respectively, p \ 0.05), as well as DO 2ren and VO 2ren (-70%, p \ 0.0001, and -28%, p \ 0.05). Whereas L-NAME further decreased DO 2ren , VO 2ren , C\muPO 2 and M\muPO 2 , and deteriorated renal function, L-NIL partially prevented the drop of DO 2ren , C\muPO 2 and M\muPO 2 (53 ± 4 vs. 39 ± 3.5 and 41 ± 1.8 vs. 30 ± 2.8 mmHg respectively, p \ 0.05)and increased VO 2ren (0.37 ± 0.04 vs. 0.13 ± 0.01 ml min g -1 , p \ 0.05) in comparison with baseline value. L-NIL also prevented deterioration of renal function. Both L-NIL and L-NAME increased O 2 ER.

CONCLUSION. This study showed that microvascular hypoxia occurs in early stage of renal I/R and demonstrates that microcirculation-targeted therapeutic strategies based on iNOS inhibition have beneficial effects on renal oxygenation and function. In contrast, inhibition of eNOS further decreased microvascular PO 2 highlighting its role in post-ischemic microvacular oxygenation regulation. Furthermore, inhibition function of NO on renal oxygen consumption during ischemia-reperfusion is described. 

Previous studies have shown that neutrophil gelatinase-associated lipocalin (NGAL) is an early predictive biomarker of ischemic and nephrotoxic AKI (1, 2) .The ability to use NGAL as a biomarker for early AKI has never been studied in critically ill adult patients with mixed etiologies. Therefore, we assessed the ability of urine (uNGAL) and plasma NGAL (pNGAL) to predict AKI development and characterize the degree of AKI in a heterogeneous adult ICU population.

We conducted a prospective cohort study of critically ill adult patients from 21st ESICM Annual Congress -Lisbon, Portugal -21-24 September 2008 S195

METHODS. Sixty two patients admitted to the ICU with a diagnosis of severe sepsis were recruited. IFNgamma, IL-10, IL-23 and IL-27 were assayed on ICU admission using QRT-PCR. IL-6 protein was assayed using ELISA. Recursive partitioning dichotomised each patients' ratio of IL-10:IFNgamma and IL-27:IL-23 into either a high and low category, with a score of 1 for high and 0 for low. A composite figure, termed a cytoscore and ranging from 0 to 2, is then calculated for each patient.

There is a direct relationship between the absolute ratio of both IL-10:IFNgamma and IL-27:IL-23 mRNA and increased mortality (p \ 0.05). The robustness of the cytoscore is described in Table 1 . The cytoscore accurately describes the temporal pattern of survival in patients with sepsis (p log rank \ 0.0001, figure A) . Addition of IL-6 protein data can add an extra layer of complexity to the score by describing a very high risk group (cytoscore 3) characterised by early ICU mortality (figure B). CONCLUSION. For those patients with a reported PCT \ 0.5 ng/ml or PCT [ 10 ng/ml, we show a good correlation between reporters. Conversely, PCT of 0.5-2 ng/ml and 2-10 ng/ml show greater variation. This is a significant finding as we know that mortality increases when PCT rises above 1 ng/ml (1) and our study has shown that 30% of BMS would have reported such results in the range outside this value (implying a more severe illness and worse prognosis). This range where there was considerable error for reporting is probably the most important of the 4 values as the transition from 0.5-2 to 2-10 also represents a change from moderate systemic inflammatory response to severe response most likely due to infection. The difficulty experienced when interpreting those bands appears to be because the band colour is frequently non monochromatic making correlation with the reference card problematic. It is imperative to realise this is only semi-quantitative and therefore it is open to individual interpretation. INTRODUCTION. In septic shock, both adrenal insufficiency and inadequate plasma concentrations of arginine vasopressin (AVP) may contribute to the failure to restore vascular tone, respectively (1, 2) . Infusion of a vasoconstrictor hormone like AVP increases arterial blood pressure in septic shock (3) . Measurement of endogenous plasma concentrations of AVP could answer the question whether such an AVP therapy reverses an endogenous deficiency or simply constitutes a pharmacological intervention to increase peripheral vascular tone.

In this prospective, controlled study, we measured AVP and copeptin plasma concentrations in 50 patients with sepsis or septic shock and in ten patients with infection but no systemic inflammation during the first seven days after intensive care unit (ICU) or hospital admission. Hemodynamic, laboratory and clinical data were recorded daily in all patients during the first seven days after ICU or hospital admission. At the same time points blood was withdrawn to determine plasma AVP (radioimmunoassay) and copeptin (immunoluminometric assay) concentrations. A mixed effects model was used for statistical analysis.

RESULTS. The AVP response was different between the three study groups (p \ 0.001), but did not change over time (p = 0.12). While patients with sepsis and septic shock had significantly higher AVP levels than patients with infection (both p \ 0.001), there was no difference in AVP concentrations between patients with sepsis and septic shock (p = 0.98).

No difference in AVP concentrations was observed between survivors and non-survivors at day 28 (p = 0.87). In patients with sepsis, serum osmolarity (p \ 0.001), arterial pH (p = 0.001) and PaO2 (p = 0.04) were indirectly associated with the course of AVP plasma levels, while it was serum osmolarity alone in patients with septic shock (p = 0.03). Plasma AVP concentrations were correlated with copeptin (r = 0.476, p \ 0.001), but this correlation was influenced by continuous veno-venous hemofiltration (p \ 0.001).

CONCLUSION. Sepsis induced a stronger AVP response than infection without systemic inflammation. However, a lacking difference in AVP plasma concentrations between patients with and without shock indicates comprehensive dysfunction of the AVP system in sepsis. Our data support the hypothesis that an impaired AVP response is at least partly responsible for the failure to restore vascular tone in septic shock. INTRODUCTION. Decrease in monocyte HLA-DR (mHLA-DR) in ICU patients has been proposed as a marker of immunity associated with poor prognosis but never evaluated in a large multicentric population. Objective: To test mHLA-DR as a marker of mortality and secondary infection risk in a large multicentric population of patients in septic shock with at least 2 organ dysfunction.

Prospective multicentric study (4 ICU units CONCLUSION. Markers of collagen synthesis and degradation are increased in patients with severe sepsis and may be used as markers of disease severity and outcome. Fibrosis may be one mechanism in the pathogenesis of organ dysfunctions.

GRANT ACKNOWLEDGEMENT. The study was supported by a grant from Instrumentarium Foundation and from Oulu University Hospital, Finland. Advances in neuro-critical care VII: 0770-0772   0770   CONSCIOUS-2: CLAZOSENTAN TO OVERCOME NEUROLOGICAL ISCHEMIA  AND INFARCTION OCCURRING AFTER ANEURYSMAL SUBARACHNOID  HEMORRHAGE (ASAH) INTRODUCTION. In CONSCIOUS-1, a dose-finding study, intravenous clazosentan (1, 5, and 15 mg/h vs. placebo) significantly and dose-dependently reduced angiographic vasospasm in patients with aSAH. A trend towards a decreased incidence of morbidity/mortality was also observed in a post-hoc analysis, using centrally-assessed computed tomography scans and a morbidity/mortality endpoint that included neurological deterioration, rescue therapy and hypodensities due to vasospasm on computed tomography. These findings support the notion that decreasing vasospasm may improve clinical outcome in aSAH patients and led us to design the CONSCIOUS-2 trial.

METHODS. CONSCIOUS-2 is a randomized, double blind, placebo controlled study to investigate the effect of intravenous clazosentan in reducing vasospasm-related morbidity and all-cause mortality after aSAH. Patients are randomized 2:1 to clazosentan (5 mg/hr) or placebo, to start within 56 h of aSAH and continuing until Day 14 post aneurysm rupture.

RESULTS. Main Inclusion criteria: (1) INTRODUCTION. Therapeutic hypothermia protects from neurological sequels and death after out-of-hospital cardiac arrest [1, 2] and it is recommended in the ERC Guidelines [3] . The Hypothermia Registry is a database on the Internet designed to enable evaluation of patients treated with intensive care after cardiac arrest.

METHODS. Data on patient and cardiac arrest characteristics, hypothermia treatment, adverse events, clinical investigations and general intensive care was registered. Outcome was documented as neurological score at ICU and hospital discharge and at six months using the Cerebral Performance Category (CPC) scale [4] : CPC 1-2 representing a good outcome and 3-5 a bad outcome. Only centres reporting all cardiac arrest patients were included.

Between October 2004 and March 2007 1109 patients from 37 centres in 7 countries were entered in the Registry of which 952 were treated with therapeutic hypothermia (TH) and 157 without. In this abstract the TH group is described. The median age was 63 years, 73 % were men and in 68% the initial rhythm was VT/VF. The time to initiation of TH was 90 minutes and the time to achieve core temperature less than 34°C was 240 minutes. Overall cooling rate was 0.6 degrees C/h. All patients, except 9, reached the target temperature. 38% developed pneumonia, 5% had a bleeding requiring transfusion and 29 % had a serious event of arrhythmia during the ICU stay. Data at 6 month was evaluated for 931 patients. 52 % were alive at follow up and 92 % of them had a good neurological outcome; in the groups with initial rhythm asystole/PEA and VT/VF good outcome was 26 % and 60 % respectively.

Our results indicate that half of the patients in an unselected cardiac arrest population admitted to intensive care for TH are alive at 6 months and they survive with a good neurological function. Patients with asystole and PEA have a better outcome than previously reported. The safety aspects compare well with previous studies on TH. Therapeutic hypothermia was feasible to use in all centres. 

To avoid complications of the volume load, used for the cerebral perfusion support in brain trauma, we tried to determine volume responsiveness in these category of patients.

We examined 30 patients with severe brain trauma. (GCS score \ 8, sedation, artificial ventilation). The stroke volume (SV) and its variability (SVV) were determined by the bioimpedans method. Additionally we determined the variability of other haemodynamic parameters: BP, HR, CI, EF and peripheral pulse, which reflect the type of central autonomic regulation (were showed in our previews works) . Especial attention were paid to the P4 (0.2-0.5 Hz) bands of SVV, which predominantly connect with breathing, as a main marker of hypovolemia. All the comparisons were made before and after increasing of PEEP to 5 and then to 10 cm H2O. These were made before and after infusion of Stabisol (app. 10 ml\kg).

RESULTS. All patients were divided in two groups: A-18 patients (responders on volume load by the increasing SV) and B-12 patients (nonresponders: not only the absents of SV increasing, but in some cases -worsening the type of central regulation of haemodynamic). In group A the increasing PEEP to 5 and 10 cm H2O before volume load showed the decreasing of BP, rising HR and the grate increasing of P4 SVV. After volume load the same changes were only if PEEP was 10 cm H2O. It's probably related with no complete compensation of hypovolemia. In group B the same PEEP increasing before volume load did not revealed any changes in haemodynamic. After volume load PEEP increasing to 10 cm H2O induced some increasing of SV.

CONCLUSION. PEEP may be used in severe brain trauma for evaluation of the range of hypovolemia and volume responsiveness prediction. PEEP-the test that allows to estimate if addition volume expansion is needed and help to avoid the exertion of regulatory systems in other cases. INTRODUCTION. The Ventilator Bundle-elevation of the head of the bed, daily sedation vacations, peptic ulcer prophylaxis and DVT prophylaxis-has been promoted by the IHI as ''a series of interventions that, when implemented together, will achieve significantly better outcomes than when implemented individually.'' However, rigorous evidence for the bundle approach is weak. In May 2006, the Welsh Critical Care Improvement Programme (WCCIP) launched implementation of care bundles at the national level. By August 2006, the Ventilator Bundle had not only been introduced in all 14 critical care sites in Wales, but WCCIPmonitored compliance rates indicated national compliance of over 95%. To date, no evaluation of patient outcomes has been attempted. Opportunistically, 13 critical care units from 12 of the 14 sites participate in the Case Mix Programme, the national comparative audit of critical care, enabling a comparison of outcomes before and after implementation and achieving high compliance.

METHODS. Data covering the implementation date were available from 10 of the 13 units. Data were extracted for all admissions, since 2002, who were ventilated during the first 24 h following admission. The effect of implementation of the Ventilator Bundle on hospital mortality (for all ventilated admissions) and on unit length of stay (for unit survivors) was assessed using multilevel models adjusted for time trends, seasonality and predicted risk of death (ICNARC model). Implementation of the Ventilator Bundle was modelled including a lag of two months following implementation, to allow for the observed delay in reaching maximal compliance.

. 10,656 admissions were included in the analysis. Units had a median of 51 months data pre-implementation (range 2-54) and 19 months post-implementation (range . Most units implemented the Ventilator Bundle between March and August 2006, although two units had implemented it previously. Crude hospital mortality for ventilated admissions was 42.0% pre-implementation and 42.8% post-implementation. Mean unit length of stay for survivors was 7.1 days pre-implementation and 6.3 days post-implementation. Implementation of the Ventilator Bundle was not associated with any significant reduction in hospital mortality (odds ratio 0.96, 95% confidence interval 0.81-1.14) or in unit length of stay for survivors (ratio of geometric means 0.98, 0.88-1.10).

CONCLUSION. Despite widespread implementation and high levels of compliance across Wales, no significant impact on patient outcomes was identified. Possible reasons for lack of impact may include: that 1.5 years is too early to detect impact; that compliance does not necessarily equate to delivery of bundle elements (consideration and rejection of an individual element equates to compliance); or that the evidence base for bundles is weak -both for individual elements and collectively. While the failure to identify any demonstrable effect on patient outcomes in our study may disappoint those that support the bundle approach to quality improvement, it should not detract from the importance of following current best practice in the management of these patients. INTRODUCTION. Unintended tube removals (UTR) are frequently observed adverse events in ICUs and potentially involving serious complications. We monitored UTR in our ICU for a 2-year period to assess the incidence, the patient conditions at the time of event occurrence to analyze the system factors associated with UTR and to evaluate the effect of targeted interventions.

METHODS. Demographic and clinical information was prospectively collected on all patients admitted in our ICU. Additional information was collected for patients who experienced an UTR. A dedicated staff analyzed the UTR bi-monthly using a system-factor approach to inform the ICU team on the rate and the potential cause of events and to propose, implement and improve care management protocols. A conditional logistic regression stratified on length of stay was used to identify risk factors of tube removals. Segmented linear regression analysis was conducted to test for the effect of the improvement interventions. INTRODUCTION. NICCIMS is a prospective regional audit study designed to analyse occurrence of adverse events in intensive care units (ICUs). An anonymous database has been installed in seven ICUs since September 2007 facilitating central analysis of all reported incidents in N.Ireland. The first NICCIMS report revealed that incidents relating to airway and central venous catheters were the events with highest potential impact as well as marked variability in frequency between the 7 units.

Following collation of data related to these events we examined if effective local policies exist. Questionnaires were sent to the lead consultants, lead nurses and educational supervisors in the participating ICUs. The questions were related to the presence or absence of policies, and their content, regarding securing of airway devices and management of central venous catheters. Responses were compared between staff within units to assess agreement.

RESULTS. 100% response rate was achieved. 5 out of 7 units claimed to have a written policy regarding endotracheal tube (ETT) fixation. However there was disagreement as to the existence of a policy between the lead clinicians and lead nurses in 4 out of 5 units. In 5 out of the 7 units, percutaneous tracheostomies are performed in more than 80% of cases, but in 2 units surgical tracheostomies are performed in 70-98%. There appears to be more adverse events reported related to tracheostomies from the latter 2 units. All units claimed to have written policies related to central venous catheters, but in 2 units there was disagreement between staff as to the existence of a policy. 57% respondents stated the policy is only followed 'most of the time'. If staff were aware of a policy they said subclavian was the primary site of choice but staff unaware of a policy said they were using the internal jugular site.

CONCLUSION. A wealth of information regarding adverse events has been collected through NICCIMS. The presence of effective policies is likely to reduce the possibility of an incident as it provides staff with an agreed management approach. However, effective policies do not appear to be present in all ICUs. Discordance between clinicians and nursing staff regarding the existence of policies suggest an ineffective policy as staff are unaware of it, and therefore are highly unlikely to achieve the aim of the policy which may give rise to increased adverse events. There is an opportunity for regional policies to be agreed between units to simplify and unify our approach to enhancing safety. Institute for Clinical Excellence in the United Kingdom (NICE) has expressed concern that ''Clinical audit has a mixed history in the NHS, and for every success story there are just as many projects that have run into the ground without demonstrating any significant contribution to quality of services.'' To improve this process, the Clinical Governance Support Team of the NHS has designed an Audit Project Assessment Tool (APAT) which assesses individual projects against national guidelines for good clinical audit. A maximum of 25 points are available and audit projects scoring more than 16 are regarded as good projects. This study examined whether the audit programme our institution was effective and associated with improved clinical governance.

Presentations from clinical audit meetings from a 24 month period were retrospectively reviewed by two independent investigators and the following questions assessed.

(1) How many of the projects were true audits? (2) Were the audit projects prospective or retrospective? (3) How many re-audits had been performed? (4) If re-audit showed improvements in performance? (5) What were the APAT scores of the projects?

RESULTS. Twenty-nine audit projects were presented during this period of which 25 were available for review. Of the projects reviewed, 22/25 (88%) were considered to be true audit projects of which 2/22 (9%) had a research overlap. 22/22 (100%) were retrospective and there were 4 (16%) re-audits during the period of investigation. Of the re-audits, 3 showed an improvement in performance and 1 showed deterioration. The APAT scores ranged from 9/25 to 21/25. Eleven projects (50%) scored greater than the 16 points required to be considered a good audit. It was noted that there was a paucity of multi-disciplinary audit projects in the cohort studied.

CONCLUSION. This study shows that a well-coordinated active audit programme was in place at the above institution. However, objective evidence of clinical governance benefit was lacking and 50% of the audits would have been rejected if the APAT tool had been used prospectively. We believe the results of this study are likely to be similar to other institutions experiences and to improve future performance, we suggest all new projects are prospectively assessed using the APAT tool at outset, and poorly scoring studies rejected. In addition, a database of past and proposed audit projects should be constructed to facilitate re-audit where performance improvement is indicated and to highlight deficits in areas (such as a lack of multidisciplinary audits). Clinical leads need to be identified to provide coherence to the process and to report to unit governance meetings. is an easy-to-understand method to monitor outcomes (1) and a powerful tool to show developments in clinical practice (2) . It has been seen as a promising tool for real-time outcome monitoring (3) . Constant data delivery is a fundamental issue for the VLAD usability within a benchmarking project. A constant reporting cycle helps the units to routinely follow outcomes compared with those of similar units.

Intensive Care Units (ICU) in Finland have measured their performance and quality of care since 1994. At present 24 Finnish ICUs compare continuously their performance within the Finnish Intensive Care Quality Consortium (FICQC). Data covering the ICU stay are delivered to the centralized benchmarking database both manually (4 units) and

with an interfacing software (20 units), presenting 15 and 85% of IC period records respectively. The hospital discharge data and long-term status data are delivered with an internet application. Primary admissions from 2007 (N = 17,517) were included. Hospital discharge status was selected as a primary outcome parameter.The reconstructed figure (Fig 1) shows the data accumulation at 1st September 2007 and the actual data delivery differences between ICUs.

Sixty four percent (64%) of HDD was recorded within 30 days and 83% within 60 days. Average delay from hospital discharge to hospital discharge data (HDD) recording was 31.2 days (95% CI 30. 6-31.8 ). Delays to register HDD were significantly shorter (24.5 days, SD 32.9) for dead patients than for survivors (32.5 days, SD 39.2). Mean HDD delivery after hospital discharge varied between units from 6.4 (SD 20.4) days to 63.7 days (SD 40.1).

CONCLUSION. Short and known reporting delays from ICU discharge and hospital discharge status recordings shorten the benchmarking reporting cycle and improve the report usability. Faster non-survivor data delivery is due to the data delivery protocol. Local possibilities to organize the data delivery cause variation between the ICUs. The recording delays and the variation between ICUs should be taken into account when benchmarking the ICU performance and outcome. CONCLUSION. IMS improved detection of instability as compared to conventional four channel monitoring in a SDU and decreased both the total number of unstable patients and the instability duration. Use of the IMS also increased the probability that MET activation would be called for serious instability. Further study will determine the relationship between improved detection and treatment approaches for unstable patients. METHODS. In a 10-bed polyvalent ICU located in a 1400 bedded community and teaching hospital a PIMS (Centricity Critical Care Clinisoft; G.E. healthcare) was introduced on June 1st 2005. Before the implementation of the PIMS fluid and medication orders were written by hand by the physicians on a daily medication chart. After the implementation, drug prescriptions were entered via the keyboard. The system was developed to supply pre-defined fluids and medication protocols, as well as default medication dosages and drug interaction warnings.To compare the frequency of medication errors before and after the implementation of the PIMS, data of two periods of 15 days each (30 to 15 days before and 60 to 75 days after) were prospectively collected and analyzed. For each period, 150 patient/days were considered. No one of the staff members, but the ICU director, was made aware of the ongoing study. Patients charts were daily revised between 7.00 and 8.00 AM by two independent physicians, not directly involved in patients care. Fluids and medication orders were revised looking for medication errors. The ICU director was immediately notified about any identified error, to allow immediate correction during the ward round. Errors were classified as Type 1 (clinically relevant and potentially dangerous) or Type 2 (registration errors without clinical relevance). CONCLUSION. the implementation of a PIMS in ICU was followed by a highly significant reduction in medication errors. Although PIMS are expensive, they may lead to a dramatic amelioration in the quality of care. The economic impact of the reduction in medication errors needs to be further investigated. (1) and improved quality of teaching (2) . In this context, it seemed relevant to create a website about echography in the ICU, accessible 24/24, recalling the fundamentals (main measures, normal and pathological values), illustrated by images and videos. This work reports the initial conditions of use.

Echorea.org is a free website, with intuitive ergonomy. It consists of files organized into three categories allowing three possible approaches: by diagnoses, by ultrasound signs, or by clinical signs. Practical clinical cases complement theory. Most assertions are justified by bibliographic references at the bottom of sheet with pubmed link. The dates, times and origins of visits has been collected day by day by an online software. CONCLUSION. The pace of consultation to echorea.org is continuing but presents 2 particular peaks which appear to correspond to the search for timely information out of senior coaching (afternoon, custody). The number of connections reported to the medical presence is higher at night and on weekends. The chosen support seems therefore to satisfy two conditions:

-A virtual companionship (3) by a modern and easily accessible support dealing with the complexity and magnitude of knowledge to be retained, particularly when people lack accompaniment, especially in custody (37.5% of connections at night and 23% at WE). Thus, 27% of visitors have already visited at least once. -Ease of access to dense but illustrated information. Internet allows a wide, fast and free diffusion (1), maintaining a rate of newcomers to the site of 73%. Connections was initiated by the hospital, the faculty, and the home. Echorea.org is visited throughout France but also abroad. The translation into English is one of the main projects in the short term. 

We first compared, for 1158 sequences of SC recordings, the mean EtCO2 with the corresponding Max EtCO2; 10 aberrant values needed to be discarded. We then calculated the differences between mean EtCO2 or maximum EtCO2 and PaCO2 over different periods.

Comparisons were performed only when mean and/or maximum EtCO2 were available shortly before and after the blood gas sampling. Mean* EtCO2-Max* EtCO2 (N = 1158) -3 ± 6 -2 (-4; -1) Mean* EtCO2-PaCO2 (N = 74) -11 ± 13 -9 (-13; -6) Max EtCO2 (5 min 

Recently a new ultrasound dilution (UD) method was introduced that uses insitu arterial and venous catheters to measure COUD and Central Blood Volume (CBV). The method is based on measurement of decrease in blood ultrasound velocity caused by injecting isotonic saline. The purpose of the study was to compare UD measurements with transpulmonary thermodilution (TT) measurements (COTT & Intrathorathic Blood Volume ITBV, parameter analogous to CBV).

METHODS. Eight adult patients (age 34-64) with hematological malignancies were included in the study. Indications for invasive monitoring were: sepsis and acute lung injury (3); ARDS (1); septic shock (2); acute congestive heart failure (1); intracranial hemorrhage (1) . For TT measurements 3 injections of 20 ml cold 5% dextrose were performed into 5Fr PiCCO femoral arterial catheter. For UD measurements, an extracorporeal AV loop was connected between insitu A-V catheters. Reusable UD sensors were clamped on the arterial and venous limbs of the loop. Pump circulated blood from the artery to the vein at 8-12 ml/min for 5-7 min. Three UD readings (COstatus, Transonic Systems Inc., USA) were obtained by injecting 20-30 ml of body temperature isotonic saline. At the end of the session, the system was flushed with heparinized saline until the next measurement session.

RESULTS. 51 pairs of averaged data were obtained for comparison. For CO, correlation was R = 0.94; COUD = 1.07*COTT-0.40 L/min; bias was 0.32 L/min and the error (2SD/Mean of TT) was 18%. For volumes, correlation was R = 0.90; CBV = 0.76*ITBV + 80 ml; bias was 401 ml (CBV \ ITBV) and the error was 21%. Loss of thermal indicator could possibly lead to overestimation of ITBV. The % error for both CO and volumes is considered clinically acceptable (\ 30%).

CONCLUSION. First clinical comparison of UD vs. TT produced close agreement in adult hematology patients. New UD method works off existing catheters; is quick to set up and involves no blood loss. INTRODUCTION. Critically ill patients are characterized by increased loss of muscle mass, partially attributed to sepsis and multiple organ failure, as well as immobilization, which may lead to the presentation to critical illness polyneuromyopathy (CIPNM).Recent studies have shown that electrical muscle stimulation (EMS) may be an alternative to active exercise in COPD and CHF patients.Our aim was to investigate the EMS effect on muscle mass preservation of critically ill patients with the use of ultrasonography (US).

In this prospective study 29 critically ill patients (age:56 ± 21 yr) (APACHE score:20 ± 5) (SOFA score:9 ± 3) were randomized after stratification upon admission to receive EMS sessions (45-min/5 days per week) of vastus lateralis, vastus medialis and peroneus longus muscles of both lower extremities (EMS-group) or to the control group (non-EMS).Nineteen patients were finally evaluated, 8 patients in the non-EMS and 11 patients in the EMS group.Muscle mass was evaluated with US, by measuring the Cross Sectional Diameter (CSD) of the quadriceps muscle (rectus femoris-vastus intermedius).US images were obtained using a GE Vivid 7 model ultrasound scanner with a 7.5 LHz linear probe.The position of the probe was selected at the midway between the anterior superior iliac spine and the midpoint of the patella and was placed ventral to the transverse plane and perpendicular to the skin.The measurements were performed on the 2nd and 7th day fo11owing admission.The MRC scale for the clinical evaluation of muscle strength was used for the diagnosis of CIPNM (cut off \48/60).

Rectus femoris CSD decreased significantly in the non-EMS group (from 1.48 ± 0.42 to 1.24 ± 0.35 cm, p \ 0.05).In the EMS group the CSD of rectus femoris decreased although not significantly (from 1.37 ± 0.43 to 1.33 ± 0.40 cm, p [ 0.05), however there was a significant between-group difference in the CSD decrease (EMS group: 0.04 ± 0.19 cm vs. non-EMS: 0.23 ± 0.11, p \ 0.05).Vastus intermedius CSD decreased in both groups, however the decrease was not significant in the EMS-group (EMS group: from 0.75 ± 0.35 to 0.73 ± 0.34 cm, non-EMS: from 1.39 ± 0.63 to 1.05 ± 0.48 cm).There was a significant difference (p \ 0.05) in the vastus intermedius CSD between EMS-group (-0.26 ± 0.19 cm) and non-EMS (0.34 ± 0.20 cm). In the non-EMS group 25% of the patients were diagnosed with CIPNM as compared to none in the EMS-group.

CONCLUSION. This is the first prospective randomized study for the evaluation of EMS for the preservation of muscle mass in critically ill patients and for the prevention of CIPNM. These preliminary data imply that EMS may preserve the muscle mass of critically ill patients and could have a decisive role for the prevention of CIPNM 

There is a need to determine the modification of the function as an indicator of the use of respiratory physiotherapy. The Method Functional Approach suggests that the action of the physiotherapist is based on the prioritization of signs and symptoms as functional markers in the interpretation of functionality.

Objective: To determine the incidence of pneumonia and atelectasis in two groups of patients in the postoperative care of cardiac surgery undergone.

METHODS. Prospective,randomized,controlled and blinded. The study was conducted at the Unit for Post-Operatório of Heart Surgery (UPC), Hospital Portuguese -Salvador/BA. 150 included in the study, patients must be 18 years old, that had hemodynamic stability. All patients were evaluated by physical therapists trained in both the pre about in the postoperative. After randomization, patients in Group I (control) were treated with NCPAP 20 cm/ H2O, 30 min of 4/4 h until the time of discharge from the UPC, the patients in group II (experimental) were treated in accordance with the routine of Physical Therapy, Method Functional Approach.

RESULTS. The length of stay in UPC, 73.52 ± 36.33 h (about three days). There was no statistically significant difference between the groups studied. 9.8% had a diagnosis of atelectasis and 5.0% were diagnosed with pneumonia and was hired in group I, 12.7% had a diagnosis of atelectasis and 5.7% were diagnosed with pneumonia and were in Group II, there was no statistically significant difference between groups. Patients in group II, 64.7% of the interventions made was kinesiotherapy, appear drills active and walking, there was a relatively low frequency with regard to the need for interventions linked to respiratory. INTRODUCTION. Threshold can be used as a physiotherapic tool in order to increase muscle strenght. This effect can be useful in weaning patients. However there are still controversies considering its advantages during weaning from mechanical ventilation (MV). This study aims to evaluate the effects of the threshold in such situation.

Patients under MV for more than 48 h and prone to weaning were randomly assigned to control or to the threshold group (trained twice daily). They were followed until extubation, tracheotomy or death. All cardiorrespiratory variables, maximal inspiratory and expiratory pressures (MIP and MEP), length of weaning and success or failure were registered. Statistical analysis was performed using ANOVA, Mann-Whitney U test and Chi-Square test, where appropriate. It was considered a significant level of 0.05.

Eighty-six patients were studied (52% men, mean age 63 ± 17 years, 48% with chronic obstructive pulmonary disease). No differences were observed when comparing initial versus final cardiorrespiratory variables in both groups, with exception of the MIP (varied from -33,72 ± 13,5 cmH2O to -40,81 ± 12,67 cmH2O in the threshold group and from -37,67 ± 10,49 cmH2O to -34,19 ± 10,85 cmH2O in the control group, p \ 0,001), MEP (varied from 25,47 ± 12,48 cmH2O to 29,65 ± 12,02 cmH2O in the threshold group and from 29,65 ± 11,97 cmH2O to 26,86 ± 11,6 cmH2O in the control group, p \ 0,05) and tidal volume (varied from 386,16 ± 236,56 ml to 436,16 ± 228,39 ml in the threshold group and from 361,91 ± 168,81 ml to 357,14 ± 121,35 ml in the control group, p \ 0,05). No differences were observed in length of weaning (1,36 days with threshold group versus 1,98 days in control group, p [ 0,05) and weaning success (83,7% with threshold group versus 76,7% in control group, p [ 0,05).

CONCLUSION. Threshold during weaning from MV can cause an increase in MIP, MEP and tidal volume. However, in this group of patients it was not associated with a decrease in length of weaning or an increase in weaning success. (1) . To counteract these effects there has been an increasing move towards early rehabilitation on Intensive Care units (ICU), with some evidence suggesting shortened lengths of stay (LOS) in response to these interventions (2) . Following an earlier study at Manchester Royal Infirmary, which highlighted a significant reduction in LOS in response to early rehabilitation (3), a structured rehabilitation programme was implemented with the aim of decreasing overall length of stay and subsequently improve outcomes such as overall mortality. This study aimed to determine the effect of the introduction of a structured rehabilitation programme for patients admitted to critical care.

METHODS. The structured rehabilitation programme was implemented at the beginning of 2004. Physical function is improved by each patient having an individualised programme that starts in ICU as soon as the patient is conscious and is adjusted accordingly throughout the remainder of his or her hospital stay. Primary outcome measures used were mean ICU and post ICU LOS, with secondary measures of mortality. Baseline data was obtained retrospectively, with annual figures presented for the three years following the introduction of the rehabilitation programme. Data were analysed using students t-test.

In the year prior to the introduction of the structured rehabilitation programme the mean LOS on ICU was 9.8 days. Initially on introduction although a mean reduction of 0.9 days was observed this was not statistically significant. However in the following two years a significant reduction in LOS was then observed, with mean LOS falling to 8 days in 2005 (p \ 0.05) and 7.7 days by 2006 (p \ 0.01). Overall mortality also significantly fell from 45% in 2003 to 35% in 2006 (P \ 0.01). INTRODUCTION. Critical tissue oxygen availability may affect jejunal motility, and this can further be modified by nitric oxide. We studied ex vivo jejunum motility after exposure to two different conditions of low oxygen availability: hypoxic hypoxia and cardiac tamponade. We hypothesized that cardiac tamponade and hypoxic hypoxia decrease jejunal motility.

METHODS. 24 anesthetized and mechanically ventilated pigs were randomized to cardiac tamponade (CT, n = 8), hypoxia (HH, n = 8) or control (C, n = 8). In CT, cardiac output was reduced in 6-h steps to reach 50, 40 and 30 ml kg -1 min -1 . In HH, FiO2 was reduced in 6 hrs to reach a PaO2 of 50-60 mmHg and then at 12 h to \ 50 mmHg. The lowest levels were maintained until 24 h or until death occurred. Cardiac output was measured by thermodilution and regional blood flows by Doppler ultrasound. At the end of the animal experiment, jejunal samples were analyzed by the tissue bath method. Acetylcholine (1-100 lM) and sodium nitroprusside (SNP, from 1 to 100 lM) effects were tested by constructing dose response curves. Maximum force and maximum relaxation values were calculated. Vascular reactivity data from the same animals are presented in another abstract.

RESULTS. Cardiac output decreased in CT from 77 ± 14 to 44 ± 12 ml/kg/min, increased in HH from 75 ± 15 to 113 ± 26 ml/kg/min, and did not change in controls (time-group interaction; p = 0.001 [ANOVA for repeated measurements]). Superior mesenteric artery blood flow decreased by 30 ± 5% in CT, and increased by 2 ± 17% in HH and 22 ± 15% in controls (p = 0.04). Acetylcholine dose response is presented in figure 1 . Maximal acetylcholine responses were 119 (105-137)%, 160 (137-165)%, and 142 (124-181)% in CT, HH and C, respectively (p = 0.06). There were no differences in sodium nitroprusside-induced relaxation between the groups.

CONCLUSION. Jejunum motility is preserved after exposure to CT and HH. Ex vivo jejunal motility is not related to tissue oxygen delivery within the range of this experiment. The role of regional mediators should be addressed in further, larger series. INTRODUCTION. Acute renal failure (ARF) is a common complication seen in 10-23% of patients admitted to the intensive care unit. When occurring, ARF leads to an increased mortality rate. The primary causes of death in this group of patients are extra renal, including infection, shock, sepsis, and respiratory failure. Post mortem examinations in these patients reveal cellular infiltration of lungs and kidneys. It is possible that ARF elicits an extra renal inflammatory reaction responsible for the increased mortality. We therefore investigated the effect on the immune system of renal ischemia/reperfusion (I/R) compared with I/R of the hind legs in mice.

METHODS. 80 mice in 4 groups A, B, C, D. A were subjected to I/R of the kidneys. B had unilateral nefrectomy. C were only anaesthetized. D had I/R to the hind legs. In each group, 2 and 24 h following the primary intervention, 10 mice were sacrificed and humoral inflammatory markers, lung and kidney tissue was obtained. The measured parameters included the adhesion molecule of CD11b, TNF-alfa, and MHCII on monocytes. In lungs and kidneys, TNF-alfa and myeloperoxidase (MPO) was measured.

RESULTS. In all the I/R groups we observed an up-regulation of CD11b and a downregulation of MHC II. As illustrated in the figure, a pronounced increase in MPO was observed in the kidneys following I/R of the kidneys as compared to I/R of the extremities. In contrast the concentration of MPO in the lungs was more pronounced in the group exposed to I/R of the extremities.

CONCLUSION. Reperfusion following ischemic injury results in increased levels of inflammatory markers. Upon I/R of the hind legs, the inflammatory response in the lungs was more pronounced than following induction of I/R in the kidneys.

GRANT ACKNOWLEDGEMENT. Danielsens foundation. INTRODUCTION. While increased arterial lactate concentrations in sepsis have been attributed both to inadequate tissue oxygen delivery and impairment of oxygen extraction capabilities, they are the consequence of decreased hepato-splanchnic perfusion when cardiac output is low. We hypothesized that tissue oxygen delivery is adequate in sepsis, and therefore, the contribution of hepato-splanchnic lactate exchange to systemic lactate flux is not increased.

METHODS. 24 anesthetized pigs were randomized to fecal peritonitis (P), cardiac tamponade (CT), or to serve as controls (n = 8 per group). Systemic (thermodilution) and regional (ultrasound transit time) flows and lactate concentrations were measured and oxygen transport and lactate exchange were calculated over 24 h.

RESULTS. Cardiac output increased in P (116 ± 37% of baseline) and decreased in CT (59 ± 18% of baseline), hepatic arterial blood flow increased in P and remained unchanged in CT, and portal flow decreased in CT and remained unchanged in P (time-group interaction: all p \ 0.03). In CT but not in P, hepatic oxygen consumption decreased (60 ± 38% of baseline and 101 ± 44% of baseline, respectively; p: 0.049). Lactate data are shown in the table. 

Increasing systemic lactate flux in peritonitis is a consequence of an early increase in regional, extra-splanchnic lactate production without signs of tissue hypoxia but with a failure to increase hepatic lactate extraction. In cardiac tamponade, regional, extrasplanchnic lactate production is accompanied by hepatic tissue hypoxia and net hepatic lactate production. INTRODUCTION. Mortality in critically ill patients with acute renal failure (ARF) is unacceptably high. Acute circulatory and respiratory failure at admission to an intensive care unit have been identified as the most important risk factors for the development of ARF. The aim of this study was to assess the effect of acute reduction of either cardiac output or arterial oxygen saturation on renal oxygen transport and kidney function.

Prospective, randomized, controlled study in 24 anesthetized, mechanically ventilated pigs (weight: 42 ± 3.2 kg). In cardiac tamponade (CT; n = 8), cardiac output and blood pressure were reduced by filling the pericardial sac with HES 6%. In hypoxic hypoxia (HH; n = 8), FiO2 was reduced. Systemic (thermodilution) and renal (ultrasound Doppler) blood flow were measured at baseline and at 6 and 12 h. 8 animals served as controls. Differences between groups were assessed by repeated measurements ANOVA [RMA], all pvalues are time-group [T 9 G] interaction).

Control animals remained stable. Cardiac index decreased in CT from 72.5 ± 14.2 to 57.3 ± 11.7 mlkg-1 min-1 and increased in HH from 71.5 ± 14.4 to 106.2 ± 27.4 mlkg-1 min-1 (p \ 0.001). Arterial blood pressure remained stable in HH, while decreasing in CT from 73.6 ± 12.7 to 58 ± 14.9 mmHg (p \ 0.06). In HH, arterial oxygen saturation decreased from 97 ± 0.0 to 68 ± 11 (p \ 0.001). Renal blood flow and oxygen delivery decreased in CT from 4.9 ± 0.7 to 3.5 ± 1.0 ml kg -1 min -1 , and from 0.64 ± 0.15 to 0.43 ± 0.12 ml kg -1 min -1 , respectively (both p \ 0.04). Renal perfusion pressure decreased in CT from 68.29 ± 13.9 to 47.14 ± 15.97 (p \ 0.025). In HH, arterial and renal venous oxygen content decreased from 118.5 ± 6.6 to 89.0 ± 17.6 ml/l, and from 82.83 ± 12.4 to 50.9 ± 12 ml/l (both p \ 0.05), while renal blood flow and oxygen delivery were maintained. In CT, creatinine clearance and urinary output decreased from 43.65 ± 8.45 to 27.25 ± 12.13 dl/hr (p = 0.09) and from 1.23 ± 0.7 to 0.48 ± 0.3 ml/kg/h (p = 0.08), respectively, and fractional sodium excretion from 2.03 ± 2.12 to 0.41 ± 0.2 (p = 0.035). These variables remained unchanged in HH and in controls.

Reduced renal blood flow and blood pressure but not reduced renal oxygen content are associated with ARF. Over the short term, severe arterial hypoxemia (SpO2 B 70%) seems to be well tolerated when systemic hemodynamics are not compromised.

GRANT ACKNOWLEDGEMENT. Supported by grant 3200BO/102268 from the Swiss National Fund and by a grant from Novartis.

In recent years there has been increasing evidence that resuscitation strategy with different fluids can have divergent impacts on the immune response and cytokines release. This study was undertaken to measure the IL-8 in the serum samples and bronchoalveolar lavage fluid (BALF) during acute normovolemic hemodilution (ANH) performed with hydroxyethyl starch (HES), normal saline solution (NSS) or gelatin (GEL).

METHODS. Twenty eight pigs were anesthetized, instrumented and randomized into four groups: Control, ANH + HES, ANH + NSS and ANH + GEL. Animals in the ANH group were submitted to acute normovolemic hemodilution to a target hematocrit of 15% with volume replacement performed with HES 130/0,4, GEL both at a 1:1 ratio or with NSS at a 3:1 ratio. The withdrawn blood was returned to the animals 120 minutes after the end of hemodilution. Cardiac Index (l/min/m2) was determined at the following time points: before ANH (Baseline), after instrumentation (INST), immediately after ANH (H), 60 minutes after ANH (60H), 120 minutes after ANH (120H), 60 minutes after blood infusion (60BI) and 120 minutes after blood infusion (120BI). The measurement of IL-8 (pg/ml) was performed with blood samples collected at the femoral vein immediately after ANH (H) and in the BALF after the end of the experiment. The cytokine IL-8 quantification was determined by Enzyme Linked Immuno Sorbent Assay (ELISA). One-way Analysis of Variance (ANOVA) was performed to evaluate differences among groups. A p value of 0.05 was considered statistically significant.

RESULTS. Cardiac Index (CI) was significant higher in groups ANH + HES (7.3 ± 1.2; p \ 0.05) and ANH + GEL (7.7 ± 1.0; p \ 0.01) when compared to Control ( INTRODUCTION. Sepsis is a major clinical problems in ICU. Many critically ill patients with severe sepsis develop ARF (acute renal failure) as a part of multiple organ failure (MOF). Sepsis is characterised by an uncontrolled immune response with release of proinflammatory and anti-inflammatory cytokines, such as Il-6, TNF-lpha, Il-10. A potential application of high volume continuous veno-venous haemofiltration is the extracorporeal removal of inflammatory mediators in septic patients, which can lead to an immunomodulatory effect. Aim of our study is to investigate the effect of high volume continuous haemofiltration on the transcriptional activity of mononuclear cells (PBMC).

We enrolled 10 medical and surgical patients from our ICU suffering severe sepsis or septic shock. High flux CVVH was carried out and blood samples were obtained before the beginning of CVVH (T0), after 12 h (T1) and daily (T2-T4). Primary outcome measure was trend of IL-6 gene expression in our patients during CVVH.

RESULTS. Figure 1 shows trend of IL-6 gene expression. We observed Il-6 mRNA drop after 12 h of treatment (T1), then a progressive increase at 24, 48, 72 h.

Our data show an improvement of transcriptional activity of PBMCs in septic patients treated with high flux CVVH. These data could represent an expression of attenuation of sepsis-related immunoparalysis. . Little attention has been given to the status of gastrointestinal tract despite gut dysfunction occurs frequently among critically ill patients (CIP).The gut has often been suggested to be one of the essential factors in the pathogenesis of MODS and, CIP are susceptible to injury of the gastrointestinal tract. The aim of this study is to identify enteric dysfunction (ED) at admission and discharge, its association to other organ dysfunctions (OD), according to OD is acute or chronic, primary or secondary, its evolution to organ failure (OF) or not, and its relations to severity and outcome. CONCLUSION. Prevalence of ED is considerable. Its influence in overcome and severity is important. ED is the most prevalent primary OD. The majority of CP with ED presented MODS. Cardiovascular, respiratory and renal OD were mostly associated to ED. The association of ED to sepsis, tumor, metabolic acidosis and invasive ventilation is clinically relevant. INTRODUCTION. The objective of this study was to evaluate efficacy of corticosteroid therapy in a broad population of patients with septic shock not only who had relative adrenal insufficiency (RAI) but also who did not have RAI according to the results of high or low dose corticotropin stimulation tests. INTRODUCTION. The hypotalamic-pituitary-adrenal axis is a major determinant of the host's response to stress and its dysfunction can be present in septic shock. Adrenal dysfunction has been associated to higher mortality (1) . However, the evaluation of the adrenal response is difficult (2) and substitutive use of low dose of steriods in septic shock is controversial (3) . The objective of the study is to evaluate the prognostic value of the adrenal response to corticotropin in patients with septic shock treated with hydrocortisone.

METHODS. Prospective, observational study in a medical-surgical ICU. All adult patients admitted in the ICU last year with septic shock were consecutively included, except those already treated with steriods. Epidemiologic and treatment data using sepsis bundles were collected. Cortisol level at To, the response to the corticotropin test (250 micrograms) and the initial lactate were measured during the first 24 h. All patients received hydrocortisone 300 mg/day until the result of the corticotropin test and it was suspended when delta cortisol was greater or equal to 9 mcg/dl. The prognostic value of the adrenal response to corticotropin was studied comparing data between survivors and non survivors by Student's t test for continous variables and X 2 or Fisher exact test for categorical variables. ROC curve and Youden's index were used to select the cortisol level at To that better discriminated mortality. INTRODUCTION. Inflammatory markers have been used in the diagnosis and follow-up of VAP, but their potential role in predicting the risk to develop VAP is unknown. This capacity has been proved in the case of community-acquired pneumonia. We prospectively assessed the evolution of inflammatory markers in mechanically-ventilated (MV) patients and their predictive and diagnostic role for VAP.

Sequential measurements of C-reactive protein (CRP), procalcitonin (PCT) and cytokines (IL-1, IL-6, IL-8, IL-10 y TNF-alpha) were done in 44 MV patients. Exclusion criteria were active infection at ICU admission and subsequent extrapulmonary infection.

Patients were divided into three categories: patients without VAP suspicion (n 24), patients with clinical suspicion of VAP but without microbiologycal confirmation (n 11) and patients with microbiologically-confirmed VAP (n 9). At admission, demographics, severity scores, and clinical and standard laboratory values were similar in the 3 groups, but serum levels of IL-6 and IL8 were higher in patients with confirmed VAP (526 ± 221 for IL-6 and 68 ± 38 for IL-8 pg/mL) compared with those without suspicion or non-confirmed suspicion of VAP (174 ± 39 and 185 ± 55, respectively for IL-6, p = 0.041, and 4 ± 2 and 15 ± 3, respectively for IL-8, p = 0.022). At onset of VAP, serum levels of IL-6 (1232 ± 343 pg/mL) and IL-8 (52 ± 11 pg/mL) were also higher in patients with confirmed VAP, compared with those without suspicion or non-confirmed suspicion of VAP (151 ± 44 and 364 ± 147, respectively for IL-6, and 12 ± 2 and 34 ± 4, respectively for IL-8, p \ 0.001 both). Serum levels of CRP and PCT, as well as CPIS and SOFA Score, were also highest in patients with confirmed VAP. The BAL levels of cytokines were similar between confirmed and non-confirmed VAP. However there is a relative paucity in underlying evidence for its use and this study aims to investigate EVLW as a prognostic marker and assess its relationship to other clinical and pulmonary parameters.

A prospective observational one-centre cohort-study in a mixed-bed ICU of a university hospital. 38 consecutive patients with SIRS and circulatory failure. Outcome parameters were time in ventilator, ICU-length of stay, ICU-mortality and 6 month-mortality.

The clinical parameters investigated were PF-ratio, PEEP, APACHE-and SOFA-score. CI, ITBVI and P-lactate were also analysed.

RESULTS. There were no differences between survivors and non-survivors in EVLWIbaseline and EVLWImax. EVLWI did not correlate significantly to the clinical parameters. When divided into subgroups according to EVLWImax, a low and a very high EVLWImax were associated with high mortality, whereas the middle group had a lower mortality. When further analysed, the non-survivors with low EVLWImax were significantly older than the rest of the study material. For patients \ 70 yo, EVLW was increased in the non-survivor group. Regardless of age, all patients with EVLW [ 20 did not survive.

CONCLUSION. EVLWI can be used as a prognostic marker in younger patients with SIRS/ sepsis and values above 20 mL kg-1 indicates a bad prognosis. In patients with advanced age, however, it should be interpreted more carefully. INTRODUCTION. Hemodynamic management,in surgical patients,is associated sometimes,to progressively decreasing of Hb oxygen saturation(SaO2Hb),because of the stagnant ischemia (surgical stress),or hemodylution,following the hyper volemic repletion [2] .Our study intended to observe the local O2 metabolism alteration [5] ,as a source for oxidative stress(OxS)production,and to verify the effects of antioxidant therapy.The study was approved by the Ethical Committee of UMF Craiova METHODS. a prospective study realized upon 70 patients from Plastic and Reparatory Surgery (Emergency Hospital Craiova) who needed of reparatory interventions to the distal part of the limbs:duration of 4 ± 1 h.Patients were divided into three groups A:good healing of wound (40cases), B:bad or nonhealing (30 cases),and control (10 cases),without surgical interventions.We noted the evolution of the local morpho-clinical aspects induced by the stagnant ischemia,after the cuff inflation at the site of intervention.We prelevated the samples of blood at (T1), preoperative at 48 h postoperative (T2) and to the outcome (T3), to determine the following values: lactate, C-reactive protein (CRP), procalcytonin (PCT), SaO2Hb [3] , oxidative stress index (OSI) and total antioxidant potential (TAOP CONCLUSION. Tissular exposure to the oxidative status induces a variability of local cellular reactions, representing the kind of their adaptability. To recognize the local mechanism of OxS production could be used to verify the adequate therapy and to predict the quality of wound healing. INTRODUCTION. Sepsis and specially patients with septic shock, have a high morbimortality, because of its quick evolution to multiorgan dysfunction syndrome (MODS). Acute renal failure (ARF) is common and continuous renal replacement therapies (CRRT) play an important role in the clinical management of septic patients. Nowadays optimal CRRT in criticallly ill patients are directed to achieve depuration doses [ 35 ml/kg/h (high volume hemofiltration-HVHF). The timing of these techniques is still controversial but many authors suggest that an early initiation of CRRT in septic patients is related to a better outcome. The objective of our study was to analyze the impact of an early CRRT strategy in patients with septic shock and ARF.

We included 32 patients with septic shock and ARF admitted to an intensive care unit (ICU) in a prospective and observational study. CRTT with HVHF was initiated in all patients. Two different groups were studied attending to the time of HVHF initiation (based on RIFLE score). Group I: HVHF initiated in early stages of RIFLE score (Risk and Injury). Group II: HVHF initiated in Failure stage of RIFLE score.

Group I: 9 patients were included (n = 9); mean age 58 ± 16 years, 5 were males, mean APACHEII 24 ± 8. 90% of these patients were on mechanical ventilation (MV) at the beggining of HVHF, 22% had disseminated intravascular coagulation (DIC) and 22% had liver failure. Mean ultrafiltrate dose was 43,5 ± 6 ml/kg/h.

Group II: 23 patients were included (n = 23); mean age 61 ± 13 years, 17 were males, mean APACHE II 29 ± 9. 100% of these patients were on MV, 61% had DIC and 39% liver failure. Mean ultrafiltrate dose was 43,7 ± 6 ml/kg/h.

We did not observe adverse effects related to HVHF. 28 day survival was 89% in group I and 30% in group II with an important statistical significance (p \ 0,004). Days on HVHF were 6 ± 3 days in group I and 7 ± 6 days in group II.

CONCLUSION. Early initiation of HVHF in septic shock patients with ARF seems related to a better survival in our study.

We need further and much larger studies to confirm our results. INTRODUCTION. Some patients with cardiogenic shock present during their evolution a systemic inflammatory response syndrome (SIRS) that can be interpreted erroneously as septic shock.In this study we attempt to establish the usefulness of procalcitonin in those patients with cardiogenic shock and associated SIRS with suspicion of sepsis.

A group of patients with cardiogenic shock and suspicion of sepsis were studied prospectively.We compare patients with cardiogenic shock and proved associated sepsis(G1) with those that they did not develop it(G2).The diagnostic of sepsis was established according sepsis criterions defined by ACCP and by microbiologic results.We analyzed :age,sex,-APACHE II and SAPS index,procalcitonina (PCT)and C-reactivates protein(PCR) levels,leukocyte count,lactate levels,type of shock according to hemodynamic parameters,focus of sepsis, isolated germens, mechanical ventilation days,ICU days and mortality.The Fisher Test and Spearman correlation was used for the study of the variables.A p \ 0,05 was considered statistically significant.

A total of 32(100%)patient with cardiogenic shock and suspicion of sepsis were studied,11(34%) had associated sepsis confirmed with positive cultures(G1),while 21 (66%)did not present it(G2).The PCT test was positive for bacterial infection in 91% (10) INTRODUCTION. Off-pump coronary artery bypass grafting (OPCAB) can be accompanied by hemodynamic and respiratory disorders and postoperative pain, requiring adequate monitoring and treatment. Epidural anesthesia (EA) using continuous infusion of local anesthetics combined with fentanyl can provide effective analgesia and reduce the number of perioperative complications in thoracic surgery. However, the role of EA as a component of multimodal patient controlled analgesia (PCA) in OPCAB still remains disputable. Thus, the main goal of our study was to assess the efficacy of high thoracic EA followed by postoperative patient-controlled epidural analgesia with infusion of ropivacaine/fentanyl mixture in OPCAB patients.

We enrolled 35 patients, who underwent elective OPCAB during anesthesia with propofol and fentanyl, in an ongoing prospective study. The patients were randomized into three groups with hemodynamics guided by PiCCOplus monitor (Pulsion Medical Systems, Germany) and postoperative pain control aiming at a visual analog scale (VAS) score in rest \ 30 mm. All groups received lornoxicame 8 mg IV before OPCAB and every 8 h postoperatively. The control group (n = 11) received fentanyl postoperatively at an initial infusion rate of 0.7 mcg/kg/h IV. The epidural infusion (EI) group (n = 13), after placing the epidural catheter at Th2-Th4 level before OPCAB, received EA with ropivacaine 0.75% 1 mg/kg and fentalyl 100 mcg followed by postoperative continuous EI of ropivacaine 0.2% and fentalyl 2 mcg/ml at an infusion rate of 5-10 ml/h. The epidural infusion/PCA (EI/PCA) group (n = 11) received the same regimen of EA followed by EI combined with PCA using a programmable infusion pump (Graseby 3300, UK) with bolus 1 ml and lock-out interval 12 min. The measurements included hemodynamics, blood gases, VAS, and drug consumption throughout 24 h after OPCAB. Data were assessed using 

Oral Presentations Biomarkers in sepsis

CYTOSCORE'' PREDICTS SURVIVAL IN SEVERE SEPSIS

We have previously demonstrated that a distinct anti-inflammatory pattern of cytokine gene expression is associated with increased mortality in severe sepsis (1,2). Based on an integral biological measure of a patients' immunological response to an infective insult, we suggest that a novel scoring system, termed a ''cytoscore'', is predictive of ICU mortality. REFERENCE(S). 1. Annane D, Bellissant E, Cavaillon JM: Septic shock

The pathogenesis of vasodilatory shock

Role of vasopressin in the management of septic shock

REFERENCE(S). 1. Germain DP. Ehlers-Danlos syndrome type IV

Anaesthetic and intensive care management of a patient with Ehlers-Danlos type IV syndrome after laparotomy

Anaesthetic implications of vascular type Ehlers-Danlos syndrome

Clinical and genetic features of Ehlers-Danlos syndrome type IV, the vascular type

Hypothermia for neuroprotection after cardiac arrest: Systematic review and individual patient data meta-analysis

Outcome from prehospital cardiac arrest in

Mild therapeutic hypothermia to improve the neurologic outcome after cardiac arrest

Treatment of comatose survivors of out-of-hospital cardiac arrest with induced hypothermia

Therapeutic hypothermia after cardiac arrest: a survey of practice in intensive care units in the United Kingdom

Diagnosis of metabolic acid-base disturbances in critically ill patients

Modern quantitative acid-base chemistry

Clinical review: Reunification of acid-base physiology

CRITICALLY ILL PATIENTS

Fialkow Critical Care, Hospital de Clínicas de Porto Alegre, Porto Alegre, Brazil REFERENCE(S). 1. Gowardman JR. Fatal metformin induced lactic acidosis: case report

Biguanide-associated lactic acidosis. Case report and review of the literature

Metformin in noninsulin-dependent diabetes mellitus

Nephrology Department, 2 Intensive Care Unit, 3 Intensive Care Unit

Leipzig, 3 Department of Anesthesiology and Intensive Care

Vakamudi Anesthesiology and Critical care

Kuntoji 1 , A. Elangovan 1 , S. Chatterjee 1 , H. Janakiraman 2 1 Critical Care Unit, 2 Nephrology, Manipal Hospital

Acute renal failure after cardiac surgery: evaluation of the RIFLE classification

Costs and outcomes of acute kidney injury (AKI) following cardiac surgery

Correlation (r) between BUNs, CG, MDRD, MDRD2 and CLcr was 0the 3 GFR estimate methods (CG, MDRD, MDRD2) Bland-Altman analysis showed a precision (standard deviation) of 53/57/55 ml/m/1.73m2 related to CLcr, and 40/42/ 47 ml/m/1.73m2 related to CLcr_bun; The mean difference (bias) ranged from -10 to -27 ml/m/1.73 m2 (CLcr) and from 9 to 26 ml/m/1

The 3 GFR estimate methods (CG, MDRD, and MDRD2) are moderately correlated to 2 reference methods, namely the pairs CG/CLcr_bun and MDRD2/CLcr_bun. Bias, precision and agreement limits are far from ideal, but the performance seemed better with MDRD2 and CLcr_bun. Generally MDRD2 formula appears to be the best GFR estimate for critical septic P

Our aim was the evaluation of clinical and microbiological features of patients developing ICU-acquired infection in our Surgical Intensive Care Unit (SICU) and the identification of predictors of infection acquisition and infection-related mortality

Prospective study of all patients hospitalized in our SICU from

Infections were: bloodstream (36.1%), ventilator associated pneumonia (VAP: 25.3%; 20.3/1000 ventilator-days), surgical site (18.7%), central venous catheter (10.4%; 7.1/1000 central venous catheter-days), and urinary tract infection (9.5%; 4.6/1000 urinary catheter-days)

Infections are an important problem in SICUs due to high rate, multiresistance, morbidity, and mortality. Recognition of clinical and microbiological features of these patients and identification of the risk factors for infection development and infectionrelated mortality are essential for prevention and treatment

Lisboa, 3 Intensive Care Unit, Hospital Garcia de Horta, Almada, Portugal INTRODUCTION. Sepsis, in particular nosocomial sepsis, is an increasingly common cause of morbidity and mortality, particularly among critically ill patients. The speed and approthe presence, severity and clinical course of sepsis. Many studies have shown that C-Reactive Protein (CRP) is a good marker of infection, however there are some controversies about its use as a prognostic marker in sepsis

During a 14 month period, we prospectively included all the patients admitted to an Intensive Care Unit who were C 18 years old and stayed for at least 48 h (N = 260). Patients were categorised for sepsis according to the ACCP/SCCM Consensus Conference criteria. Acute Physiology and Chronic Health Evaluation II (APACHE II)

In the subgroup of patients with documented sepsis (N = 76) there were also no differences between survivors and nonsurvivors in what regards CRP levels, WBC, temperature and APACHE II score. Only the SAPS and SOFA score were significantly higher in nonsurvivors. REFERENCE(S). 1. Lortholary O

Survival of Acinetobacter on three clinically related inanimate surfaces

Surveillance cultures and duration of carriage of multidrug-resistant Acinetobacter baumannii

Vijeta kamath and our patients. 0998 PNEUMONIA DOES NOT AFFECT MORTALITY

Intensive Care Unite, CICRET-Rosary Clinic-IPS U de A, 2 ICU, UPB, Medellin, Colombia REFERENCE(S). 1. ICM

Clinical and economic implact of introducing a BMS

Bahloul Service de réanimation, Centre hospitalier général, Salon de provence, France INTRODUCTION. Laboratory abnormalities are frequent, but non-specific

The study included 49 patients admitted to the ICU during a four-years period. Retrospective chart review recorded admission's value of platelets and WBC numbers, pH, CO2 arterial pressure, bicarbonates, cratinin, ASAT, ALAT, glucose and proteins serum levels. For each of the criteria studied, the mortality rate was assessed and compared to the overall mortality

The overall mortality rate was 34.7% in this study

IU/L) serum level, hypoprotidemia (\ 60 g/L), leucocytosis ([ 12 000/mm3) and acidemia (pH \ 7.36) are the most common abnormalities. These were found in 36, 24, 19, 21, 18 and 18 patients respectively. However, bicarbonates serum level under 15 mmol/L was found only in 7 patients. ASAT serum level over 120 IU/L was found in eight patients. The mortality rates for each of these abnormalities were: thrombocytopenia (63.8%), increased creatinin (37.5%), ALAT (62.5%) and ASAT (52.6%) serum level

No differences were observed in hemodynamic data. Pulmonary artery catheter showed no differences except for systemic vascular resistance index (SVRI) at 24 h of admission (1814.6 ± 418.6 vs. 1242.3 ± 214, p = 0.02). Major doses of norepinephrine were used in non-survivors patients

Early StO2 determination by NIRS in the forearm region is a predictor of outcome, outperforming conventional hemodynamic data

A MORTALITY PREDICTOR IN SEPTIC PATIENTS

Early goal-directed therapy in the treatment of severe sepsis and septic shock

Tissue oxygen saturation measured by near-infrared spectroscopy as a non-invasive predictor of low central venous oxygen saturation: a useful parameter in our resuscitation algorithm? (accepted for presentation at

Teboul Medical Intensive Care

Vincent Department of Intensive Care

ScvO2 (%) Non-survivors 64

Sublingual capnometry could be used to identify and further resuscitate patients with persistent microcirculatory alterations

Jakob 1 1 Intensive care medicine, 2 Anesthesiology, Bern University Hospital -Inselspital and University of Bern, 3010 Bern, Switzerland oxygen content and 2) to test afterwards the in-vitro vascular reactivity of gutand liver-supplying vessels to vasoconstricting and vasodilating drugs

At the end of the animal experiment, vascular samples from hepatic (HA) and superior mesenteric arteries (SMA) were analyzed by the tissue bath method. Contractility was assessed using norepinephrine (dose-response curve), and relaxation using adenosine and sodium nitroprusside (percent of peak contraction, baseline 100 ± 0%). Dose-response curves were constructed for norepinephrine, sodium nitroprusside (SNP) and adenosine. Variables were compared using one-way and repeated measures ANOVA

Cardiac output decreased in CT from 77 ± 14 to 44 ± 12 ml/kg/min, increased in HH from 75 ± 15 to 113 ± 26 ml/kg/min, and did not change in controls (time-group interaction

In-vitro norepinephrine maximal contraction was similar in all groups. SNPinduced relaxation was 64 ± 26%, 61 ± 30% and 47 ± 23% in CT, HH and C, respectively, in the hepatic artery (p = 0.002) and 69 ± 19%, 62 ± 22% and 51 ± 32% in CT, HH and C, respectively

CT, HH and C, respectively (hepatic artery, NS), and 24 ± 23%, 14 ± 19% and 8 ± 15% in CT, HH and C, respectively (SMA

Both CT and HH were associated with increased fractional hepatic arterial perfusion, whereas for SMA, this was only the case in CT. In SMA, maximal relaxation in response to both SNP and adenosine was increased, whereas in the hepatic artery this was only the case for SNP. Neither CT nor HH modified the maximal contraction to norepinephine. Whether the in-vitro findings have implications for the administration of vasodilating drugs in vivo

Supported by grant 3200BO/102268 from the Swiss National Fund and from Novartis grant

Noninvasive ventilation reduces mortality in acute respiratory failure following lung resection

Treatment and outcomes of oesophageal perforation in a tertiary referral centre

Boerhaave's syndrome: a review of management and outcome

WHAT IS THE ADEQUATE TIMING FOR EXTUBATION MAXILOFACIAL SURGERY? REFERENCE(S). 1. Jansen JP et al. Postpneumonectomy syndrome in adulthood. Surgical correction using an expandable prosthesis

Postpneumonectomy syndrome: Diagnosis, management and results

In univariate analysis, TC, CURB-65 and APACHE II, but not the presence of AI, were predictors of death. The discriminative ability of TC [Area under ROC curve = 0.77(95%CI, 0.65-0.90)] for in-hospital mortality was better than those of usually employed severity predictors as APACHE II, CURB-65, Ddimer or C-reactive protein. In ROC curve analysis, the best cutoff for TC was 25.7mcg/dL. CONCLUSION. Baseline TC levels are better predictors of severity and outcome in severe CAP than routinely measured laboratory parameters or scores as APACHE II

SAPS II 50 ± 18, duration of mechanical ventilation 23 ± 22d, length of ICU stay 30 ± 28d, ICU-mortality 30%. Rate of patients with ICU-acquired infection was significantly higher during the 96 h following remifentanil withdrawal than the 96 h preceding remifentanil withdrawal (23 vs. 8%, p \ 0.001), as well as rate of patients with ventilatorassociated pneumonia (10 vs. 3%, p = 0.004) and ICU-acquired bacteremia (6 vs. 1.5%, p = 0.012). Incidence rate of ICU-acquired infection was significantly higher during the 96 h following remifentanil withdrawal than the 96 h preceding remifentanil withdrawal (94 vs. 29/1000 ICU-days, p \ 0.001), including incidence rates of ventilator-associated pneumonia

In addition, incidence rate of ICU-acquired infection was significantly higher during the 96 h following remifentanil withdrawal than all ICU stay (94 vs. 40/1000 ICU-days

Remifentanil withdrawal is associated with higher rates of ICU-acquired infections, including ventilator-associated pneumonia and ICU-acquired bacteremia

Pulmonary coagulopathy is a hallmark of pneumonia and is important for its pathogenesis

We investigated effects of local administration through nebulization of APC and AT in male Sprague-Dawley rats with Streptococcus pneumoniae (SP)-or Pseudomonas aeruginosa (PA)-pneumonia. Rats were intratracheally challenged with either S. pneumoniae or P. aeruginosa, and randomized to local treatment with normal saline, recombinant human APC (5 mg/kg) 30 minutes before induction of pneumonia and was repeated every six h. Human plasma-derived AT (500 IU/kg) was repeated after 24 h because of its longer pharmaceutical half life. Non-infected rats served as controls. Rats were sacrificed at 16 h (PA-pneumonia) or 40 h (SP pneumonia) after the bacterial challenge, for bronchoalveolar lavage. Data are presented as means (SD)

While nebulized APC or AT had no systemic effects, they attenuated the local rise in TATc-levels

Bacterial outgrowth of SP was attenuated in after AT treatment (P \ .05) but not after APC treatment

Local treatment with APC or AT impairs pulmonary activation of coagulation in pneumonia; nebulized AT attenuates outgrowth of SP

Bozza Intensive Care, Instituto De Pesquisa Clínica Evandro Chagas, Rio De Janeiro, Brazil INTRODUCTION. HAART led to a great decline in lethality of HIV

We conducted an observational study of all HIV infected patients admitted on our ICU between 07/01/2006 and 12/31/2007, collecting data regarding age, gender, reason for admission, time since AIDS defining condition and CD4 count, besides length of ICU stay, opportunistic infection at admission and sepsis during ICU stay. Clinical severity was evaluated by expanded SAPS

Multivariate analysis examined organ dysfunctions independently related with length of stay and lethality

CD4 count was lower than 100/mm3 in 71% of patients. Respiratory (31%) and neurological (22%) diseases, besides sepsis (23%), were the most frequent reasons for admission. Many admissions were motivated by non-infectious causes (27%), mainly metabolic and cardiovascular. Higher SAPS II and SOFA scores were highly associated with death (p \ 0.0001). These scores had good performance in predicting death, as analyzed by area under ROC curves: SAPS

One quarter of ICU admissions were motivated by non-infectious causes. SOFA and expanded SAPS II were useful tools for prognosis evaluation in this population. Sepsis and combined renal and hematological dysfunctions were associated with higher lethality and respiratory or cardiovascular dysfunctions were associated with a prolonged ICU stay

Time to Positivity as a Novel Predictor of Outcome in ICU Patients with

REFERENCE(S). 1. Davidson et al. Clinical practice guidelines for support of the family in the patient-centred intensive care unit: American College of Critical Care Medicine Task Force

S. Tafelski*, I. Nachtigall, C. Spies Anaesthesiology and Intensive Care Medicine, Charite Universitätsmedizin Berlin, Berlin, Germany

based on the TNM classification system consisting of 4 items: P for predisposition, I for infection, R for response, O for Organdysfunction. This study evaluates the practical use of PIRO in a University-hospital setting.METHODS. This observational study was approved by the local Ethics Review Board and the data safety authorities. 474 patients were included who experienced treatment on 5 ICUs (1 cardiosurgical, 1 neurosurgical, 2 interdisciplinary, 1 intermediate care) for at least 2 consecutive days. Data were analysed retrospectively for PIRO -criteria: Predisposition [age, gender, immune suppression] , infection [SIRS of non-infectious origin, low-risk, high-risk, infection with special pathogens], response [adapted, anergia, hypergia] , organdysfunction [using SOFA-Score]. Univarate statistic was calculated using Chi2-test (alpha = 0.05).RESULTS. All patients were classified using the PIRO criteria. Total mortality was 9.5% [45/474]. Significant differences in mortality were demonstrated between the categories ''infection'' (I0 2.9%, I1 3.8%, I2 14.1%, I3 24.4%; p \ 0.001) ''response'' (R0 2.5%, R1 7.5%, R2 35.1%; p \ 0.001) and ''organ dysfunction'' (O0 3%, O1 3.2%, O2 25%; p \ 0.001). Mortality in ''predisposition'' groups did not differ significantly in our population.CONCLUSION. Expected differences in mortality in the PIRO categories were reproduced except for predisposing factors. In concordance to other publications the PIRO system reflects qualitative differences in individual courses of infections. Therefore it may lead to a better distinction of patients with sepsis and could help to differentiate responders for therapeutic options. More analysis for concrete parameters is necessary for assessing factors to establish a consented system for research projects.Poster Sessions Acute kidney injury: 0967-0980 0967

INTRODUCTION. Acute Kidney Injury (AKI) is an independent predictor of mortality in the critically ill (1) . We conducted this prospective study to look at the incidence of and outcomes from AKI in our multidisciplinary intensive care unit.

We prospectively studied all patients with acute renal failure admitted to our unit during a 4 month period between August and November 2007. Patients were classified according to the RIFLE criteria (2) . The APACHE II score was used to assess baseline severity of illness. Other data collected included ventilation days, ICU days, days on renal replacement therapy (RRT) and ICU survival.RESULTS. There were 75 patients who had AKI by the RIFLE criteria out of 502 admissions during the study period (14.9%). Fourteen patients were at ''Risk'', 23 had ''Injury'' and 38 were in ''Failure'' by the RIFLE classification of AKI. Twenty one patients underwent RRT. Sixteen patients underwent Slow Low Efficiency Daily Dialysis (SLEDD) for 6 to 8 h at a time, 4 patients underwent conventional haemodialysis for 4 h at a time and one patient underwent continuous veno-venous haemodiafiltration. Fifty two (69.3%) patients survived. Out of the 21 patients who underwent RRT, there were 15 (71.4%) survivors. Mean duration of RRT was 7.9 ± 5.4 days. Mean APACHE II score of all patients was 20.3 ± 8.9. Eighteen patients (24%) required mechanical ventilation; the mean duration of mechanical ventilation in these patients was 6.7 ± 5.9 days. On multiple logistic regression analysis, increasing APACHE II score was related to increased mortality (odds ratio: 1.16, 95% confidence interval: 1.07 to 1.27 p \ 0.0005). Increased severity of AKI by the RIFLE criteria was not associated with increased mortality (odds ratio: 0.86, 95% confidence interval: 0.42-1.76, p = 0.68).CONCLUSION. AKI was common in our multidisciplinary intensive care unit, and carried a significant mortality. SLEDD was the most commonly employed modality of RRT. Increasing APACHE II score was related to increased mortality. However, worsening kidney injury by the RIFLE criteria did not predict increased mortality. INTRODUCTION. Acute kidney injury (AKI) is a recognized complication of cardiac surgery. For this group the mortality range is 10-83% depending on the different series, using different criteria for diagnosis and classification of AKI. Mortality exceeds 50% when AKI is associated MODS or dialysis. The Acute Dialysis Quality Initiative (ADQI) proposed consensus-based classification criteria for AKI based of incremental changes in maximum serum creatinine or hourly urine output. The AKI classification is defined by the RIFLE acronym: Risk, Injury, Failure (severity stages), Loss and End-stage Kidney disease (prognostic stages). This study analyzed the incidence of AKI in a cohort of patients following cardiac surgery, the RIFLE categories, the association with mortality, length of stay and DRG-based costs.

We have carried out a retrospective cohort study of 1552 patients undergoing cardiac surgery with cardiopulmonary bypass, admitted to our ICU in the period from 2000 to 2006. We analysed the incidence of AKI, in base of RIFLE classification, lenght of stay, mortality and associated costs. Patients with preoperative dialysis, cardiac surgery without cardiopulmonary bypass and the RIFLE categories loss and end-stage renal disease were excluded. The difference between the highest creatinine in ICU and baseline creatinine defined AKI ICU. We use score APACHE II to determine the severity. Cuantitative variables were studied with T Student, cualitative variables with Chi-square test and logistic regression for statistical significants results. Costs were calculated using DRGs.

We analyzed 1552 patients finding AKI at 20.1% of them (11.5% risk, 34 .5% injury and 4.1% failure). The APACHE II score was higher in the patients with AKI. We are observed mortality of 10%, (3, 8% in ICU and 6, 2% hospital mortality) . The mortality increased correlated with the severity of AKI determined by RIFLE stages just as the length of stay (LOS). Kaplan-Meier survival curves according to RIFLE classification were done. INTRODUCTION. The cytochrome P450 3A enzyme system (CYP3A) plays a significant role in drug metabolism. Studies suggest that chronic renal failure impairs hepatic CYP3A drug metabolism, an effect not reversed by haemodialysis. Midazolam may be used as an invivo probe of CYP3A activity. The impact of acute kidney injury (AKI) on hepatic drug metabolism is untested. This pilot study demonstrates an effect of AKI on midazolam metabolism in critically ill patients.METHODS. Patients admitted to the adult intensive care unit were excluded if they had received a benzodiazepine in the previous 24 h, were diagnosed with acute or chronic hepatic failure, were pregnant or were receiving major CYP3A inhibitors. 1 mg midazolam (i.v.) was given time 0 and serum collected at 4 h. Serum midazolam concentration was determined by mass spectrometry (HPLC/MS/MS). AKI was defined by an acute change in the eGFR prior to entry into the study. The estimated glomerular filtration rate (eGFR) was determined using the modified diet in renal disease (MDRD) equation.RESULTS. 16 patients (10 male). Age 75(53-87). Median APACHE 18. Patients were separated into two groups; normal renal function, (8 pts) or those with AKI. Median (IQR) 4 hr midazolam concentrations were 2.86 (1.5-3.5 ) ng/ml and 3.97 (3.18-5.07 ) ng/ml respectively and differed significantly (t-test) p = 0.028.CONCLUSION. This pilot study suggests that in the critically ill patient, AKI causes a reduction in the metabolism of midazolam by the CYP3A enzymes. A bigger, more comprehensive study needs to take place to confirm these findings and also investigate the relative importance of AKI on CYP3A activity in critically ill patients. J. P. Baptista* 1 , H. Sá 2 , E. Sousa 1 , J. Pimentel 1 1 Intensive Care Medicine, 2 Nephrology, Coimbra University Hospital, Coimbra, PortugalINTRODUCTION. An estimate of Glomerular Filtration Rate (GFR) has not been validated in critical patient, and a ''gold standard'' method is not always available in the ICU. The aim of this study was to compare between 5 evaluation parameters of renal function in critical septic patients: Creatinine (Cr), Blood Urea Nitrogen (BUN), Cockroft-Gault (CG), 4-variable (MDRD) and 6-variable (MDRD2) ''Modification of Diet in Renal Disease'' formula.

This study was carried out on 72 critical septic patients (P), 40% with septic shock (28P), 54 men (75%), average age 51 ± 21 yrs, average APACHE II and SAPS II of 16. 3 

A double blind randomised pilot study was conducted in a mixed surgical/ medical Critical Care Unit. Within 24 h of presenting in ADCS, patients were commenced on a 24 h infusion of levosimendan or enoximone. 12 patients were recruited, 6 per group. Cardiovascular parameters were measured at time 0, 30 mins, 1, 4, 6 and 24 h, along with baseline demographics and survival outcomes.RESULTS. The groups were well matched for age, sex and illness severity. In each group, 5 patients completed the 24 h infusion, with 1 patient in each group withdrawn due to clinical deterioration suspected to be due to the trial drug. In both groups, the mean increase of Cardiac Index at 6 h was 37% (Enoximone) and 38% (Levosimendan).There were no significant differences between the groups for other cardiovascular parameters. Similarly, there were no survival differences (Fig 1) .CONCLUSION. Levosimendan could be considered as an alternative to standard pharmacological agents in ADCS. It appears to be as well tolerated as Enoximone. 

This prospective, open, controlled study was conducted in post cardiac surgery patients receiving 1:1 nursing care, in seven Australian tertiary ICUs. Following admission to the ICU, patients were randomized by centralized telephone system to receive either: clinical guidance using Navigator or standard hemodynamic care. Patients remained connected to Navigator until CO monitoring was ceased by the medical team. CO was measured continuously (Edwards, Vigilance PAC) in both groups. The Navigator system provided guidance to staff for patients in the treatment arm in the intervention period but not the control group. The primary endpoint of the study was a measure of the difference between the targeted values and the actual values for CO and MAP averaged over the study period -the Average Standardized Distance (ASD). The lower the ASD, the better the therapeutic control.RESULTS. 112 patients were randomized. Primary outcome data was available for 105 patients (57 Nav; 48 Control). The mean ASD for the treatment group was 1.73 and the control group 1.95, an improvement in control in the treatment group of 11.5% (95% CI -4.5% to +27.5%, p = 0.157). The lowest ASD recorded in the total population was 0.89. The treatment effect was stronger in the early period of monitoring: ASD for the first 15 h was 1.65 in the treatment arm and 1.93 in the control arm, an improvement of 14.3% (p = 0.064) and was significant when adjusted for differences between centers 16.3% (p = 0.049). In the first 15 h, the proportion of time CO and MAP readings were in the target range was 39.5% in the treatment arm and 32.5% in the control arm (improvement: 21.5%, p = 0.064). SOFA scores, atrial fibrillation, and serious adverse event rates were not different between the groups. There were no device failures. There was one death (treatment arm).

We have demonstrated in a multi-centre post-operative cardiac patient group that continuous automatic therapeutic guidance is possible and Navigator-guided hemodynamic control is the same or better than standard care, with post hoc evidence of superiority. Navigator-guided hemodynamic control was safe. CONCLUSION. Abnormal laboratory findings are frequent during community acquired bacteremia. Thrombocytopenia, leucocytosis and elevated ALAT serum levels at admission were associated with higher mortality rates in adult patients with community acquired bacteremia. Acute renal injury did not impact the prognosis of our patients. CONCLUSION. Patients with EVD were less than 2% of the total admissions in the ICU. They were younger than the average, with predominance of the medical aetiology. They had long stay, high instrumentation and high nosocomial IR.

GRANT ACKNOWLEDGEMENT. Aventis. INTRODUCTION. Microcirculatory derangements drive to situations of dysoxia and to acute organ dysfunction in critical ill patients. The development of new techniques to determine tissue oxygenation, such as the near-infrared spectroscopy (NIRS) is a non invasive way to infer the grade of alteration in microcirculation. Our hypothesis is to determine whether NIRS technique allows to recognize alterations in the muscular tissue saturation of oxygen (StO2) in the critical care patients. Our objective is to determine the utility of NIRS in the evaluation of StO2 in critical care settings.

Observational, prospective and controlled study conducted in a twenty-two bed, university hospital Intensive Care Unit (ICU). Reference values were obtained from StO2 measurements by NIRS in healthy volunteers in the forearm zone. These measurements were compared with those obtained in critical patients. Basal measurements at admission in ICU and during the first 48 h were obtained, as well as conventional hemodynamic indexes. Mortality was considered as the main outcome. Continuous variables are expressed as mean ± sd. Student's t-test was used to compare data. Statistical significance was defined as p \ 0.05.

The near infrared spectroscopy (NIRS) assesses the haemoglobin saturation in the tissue (StO 2 ). Induction of transient ischemia followed by hyperaemia (forearm occlusion and release) provides additional information on the ability of microvessels to be recruited through the analysis of the StO 2 re-ascension slope. Activated protein C (APC) is an anti-thrombotic and anti-inflammatory agent used as an adjuvant therapy of patients with septic shock. One previous study showed an improved sublingual microperfusion with APC (1).OBJECTIVE. To examine if APC was able to affect microcirculation assessed by NIRS parameters in septic shock patients.

We included 11 patients with septic shock and at least two organs failures considered for administration of APC and equipped with a PiCCO'' monitoring system. The thenar muscle StO 2 was continuously measured with Inspectra'' model 650 (Hutchinson Technology) before and during pneumatic arm cuff inflation (until StO 2 by 40% is reached) and after deflation, which allowed calculating the StO 2 re-ascension slope. Mean arterial pressure (MAP), cardiac index (CI), central venous oxygen saturation (ScvO 2 ), dose of norepinephrine (NE) and StO 2 measurements were performed before and 4 h after starting APC infusion (24 lg/kg/h). We also collected MAP, StO 2 and StO 2 re-ascension slope in 15 healthy volunteers. INTRODUCTION. Definitive diagnosis of mesenteric ischemia usually requires invasive testing, e.g., splanchnic angiography. However, angiographic findings frequently are not conclusive and do not correlate with clinical symptoms. The aim of the study was to evaluate the potential of reflectance spectrophotometry to assess non-invasively oxygenation in splanchnic mucosa (lHbO2) for the detection of clinical important mesenteric ischemia.

METHODS. Five patients presenting with hitherto unexplained abdominal pain and/or weight loss were enrolled consecutively in a pilot study. lHbO2 was assessed in the upper gastrointestinal (GIT) tract to diagnose and grade mesenteric ischemia preoperatively, as well as during and after revascularization; the values were compared to clinical, angiographic and endoscopic findings.RESULTS. In 4 patients lHbO2 was B 60% at least in some parts of the GIT before revascularization and increased to values [ 60% after surgical therapy. These results correlated well with the disappearance of clinical symptoms of mesenteric ischemia. The 5th patients did not undergo revascularization because he presented with clinically not clear cut findings and showed only in 1 of 5 measuring sites lHbO2 \ 60%.CONCLUSION. We suggest that lHbO2 values [ 60% probably represent the cut-off point of physiologic lHbO2 as values below this range were associated with clinical symptoms of mesenteric ischemia. Moreover, lHbO2 values but not endoscopic findings correlated with clinical signs of mesenteric ischemia. We suggest that reflectance spectrophotometry has the potential to serve as an appropriate additional assessment tool for diagnosing and grading intestinal ischemia, and may be useful to identify the indication of surgical revascularization.GRANT ACKNOWLEDGEMENT. Departmental funds. 

Pathophysiological effects of hyperoxia such as breath depression, acute hyperoxic lung injury, atelectasis induction or cardiac output decrease have been investigated in details many times. Effect of normobaric hyperoxia on intestinal microcirculation was intensively investigated last years with some evidence of positive effects on micovascular flow index (MFI) and functional capillary density (FCD) especially in model of I/R injury (1) . The goal of this study was to evaluate changes in FCD and MFI of sublingual and ileal mucosa induced by short-term hyperoxia and to compare changes in FCD of sublingual and intestinal microcirculation revealed by Sidestream dark-field imaging (SDF).

Microcirculatory network of sublingual mucosa and of terminal ileum was visualized via an ileostomy by SDF imaging in 8 anesthetized pigs. Four sequences of microcirculatory status from sublingual and intestinal mucosa were recorded on-line at baseline conditions, after surgical preparation and stabilization (FiO2 0,40), immediately after preceeding 15 min hyperoxia (FiO2 1,0) and 60 minutes after the end of hyperoxia(hyperoxia group, n = 4). The same measurements were performed in control group, n = 4, with constant FiO2 0,40 throughout the study. Macrohemodynamic values were recorded continuously, microcirculatory parameters were analyzed off-line.

No difference has been detected between the groups at the baseline. There was statistically significant increase (p \ 0,001) of FCD and MFI in the ileal mucosa after 15 minutes of hyperoxia in contrast to sublingual area where significant decrease in FCD was observed and MFI remained unchanged (p \ 0.01). These changes were not observed in control group. Thus, no correlation between FCD values from sublingual and ileal mucosa was detected in terms of hyperoxia. Microcirculatory effects of hyperoxia were still significant 60 minutes after the hyperoxia was finished.

We have demonstrated positive effect of short-term hyperoxia with 100% oxygen on intestinal microcirculation in healthy anesthetized mechanically ventilated pigs. These findings correspond with previous studies assuming redistribution of cardiac output into the splanchnic region and attenuation of inflammatory pathways with maintainig microcirculation patent. Positive effect of supplemental oxygen on tissue oxygen tension in healthy and anastomotic colon was described previously, this effect was not reached using aditional krystalloid fluid. Trends and values obtained from ileal mucosa does not correspond with those obtained from sublingual mucosa, where decrease in FCD is propably due to oxygen induced vasoconstriction. INTRODUCTION. The aim of this study was to assess microvascular dysfunction in trauma hemorrhagic shock (HS) patients by recording near-infrared spectroscopy (NIRS) signal during arterial occlusion tests.

After approval from our local IRB, trauma patients with HS were included in this prospective study. As soon as hemodynamic stabilization was obtained, NIRS-derived StO2 signals (InSpectraÒ, Hutchinson, USA) were recorded on the thenar eminence during 3 min brachial artery occlusion ( Figure) . The same trial was applied in ten healthy volunteers. Demographic, hemodynamic data, SAPSII and ISS scores were gathered. Data are median [IQR] are were analyzed with Mann-Whitney and Spearman tests. RESULTS. Procedure duration: 11 h (7-14) with a large fluid loading (15.5 ml/kg/h, range 7.1-28) with crystalloids (1/3 0.9% saline, 2/3 ringer lactate), comparable in both anaesthesia techniques. 3 patients required norepinephrine infusion. Hyperthermia: 1) under GA, BP, CO, ScvO2 or StO2 did not change;2) under GA + TEA, pre-hyperthermia CO tended to be higher than in GA (median TEA = 8.0 vs. GA = 6.6 L/min, p = 0.08); TEA modified the response to hyperthermia with PB decrease (p \ 0.05) without change in CO (intense vasodilatation) but with more reflex tachycardia (80 to 98 b/min, p \ 0.05). This sympathetic stimulation targeted the area unblocked by TEA (thenar eminence). The decrease of StO2 during arterial occlusion with TEA (median -0.30 to -0.45 %/sec, p \ 0.05) versus GA (-0.37 to -0.41 %/sec, NS) was faster (oxygen demand). The StO2 reperfusion slope (vascular reserve)did not differ between GA and TEA. In the TEA group, the more rapid was the drop of occlusion slope (rapid ischemia), the more rapid the reperfusion slope (vascular recruitment) (regression R = 0.67, Spearman test p = 0.04).

CONCLUSION. Before hyperthermia, TEA + GA induced hyperkinetic syndrome. Only in TEA + GA, hyperthermia induced a large vasodilatation that mainly concerns the unblocked area, as suggested by the remarkable rapid ischemia during arterial occlusion, which determined the vascular recruitment during reperfusion. The benefit of TEA associated with GA compared to GA alone remains to be demonstrated in postoperative period especially for inflammatory parameters.

S. Vamadevan*, C. Barrera-Groba Intensive Care, Royal Sussex County Hospital, Brighton, United Kingdom

Lactic Acidosis 1 and decreased lactate clearance in early critical care 2 have been demonstrated to be prognostic of increased mortality from critical illness. Fluid resuscitation that aims to avoid lactic acidosis may create a hyperchloraemic metabolic acidosis but its impact on outcome is controversial 3 .METHODS. This is a cohort analysis of a prospectively collected observational database on all-comers to a single mixed ICU. We evaluated the prognostic value (on ICU mortality) of the maximum chloride (Cl) and maximum lactate levels in the 2nd 24 hrs of their stay. CONCLUSION. An apparently mild peak lactate of just C 2 at any point in the second twenty four h of stay confers a profound survival implication. Also failure to clear serum lactate to below 2 in the first 24 h, and (albeit for some confounders) a maximal day 2 [Cl] of C 115 are associated with a significant increase in mortality. Therefore, a mild lactic acidosis and hyperchloraemic acidosis in the second twenty four h are both prognostic indicators of mortality in critically ill patients.

M. E. van Genderen*, J. van Bommel, C. Ince, D. Gommers Intensive Care, Erasmus MC, Rotterdam, Netherlands

Esophagectomy with gastric tube reconstruction is the surgical treatment for patients with esophageal cancer and is associated with high morbidity and mortality rates.Complications of these operations are leakage and stenosis of the distal anastomosis, which may be due to compromised local microvascular blood flow. We have shown that the sublingual microvascular perfusion is impaired on the first day after esophagectomy and therefore the aim of this study was to determine if this impairment also occured after Whipple pancreas resection, and if these changes could be prevented with dobutamine.

We included a total of 30 patients in this study. We randomized the 20 esophagectomy patients over 2 groups: an intervention and a control group. The intervention group received dobutamine 2-3 mcg/kg/min directly postoperative until the second day after surgery whereas the control group did not. On the day before surgery (T-1), the day of surgery (T0), and two days after surgery (T1, T2) hemodynamic data were collected and the sublingual microcirculation was analyzed using a Sidestream dark-field (SDF) imager. Similar measurements were performed in 10 patients undergoing a Whipple pancreas resection. SDF data was analyzed according to the semi-quantitative analysis described by Boerma et al.1 and expressed as Microvascular Flow Index (MFI).

The results are presented in the figure.CONCLUSION. Sublingual microvascular perfusion is significantly reduced on the first day following esophagectomy, whereas not after Whipple pancreas resection, and the administration of dobutamine prevented this decrease. INTRODUCTION. Microvascular alterations can be visualized at the bedside but quantification remains challenging. In the sublingual region, two semi-quantitative scores are used: Proportion of perfused vessels and the Mean Flow Index (MFI). We compared the scores in patients with severe sepsis.

METHODS. Using Sidestream Dark-Field imaging (Microvision Medical, Amsterdam, Netherlands) we evaluated the sublingual microcirculation in 43 patients with severe sepsis. At each evaluation, 5 sequences of 20 s each were recorded. An investigator blinded to the patient's clinical course and sequence order calculated the scores for each image, focusing on vessels less than 20 lm in diameter. Perfused vessel density was calculated as the proportion of perfused vessels crossing 3 vertical and 3 horizontal lines. For the MFI score, the image was divided into 4 quadrants and the predominant flow type in each quadrant assessed (0 = absent, 1 = intermittent, 2 = sluggish, 3 = normal); the values of these 4 quadrants were then averaged. Linear regression was used to evaluate the relationship between the scores.

We evaluated 490 video sequences. Globally there was good correlation between the scores (MFI = 0.0447 x -0.974; R 2 = 0.75, p \ 0.001)( Figure 1 ). However, there was great variability, as an MFI of 3 (normal) was associated with perfusion ranging from 70 (markedly abnormal) to 98% (normal).CONCLUSION. Although related, the 2 scores are not equivalent. Measuring the proportion of perfused vessels allows a more precise evaluation of the microcirculation. 

Nineteen consecutive CTEPH patients, in whom spiral CT excluded a distal pattern, underwent 2 right heart catheterizations: preoperatively to assess basal hemodynamics and response to iNO and postoperatively to assess outcome of PEA.RESULTS. At basal hemodynamic assessment, Cardiac Output (CO) was 4.3 ± 1.6 l/min, mean Pulmonary Artery Pressure (mPAP) 45 ± 14 mmHg and Pulmonary Vascular Resistance 826 ± 371 dynes*s*cm-5. Inhaled NO decrease mPAP to 37 ± 10 (-18%) and PVR to 707 ± 304 (-15%). Four patients (21%) were high responders to iNO and 15 patients (79%) were low responders according to a PVR decrease C or \ 200 dynes*s*cm-5, respectively. High responders had also higher basal PVR (1213 ± 222 vs. 723 ± 335; p \ 0.01) but not significantly higher PVR during NO inhalation (839 ± 167 vs. 672 ± 327) than low responders. PEA decreased mPAP to 25 ± 8 (-44%), PVR to 248 ± 103 (-63%) and increased CO to 5.4 ± 2.0 (+26%). The 4 high responders had worse PEA outcomes: 1 patient experienced severe postoperative pulmonary hypertension and right heart failure, required ECMO and finally died 4 days after the operation; in the other 3 high responders, mean postoperative PVR was significantly higher than that observed in low responders (370 ± 97 vs. 223 ± 87; p \ 0.005).CONCLUSION. In CTEPH patients, a preoperative high response to NO inhalation may suggest that a significant portion of pulmonary vascular resistances is not correctable by pulmonary endarterectomy. INTRODUCTION. Lipid emulsions are an essential part of parenteral nutrition (PN), both as a part of energy supply, and as a source of essential fatty acids. It has been shown that the fatty acid composition of cell membranes is influenced by the fatty acid profile of dietary lipids, and may therefore be responsible for modulation of immune response (1). The aim of this study was to assess the effects of a new lipid emulsion based on soybean oil, mediumchain triglycerides (MCT), olive and fish oil (SMOFlipidÒ) compared with a lipid emulsion based on olive and soybean oil (ClinOleicÒ) on the inflammatory response in postoperative intensive care unit (ICU) patients.

A prospective randomised study. After approval of the ethical committee, 44 postoperative surgical patients with an indication for PN were included in this study. Nonprotein calories were given as 60% glucose and 40% lipids. The total energy intake per day was 25 kcal/kg body weights. Sedation regimen was standardised, propofol (lipid emulsion as a carrier) was avoided. Patients were thus allocated to one of two nutritional regimens: Group A (n = 22) received SMOFlipidÒ 20% (SMOF group), group B (n = 22) a lipid emulsion based on olive and soybean oil (ClinOleicÒ 20%, control group). PN including lipids were given for 5 postoperative days. Interleucin-6-(IL-6), tumor tecrosis factor-alpha-(TNF-a), leukotriene B5 (LTB5), and soluble E-selectin-levels (sE-selectin) were measured before the start of infusion (d0), at day 2 (d2) and day 5 (d5) after the start of administration. The significance level was defined at p \ 0.05.RESULTS. There were no significant differences between the two groups in the proinflammatory response at d0 and d2, but at d5 significantly lower IL-6-(SMOF group ( CONCLUSION. The administration of SMOFlipidÒ within a parenteral nutrition regimen led at the final study day (d5) of the nutrition regimen to a significantly reduced proinflammatory response and at d2 and d5 a significantly increased anti-inflammatory response compared with a lipid emulsion based on olive and soybean oil. 

frequently a triad of lung, brain and skin dysfunction due to fat emboli. FES commonly occurs in trauma patients with long bone fractures, after intramedullary nailing of long bones, however occasionally cases in patients with pancreatitis, burns and on parental nutrition and very rarely after liposuction or dermolipectomy are reported. With increasing popularity of cosmetic surgeries, the number of patient undergoing body sculpture surgery is increased and up to 8% of the deaths are blamed to FES. FES can easily be confused with MODS due to other etiologies, hence it is essential that critical care physicians are aware of this entity.

Case report -A 46-years-old female suffering from hypertension, diabetes mellitus and pendulous abdomen, underwent dermolipectomy and abdominoplasty. On the 2nd postoperative day, she complained of shortness of breath and drowsiness.RESULTS. Her O2 saturation dropped to 69% and required intubation. Pulmonary embolism (PE) was suspected. Chest X-ray and spiral CT chest showed no evidence of PE, but bilateral basal lung collapse and consolidation. She was tachycardic (120 to 130/min), febrile (39.5°C), oliguric for 8 h and showed signs of ARDS. By day 5, her chest improved, but she was stuporous and remained in this state for almost 18 days. MRI showed periventricular scattered white matter hyperintense lesions on T2 weighted images (Fig.) . Septic work-up was negative and FES was suspected. By day 23, her neurological status started to improve; she was weaned off the ventilator and could be discharged from ICU. She made a complete neurological recovery and was discharged from the hospital on day 80.CONCLUSION. FES post-dermolipectomy is a rare but life-threatening condition. The lack of specific diagnostic test might complicate and delay the diagnosis. INTRODUCTION. Thoracic Epidural Analgesia (TEA) can be safe and effective for providing perioperative analgesia but is not without risk. Inadvertent intravenous administration of bupivacaine may induce changes to cardiac rhythm and contractility (1) . We have prospectively evaluated the safety of commercially prepared levobupivacaine (a potentially less cardiotoxic local anaesthetic) since its introduction in June 2003.

A multidisciplinary medication safety initiative (MSI) grouped was formed to review current practice in the critical care environment including; reported TEA-related adverse events, the variety of opioid and local anaesthetic presentations, clinical storage, prescribing practices, initiation of the infusion, infusion devices and staff education.

Only two distinct presentations of infusates are identified on the pre-printed prescription and available in our institution. Drug dilution errors (2001 n = 2, 2002 n = 1) have not been reported since the introduction of levobupivacaine. The adverse incident rate for epidural analgesia has fallen from 7 reports in 2001 (43% amber rating indicating significant patient impact using NPSA criteria (2) ), 6 reports in 2006 to 3 reports in 2007 (no patients experiencing any significant impact). Storage was revised with the sole opioid infusate relocated to a dedicated cupboard. A competency based programme for all clinical staff was initiated with the need for annual updates identified. All medication-related adverse events are disseminated quarterly to all staff.CONCLUSION. Identifying and managing risks related to TEA can reduce the incidence and severity of adverse events. A multi-disciplinary approach can improve documentation, storage and administration. INTRODUCTION. Despite being well studied molecules, little is known about the utilities of brain natriuretic peptides(BNP) and C-Reactive Protein (CRP) as biological markers of right cardiac overload and inflammation in thoracic surgery. The aim of this study is to find out the patterns and activity of such these molecules after large pulmonary resection (lobectomy or pneumonectomy).METHODS. Prospective and observational study including 32 consecutive patients. We measured sequentially changes in plasma Pro-BNP levels using quimioluminiscence technics (HELECSYS 2010, Hitachi. Japan) and CRP levels, before and just after pulmonary resection and also 6, 24, 48 CONCLUSION. We observed a remarkable elevation of both, proBNP and CRP since 24 h after surgery. Plasma proBNP concentrations showed the highest increase at 72 h, possibly owing to increased right ventricular wall stress as a result of an increment of pulmonary vascular resistance. An inflammatory level peak (CRP) was found 48 h after surgery. Largest studies are necessary to analyse correlation between right ventricular function and inflammation with proBNP and CRP levels in such patients. It may be a potential predictor for both, prognostic and postoperatory complications. (1) found that, in selected patients with acute respiratory failure (ARF) following lung resection, Noninvasive Ventilation (NIV) decreases the need for endotracheal mechanical ventilation (ETMV) and improves clinical outcome.

We prospectively evaluated NIV use for ARF after lung resection during a 4year period in the setting of a medical and a surgical ICU of a university hospital. We documented demographics, initial clinical characteristics and clinical outcomes. NIV failure was defined as the need for tracheal intubation.

Among 690 patients at risk of severe complications following lung resection,113 (16.3%) experienced ARF, which was initially supported by NIV in 89 (78.7%), including 59 with hypoxemic ARF (66.3%) and 30 with hypercapnic ARF (33.7%). The overall success rate of NIV was 85.3% (76/89). In-ICU mortality was 6.7% (6/89). The mortality rate following NIV failure was 46.1%. Predictive factors of NIV failure in univariate analysis were age (p = 0.046), previous cardiac comorbidities (p = 0.0075), postoperative pneumonia (p = 0.0016), admission in the surgical ICU (p = 0.034), no initial response to NIV (p \ 0.0001) and occurrence of non infectious complications (p = 0.037). Only two independent factors were significantly associated with NIV failure in multivariate analysis: cardiac comorbidities (odds ratio, 0.066; 95% confidence interval, 0.008-0.557; p = 0.015) and no initial response to NIV (odds ratio, 0.008; 95% confidence interval, 0.001-0.092; p \ 0.0001).CONCLUSION. this prospective survey confirms the feasibility and efficacy of NIV in ARF following lung resection.

N. Skampas* 1 , E. Tripodaki 2 , L. Vastardis 3 , T. Sakellaridis 3 , V. Gerovasili 2 , K. Kritikos 2 , A. Tasoulis 2 , M. Argiriou 3 , C. Charitos 3 , S. Nanas 2 1 ICU Department, 3rd IKA Hospital, 2 1st Critical Care Department, National and Kapodistrian University of Athens, 3 2nd Department of Cardiac Surgery, Evangelismos Hospital, Athens, Greece INTRODUCTION. The purpose of this study was to compare microcirculation parameters, as assessed by the non-invasive method of Near Infrared Spectroscopy (NIRS), with haemodynamic indices, evaluated by right catheterization. The peripheral microcirculation of cardiac surgery patients is post-operatively altered as a result of the extracorporeal circulation, the anaesthesia and the hypothermia they undergo during surgery. A non-invasive monitoring technique of the peripheral microcirculation would be very useful in aiding post-operative monitoring and in helping ensure adequate tissue perfusion in these patients.

We studied 10 cardiac surgery patients (8 males, 2 females, 59 ± 17 years) during the immediate postoperative period. All patients had undergone extracorporeal circulation and were treated with inotropic and/or vasopressor agents (7 norepinephrine, 6 dobutamine, 2 adrenaline, 3 levosimendan). A Swan-Ganz catheter was used to assess the haemodynamic parameters. Tissue oxygen saturation (StO2%) of the thenar and masseter muscles was obtained with the help of In Spectra Model 325 Near-Infrared Spectrometer (NIRS). The brachial artery occlusion technique was utilized to better assess the endothelial function, with the calculation of the oxygen consumption rate, the reperfusion rate and the vascular reactivity.RESULTS. The haemodynamic parameters of our patients were: MAP 73 ± 15 mmHg, CVP 8 ± 4 mmHg, PCWP 12 ± 5 mmHg, MPAP 26 ± 6 mmHg, CI 2.6 ± 0.7 L/min/m2, SVR 1161 ± 317 dyne x s/cm5, PVR 279 ± 146 dyne x s/cm5, HR 104 ± 11 bpm, Hb 10.7 ± 1.7 g/dl, lactate 3.5 ± 2.4 mg/dl. We found a significant correlation between cardiac index and reperfusion rate (r = 0.39, P = 0.016), as well as with vascular reactivity (r = -0.39, P = 0.027). CI also correlated significantly with masseter StO2 values (r = 0.64, P = 0.001). We also found a significant correlation between masseter StO2 values and SvO2 (r = 0.53, P = 0.011), as well as with ScvO2 (r = 0.52, P = 0.014).CONCLUSION. By utilizing the occlusion technique with NIRS technology we found a correlation of NIRS derived parameters and haemodynamic indices. As adequacy of tissue oxygenation and not simple maintenance of ''ideal pressures'' is the physiologic role of haemodynamic monitoring, the potential introduction of a non-invasive method to postoperative cardiac surgery monitoring could help ensure the optimization of patient management. INTRODUCTION. High mobility group box 1 (HMGB 1) is a well known proinflammatory cytokine that initially described as DNA binding protein responsible for transcriptional events and gene expression. Further investigation focused HMGB 1 as a late appearing inflammatory mediator that persist for longer duration after initiation of inflammatory response. Previous studies have shown that cardiac surgery with cardiopulmonary bypass (CPB) triggers release of proinflammatory cytokines and provokes inflammatory response. The present study was designed to investigate the changes in serum HMGB-1 levels in cardiac surgery using cardiopulmonary bypass (CPB).

Eighteen cardiac valvular disease patients were included in our study; AVR8, MVR6, MVR + CABG1, MVP3. Study protocol was approved by research committee of our university. HMGB-1 levels were measured before surgery (immediately after tracheal intubation,T1), 1 hr. after start of CPB(T2), 1 hr. after removal of CPB(T3), 1 hr. after ICU admission (T4), ICU day-2(T5), ICU day-3(T6). Simultaneously, hemodynamic variables, required doses of catecholamines, serum lactate concentrations and other laboratory data were recorded. Repeated measure ANOVA was used to analyze the changes in Tl-T6 values and values at T2-T6 were compared with T1 value (Dunnet's test).RESULTS. Serum HMGB-1 at T1 was 5.99 ± 2.4. It significantly increased at T2 (16.76 ± 9.16, P \ 0.01) and gradually decreased at T3(9.71 ± 6.11), T4(6.51 ± 2.86), T5(4.98 ± 1.57) and at T6(5.54 ± 1.97) while IL-6, white blood cell count (WBC), C-reactive protein(CRP) elevated postoperatively. Patients' postoperative course was uneventful and all patients discharged from ICU within 3 days.CONCLUSION. 1. Serum HMGB-1 level increased transiently during cardiac surgery using cardiopulmonary bypass. 2. Patients' postoperative course was uneventful in spite of high HMGB-1 level during cardiac surgery. 3. This is the first report showing an increase in serum HMGB-1 level during cardiac surgery. 4. High HMGB-1 level might be due to the combination of an increased HMGB-1 release from ischemic organs and a decreased HMGB-1 elimination during CPB.

CRITICAL CARE OUTCOME FOLLOWING RUPTURE OESOPHAGUS S. Hussain* 1 , H. El Sayed 2 , R. Vennila 1 , M. J. Shackloth 2 1 Cardiothoracic Anaesthesia, 2 Thoracic Surgery, The Cardiothoracic Centre, Liverpool, United Kingdom INTRODUCTION. Rupture oesophagus is a thoracic surgical emergency with significant morbidity and mortality [1] . We present our experience in managing these patients in a tertiary care cardiothoracic critical care unit.

We conducted a retrospective clinical review of patients who were admitted to the Cardio thoracic intensive care unit following rupture oesophagus over a period of last five years (2002) (2003) (2004) (2005) (2006) (2007) .

We had 24 admissions following isolated rupture oesophagus to our critical care unit. Out of which 71% were males, 29% females. The median age was 64 years (range 22-82). The aetiology was spontaneous (Boerhaave's syndrome) in 87.5% and iatrogenic in 12.5%. The diagnosis was established early (\ 24 hrs) in 71% and delayed ([ 24 hrs) in 29 %. 19 patients (83.3%) were admitted to ITU preoperatively and all 24 (100 %) were admitted postoperatively. In preoperative admission group, 63.0% had shock requiring ionotropes and 42.0 % required ventilation. Primary surgical repair was done in 74.4 %, a 2-stage repair in 8% and conservative management in 16.6%. Postoperatively, 71% were electively ventilated and nearly half of them (47%) were extubated within first 24 hrs. Mean ITU length of stay was 12 days (range 1-45) and mean hospital stay was 35 days (range . Over all in hospital mortality was 16.6%. In 75% of the non-survivors, there was a delay in diagnosis ([ 24 hrs) along with shock needing ionotropes on initial admission and they also required preoperative mechanical ventilation. Positive bacteriology was obtained in more than 50 % of patients (Gram Negatives) from one or more specimens and fungal organisms in 41.5%. All the non-survivors had yeasts from one or more specimens.CONCLUSION. Our review confirms the early diagnosis and management is crucial for successful outcome in patients with rupture oesophagus. However, shock on admission and need for preoperative ventilation by them selves are not indicative of poor outcome. Microbiological surveillance and aggressive management of bacterial and fungal infection are crucial for improving outcomes. Our mortality figures are marginally lower than other centres [1, 2] , possibly because of higher volume and also early critical care interventions. RESULTS. 1. Age [ 70 yrs comprised 37.3% (n = 210/565) of the total patients. 2. Cardiovascular disease was the most common co morbidity seen in 86.6% (n = 489) of patients with 73.3% (n = 414) hypertensive, 6.66% (n = 39) coronary artery disease and 4.44% (n = 25) rheumatic heart disease and obesity was the second most common 35.5% (n = 200). Other comorbid conditions were diabetes mellitus and bronchial asthma in 13.3% (n = 75) each, hypothyroidism 10.6% (n = 61), depression 9.3% (n = 52), nephropathy 6.6% (n = 37), anaemia 4 % (n = 22) and dyslipidemias 2.6%(n = 14). 

We usually consider delaying the extubation of maxillofacial surgery more than other postoperative, because of the anatomy of face and neck, the inflammatory response after surgery and the length of surgical procedure. Our objective is to detect complications in a cohort of patients after maxillofacial surgery trying to relate with early or late extubation.METHODS. Observational, prospective study of a cohort of 104 patients after maxillofacial surgery and admitted in our ICU between February 200 to March 2008. We analyzed demographic and clinic variables, preoperative situation, surgical risk (measured with American Anestesiology Scale risk), difficult to intubation, length of surgery, complications in ICU and time under mechanical ventilation and complications due to extubation moment. We considered early extubation before than 6 h after admitted in ICU. Dates are shown as mean and standard desviation or percentages and absolute number. We used SPSS package for the statistical analysis.

We collected 104 patients (59,6% male). Mean age 52,6 ± 18,9 years, urgent surgery 5,8% (6), Lengh of surgery 4,41 ± 1,5 h, difficult expected intubation 26,1% (24) and mehanical ventilation in ICU 8,64 ± 11,8 h. We registered complicatios in 13 patients (12,5%), most frequents respiratory (5,8%), hemodynamic (2,9%) or combinations. One patient died and 4 patients were reoperated. 38 patients were intubated six or more h and this group had more rate of complications than the group extubated prior to 6 h. (p = 0,02) 3,78 IC95% (1, 31) . ASA risk and difficult expectation to intubate were not associated with more complication. Univariate analysis showed that cervical ganglionic drainage was associated with more complications after extubation. (p = 0,012. 4,4 IC95% 1,2-13,1). Multivariate analysis reported as predictors of more rate of complications: Extubation after 6 h in ICU(p = 0,049. 3,3 IC95% (1-11,2)) and tumoral surgery which needs cervical ganglionic drainage (p = 0,032. 3,6 IC95% (1,1-12,1)).CONCLUSION. Premature tracheal extubation can be hazardous in this surgery due to oedema and haemorrage and the proximity to the airway. We usually leave intubated more time to patients who are difficult to intubate after a surgical procedure. In spite of a not very high rate of complications (12,5%), our study revealed that delayed extubation and cervical ganglionic drainage are associated with more complications and rate of reintubation. RESULTS. RESULTS. Neither germs of the physiological flora nor the colonizing pathogens of the oropharynx of the operated patients were transmitted through the filters. No organisms were cultivatet from the condensation water or in the ventilator. In opposite, all anesthesia bags surfaces were highly contaminated.CONCLUSION. CONCLUSION. The HME-F protects the BCS as well as the ventilator from the respiratory flora of the patient, allowing to use the BCS for more than one patient and extend the change interval up to 7d, which proved to be cost effective. In order to minimize cross infections, the external surface of the BCS as well as the anaesthesia bag must be disinfected after each patient.GRANT ACKNOWLEDGEMENT. 

A 64-year-old male with 45 pack-years was evaluated for arthropathy. After a positive Mantoux skin test chest x-ray revealed a right central tumor. Pulmonary function studies showed FEV1 2,96L (91% predicted) and FEV1/VC 0,55. PET-CT showed a PET positive lesion apical in the right inferior lobe and a positive lymph node. Bronchoscopy, mediastinoscopy and EUS were tumor-negative. Right-sided pneumonectomy was performed for squamous cell bronchuscarcinoma T1N0M0, stage IA. After surgery he developed ARDS, followed by stump leakage for which revision was performed. Intercurrent airway infections were treated with antibiotics. Despite this, the patient had major problems weaning from the ventilator. CT scan showed compression atelectasis of the left inferior lobe, reason to consider the possibility of PPS. Despite absence of mediastinal shift, surgery was performed 4 months later to implant expandable prosthesis in the pleural cavity. In spite of this intervention, weaning from the ventilator was hampered by muscle weakness and swallowing difficulties leading to recurrent aspiration.RESULTS. PPS is a rare complication of pneumonectomy. Jansen et al. reported one case in a series of 640 pneumonectomies (incidence 0,2%) and just over 20 cases are reported in literature. Clinical symptoms are progressive dyspnoe, cough, inspiratory stridor, recurrent pneumonia and bronchiectasis. In reported cases the symptoms developed 5 months to 35 years after pneumonectomy. In all described patients a shift of the heart and mediastinum towards the side of the pneumectomy, causing bronchial obstruction, was found. As PPS can lead to respiratory insufficiency and death this diagnosis should be considered in every postpneumonectomy patient presenting with corresponding symptoms. Confirmation of the clinical diagnosis PPS is accomplished with CT-scan and bronchoscopy, showing a shift of heart and mediastinum towards the pneumectomy-side with hyperinflation of the contralateral lung, stretching of the distal trachea and left main bronchus and lobair bronchus. The stretched airways can be compressed by pulmonary arteries, aorta and spine leading to tracheabroncho-malacia. Current surgical treatment consists of mediastinum repositioning and implantation of expandable prostheses with an additional role for airway stenting.CONCLUSION. PPS, although rare, may have contributed to the weaning problems in this patient and should be considered after pneumonectomy. INTRODUCTION. Cortisol levels are of prognostic value in patients with communityacquired pneumonia (CAP). However, the predictive value of total cortisol (TC) and of the frequency and impact of adrenal insufficiency (AI) in severe CAP remains to be thoroughly evaluated.

The main objective was to investigate the predictive value of baseline and post-ACTH (250mcg) TC levels at ICU admission in patients with severe CAP in a prospective cohort study of patients with severe CAP admitted to the ICU. TC at baseline and after ACTH test, C-reactive protein, D-dimer, clinical variables, SOFA, APACHE II and the CURB-65 score were measured. The major outcome measures were hospital mortality and disease severity.

J. Tan, E. Ng*, J. Foong, R. Sim, K. Lee, T. Chong, T. Chua, A. S. L. Wong, T. Koh Cardiology, National Heart Centre, Singapore, Singapore, Singapore (2). Since 2004, and after SARS, the hospital campus has initiated a influenza vaccination program, at no cost, to all HCWs. Despite the availability of this program,the take up rate is a low 58%. We aim to explore the perceptions and decision making factors towards vaccination amongst HCWs at our centre.

METHODS. The study was approved by Singapore General Hospital ethics committee. All employees except doctors were asked to participate in the survey. Employees were asked for signed consent before the survey was administered. Statistical analysis was performed with JMP 7.

. 436 completed questionnaires out of a total of 668 employees were gathered. 59% of the cohort were nurses, 41% were outpatient staff. 89.2% of respondents were female and the median age group was between 21 to 30 years old. 75% of this cohort did participate in the last vaccination program with 78% agreeing that HCWs should be vaccinated but only 30% feel it is ethically wrong for HCWs not to do so. The strongest reason for vaccination was for protection against a possible flu pandemic (73%) followed by the wish to protect our patients (15%) and the main reasons for opting out was due to forgetfulness or inconvenience (37%), fear of needles or side effects (21%) and the notion that it was ineffective (20%). Only one respondent was unaware that it was available in campus and free of charge. 53% of respondents developed side effects post vaccination, the most common was Flu-like symptoms ( 

A recent animal study suggested that morphine withdrawal lowers host defense to enteric bacteria resulting in higher rates of infection [1] . We hypothesized that remifentanil withdrawal would be associated with higher rates of intensive care unit (ICU)acquired infections.METHODS. Prospective observational cohort study performed in a 30-bed ICU during a 12month period. All patients requiring intubation, mechanical ventilation, and sedation for [ 3d were eligible. Patients who died before sedation withdrawal were excluded. Sedation was based on a written protocol including remifentanil with or without midazolam. The bedside nurse adjusted sedative infusion according to physician's prescription. Ramsay score was used to evaluate consciousness. [1] . It has been proven that inhalation injury and fluid resuscitation do not influence extravascular lung water index (EVLWI) [2] . Data in the literature suggest that elevated EVLWI may be a sign of the developing bacterial sepsis in burned patients [3] . The aim of our retrospective study was to analyse EVLWI regarding presence of infection, sepsis and mortality. The correlation between serum procalcitonin (PCT) level and EVLWI has also been studied.

Daily data from 38 patients admitted with burn injury (Total Body Surface Area [ 15%) were analysed (laboratory parameters and thermodilution monitoring). The day when microbiological samples were taken due to suspected developing bacterial infection was determined as day 0, the onset point of infection and /or sepsis. Data collected three days before and after onset of sepsis were studied. Patients were divided into survivors and nonsurvivors according to the outcome. Laboratory parameters were sorted into two groups using EVLWI, with a cut-off point of 8 ml kg-1. Correlations between variables were analyzed with the Spearman test and variables were compared with repeated measure ANOVA.RESULTS. EVLWI significantly correlated with PCT (r = 0,556; p \ 0.001), Baltimore Sepsis Scale (BSS) (r = 0,275; p \ 0.01), Intrathoracic Blood Volume Index (ITBVI) (r = 0.334; p \ 0.001) and Positive End Expiratory Pressure (PEEP) (r = 0.211; p \ 0.05). Cardiac Index (CI), ITBVI, PEEP, PCT and BSS were significantly higher in the high-EVLWI group (p \ 0,05). EVLWI showed a marked elevation on day -1 in both groups (p \ 0.05) compared to day -3. Elevated PCT level was first detected (p \ 0.05) on day 0 in both groups. PCT started to normalise both in the survivor and non-survivor groups after administration of antibiotics (i.e. after day 0) but its level still remained elevated on day +3 in non-survivors. In contrast to PCT, EVLWI remained in higher range or further increased among non-survivors on day +3 (p \ 0.01) but no significant difference could be found in the survivors on day +3 compared to the value measured on day -3.

Our data suggest that EVLWI may be an early warning sign of developing bacterial infection and its continuously increased level may predict poor prognosis in burned patients. 

The morbidity and mortality of HIV-infected patients admitted to the ICU has improved significantly. Many ascribe this improvement to the increased use of highly active retroviral therapy (HAART) (1, 2) whereas others have found it has no influence on outcome (3) . The aim of this study was to describe the demographics and outcomes for HIV infected patients admitted to a London Teaching Hospital Intensive Care Unit.METHODS. The clinical information system of our ICU was interrogated to identify all HIV-infected patients treated between 2002-6. Standard demographic data, reason for ICU admission, severity of illness scores, use of HAART, ICU and hospital length of stay (LOS) and mortality were recorded.RESULTS. 106 (61male:45female) HIV-infected patients had 113 patient episodes. ICU mortality was 22.1% and hospital mortality 25.9%, with median ICU and hospital LoS of 5 and 16 days respectively. Mortality for known HIV patients was 28% and that of newly diagnosed patients 16% (Table 1) . Admissions for HIV related illnesses were associated with a higher mortality than those for non-HIV related illnesses (29% vs. 10%, P \ 0.05). Admission diagnoses were mainly HIV related with associated mortalities as shown in Table 2 . CONCLUSION. In line with recent studies (1, 2, 3) mortality rates in our cohort of HIVinfected patients are similar to those of the general ICU population. Unlike recent reports (4), the most common presentations are HIV related. New HIV diagnosis also has a better outcome in our cohort, as do those patients admitted for non-HIV related illness.

Interpretation of pleïocytosis in the context of SAH is often difficult because of the presence of blood in the CSF. The aim of the study was to evaluate classical diagnostic criteria of meningitis for patients needing an external ventriculostomy drain (EVD) after SAH.METHODS. Inclusion criteria : All consecutive patients admitted to our ICU for SAH and needing an EVD between 2005 and 2007. Collection of all CSF values available. CSF samples were obtained if there was a clinical suspicion of VRI (fever [ 38°C and (or) altered consciousness). Proved VRI (PVRI) was defined when there was a positive culture of CSF, and suspected VRI (SVRI) when culture was negative but treatment was prescribed [ 5 days by the attending physician. Sensibility, specificity, positive and negative predictive values were obtained for the following variables and their association : CSF WBC [ 100/mm3 and PMN [ 50%, CSF WBC/RBC [ 1/500, hypoglycorachia defined as glycorachia (G).

Among 169 SAH, 54 needed an EVD (32%). Their WFNS score was 4 ± 1, Fischer 4 ± 1, GCS 9 ± 4. Among the 54 patients, 223 CSF were analysed, 7 PVRI (13%) and 8 SVRI (15%) were identified. Staphylococcus was found in 71% and a gram negative bacteria in 29% of PVRI. Diagnostic criteria of VRI for the 223 CSF are summarized in Table 1 . The univariate analysis identified an EVD [ 20 days (47% vs. 21%), a co-infection (67% vs. 37%) and the number of CSF withdrawn during ICU stay (10 ± 5 vs. 6 ± 5) as significant factors associated with VRI (p \ 0,05 for PVRI + SVRI vs. non VRI respectively), whereas physiological scores, surgery and number of EVD placed were not. By multivariate analysis, among the 3 significant variables tested, none was independently associated with VRI. 

Inadequate antibiotic therapy predicts poor outcome from sepsis, but there is no simple test of adequacy. We suggest that a time-to-positivity assay (Tpos) might act as a surrogate for antimicrobial activity and predict outcome from sepsis in ICU. 1 We conducted a prospective trial to test this hypothesis.

In this phase II exploratory study 47 sequential ICU patients with onset of sepsis who had not had antibiotics for [ 3 days were recruited. Sepsis was defined as evidence of infection plus C 3 criteria for SIRS. Sera taken at 24 h post empiric antibiotics were inoculated into culture bottles containing standardised bacteria, incubated in an automated system and time to positivity noted. The primary clinical endpoint was days in ICU. We also looked at 28 day survival.

. 48.9% were female with an average age of 59.7yrs. Cultures that are negative at 5d incubation indicate adequate antimicrobial therapy; cultures that become positive in \ 5d indicate inadequate therapy. The median SOFA scores on day 1 in the negative group was 9 and in the positive group 10(p [ 0.05).There was no statistical association between Tpos1 and 28 day survival(p [ 0.05): consequently, the area under the ROC curve was 55.9% (SE = 9.4%) and sensitivity 50%, specificity 65% for a cutoff of 9 h. Figure 1 shows the length of stay in ICU was significantly predicted by time for Tpos1 (p \ 0.05).The group where it remained negative at 5 days (ie indicating adequate antimicrobial therapy) were more likely to have a shorter stay in ICU (median 7.8d vs. 14.4d).CONCLUSION. These data show that patients whose Tpos1 remains negative are more likely to leave ICU early. The data also show that if Tpos1 has not become positive after 9 h the patient may do worse. At this point a change in therapy could be considered.

In vivo, lipopolysaccharide (LPS) induces lung injury, causing inflammation, alveolar-capillary leak and epithelial cell death. In contrast to this observation, several in vitro studies have shown pulmonary epithelial cells to be hyporesponsive to LPS stimulation. A possible explanation is that in vivo alveolar macrophages act as a first line of defense to infections, thereby regulating the epithelial response to LPS; however the mechanisms and effects of interaction remain unclear. Therefore, the aim of this study is to further understand and characterize the pulmonary epithelial response to LPS through macrophages. It is hypothesized that the effects of LPS on lung epithelial cells occurs indirectly through the mediators released by activated macrophages.METHODS. THP1 differentiated macrophages were stimulated with LPS or vehicle and the resulting conditioned media (CM) was subsequently used to stimulate pulmonary epithelial A549 cells. Additionally, polymyxin B (PMX-B) was added to LPS CM prior to A549 cell stimulation to bind and sequester residual LPS. Epithelial cells were also stimulated with LPS alone. After 24 hrs, the epithelial production of IL1b, VEGF, and TNFa (ELISA) and the surface expression of ICAM-1 (IF) were determined.RESULTS. LPS (10ug/ml or 100ug/ml) and control CM had no direct effects on A549 cells compared to the vehicle. Incubation with LPS stimulated macrophage CM resulted in increased production of IL1b (27.45 ± 0.94 pg/ml) and VEGF (1085 ± 121 pg/ml) versus control CM (0 and 584 ± 36 pg/ml respectively). TNFa, which is not produced by A549 cells, in the LPS CM was reduced by 75% following incubation with A549 cells, suggesting internalization of this molecule by pulmonary epithelial cells. Additionally, there was an upregulation of ICAM-1 surface expression (representative figure below) by LPS CM versus controls. All observed effects were unaltered by the presence of PMX-B.CONCLUSION. Inflammatory mediator production and adhesion molecule expression of A549 cells can be induced by LPS indirectly through macrophages. Furthermore, LPS sequestration did not alter this response, implying LPS does not enhance the effects of the macrophage mediators. By understanding these cellular mechanisms involved in LPS-induced pulmonary injury novel therapeutic strategies to prevent acute lung injury can be developed. CONCLUSION. Patients with DER are less than 4% of the total admissions in the ICU. Nonetheless they consume many resources due to the long stay and mortality. Their IR is higher than the overall IR in the ICU and this could be explained by the associated extrinsic and intrinsic risk factors.

GRANT ACKNOWLEDGEMENT. Aventis.

K. Ikeda*, T. Ikeda, H. Taniuchi, T. Ueno, S. Suda Critical Care And Subacute Medicine, Hachiouji Medical Center Of Tokyo Medical University, Tokyo, Japan INTRODUCTION. Introduction: Danaparoid sodium (DA) is a low molecular weight heparinoid with a mean molecular weight of 6000 daltons. It consists of mainly of heparin sulfate, and high-affinity fraction of heparin sulfate inhibits factor Xa by catalyzing its binding to antithrombin (AT). In animal experiment,DA was proved to have both anticoagulant and anti-inflammatory activities.The aim of this study was to investigate the potential benefits of combination therapy using DA with AT compared with the use of AT alone for the patients with SAC.METHODS. METHODS: The subjects were 22 patients with SAC in our ICU. The patients were divided into two groups, including the AT group: intravenous (iv) administration of 1500 units of AT alone and the AT + DA group: iv administration of the same dose of AT and 2500 units of DA for three days. We measured WBC, CRP, the platelet counts, the values of AT, fibrinogen, FDP, % PT before, 1 and 3 days after the administration of these medicines. Furthermore we examined the changes in scores of APACHE II, SOFA, SIRS, DIC in this study. All data are expressed as mean ± SD.RESULTS. APACHE II and SIRS scores were decreased significantly in the both groups. SOFA and DIC scores before administration of medicine were significantly high compared with the AT group and significantly decreased at 3 days after the administration in the AT + DA group. The platelet counts decreased significantly in the AT + DA group. The ATIII values increased significantly in the both groups. The FDP values before the administration were significant high in the AT + DA group compared with the AT group and decreased markedly in the AT + DA group. We did not observe significant change of fibrinogen in the both groups. We observed significant increase of % PT only in the AT group.CONCLUSION. The combination therapy of using danaparoid with antithrombin could reduce SOFA and DIC scores. This therapy may have the potential benefit to improve coagulation abnormalities for the patients with SAC reducing FDP.GRANT ACKNOWLEDGEMENT. Japanese society of intensive care medicine. (1). Patient/Family-Controlled Sedation(PFCS) with midazolam has not been described before.

A prospective protocol was established and a singe sedative was used. Inclusion criteria were severe pain (VAS [ 7), age [ 18, written pt or family consent, continuous presence of a competent adult at bedside and limited life expectancy, based on a PPI [ 7(2)(table1). Midazolam was administered IV in a Pt Control Analgesia mode. The dose was adjusted to pt comfort and the pump was activated as-needed by the pt or a first degree relative. Sedation was rated as: 1) awake 2) arousable to voice 3) arousable to light pain or 4) unarousable. Family satisfaction was rated as: 1) good, 2) fair, 3) poor, or 4) unacceptable.RESULTS. PFCS was applied in 7 pts in a 3-years period. Respiratory distress, when present, improved. Death occurred 1-6 days after PFCS started; Median sedation score was in the range 1-3 based on three measurements/day (Table 2) . Pt demographics, family satisfaction and daily midazolam use are summarized in Table 1 . 1  3  3  2  death  2  3  1  death  3  3  death  4  3  3  3  death  5  3  3  2  death  6  1  2  2  3  3  death  7  2  2  2  3  2  3 CONCLUSION. Patient/Family-Controlled Sedation with Midazolam is very effective in providing comfort in terminal cancer pts, by allowing titration of sedation to individual pt need. INTRODUCTION. End stage heart failure (NYHA class IV) is associated with frequent and prolonged admissions in the hospital. It has been shown that non invasive ventilatory support (continuous positive airway pressure-CPAP and bi-level airway pressure-BiPAP) improves clinical parameters and alleviates breathlessness in these patients. The aim of our study was to assess the role of non invasive ventilation (NIV) in end-of-life care of critically ill patients with end stage heart failure.

The study was carried out in the Cardiology Department of a District General Hospital in cooperation with the Intensive Care Unit. Patients with acute respiratory failure in the context of refractory heart failure (functional capacity IV) were enrolled. None was suitable for heart transplantation due to malignancy, multiple organ failure, diabetes or age [ 75 years. Standard medical treatment was initially offered to all of them (oxygen, diuretics, digoxin, angiotensin converting enzyme inhibitors, b-blockers, levosimendan, inotropic agents and morphine). NIV was implemented additionally to conventional treatment in order to relieve patients' distress. Dyspnea intensity was assessed independently by two physicians on a 3 hourly basis. The assessment was based on a scale, ranging from 0 to 2 (0 = no effect, 1 = medium improvement, 2 = maximum improvement).RESULTS. 24 patients were included. The mean age was 69 years and 16 were males. In relevance to dyspnea the score was 1 for 14 patients (58.33%) and 2 for 2 patients (8.33% INTRODUCTION. Among the recommendations for end-of-life care, one of the most important methods of comfort is analgesia and sedation. In a hostile and painful environment as intensive care units, the concern with the reduction of the suffering must be present from the admission of the patient and during through the entire hospitalization. In this work we analyze the prescribed analgesics and sedatives in patients who progressed to death.

We retrospectively analyzed the prescriptions of patients who died in the year 2007 with more than 24 h of hospitalization, in a general ICU with 23 beds. We collected patients informations about age, length of stay and the presence of regular analgesics and sedatives prescription within 24 h preceding the death.RESULTS. There were 1043 admissions and 139 patient deaths. Twenty-five patients were excluded because died before the first 24 h. The mean age was 65.34 years (SD 18.23), with an average length of hospitalization of 13.49 days (SD 12.5) . In the select group, 63 patients (55.26%) were in regular use of analgesics within 24 h preceding death and 67 patients were receiving sedatives (58.77%). Seventy-two patients (63.16%) used or analgesia or sedation, and 48 patients (42.11%) received both. A research previously conducted about end of life care with the physicians group (n = 21), none was favorable of the suspension of sedatives or analgesics. In the group of nurses (n = 39) 12.82% agree with the suspension of analgesia (n = 5) and 17.97% of sedation (n = 7).CONCLUSION. An approach with quality must ensure routine sedation and analgesia of all hospitalized patients, regardless of their prognosis. The speech of maintaining such care, even when death is inevitable, needs to be checked daily since the start of hospitalization. Our goal is that all of our patients receive comfort, and not just a piece, as shown in this study. 

Withholding and withdrawal of life sustaining treatment is common in the intensive care unit (ICU), but little is known about the type of patients in which treatment is withdrawn. The objective of this study was to evaluate the causes of death, and the incidence of withdrawal of treatment in patients who died in the ICU.

A retrospective observational, single centre study in the mixed ICU of a university hospital. All patients admitted to the ICU between November 1, 2006, and October 31, 2007 were included. Age, sex, length of stay, APACHE II score and diagnosis, and SOFAscore were collected for all subjects. Additional data on cause of death and withdrawal of treatment were collected for patients who died in the ICU. Data were extracted from our patient data management system, electronic patient dossier, and handwritten medical charts. INTRODUCTION. Effective communication with the relatives or visitors of critically ill patients is essential for clinical, ethical and legal reasons, and provides relatives with an insight into the patients' diagnosis, current condition, management plan and likely outcome. This allows family to feel part of the management and reduces fears and anxieties which arise due to misunderstandings.

We conducted a prospective audit over 6 months at our ICU department. Next of kins of 76 patients admitted were questioned using a pre-prepared questionnaire based on ''Proposed standard or target for best practice'' with regards to ''Communication with relatives or visitors of ICU patients'' as described by the Royal College of Anaesthetists audit recipe book.RESULTS. 1. 40% of relatives were acknowledged by the ICU nurses within 2 minutes.2. 36% were given information within 15 minutes of arrival and 24% within 15-30 minutes.3. 50% were interviewed by the ICU doctor in less than 30 minutes and 20% had no interview at all. 4. On 32% of days there were formal interviews with doctors; 24% of whom were consultants. 5. 98% of the time the interviews were documented in patient notes and the consistency between medical and nursing notes were 46%. Nurses documentation towards the end of their shift and incomplete information were the main reasons for the inconsistency. Essential information were however consistent.CONCLUSION. Inspite of the observed results being significantly less than the 100% targets set by the Royal College, a similar survey in our ICU department (in 2006) showed that 92% of relatives were satisfied with the overall level of communication. Following the results of our audit we conducted a telephone survey of 70 ICU departments across UK, questioning a senior member of the nursing staff. The results showed only 10% were aware of the national guidelines; 26% acknowledged relatives within 2 minutes of their arrival into the ICU; 10% see them within 15 minutes of arrival of the patient; 24% have a daily formal interview with a doctor; 6% said relatives are seen by a consultant daily; and 100% said all communication is documented in case notes. If proposed guidelines were strictly adhered to the satisfaction of relatives would likely be significantly higher than our survey results however the reality is that ICU departments are frequently unpredictable and busy. Common reasons for failure to reach these targets are, insufficient medical or nursing staff available to talk to relatives, preoccupation of staff with emergencies and family being unable to travel to the hospital. We contend that the proposed standards are unrealistic given the busy and unpredictable nature of the intensive care environment. We suggest that overall satisfaction is ultimately the best measure to evaluate communication between health care staff and relatives of critically ill patients. METHODS. This retrospective study was performed within our 15-beds intensive care unit. We analysed the patient files and ethical meeting reports from 2005 and 2006: 52 items were collected, including the number and qualification of participants to the meeting, whether an external practitioner had given advice, the implication of the family, the modalities of application.RESULTS. 1137 patients were admitted in the study period: 280 died (23.8%) and 94 withdrawing or withholding life sustaining treatments (WWLST) were decided (8% of admissions). Patient with WWLST tended to be elderly (61 mean age vs. 55.5) and more severe (SAPS 56.9 vs. 46)as compared with overall patients. The average delay between admission and ethical meeting was 12.9 days; the idea of WWLST appeared in the file in 37.2% of cases, on average 2 days before the decision. Patients will were expressed in 26.6% of cases, by the patient himself in 56% of cases. Relatives were known in 97% of cases, 76% of which had given their opinion. 

A WWLST decision is made on average once a week, more often in older patients, and those whose SAPS is 25% higher than the average. These weekly ethical meetings implying all persons in charge of the patient are important events in the life of the unit. Reaching a consensus may delay the moment of the decision. The fact that more than one for five of these patients leaves the unit alive proves that nothing is done to accelerate their death. Advice from an external practitioner often remains difficult to obtain, but this is improving with time. . We sought to conduct a comparative analysis of the needs among family members of patients on the basis of the severity of the illness.

We conducted a multi-center prospective questionnaire based study. All patients admitted to the ICU of the three Fortis Hospitals (tertiary care multi-specialty hospitals) in the North India (Noida, Vasant Kunj & Faridabad) for at least 72 h were included. The nearest relative of each patient were administered an Indian customized version of the modified Molter's questionnaire, also known as the Critical Care family Needs Inventory. The total number of domains (namely Information, Comfort, Support, Assurance, and Proximity), the total number of questions (n = 45) as well as the choice of responses (1: 'not important' to 4: 'very important') were as in the original questionnaire. The responses were converted into a binomial variable by merging 1&2 as one category and 3&4 as another category. The frequency of each response was then compared between the groups formed on the basis of APACHE II score (with a cut-off of 10) and need of mechanical ventilation. INTRODUCTION. It could be argued that the structure of an intensive care unit (ICU) is designed to suit the functions of the medical and allied professions. Within such a framework, it is important that we also recognise the needs of the patient's family. Guidelines for support of the family in the patient-centred intensive care unit have recently been published. [1] METHODS. Part A: We circulated a post-discharge questionnaire to the next of kin of patients discharged from our ICU during the month of October 2007. Part B: A telephone survey was also conducted in order to ascertain arrangements for family support across all of the ICUs in Northern Ireland.