key: cord-015024-2xzc0uc5
authors: nan
title: ESICM 2010 WEDNESDAY SESSIONS 13 October 2010
date: 2010-08-31
journal: Intensive Care Med
DOI: 10.1007/s00134-010-2001-7
sha: 
doc_id: 15024
cord_uid: 2xzc0uc5

nan

Power spectrums for Vt and Eadi are shown in Fig. 1 (PS and NAVA) for a typical patient. The enlarged section highlights how changes in Eadi are highly synchronized with NAVA ventilation, but less so for PS. Table 1 ) and complications of mechanical ventilation ( Table 2 ) did not differ significantly between the two studied groups. INTRODUCTION. High tidal volumes in mechanically ventilated patients with ARDS lead to baro/bio-trauma and increase mortality. Also, it was recently shown that ventilation with high tidal volumes is a risk factor for ''acquired ARDS'' in a medical population.

OBJECTIVE. We evaluated the impact of high tidal volumes after cardiac surgery.

METHOD. We analysed the prospectively recorded data of 3,434 consecutive patients who underwent cardiac surgery from 2002 to 2005. We predefined 3 groups of patients based on the tidal volume delivered immediately after surgery: (1) low: 7-9.9, (2) ''traditional'': 10-12.9, (3) high: above 13 ml/Kg of predicted body weight (PBW). We assessed the risk factors for organ dysfunction (prolonged mechanical ventilation, hypoxemia, hemodynamic failure and renal failure) by univariate and multivariate analysis, including the initial tidal volume in the models.

Mean tidal volume/actual weight and tidal volume/PBW was 9.2 ± 1.3 and 11.1 ± 1.5 in men (P \ 0.0001), 9.1 ± 1.4 and 12.5 ± 2.2 in women (P \ 0.0001). 411 patients (12%) were ventilated with low tidal volumes, 2,194 (63.9%) with ''traditional'' TV and 829 (24.1%) with high TV. The mean body mass index in the 3 groups was 23.8 ± 4.0, 27.0 ± 4.1 and 31.5 ± 5.4 respectively (P \ 0.0001). With increasing BMI, the tidal volume/ actual weight decreased while the tidal volume/PBW increased (Figure) . The percentage of women was 12.2, 20.8 and 52.2% respectively for low, ''traditional'' and high TV (P \ 0.0001). High tidal volumes were associated with prolonged intubation ([48 h) (1.5 vs. CONCLUSION. Traditional and very high tidal volumes are associated with prolonged mechanical ventilation and organ dysfunction after cardiac surgery and use of high tidal volumes is an independent risk factor. ''Prophylactic'' protective ventilatory strategy should be provided in this population with inflammatory state at risk to develop ventilator induced pulmonary edema. Women and patients with high BMI are more at risk to be ventilated with injurious tidal volumes. INTRODUCTION. Evidence shows that clinicians' non-technical skills (behavioural and cognitive skills) have a significant impact on teamworking, patient safety, efficiency of care provided and potentially patient outcomes (1) . Such skills are key for Cardiac Arrest Teams (CATs), which are multi-professional (Anaesthetists, Physicians and Nurses) and normally function under high pressure. To date, most tools to assess nontechnical skills in healthcare have focused on surgery (2) and anaesthesia (3) . No validated, robust tools are currently available for assessing non-technical skills in CATs.

OBJECTIVES. To develop and validate an Observational Skill-based Clinical Assessment tool for Resuscitation (OSCAR). This should be psychometrically robust for use in both training and assessment contexts.

METHODS. OSCAR was based on a well-validated tool for surgery (OTAS) (4) and was developed in phases. Six behaviours were included in the assessment: Communication, Cooperation, Coordination, Monitoring, Leadership and Decision-Making. Observable behavioural exemplars were derived for each one of these behaviours across the three CAT subteams-Anaesthetists, Physicians and Nurses (Phase 1). Quantitative expert consensus methodology was employed to assess content and face validity and observability of the exemplars (Phase 2). Two clinician observers used OSCAR to blindly rate eight CATs performance in a series of simulated cardiac arrests. Psychometric analyses of these ratings were used to determine observable behaviour applicability, internal consistency, and inter-rater reliability (Phase 3).

. 15 of 18 OSCAR behaviours demonstrated high internal consistency (Cronbach a = 0.736-0.970). Psychometric analyses dictated removal of three behavioural exemplars (two in Anaesthetic group; one in Physician group) to significantly improve internal consistency. Inter-rater reliability was also high (inter-observer Pearson r = 0.661-0.911, all p \ 0.005). Inter-observer reliability analyses revealed a learning curve between the two observers, with significant reduction in scoring discrepancies from the first to the eighth observed resuscitations.

CONCLUSIONS. OSCAR is a psychometrically robust (reliable, content-and face-valid) tool for the assessment of teamworking skills in cardiac arrest events. The tool is feasible to use and can be employed for both training and assessment purposes. INTRODUCTION. Different educational methodologies are used to teach basic skills in emergency medicine. High-fidelity patient simulation offers an ideal venue for presentation of critical events that can be managed by medical students without risk to a patient. Therefore full scale simulation training could be superior to paper case based seminary rounds to achieve these specific educational objectives.

OBJECTIVES. The aim was to compare simulation to a standard education measured by multiple choice questionnaire.

After written informed consent and approval of the institutional research ethics board 160 fifth year medical students were included in the survey. They took part in the compulsory emergency medicine curriculum of Charité Universitätsmedizin Berlin. The students completed a basic multiple question tests on day 1 including 25 questions concerning the topic of ''acute coronary syndrome'' (ACS). On day 2 for the topic ''ACS'' half the group was assigned a 45 min session simulation training while half the group was assigned a 45 min session paper case training. On day 3 groups were reversed and the topic ''AIC'' was taught in either simulation training or paper case seminary round. The test of day 1 was repeated after each training sessions. Results of the tests were evaluated using SPSS(TM) 18. The Mann Whitney U test was used to show any significant differences in reaching educational objectives in the test (a \ 0.05 was considered significant).

There was an even distribution of men and women among the two groups. The test results showed no significant difference between the two groups on day 1. On day two for the topic ''ACS'' the group with simulation training achieved significantly better test results. For the topic ''ACS'' on day 3 there was no difference while students received further training in ACS not using a high fidelity simulator. The results were not linked to specific teachers. INTRODUCTION. Rapid sequence induction (RSI) involves loss of spontaneous breathing and mandates airway control. Steps to reduce adverse incidents include adherence to minimum monitoring standards, appropriate drug selection, access to difficult airway equipment and presence of skilled anaesthetists. There is substantial evidence that appropriate monitoring reduces risk by detecting the consequences of errors, and by giving early warning of patient deterioration. OBJECTIVES. To assess conduct of emergency anaesthesia (monitoring and drugs) for critically ill patients not in an operating theatre (OR) administered by intensive care doctors.

METHODS. Prospective analysis of RSI for critically ill patients in a UK NHS Acute Hospital over 1 month. OR based practice was excluded. Reason for anaesthesia, location, drugs administered, monitoring modalities, adverse events and access to airway equipment were recorded. RESULTS. Data from 26 patient episodes were collected: predominantly in the emergency department (38%) and intensive care unit (27%) for respiratory failure (46%), reduced consciousness (19%) and to facilitate investigations (19%) . The most common induction agent was propofol (52%); thiopentone (26%) and etomidate (11%) were less frequently used. Suxamethonium (73%) was preferred for initial neuromuscular blockade. During induction most doctors used pulse oximetry, electrocardiography and blood pressure monitoring. Only 24% used capnography. No doctor used minimum monitoring to Association of Anaesthetists of Great Britain and Ireland (AAGBI) standards. 1 Rescue airway equipment immediately available is shown in Fig. 1 . Complications occurred in 12 cases (Fig. 2) . Patients that had a hypotensive episode during induction all had thiopentone or propofol used as induction agents. 15% of patients had a period of desaturation, and 23% required more than one attempt for successful intubation. In cases with complications, rescue airway equipment was unavailable in[60 and 75% did not achieve UK minimum monitoring standards. CONCLUSIONS. Shortcomings during emergency anaesthesia were recorded including monitoring, access to rescue airways and physiological disturbance. Procedural guidelines and training are to be developed for emergency anaesthesia; access to capnography and alternative airway equipment will be assured. These issues are unlikely to be unique to our Trust and assessment of practice is recommended. INTRODUCTION. Critical care echocardiography (CCE) is performed and interpreted by the intensivist at the bedside to establish diagnoses and guide the management of patients with circulatory or respiratory failure in the ICU. Competence in basic and advanced CCE has been recently defined [1] , but no curriculum to reach the required cognitive and technical skills has yet been elaborated. OBJECTIVES. To assess the efficacy of a limited, tailored training program for noncardiologist residents without experience in ultrasound to reach competence in basic CCE. METHODS. Six noncardiologist residents (Anaesthesiology: n = 5, Pneumology: n = 1) without previous experience in ultrasound participated to the study during two 3-month periods. The curriculum consisted in 4 h of didactics, 2 h of interactive clinical cases and 6 h of tutored hands-on. Color Doppler mapping was excluded from the training. After completion of the training program, all eligible patients underwent subsequently a transthoracic echocardiography (TTE) performed in random order by a recently trained resident and an experienced intensivist with expertise in CCE who was used as a reference. In each patient, the resident and the experienced intensivist answered binary ''rule in, rule out'' clinical questions covered by basic CCE [1] : global left ventricular (LV) size and systolic function (eye-ball evaluation of ejection fraction), homogeneous or heterogeneous LV contraction pattern, global right ventricular (RV) size and systolic function, identification of pericardial fluid and tamponade, and assessment of both the size and respiratory variations of the IVC. In case of undetermined interpretation, the corresponding clinical question was considered not addressed. The agreement between responses to clinical questions provided by the two investigators who independently interpreted the TTE study at bedside was used as an indicator of effectiveness of the tested curriculum. proportion of graduates to work within a ''Critical Care'' setting. The level of support available to trainees may vary with local resources but risk management and national guidelines stipulate that close supervision is provided to junior doctors in high stake decisions and procedures until deemed competent at the relevant tasks 1 . Furthermore, substantial ongoing reduction in working hours places further limitations on training; both majors can impact adversely on junior doctors service output and experience. A modified Delphi method was used 2 years ago to design a task focused single-day course on the theoretical basis of critical care and provide lab-based training in Delphi identified high risk procedures and interventions 2 .

OBJECTIVES. Assess the impact of the course on the following:

Trainee confidence and the start of the ''novice'' Critical Care post Trainee performance in comparison to peers Perceived educational benefit from their training post compared to peers METHODS. 30 junior doctors attending the course were enrolled in the study and matched for graduation year and medical school to 30 junior doctors who did not attend similar training prior to commencing their post. Data was collected through anonymous standardized forms on the day of the course, first day of the job, end of week 1, week 6 and 4 months into the post. Trainee confidence and self perceived competence were assessed on a ten point scale. In addition, trainees were requested to maintain a log of 3 interventions: 1-ultrasound guided central venous catheter insertion, 2-arterial catheter insertion, 3-ventilation problem solving.

Candidates attending the course demonstrated greater confidence at multiple points within their post as well as higher performance, satisfaction and educational value scores.

CONCLUSIONS. Critical Care trainees benefit from a task focused orientation to the fundamentals of critical care before commencing first post in this setting.

. 100 enrolled 50 patients in each group. No differences in age and gender. Incidence of VAP-study group 8.8% compared to control group 28.8% p value 0.004. VAP per 1,000 hospital days: control-29.5% compared to study-10.2% p value 0.06; average days in ICU control-14.97 compared to study-11.8, p value 0.151; average ventilated days, control-12.68 compared to 8.9, p value 0.044; average antibiotic use in days control-9.7 compared to study-7.2, p value 0.413. INTRODUCTION. Nosocomial infections are the most common in-hospital complications with high morbidity and mortality. Educating healthcare professionals is an important prevention measure.

OBJECTIVE. To analyze the impact of a nurse consultant team on nosocomial infections prevention in the ICU, the improvement in prevention knowledge of the nurse staff, and its impact in the application of the prevention measures in the daily practice.

METHODOLOGY: The nurse referent team was constituted by 8 nurses. The study subjects were all the staff ICU nurses and all the patients admitted during pre and post-intervention phases. The study was conducted in our medical-surgical ICU (16 beds) in 3 phases:

1 Pre-interventional (15/01/09-15/02/09) Observational. Record of the accomplishment of 12 CDC recommended variables about mechanical ventilation associated pneumonia (VAP) and catheter related bloodstream infection (CR-BSI) prevention measures.

2 Interventional (1/03/09-1/10/09) Eight educational meetings with the nurses staff groups to teach the most important aspects of the nosocomial infections prevention. Before and after lectures every nurses answered an anonymous questionnaire about their knowledge in those subjects. A poster with the most important reminders was place in every ICU patient room.

3 Post-interventional (15/01/10-15/02/10) Observational. New record of the same 12 CDC pre-interventional variables. We compared the accomplishment of these variables before and after the interventional phase as well as the number of correct questionnaire answers. Statistics were made with SPSS software. RESULTS. During the interventional phase 70% of the staff nurses attended the educational meetings. The number of correct answers increased significantly after the conference (77.1 vs. 55.2% p \ 0.0001).

Regarding to the daily practice, we observed a significant increase in the accomplishment in most of the variables (see table below), while in 3 of them no improve was observed and in 2 the improvement was not statistically significant. During the study period we observed a decrease in the incidence of VAP (6.7-14.9 episodes/1,000 MV days) and CR-BSI (2. INTRODUCTION. Glucose variability has been found to be associated with mortality in critically ill patients, independent of mean glucose concentration [1] .

OBJECTIVES. The aim of this analysis was to assess the impact of real time continuous glucose monitoring (CGM) on glucose variability in critically ill patients receiving intensive insulin therapy (IIT).

METHODS. This is the post-hoc analysis of a prospective, randomized, controlled trial [2] . Data of 124 patients admitted to the ICU either receiving IIT according to a real time CGM system (Guardian Ò , Medtronic, Northridge, CA, USA) (n = 63) or according to an algorithm (n = 61) with selective arterial blood glucose measurements (simultaneously blinded CGM) for 72 h were analysed. Insulin infusion rates were guided according to the same algorithm in both groups. Mean glucose and standard deviation, as a marker of glucose variability, were calculated for the first 24 h (GluMEAN1, GluSD1) and for the whole study period (Glu SD ). Statistical comparison of parameters between study groups and between ICU survivors (n = 94) and non-survivors (n = 30) was performed using student's t test.

RESULTS. The variability of sensor glucose during the entire study period was comparable between the real time CGM group and controls (21.51 ± 1.10 vs. 23 INTRODUCTION. In the gastrointestinal tract, the gut flora which comprises several hundred grams of bacteria is crucially involved in host homeostasis through their metabolic, trophic, and protective activities. However, the immediate changes in the gut flora in critical illness following severe insults are unknown.

OBJECTIVES. To investigate the changes in the gut flora at an early phase of severe insult in critically ill patients.

METHODS. Fifteen patients who experienced a sudden and severe insult including trauma, out-of hospital cardiac arrest, and cerebral vascular disease were studied, along with 12 healthy volunteers as the control group. Two fecal samples were acquired from the subjects by swabs of the rectum within 6 h after admission to the emergency room (day 0). Samples were serially collected from patients on day 1, 3, 5, 7, 10, and 14. Samples were collected from control subjects.

RESULTS. Total bacterial counts, especially various obligate anaerobes and total Lactobacillus, significantly decreased in comparison to those of the control subjects on day 0. In addition, on day 0, the total organic acid levels of the patients were significantly lower than those of the control subjects; particularly acetic acid, propionic acid, and butyric acid. The levels of these acids remained low throughout the 14 days period of study. The total bacterial counts did not recover to normal levels during the 14 day study period. Obligate anaerobe counts of the patients did not improve until day 14. Total Lactobacillus counts were low on day 0 and increased gradually thereafter, but did not attain the levels found in controls. The counts of pathogens (Enterococcus and Pseudomonas) increased during the study period.

CONCLUSIONS. Gut flora in critically ill patients can change drastically immediately after a severe insult, and may not recover for up to 14 days. At the same time, the number of harmful bacteria can increase. Total bacteria 10.1 (9.8-10.4) 7.5 (6.7-8.0) .001 Obligate anaerobes Clostridium cocades group 9.4 (9.0-9.7) 6.1 (5.4-7.0) .001 Clostridium leptam subgroup 9.2 (9.0-9.7) 6.5 (6.1-7.3) .001 Bacteroides fragilis group 9.4 (9.0-9.8) 6 .8 (5.9-7.6) .001 Bifidobacterium 9.2 (8.9-9.4) 6.0 (\5.0-6.9) .001 Atopobium cluster 9.0 (8.5-9.5) 6.0 (\5.1-7.1) .001 RESULTS. Mean serum 25(OH)D level was 18.9 ± 10.9 ng/ml. By current definitions the majority of patients (62.5%) were vitamin D deficient (\20 ng/ml) and 24.9% were vitamin D insufficient (C20 and \30 ng/dl). Normal 25(OH)D levels ([30 ng/ml) were present in 12.6%. Table 1 provides information on clinical and laboratory findings in the three 25(OH)D groups. Both lower 25(OH)D tertiles were associated with increased hospital mortality after adjustment for age, sex and SAPS II. For 330 patients both 25(OH)D and PTH levels were available. Adjusting the Cox regression analysis also for PTH and dialysis status increased the HR for hospital mortality to 2.3 (1.1-4.7) and 2.1 (0.99 and 4.4) for the two lower 25(OH)D tertiles. In addition tertiles of PTH and serum calcium levels suggested higher mortality rates for patients in the highest PTH (P = 0.09) and those in the lowest calcium tertile (P = 0.12). 

Our results demonstrate that independent of baseline SAPS II, age and sex, critically ill patients with low 25(OH)D levels seem to be at increased risk for hospital mortality. Whether a rapid correction of vitamin D status may be beneficial in the ICU setting remains to be further explored in randomized controlled trials.

• The autonomic storm after brain death must be early diagnosed and treated with a standardized protocol including hormone therapy INTRODUCTION. The use of filling pressures of the right atrium and left atrium is normal in the monitoring of critically ill patients undergoing mechanical ventilation. This monitoring is done through an invasive catheter placed in the superior vena cava and pulmonary artery, which is not free of complications. The ability to make measurements of these parameters in a non invasive way, makes the echocardiography an useful and essential tool when monitoring critically ill patients OBJECTIVES. We focus the study on validate the reliability of noninvasive measurements by echocardiography and invasive measurement catheters of filling pressures METHODS. We conducted a prospective observational study relating the filling pressures, between central venous pressure (CVP) with the diameter of the inferior vena cava and left atrial pressures with the values of the ratio E/e 0 . The filling pressure variables were only discriminated as high or low. Low values were accepted when invasive measurement of CVP was \6 and \4 mmHg in the LAP; and by echocardiography when the diameter of the IVC was\12 mm and the ratio E/e 0 \8. High values were accepted when the measurement of CVP was higher than 10 and 8 mmHg in LAP and in echocardiography when the diameter of the IVC [16 mm and the ratio E/e 0 [15. We collected data from 38 patients in the immediate postoperative period, under mechanical ventilation (Vt 8-9 ml/kg, FiO 2 50%, PEEP 3), sinus rhythm, good cardiac function and without postoperative drug support. All of them had a central venous line and right atrium catheter as habitual monitoring of postoperative cardiac patients. We performed an echocardiography when the patient presented hypotension, with low values of CVP and LAP, and we repeated the measurements after the infusion of the habitual fluid protocol (500 ml HES 6% in 30-45 min). The data we record were: diameter of IVC and ratio E/e 0 by echo and CVP and LAP values by invasive catheters. 23rd ESICM Annual Congress -Barcelona, Spain -9-13 October 2010 S333 INTRODUCTION. An attenuated cardio-hemodynamic response to dobutamine is associated with a poor outcome in established human sepsis [1, 2] . Establishing a sensitive method to identify early cardiac dysfunction in both experimental and human sepsis would be a useful tool to explore timesensitive mechanisms further.

OBJECTIVES. To assess myocardial responsiveness to dobutamine in early sepsis.

METHODS. All procedures were in accordance with UK Home Office laboratory animal legislation. Under isoflurane anaesthesia, male adult Wistar rats underwent left common carotid and right internal jugular venous cannulation for blood sampling/continuous BP monitoring and fluid administration respectively. Rats received either 1.8 ml caecal slurry (Sepsis; n = 8) or 1.8 ml saline (sham; n = 8) ip, before fluid resuscitation (0.9% saline 10 ml/kg/h) and conscious monitoring was commenced. After 4 h, rats were re-anaesthetized with isoflurane and transthoracic echocardiography was performed. Stroke volume was optimised with saline boluses prior to an incremental dobutamine infusion (1.25-20 mcg/kg/ min). Data are presented as mean (SD); analyzed with 2-way ANOVA and post-hoc Tukey test.

RESULTS. Figure 1 summarizes hemodynamic changes after sepsis, fluid resuscitation and dobutamine infusion.

Baseline parameters were similar after echocardiography-guided fluid resuscitation, with contractility and stroke volume restored in septic rats to sham values. Septic rats demonstrated an enhanced chronotropic response to dobutamine compared to sham (p \ 0.002). Both peak velocity and cardiac output were attenuated by C25% in sepsis (p \ 0.0001). In sepsis, baseline MAP was higher but neither sham nor septic MAPs were affected by dobutamine infusion. CONCLUSIONS. Dobutamine stress echocardiography is a sensitive, reproducible, dynamic physiological probe that reveals early cardiac dysfunction in septic rats with apparently similar baseline cardiovascular physiology. INTRODUCTION. The evaluation of right ventricular (RV) function is clinically useful in patients with acute respiratory distress syndrome (ARDS) because the presence of RV failure has large prognosis implications. The purpose of the current study was to compare right ventricular myocardial strain imaging parameters with conventional echocardiographic indices evaluating right ventricular function during ARDS.

OBJECTIVES. We hypothesized that peak systolic strain would be more sensitive than conventional echocardiographic parameters in detecting subclinical right ventricular systolic dysfunction in patients with ARDS.

METHODS. In total, 10 patients with ARDS and with normal right ventricle function assessed by two dimensional echocardiography and 10 age matched subjects under mechanical ventilation without heart or pulmonary disease were included in the present study. Conventional echocardiography parameters for RV function assessment like RV fractional area change (RVFA) or the tricuspid annular plane systolic excursion (TAPSE) were measured and compared to Tissue Doppler imaging parameters with strain value obtained from the right ventricle free wall.

. Strain values were reduced in the RV free wall of the patients with ARDS compared with the control group (23.45% ± 3.8 vs. 31.8% ± 2.8 p = 0.005) Moreover no significant difference was observed in conventional two dimensional parameters evaluating RV systolic function between these two groups of patients. In patients with ARDS a significant relationship was shown between peak systolic strain at basal free wall and arterial carbon dioxide tension (rho = -0.79 p = 0.017) and with the end inspiratory pressure (rho = -0.67 p = 0.04).

CONCLUSIONS. During the ARDS, Doppler tissue imaging parameters can determine RV dysfunction that is complementary to conventional echocardiographic indices and is correlated with respiratory parameters. On Doppler Tissue imaging, patients with ARDS exhibit abnormal RV systolic function even in patients with normal RV function assessed with conventional echocardiographic parameters. OBJECTIVES. Studying the effect of OLV on RV outflow impedance during inspiration and expiration using transesophageal echo-Doppler in a trial to differentiate the RV consequence of increasing lung volume from those secondary to increasing airway pressure during mechanical ventilation.

METHODS. Thirty stable patients on mechanical ventilation because of different causes were enrolled prospectively in this single center, cross sectional clinical study. Each patient was firstly subjected to conventional ventilation (CV) with volume controlled ventilation, followed by open lung concept (OLC) ventilation by switching to pressure controlled mode, then recruitment maneuver applied until PaO2/FiO2 [375 torr. Hemodynamic (mean arterial pressure ''MAP'', central venous pressure ''CVP'' and heart rate ''HR'') and respiratory (total and intrinsic PEEP, peak, plateau and mean airway pressure and total and dynamic lung compliance) measurements were recorded before, 20 min after a steady state of CV and 20 min after a steady state of OLC ventilation. Also, transesophageal ECHO Doppler was performed at end of inspiration and end of expiration to calculate the mean acceleration (AC mean ), as a marker of the RV outflow impedance, 20 min after a steady state of CV and 20 min after a steady state of OLC ventilation.

RESULTS. During inspiration, AC mean was significantly lower during CV compared to OLC ventilation (p value .001). Inspiration didn't cause a significant decrease in ACmean compared with expiration during OLV (p value.001) but did do so during CV. In comparison to baseline and CV, OLC ventilation was associated with a statistically significant higher CVP (p value .001 for both), higher total quasi-static lung compliance (p value .001 for both) and dynamic lung compliance (p value 0.001 for both). Moreover, PaO 2 /FiO 2 ratio of OLV was significantly higher than in baseline and CV (p value .001 for both).

CONCLUSIONS. OLC ventilation does not change RV afterload during inspiration and expiration as RV afterload appears primarily mediated through the tidal volume. Moreover, OLC ventilation provide a more stable hemodynamic condition and better oxygenation and lung dynamics. INTRODUCTION. Among indices provided by the analysis of aortic blood flow through esophageal Doppler, mean acceleration (Acc) is supposed to reflect the left ventricular (LV) systolic function, but this has been poorly validated. In particular, Acc could be influenced by loading conditions of the LV.

OBJECTIVES. To test whether Acc actually behaves as an indicator of LV systolic function by testing if 1. it increased with inotropic stimulation, 2. it was not altered by fluid loading, 3 . it correlated with the echographic LV ejection fraction (LVEF) and it reliably tracked the changes in LVEF during therapeutic intervention.

In 37 patients with cute circulatory failure (SAPSII 59 ± 18, age 64 ± 10 years, 26 receiving norepinephrine), we administered either a volume expansion (500 mL saline over 20 min in 24 patients) or dobutamine (5 lg/kg/min in 13 patients). We simultaneously measured Acc (CardioQ, Deltex Medical) and LVEF at baseline and after therapeutic intervention.

RESULTS. Volume expansion significantly altered neither LVEF (from 56 ± 12 to 58 ± 12%) nor Acc (from 9.5 ± 3.2 to 9.9 ± 3.1 cm/s 2 ) while dobutamine infusion significantly increased LVEF by 27 ± 15% and Acc by 44 ± 27%. Considering the 74 Acc/LVEF pairs of measurements, an Acc \8.2 cm/s 2 predicted a LVEF B 45% with a sensitivity of 77% (95% CI [56-91%]) and a specificity of 68% (95% CI [52-80%]). The changes in LVEF and in Acc during fluid and dobutamine administration were significantly correlated (r = 0.64, p \ 0.05).

CONCLUSIONS. Acc fulfilled the criteria required from a clinical indicator of LV global systolic function. A given value of Acc allowed detecting a low LVEF with a modest accuracy. By contrast, the treatment-induced relative changes in Acc were reliable for tracking the treatment-induced relative changes in LVEF. OBJECTIVES. To compare the relationship between systolic or diastolic dysfunction at ICU admission and the incidence of cardiologic complications and mortality at sixth months.

METHODS. Prospective study of forty consecutive patients diagnosed of Acute Myocardial Infarction (AMI) (23 NSTEMI, 17 STEMI) who were admitted in the ICU of University Hospital Puerto Real (Cadiz, Spain) from 1st May 2009 to 30th September 2009. Studied variables: Age, gender, type of AMI (NSTEMI, STEMI), Left Ventricular Ejection Fraction (LVEF) by biplanar Simpson's rule, Diastolic function (ratio E/E 0 of the mitral annulus included), incidence of cardiac complications (acute pulmonary oedema, atrial fibrillation with hemodynamic instability and cardiogenic shock) and mortality at sixth month. Echocardiographic studies were performed with a GE Vivid 7 Pro(R) by an intensivist who had performed up to 500 Doppler studies in critical patients. All studies were remeasured by a second observer in an echocardiographic workstation with no statistical difference in measured velocities. Patients were classified according to their LVEF in (a) preserved ([55%), (b) mildly depressed (45-55%), (c) moderately depressed (30-45%) and (d) severely depressed (\30%); and according to their E/E 0 ratio in (a) normal E/E 0 ratio (\10) and (b) elevated E/E 0 ratio (C 10). The results were statistically analysed with Chi-square test and odds ratio calculus.

RESULTS. Diastolic dysfunction measured with E/E 0 ratio was associated with high incidence of cardiac complications (Chi 2 test CL 95% p \ 0.001, OR 24). Systolic dysfunction measured by LVEF was also associated with more complications but with less strength of statistical association (Chi 2 test CL 95% p \ 0.05, OR 8. 25 ). There were no significative statistical difference between LVEF and E/E 0 ratio in mortality at sixth month.

CONCLUSIONS. In our study, diastolic and systolic dysfunctions in patients with AMI at ICU admission were associated with high incidence of cardiac complications, with more strength of statistical association in patients with diastolic dysfunction. The small sample volume didn't allow us obtaining significative statistical differences in mortality at sixth months. 

A new method has been developed to assess global end-diastolic volume (GEDV) and extravascular lung water (EVLW) from a transpulmonary thermodilution curve. Our goal was to compare this new method to the established method currently in clinical use, over a wide range up to extreme pathophysiological conditions. OBJECTIVES AND METHODS. 11 anesthetized and mechanically ventilated pigs (90-110 kg) were instrumented with a central venous catheter and a right (5F PulsioCath, Pulsion, Munich, Germany) and a left (5F VolumeView, Edwards Lifesciences, Irvine, CA) thermodilution femoral arterial catheter. The right femoral catheter was connected to a PiCCO2 monitor (Pulsion) and used to measure COp, GEDVp and EVLWp using the old method based on the equation: GEDV = COp 9 (MTt -DSt). The left femoral catheter was connected to the new EV1000 monitor (Edwards) and used to measure COe, GEDVe and EVLWe using the new method based on the equation: GEDVe = f (S2/S1) 9 COe 9 MTt, where S1 and S2 are respectively the maximum up-and down-slopes of the dilution curve, respectively. 137 measurements were done during inotropic stimulation (DOBU), during hemmorhage (HYPO), during fluid overload (HYPER), and after inducing oleic acid-acute lung injury (ALI).

Overall, COp and COe ranged from 3.1 to 15.4 and from 3.4 to 15.1 l/min, respectively. COp and COe were closely correlated (r 2 = 0.99), mean bias (± SD) was 0.18 ± 0.29 l/min and %error was 7%. GEDVp and GEDVe ranged from 701 to 1,629 and from 774 to 1,645 ml. GEDVp and GEDVe were closely correlated (r 2 = 0.79), mean bias was -11 ± 78 ml and %error was 14%. EVLWp and EVLWe ranged from 507 to 2,379 and from 495 to 2,222 ml. EVLWp and EVLWe were closely correlated (r 2 = 0.97), mean bias was -5 ± 72 ml and %error was 15%. Parameters over the study period are presented in the table (*p \ 0.05 Intervention vs. BASE or HYPER). INTRODUCTION. Fluid resuscitation is a major therapy in ICU. Various mechanisms are involved in the regulation of the microcirculation and the macrocirculation.

OBJECTIVES. The goal of this study is to assess the sublingual microcirculatory changes in response to fluid challenge in preload-responsive and non preload-responsive patients.

After approval by our local Institutional Review Board, 18 patients in surgical ICU have been included in an observational study. Each patient was monitored by an arterial catheter and an oesophageal Doppler. The decision of fluid infusion was taken by the physician in charge of the patient. Preload-responsive patients were defined by variations in cardiac index (CI) C 15%. Sublingual microcirculation videos were obtained using the orthogonal polarized spectral (OPS) imaging technology. Functional capillary density (FCD, cm cm -2 ) and microcirculatory flow index (MFI) were collected. The macrocirculatory and microcirculatory measurements were obtained before, during and after the infusion of 500 mL of saline. Five sublingual sites were recorded before and after the fluid resuscitation. The ventilator settings and sedative and vasoactive drugs infusion rates were kept constant throughout the procedure.

RESULTS. Patients were admitted in ICU for acute brain trauma (n = 5), hemorrhagic shock (n = 4), septic shock (n = 4), acute brain hemorrhage (n = 3) and acute pancreatitis (n = 2). The average age of the patient was 44 ± 17. The mean values of CI and mean arterial pressure (MAP) before the fluid therapy were respectively 2.7 ± 1.2 L/min/m 2 and 80 ± 15 mmHg. Nine patients responded to fluid infusion (CI C 15%.). About the microcirculation, there was no significant difference between responders (R) and non-responders (NR) concerning the variations of MFI (0. INTRODUCTION. Passive leg raising (PLR) was shown to discriminate hemodynamically unstable patients who will benefit from subsequent fluid administration or not. 1 Concerned by the possibility of harmful hypotension starting the PLR maneuver from a 45°semirecumbent position, in a previous study, we found that raising patients' legs from a supine position, we were not able to predict fluid responsiveness in a heterogeneous cohort of medical intensive care unit (ICU) patients. 2 OBJECTIVES. To investigate whether starting PLR maneuver from a 45°semirecumbent position would better predict volume responsiveness without harmful hypotension in spontaneously breathing critically ill medical ICU patients.

METHODS. Fluid responsiveness was tested in 27 consecutive patients (14 sepsis, 4 respiratory failure, 5 heart failure, 4 others) with a mean arterial pressure (MAP) \60 mmHg and/or a cardiac index (CI) \2.4 L/min/m 2 . Heart rate (HR), mean arterial pressure (MAP), global end-diastolic volume index (GEDVI), cardiac index (CI) and stroke volume index (SVI) were recorded using the PiCCO method. Patients were stable in a semirecumbent (45°) position when first measurements were taken (baseline 1). For the PLR maneuver, patient's bed was tilt to have the lower limbs raised to a 45°angle while the patient's trunk was then in a supine position. Changes after 2 min were recorded. The patient was then brought into a supine position, and heamodynamic measurements were recorded when stable (baseline 2). Thereafter, 500 mL of 0.9% NaCl were administered over 15 min. Positive predictive values (PPV) and negative predictive values (NPV) of the PLR maneuver were calculated using a cut-off value of 15 % increase for CI and SVI and 10 % increase for MAP.

RESULTS. Patients' median age was 60 (29-82) years and their SAPS score 59 (16-90). All patients received vasopressors and/or inotropes. Baseline hemodynamics and changes after PLR and fluid challenge are shown in Table 1 . Results are given as median (range); n/a = not available, *p \ 0.05 versus baseline. PPV and NPV for CI were 50 and 86%, for SVI 20 and 77% and for MAP 12 and 80%, respectively. CONCLUSIONS. In our hands, PLR was not useful identifying fluid responders in this heterogenous population of severely ill medical ICU patients, the starting semirecumbent position being associated with a potentially harmful decrease in MAP. However, it was helpful to detect patients who will not benefit (or even suffer harm) from further fluid administration. Recently, some studies suggested that an impaired diastolic function is a predictive factor of mortality in patient with shock. It is not already known whether fluid infusion could improve diastolic function.

OBJECTIVES. The aim of the study was to determine the impact of rapid fluid infusion on diastolic function.

After acceptance by the local ethic committee, 94 ICU patients were prospectively included. Volume expansion (VE) by 500 ml of saline was performed by the intensivist in charge. Transthoracic Doppler echocardiography was performed before and after fluid infusion. Stroke volume (SV), early diastolic transmitral velocity (E), early diastolic mitral annular velocity (Ea) and E/Ea ratio (reflect of LV filling pressure) were studied. Patients were divided in 2 groups according to their SV' increase: Responders (R) (those who increased their SV by at least 15 %) and Non-Responders (NR). Wilcoxon Rank sum test was performed to compare data before and after VE. Data are presented in median (IQR)

RESULTS. Fifty-three (56 %) patients were R and 41 (44%) were NR. In the overall population, Ea increased significantly with VE [from 12.8 (4.5) to 13.9 (5.3) cm/s, p = 0.002]. In the R group Ea increased significantly [from 12.6 (4.1) to 14.6 (4.9) cm/s, p = 0.01] and E/Ea did not change significantly [from 6.1 (2.1) to 6.3 (2. 3), p = 0.4]. However in the NR group, Ea did not change significantly [from 12 (5) to 12.7 (6.6) cm/s, p = 0.47] while E/Ea increased significantly [from 6.7 (2.8) to 8.5 (3.9) cm/s, p = 0.008].

CONCLUSIONS. According to these results, adequate fluid infusion seemed to enhance LV relaxation without increasing LV filling pressure while inadequate fluid infusion did not affect relaxation but increased LV filling pressure. OBJECTIVES. The aim of our study is to compare the rapid variation of CO measured by Vigileo-Flotrac Ò with Doppler-echocardiography which is considered as a reference method.

During the 36 first hours of hospitalisation, we studied 16 mechanically ventilated patients receiving norepinephrine who underwent arterial pressure monitoring via a radial artery catheter. The Flotrac Ò pressure sensor and the Vigileo Ò monitor were connected to the arterial line. At each fluid expansion or norepinephrine dose modification a transthoracic Doppler-echocardiography was performed and CO was calculated. Variations for CO measured by each method were compared. Results are presented as median (IQR). Linear regression and the Bland-Altman method were used for statistical analysis. METHODS. For the in vitro experiments blood of 20 healthy donors was incubated (in the ratio 5:1) with one of the following solutions: Ringer solution, Ringer-lactate solution, modified gelatin (Gelofusin); hydroxyethyl starch (HES) 130/0.42. After incubation, the following parameters of erythrocyte aggregation were measured: T1 and T2-characteristic times of spontaneous erythrocyte aggregation; b-hydrodynamic strength of aggregates; I 2.5 -index of strength of the largest aggregates at shear rate 2.5 s -1 . RBC deformability at various shear stresses was determined by ektacytometry. In vivo study on 60 patients with trauma treated randomly with either only crystalloids (group 1; n = 20), or crystalloids + HES 130/0.42 (group 2; n = 20) or crystalloids + Gelofusin (group 3; n = 20) over 7 days, the same parameters as in vitro study were determined at day 1-7. Twenty healthy men and women were included as controls. For statistical analysis the statistical package SPSS version 15.0 was used. Statistical significance was considered at p \ 0.05.

In vitro study in the final analysis effects of different colloids on RBC aggregation and deformability were considered as increasing impact (:), decreasing impact (;) and no impact (-) ( Table 1) . In vivo study significant microrheological disturbances were detected at day 1 after admission. Deformability index was lower in patients compared with controls (0.39 ± 0.012 vs. 0.44 ± 0.016; p = 0.017). Simultaneously, the patients showed erythrocytes hyperaggregation compared with control (;T1, ;T2; :I 2.5 , :b). In the first group (crystalloids) described violations persisted throughout the study time. In group 2 (crystalloids + HES), the deformability was higher than in the 1st group, from 3 days till the end of the study, attaining the normal range, and also higher than in the third group (crystalloids + gelofusin). In the third group, deformability index was not significantly different from group 1.

According aggregatometrical data in the first group hyperaggregation syndrome remained the entire period of observation. HES adding (group 2) decelerated aggregate formation (:T1, :T2; ;I 2.5 ). In contrast, modified gelatin adding enhanced erythrocyte aggregation (;T1 :I 2.5 , :b).

CONCLUSIONS. Crystalloid solutions are not able to improve microrheological parameters. HES 130/0.42 increases RBC's deformability and reduced RBC's aggregability. Gelofusin increases erythrocyte aggregation and no effect on deformability. INTRODUCTION. Trauma patients often require norepinephrine (NE) infusion and fluid challenge to keep normal blood pressure values. The reliability of dynamic predictors of fluid responsiveness during vasopressors therapy is under debate. We investigated the impact of norepinephrine (NE) infusion changes on pulse pressure variation (PPV) assessed with the MostCare system (Vytech Health, Laboratoires Pharmaceutiques Vygon, Ecouen, France) in intensive care unit patients. This device is a pulse contour method that provides cardiac output and fluid responsiveness variables and does not need any kind of calibration or preloaded data.

METHODS. 15 trauma patients (8 female, 7 male, mean age 53 ± 22) admitted to a 13-bed university hospital medico-surgical ICU were prospectively enrolled. Inclusion criteria were: mechanically ventilated patients (tidal volume [8 ml/kg and constant respiratory rate); invasive arterial blood pressure monitoring; NE infusion. PPV values were recorded continuously during three different haemodynamic states: at baseline (T1), 10 min after a 0.05 lg/kg/min NE increase (T2), 10 min after a further 0.05 lg/kg/min NE increase (T3), 10 min following the reduction of NE to T2 dosage (T2 0 ) and 10 min after setting NE to baseline value (T1 0 ). During the study neither fluid challenge nor other vasoactive/inotropic drug changes were done. ANOVA test was applied.

RESULTS. See data in Table 1 . At T1 NE mean dosage was 0.12 lg/kg/min (range 0.05-0.2 lg/kg/min). The mean PPV was: at T1 10.69 ± 6.9%, at T2 8.78 ± 5.9%, at T3 7.77 ± 6.1%, at T2 0 8.73 ± 5.5%, at T1 0 12.07 ± 8.2% (p \ 0.05). CONCLUSIONS. Our findings demonstrated that PPV was significantly affected by changes in NE: the higher the NE dosage the lower the PPV. Changes in arterial tone due to NE infusion can impair PPV reliability in assessing fluid responsiveness in trauma patients. INTRODUCTION. In mechanically ventilated patients respiratory variation in the arterial pulse pressure (DPP) is a reliable predictor of fluid responsiveness 1 . Respiratory variation of pulse oximetry plethysmographic waveforms correlate to DPP 2 and can be calculated automatically in real time (Heart-Lung Index [HLI Ò ] from Hamilton Medical). This prospective study evaluates the relationship between DPP and HLI Ò to predict fluid responsiveness.

Mechanically ventilated patients were investigated; all connected to an Hamilton G5 ventilator and ventilated in Adaptive Support Ventilation (ASV), paralyzed and none had severe cardiac dysrhythmia. 15 were eligible for fluid expansion. DPP, HLI Ò (obtained from a finger probe pulse oxymeter integrated to the ventilator) and Cardiac Index (CI from transthoracic echo-doppler), were obtained before and after fluid expansion (8 ml/kg of HEA over 20 min). CI-responders were defined by 15% increase from baseline.

RESULTS. 10 out of the 15 patients were CI-responders and had significantly higher HLI Ò before volume expansion (21% ± 6 vs. 9% ± 4, p \ 0.02). Before fluid expansion HLI Ò was correlated with DPP (r 2 = 0.68, p \ 0.01, Fig. 1 ). HLI and DPP were significantly correlated with change in IC induced by fluid expansion (r 2 = 0.53 and r 2 = 0.73, respectively). OBJECTIVES. The primary end point of this study was to evaluate the RVD of the IVC in ICU patients with spontaneous breathing.

METHODS. ICU patient with spontaneous breathing and signs of hypoperfusion (oliguria, mottles, serum lactate level [2 mmol/l) were eligible after the approval of the local Ethics Committee. We excluded patients with acute heart failure with pulmonary edema, moribund and arrhythmic patients. The trans thoracic echocardiographic (TTE) evaluation was done by 4 confirmed intensivists (level [2 in echocardiography). The aortic diameter measured at the LV outflow chamber and the TVI were measured. The vena cava inferior diameters at inspiration and at expiration were measured on the sub costal view. The RVD of the IVC was defined as the (maximal IVC diameter -minimal IVC diameter)/maximal IVC diameter. These measures were realized at T0, before fluid challenge, and after a fluid challenge of 500 ml of HES 6% (130.9/1) over 30 min (T30). Patients with an increase of TVI of more than 15% were considered as responders to the fluid challenge. The measures of TVI and of the RVD of the IVC were validated by an experimented intensivist and echographist (level 3) after blinding the patient' name and of the times of measurement. ROC curves were constructed, and the cut off was determined as the closest point of the ROC curve to the ideal point (sensibility = specificity = 1). The values are expressed as median and extremes. OBJECTIVES. Our objective was to test whether non invasive assessment by trans thoracic echocardiography of sub aortic velocity time index (VTI) variation after a low volume of fluid infusion (100 ml of hydroxy ethyl starch, HES) can predict fluid responsiveness.

METHODS. Sub aortic VTI was measured by transthoracic echocardiography before fluid infusion (baseline) in 39 sedated patients with acute circulatory failure and low tidal volume mechanical ventilation in whom volume expansion was planned. Then, VTI was recorded after 100 ml of fluid infusion over 1 min, and after an additional infusion of 400 ml of HES over 1 min. We measured the variation of VTI after 100 ml of fluid (DVTI 100 ) for each patient. Receiver operating characteristic (ROC) curves were generated for DVTI 100 in all patients. When available, ROC curves were also generated for pulse pressure variation (PPV) and central venous pressure (CVP). , volumes (GEDVI) and variabilities (SVV, PPV) have been suggested to predict volume responsiveness (VR). The final classification of a patient as ''volume responsive'' is usually made by a volume challenge (VC) with an infusion of a pre-defined amount of fluid over a certain time. Among many variations of VCs, the infusion of 7 ml crystalloid over 30 min is one of the most established. Despite superior predictive capabilities of SVV, PPV and GEDI compared to CVP and PAWP in a number of studies, they fail to predict VR in a substantial number of patients. Furthermore, the use of these parameters is limited due to femoral access of the CVC (GEDI; CVP) or the absence of controlled ventilation and/or sinus rhythm (SVV, PPV). Repeated ''exploratory'' VCs with 7 ml/kg might result in volume overload in some patients.

OBJECTIVES. Therefore, we investigated the usefulness of a ''small VC'' with 3.5 ml/kg crystalloid over 15 min compared to a standard VC with 7 ml/kg over 30 min.

In 48 patients equipped with PiCCO hemodynamic monitoring we performed a 30 min VC with 7 ml/kg of crystalloid. During the VC transpulmonary thermodilution (TD) was performed at 0, 15 and 30 min to obtain TD-derived CI (CI td ). Additionally pulse contour CI (CI pc ) was recorded in intervals of 5 min. INTRODUCTION. The prevalence of obesity, defined as a Body Mass Index (BMI) C30 kg/ m 2 , reaches epidemic proportions. It is not only a risk factor for health problems, but also exacerbates illness progression. Consequently, the number of obese patients on the Intensive Care Unit (ICU) has increased enormously. Caring for obese patients can be quite challenging due to the weight and size of this person. The extent of and specific problems associated to the care of obese ICU patients are unknown. The aim of this study is to identify and quantify problems nurses face in caring for obese patients on the ICU.

This study was performed on the ICU at the Radboud University Nijmegen Medical Centre and contained two parts. In the first part a selection was made of 44 obese patients admitted between 2004 and 2009; these patients were matched with 44 normal weight patients (BMI 18.5-24.9 kg/m 2 ). Patients were matched on gender, age, length of ICU stay and APACHE-II score. All patient files were screened for the presence and intensity of problems in caring for these patients.

In the second part 25 nurses were asked in a survey to share their experiences in caring for obese patients. They were asked about the nature, frequency and intensity of the problems they faced.

In total, 94 problems were identified in the 88 screened patient files. Seventy-two problems (76.6%) occurred in care for obese patients and 23 (23.4%) in care for normal weight patients. In both groups, most of the problems were related to activities of daily living (ADL) such as (re)positioning in bed, transfers and personal care. Surprisingly, the intensity of the problems was similar in both groups. Most of the problems were moderate (hardly to solve by one person) or severe (only to solve with two persons or special equipment). Moderate problems occurred in 32.3% of normal weight patients and in 33.3% of obese patients; severe problems 29.0 and 31.9%, respectively. This result was also confirmed by the survey. The nurses qualified most of the problems they were asked about as moderate or severe, and the frequency of the experienced problems was much higher. From the files it appeared that in 31.8% of the obese patients nurses had ADL problems. Strikingly, in the survey nurses reported that they frequently (72.0%) or even always (28.0%) experienced ADL problems in obese patients.

Nurses reported and experienced more problems in daily care for obese ICU patients compared to normal weight ICU patients. Although the intensity of the problems with obese patients did not differ from normal weight patients, the frequency in which they occur was much higher. Differences between reported problems and the survey suggest an underestimation of problems that can be solved by performing a prospective study. Nevertheless, based on these results, and taking into account that obesity will increase in the future, we recommend anticipating to the needs of the nurses whenever possible.

INTRODUCTION. Worldwide the number of obese patients (BMI [ 40) is increasing rapidly (1); this also includes patients admitted to the Intensive care units (ICU). This raises special demands on the staff, the surroundings and the equipment (2) . Often the obese patient is not mobilised according to the clinical standard this causes complications to breathing, circulation and skin etc. Furthermore the length of stay in the ICU increases and the mortality rises.

OBJECTIVES. The aim of this study therefore was to make clinical guidelines and recommendations for mobilisation of the obese ICU patient based on evidence. This will increase the knowledge and importance of mobilisation between staff and on longer term improves the daily average number of mobilisations performed with these patients. A secondary aim is that increased knowledge on this topic will improve the interdisciplinary work between the different professions based on the same overall aim.

A systematic review of the literature concerning mobilisation of the obese ICU patients was made in the year 2009-2010. The study is still work in progress analysing the literature to make guidelines and recommendations based on evidence. Furthermore evidencebased education of special trained staff in mobilisation has been conducted in January/ February to improve their knowledge of the impacts mobilisation has on the respiration, circulation and skincare etc. The education was planned to aiming at a interdisciplinary audience.

RESULTS. The preliminary results shows that it is more difficult to care for and mobilise the obese ICU patient, because there is lack of space, non-availability of the correct equipment, too few available staff members and a significant negative attitudes among the staff towards the obese patient. Recommendations are made within airway, breathing, circulation, nutrition, pain, equipment and patient experience according to the procedure of mobilisation of the obese ICU patient. The recommendation was implemented in the already performed education and resulted in a changed attitude among the participant and improved the status of mobilisation in the daily prioritization. This knowledge was obtained in the evaluationinterview conducted approximately one month after the seminar.

CONCLUSIONS. According to the literature mobilisation of the obese ICU patient needs special attention towards a safe clinical practise based on evidence with focus on both the patient and the staff. Special attention towards this group of patient is created by performing evidence based research resulting in clinical guidelines that has to be implemented through theoretical and practical education on an interdisciplinary level. Nurses are constantly exposed to the pain and suffering of those in their care 3 .

The primary aim of this study was to investigate the risk of secondary traumatic stress/compassion fatigue (STS/CF-the trauma suffered by the helping professional) and burnout (BO-emotional exhaustion, depersonalization, and reduced sense of personal accomplishment), and the potential for compassion satisfaction (CS-the fulfillment from helping others and positive collegial relationships) among nurses working in ICU. An additional goal was to test the relationship of these three constructs to each other. (1) . The use of closed suction circuits has been suggested beneficial as a prophylactic measure (2) .

OBJECTIVES. The aim of this study was to compare the incidence of VAP and the occurrence of desaturation during suction using either OSS or CSS. We also investigated contamination of the closed suction circuit and the occurrence of adverse events.

METHODS. CSS were a new product in our clinic. All staff underwent a user course supervised by the manufacturer of the closed circuit. After this, data were collected during four periods in 2009, 2 month CSS followed by 2 months OSS which was repeated twice. During the summer period CSS were used without any data collection and then followed by two periods of CSS and OSS. All mechanical ventilated patients were consecutively included. A culture of deep endotracheal aspirate and a blind microbiology brush was taken in association with the intubation, after 72 h and every Monday. After changing CSS and in case of extubation, the tip of the catheter was sent for culture. Demographic data were retrieved from the hospital database. Data were analyzed with descriptive methods.

RESULTS. The incidences of VAP were higher in the CSS group (Table 1) . Both suction systems showed almost no desaturation during and after suctioning. Positive cultures were obtained in 55% of all the retrieved CSS catheters. The microbiological flora resembled the species found in the airway cultures. There were no inter patient contamination and neither did the bronchoscopy frequency differ between OSS and CSS patients. In the CSS group six adverse events were seen; three tube occlusion and three incidences with secretion clogging.

CONCLUSIONS. The use of a CSS did not prevent VAP, in our study. There were no benefit with CSS other than maybe to protect the staff and our finding of positive culture in 55% of the cases is in line with earlier studies. OBJECTIVES. The aim of this study was to determine which intensive care patients the nurses defined as 'difficult' and their experiences in coping with such patients.

The study was carried out as a qualitative design with 18 voluntary nurses employed in five intensive care units of a research and training hospital. The data were collected using demographic characteristics form and a semi-structured interview form. Interviews with nurses were made individually and face to face. The data were evaluated by using Colaizzi's phenomenological data analysis method.

As a result of data analysis into two categories and two themes were identified. The categories were (1) difficult patient definition of the nurses, (2) the effect of difficult patients on their care, and (3) how the nurses are affected and cope with difficult patients. The nurses listed their reasons for defining some persons as difficult as difficult physical care of the patients, and the difficulty in communicating due to dementia, agitation, Alzheimer's disease or the patient's personal characteristics. The nurses said that they found taking care of patients they found difficult physically and psychologically demanding. They used methods such as finding out the patient's problem and taking appropriate measures, increasing communication with the patient and providing explanations, trying to obtain spiritual satisfaction and transferring the patient's care to another nurse when communication problems were impossible to overcome.

Intensive care nurses have difficulty in caring for and communicating with some intensive care patients due to the characteristics of the disease, physical/psychological factors and personal characteristics. We found that nurses continued the care of these ''difficult'' patients by focusing on solving their problems, transferring the care to another nurse when necessary or by trying to obtain spiritual satisfaction. METHODS. Teams of three delirium experts visited ten ICU's in the The Netherlands in which the CAM-ICU was incorporated in daily practice, twice. These teams consisted of two consultants in either psychiatry, clinical geriatrics or neurology, and either a research-physician (MMJvE) or a research-nurse (MvdB). Based on cognitive testing, inspection of the files and DSM-IV criteria for delirium, the teams classified patients as awake and not delirious, or delirious or comatose. This classification served as gold standard to which the CAM-ICU as performed by the bed-side nurses was compared. A simple 2 9 2 table was used to calculate the sensitivity and specificity. RESULTS. 15 delirium experts performed 333 assessments. 181 (54%) of these patients were assessable for delirium, 126 (38%) patients were excluded because the level of consciousness was too low, and 27 (8%) patients were non-assessable due to other reasons. Overall, we found a sensitivity of 47% (95% CI 35-58%) and a specificity of 98% (95% CI 92-99%). The strengths of this study include the large numbers, the multicentre design, the extensive evaluations by teams of various delirium experts and the independent assessments of delirium experts and bed-side nurses. A limitations is the time interval between the expert assessment and the administration of the CAM-ICU (mean 125 min; standard deviation 118 min). There were striking differences in implementation strategies of the CAM-ICU between the centres. Tables 1, 2 . 23rd ESICM Annual Congress -Barcelona, Spain -9-13 October 2010 S341

INTRODUCTION. Presence of expiratory ineffective efforts in mechanically ventilated patients is a common problem associated with increased duration of mechanical ventilation, length of stay and also a higher cost and mortality. Nowadays, identification and categorization of expiratory asynchronies can only be done at the bedside with the continuous observation of the ventilator interface. Nurses must be skilled to understand non appropriate situations of anomalous patient-ventilator interactions.

OBJECTIVES. We tested the hypothesis that after specific training nurses would acquire enough skills to detect expiratory efforts as intensive care expert physicians would do.

Training Phase: Nurses were provided with selected bibliography on patient ventilator interaction and afterwards trained by intensivists with expertise on mechanical ventilation (2 h/day during 20 days) on airway pressure, flow and volume waveforms identification and eye interpretation of early and late ineffective expiratory efforts during expiration. Validation phase: Airflow and airway pressure waveforms were obtained from 8 different ICU mechanically ventilated patients using and acquisition and processing biomedical signal software (Better Care Ò ). One thousand and seven breaths were randomly selected from a total of 2,608,800 breaths. Subsequently, selected breaths were blindly analyzed by 2 trained nurses and 3 intensivists to identify ineffective expiratory efforts. INTRODUCTION. Several publications indicate that manual hyperinflation is a widely used measure in the ICU, but more important is the fact that there is no uniformity in the implementation of this measure. This is also on my ward. In literature there are a number of reasons given to start manual hyperinflation: abolish mucus retension, improve oxygenation and removal of atelectasis. The positive effects are improved compliance, improved oxygenation and a decrease in the number of VAP's (Ventilator Associated Pneumonia). The negative effects are a decrease in cardiac output due to high peak pressures, an increased risk of baro-/volutrauma and the risk of giving too much tidal volumes. The risk of barotrauma increase with pressures above 40 cmH 2 O. Other side effects include the development of a pneumothorax and increased ICP (intra cranial pressure).

OBJECTIVES. Creating more awareness of the procedure with lower peak pressures as a result.

METHODS. Through literature review, clinical courses and the introduction of a pressure gauge achieve greater uniformity and awareness of the procedure. We used a flow analyzer of IMT Medical, a laptop with Flowlab software version 4.1.1 and an artificial lung to demonstrate how much pressure and volume is generated during manual hyperinflation. CONCLUSIONS. Compliance with BTS Guidelines could be improved. Unsurprisingly co-morbidities were frequent, but did not seem to affect outcome. Use of a pneumonia severity assessment tool was sub-optimal, however mean CURB-65 score didn't correlate with that recommended to prompt critical care assessment. Apart from functional status, we are unable currently to identify any factors in this age group which can be used to guide critical care admission decision making. CONCLUSIONS. In our study the incidence of complicated pneumonia was 316/10,000 patients admitted in PICU. In necrotizing pneumonias the blood cultures were more positive than in non-necrotizing patients. Although the surgical approach in necrotizing pneumonia is controversial, it resulted in a insignificantly lower mortality rate, comparing with non-necrotizing pneumonias. BACKGROUND. Community-acquired pneumonia (CAP) of mixed etiology has frequently been described in the literature, but its clinical significance remains unknown. The aim of this study was to describe the prevalence, clinical characteristics, and outcome of severe CAP of mixed etiology in ICU patients.

A 5-year prospective study was conducted on consecutive patients with severe CAP admitted to ICU in whom an extensive microbiological investigation was performed.

RESULTS. 316 patients were included. A single pathogen was detected in 128 (40.5%) cases, while two or more pathogens in 38 (12.0%) cases. The most frequent pathogens' combinations were those of two bacteria (28.9%) and bacterium plus virus (21.1%). Compared with patients with monomicrobial pneumonia, patients with mixed pneumonia were older, had higher severity score (PSI) and were more likely to have previous chronic pulmonary disease (see Table 1 below). Moreover, mixed CAP patients showed similar clinical and analytical data at admission but increases in the frequency of respiratory distress and in length of stay and a trend to higher orotracheal intubation and mortality rates.

A mixed etiology was detected in 12% of cases with CAP requiring ICU hospitalization and was associated with older age and increased severity. Despite similar radiological features (n of involved lobes, pleural effusion) at admission, CAP with mixed etiology showed a trend to worse clinical course and outcomes than monomicrobial pneumonia. OBJECTIVES. To assess the incidence and aetiology of pneumonia in a mixed medicalsurgical ICU, in order to develop local epidemiologically guided protocols to reduce antibiotic resistance selection in patients with pneumonia.

METHODS. Retrospective observational study on prospectively collected data in a mixed medical-surgical ICU of a secondary care Italian hospital. At our institution, epidemiological data on infections and data on antibiotic use are recorded since 1992; in 2005 a new electronic recording of ICU infections was introduced. Type of infection, germ characteristics, clinician diagnosis and antibiotic use were prospectively collected in an electronic database and retrospectively reviewed. Antibiotic exposure index was calculated as each antibiotic total amount administered divided by its defined daily dose times total days of admission.

Between 2005 and 2009 a total of 690 patients were admitted to our ICU. Pneumonia was the commonest infectious disease at admission (148 cases, 21% of patients), and the commonest infectious complication during ICU stay (38 new occurrences, 21% of total pneumonia patients). Table 1 shows major epidemiological findings in the study population. The incidence of acquired pneumonia was remarkable: 9.7 cases every 1,000 days of mechanical ventilation. The most frequent isolated organisms were S. aureus (18 patients) and P. aeruginosa (12 patients). Methicillin-resistant S. aureus (MRSA) accounted for 61% of pneumonia caused by S. aureus, and its prevalence matched closely the exposure index to vancomycin. Such a high incidence of MRSA is consistent with other records in Mediterranean countries. Carbapenem-resistant P. aeruginosa was somewhat less of a problem (33% of pneumonia by P. aeruginosa), and was not apparently associated with antibiotic exposure, at least within the unit.

CONCLUSIONS. In our retrospective observational study we found a high incidence of pneumonia at our institution, as well as a high percentage of MRSA, the latter with strong relationship with exposure to vancomycin. New protocols for infection containment and antibiotic usage are urgently needed. INTRODUCTION. Community-acquired pneumonia (CAP) carries a high morbidity and mortality. A major problem is the insufficient monitoring of CAP by standard chest radiography, as the evaluation depends highly on the observer and the extent of pulmonary infiltration cannot be assessed properly (1).

OBJECTIVES. The aim of our study was to compare the process of inflammation in CAP measured by alveolar Nitric oxide (NO)-analysis (2) in exhaled breath and the extent of the inflammatory infiltration by electrical impedance tomography (EIT) (3) in spontaneously breathing patients.

After approval of the local ethic committee and obtained written informed consent 24 patients with CAP were included in the study. All patients showed an acute pulmonary infiltration in chest X-ray, pulmonary symptoms (coughing, shortness of breath), positive findings in auscultation, leukocytosis, elevated CRP and a pneumonia severity index C4. NO analyses (Analyser CLD88sp, ECO MEDICS, Dürnten, Switzerland) were performed at t0 (up to 24 h after admission), t1 (2 days after admission) and t2 (4 days after admission. EIT measurements (EIT Evaluation KIT, Dräger Medical, Lübeck, Germany) were performed at t0 and t2 and inhomogeneity of ventilation was assessed by offline analysis. All measurements were made at beside in sitting position. Data were compared by t test and regression analysis.

RESULTS. There was no significant correlation between the alveolar NO concentration and the extent of inhomogeneity of the local infiltration measured by EIT. Also during the study the time course of the inhomogeneity index was not correlated with change in exhaled NO. The right/left distribution of the pulmonary infiltration in the chest X-ray and the EIT measurement showed a positive correlation (p \ 0.0001; r 2 = 0.69).

CONCLUSIONS. Pulmonary regional infiltration in CAP measured by EIT can not predict the actual alveolar process of inflammation in the lung. Nevertheless the monitoring devices give additional information to better evaluate the time course of inflammation and the dimension of the respiratory dysfunction in diseased lung.

Organizing pneumonia (OP) presenting as acute respiratory failure (ARF) is a relatively rare disease, and was only previously specifically reported in 2 small series [1, 2] , with mortality up to 90%. These studies were performed before the publication of international consensus classification of idiopathic interstitial pneumonias in 2001 [4] .

OBJECTIVES. To compare clinical features and prognosis of patients with OP with those of patients presenting diffuse alveolar damage (DAD), during ARF.

Design: Retrospective monocentric study in a university hospital conducted during an 11 yr-period. 

To determine predictors of NIV failure in patients who were intubated for respiratory failure and extubated directly to NIV.

METHODS. This is a retrospective analysis of prospectively collected data from January 2008 to Dec 2009. Patients with respiratory failure were mechanically ventilated in a university hospital's medical intensive care unit (ICU) and subsequently extubated to NIV. Physiological and biochemical parameters, using arterial blood gas measurements, were collected at the end of the spontaneous breathing trial and 1 h after the application of NIV. Failure of NIV was defined as respiratory failure requiring re-intubation within 48 h.

Out of 175 patients, 78.9% were successfully extubated to NIV. Success rates were 70.6% in patients with chronic obstructive pulmonary disease (COPD) and 80.9% in other patients (p = 0.24). Patients who failed NIV were more tachypnoeic, acidaemic and hypercapnic pre-NIV, and more tachycardic, hypotensive, acidaemic, hypercapnic and hypoxaemic post-NIV (p all.05). On logistic regression analysis, three physiologic parameters predicted NIV failure: pre-NIV respiratory rate (OR 1.57, 95% CI 1.07-2.31 per 5 breaths increase), post-NIV heart rate (OR 1.28, 95% CI 1.08-1.52 per 5 beats increase) and post-NIV systolic blood pressure (OR 1.15, 95% CI 1.02-1.29 per 5 mmHg decrease).

CONCLUSIONS. Physiologic parameters, including the respiratory rate pre-NIV, and heart rate and systolic blood pressure post-NIV, independently predict NIV failure post-extubation. These parameters should be taken into account in the decision to extubate directly to NIV. INTRODUCTION. Discontinuation of mechanical ventilation in critically ill patients is a challenging task and involves a careful weighting of the benefits of early extubation and the risks of premature spontaneous breathing trial (SBT). Only a few studies have explored indices derived from both heart rate and breathing pattern variability analysis for the estimation of weaning readiness.

OBJECTIVES. To investigate heart rate (HR) and respiratory rate (RR) complexity in patients with weaning failure or success, using both linear and nonlinear techniques from signal processing theory.

METHODS. Forty-two surgical patients were enrolled in the study. There were 24 who passed and 18 who failed a weaning trial. Signals were analyzed for 10 min during two phases: Despite of passing the protocol the decision to extubate was postponed in some patients.

To gain insight on the physicians reasons for continuing mechanical ventilation after passing the wean screen protocol.

A wean screen protocol was introduced at a mixed medical (neuro-)surgical ICU of a teaching hospital in December 2009 to April 2010. Ventilation practitioners assessed ventilated patients and recorded the physicians reasons for continuing mechanical ventilation despite of passing the wean screen protocol.

. 223 patients were ventilated in this period. 135 daily screens were performed, 111 screens were successful. Only 24 passed wean screens resulted in extubation. The rate of extubation was 22%. 78% screens did not lead to liberation from mechanical ventilation. The extubation rate does not correspond with the findings of the ABC trial with an extubation rate of 54%. Table 1 shows the physicians' reasons to continuing mechanical ventilation. It should be noted that all patients with an unsafe airway were patients with a Glasgow Coma Scale (GCS) of B 8 [intracerebral haemorrhage (68%), cerebral infections (13%), post-CPR encephalopathy (13%) and severe brain injury (6%)]. We accomplished a reduction in the use of sedatives (-12% midazolam and -35% propofol) and morphine (-27%) ( Table 1 ). The amount of time spend on ventilators decreased, albeit not significantly (p = 0.902). This was probably due to the VAP-ventilatorbundle (introduced last year), the heterogeneity of our cohort and the already short MV-duration. . Non-invasive ventilation (NIV) has been utilized in selected patients with hypoxemic ARF to avert endotracheal intubation, which is related to life-threatening complications. NIV has been also proposed to facilitate weaning and extubation in patients with hypercapnic ARF. So far, no controlled randomized study has investigated the potential role of NIV in weaning patients with hypoxemic ARF. OBJECTIVES. We designed this pilot study to assess safety and feasibility of NIV to wean hypoxemic ARF patients. Twenty mechanically ventilated patients with hypoxemic ARF were randomized to receive early extubation followed by NIV application via helmet (Helmet group) or conventional weaning through the endotracheal tube (Tube group). Primary outcomes were the duration of invasive mechanical ventilation and the adherence to the study protocol. Secondary outcomes were protocol failure (i.e. need for re-intubation), ICU and hospital mortality, rate of tracheotomy, duration of continuous intravenous sedation, weaning time, and septic complications. Table 1 . Weaning through helmet by NIV application following early extubation was safe and feasible. Overall the adherence to the study design was 90%. In addition, in the helmet group, there was a significant reduction in the rate of tracheotomy and a trend toward a lower rate of protocol failure, and fewer days on invasive ventilation. There was no difference with respect to days of continuous sedation, ICU and hospital mortality, weaning time and septic complications. (1) . Delirium is a common occurence on the ICU and is associated with increased length of stay (LOS) and poor outcomes (2) . OBJECTIVES. We developed a combined daily sedation hold, delirium management, and weaning (SDW) protocol and implemented this to reduce ICU LOS and improve outcomes. METHODS. A SDW protocol was implemented in 2009. We prospectively audited all patients from January to March 2010. Delirium was measured using the ICDSC. Data was analysed using Graphpad statistical software. RESULTS. 109 consecutive patients were analysed. The incidence of delirium was 35% (38 pts). Of these, 63% (24) had risk factors for delirium. There was no difference in onset of delirium between sexes, age, type of admission, or severity of illness. However, in patients with delirium, duration of mechanical ventilation (MV) and ICU LOS were significantly longer and there was a trend towards increased hospital LOS ( CONCLUSIONS. Measuring the linear dependence of variables through time by K 2 and ø may be used to determine non-linear behavior between the variables of the EMMV. Non-linear behavior during weaning perhaps indicates the dependency of, either the resistance or compliance of the respiratory system, on the ventilatory support (i.e. Pi). Accordingly, K 2 and ø, estimated at the frequency interval form 6 to 1 (h) -1 , can provide information concerning to the dynamics of the respiratory system that can be used as a complement to determine the suitability of the MV withdrawal. OBJECTIVES. To study the potential superiority of APRV on CMV in a subgroup of patients with severe ARDS.

METHODS. Retrospective observational study on 39 patients severe ARDS who were admitted between July 2008 and January 2010 to Mafraq Hospital ICU in UAE. The diagnosis of ARDS was based on presence of bilateral infiltrates in CXR and P/F ratio of less than 200 in absence of evidence of elevated left atrial pressure. All patients were managed according to ARDSnet guidelines using low tidal volume CMV and IV steroids. Criteria for transition to Airway Pressure Release Ventilation (APRV) included failure to wean down FiO2 below 60% after 24 h, hemodynamic instability due to high PEEP, and failure to maintain plateau airway pressure below 30 cmH2O. Initial settings of APRV were Ph 26, Pl 2, Th 5, and Tl 0.5 with titration of FiO 2 as required keeping PaO 2 more than 60 mmHg. We compared the outcome of CMV and APRV groups with special concern to the duration of mechanical ventilation, requirement for tracheostomy, and survival to ICU discharge.

Twenty four male and 15 females were included in the study with a mean age of 42 years (± 24). Fourteen out of them fulfilled the criteria and were shifted to APRV within 24 h of initiating mechanical ventilation. Ten out of 14 (70%) patients in the APRV arm survived to ICU discharge versus 16 out of 24 (67%) patients in CMV group (p 0.45). Survivors in APRV group spent significantly shorter periods of mechanical ventilation compared to survivors in CMV group (9.6 vs. 12.1 days p 0.03). While 8 out of 16 (50%) survivors in CMV required tracheostomy for prolonged intubation or recurrent lavage, only 2 out of 10 (20%) survivors in APRV group required tracheostomy tube placement (p 0.02).

We concluded that APRV can be effectively used as rescue measure of ventilation in patients with severe ARDS. Although our study does not show any mortality benefit of using APRV over CMV, there was a shorter ventilation days and ICU stay using APRV. We strongly recommend further studies to investigate the probability of using APRV as initial mode of ventilation in this subset of patients. 

Weaning from mechanical ventilation is a common daily procedure when caring for critically ill patients, and a lifesaving practice on which nurses are taking an increasing role with the introduction of nurse-led protocols. The literature supports that nurse-led protocols facilitate weaning and increase nurses' input in decision-making. On the other hand, decision-making is a complex function affected by the nature of the task, the decision environment and the characteristics of the decision maker. Although the cognitive process of clinical decision-making has been investigated with many different methodologies, little is known about the decision environment and its impact on decisions' during the weaning process.

OBJECTIVES. This paper aims to address one of the factors of the clinical environment and its impact on the decisions when discontinuing mechanical ventilation.

METHODS. This paper is part of a large comparative ethnographic study looking at nurses' input during the weaning process of mechanically ventilated patients. Participant observation of critical care nurses took place in an 12-bedded ICU in Greece and an 18-bedded ICU in Scotland for 6 months each to examine nurses' involvement in the decisions made. In-depth semi-structured interviews with the nurses followed focusing on how nurses perceived their participation in the decisions made. Data from field notes and interview transcripts were analysed thematically using the Qualitative Data Analysis software NVivo, version 8.

Inter-personal and inter-professional relationships were considered revealing influences of nurses' input in decision-making. Clinicians' personality played a significant role in their involvement in decisions, whereas trust and appreciation, the sense of support and the sense of accountability were also considerable dynamics of inter-professional relationships and predisposed decision-making.

Clinical decision-making is a multi-dynamic process specifically in complex clinical long-term situations such as weaning. Aspects of the decision environment, such as the interprofessional relationships should be acknowledged when introducing methods to enhance nurses' role in teamwork and collaborative decision-making in order to improve the weaning process of ventilated patients and their outcome. OBJECTIVES. The objective of our study was to analyze the temporal trends and outcomes of two cohorts of patients ventilated with PSV and PAV+.

A cohort of 16 consecutive patients who were ventilated with PAV+ and another cohort of 15 consecutive patients who were ventilated with PSV were compared. All patients had the same inclusion criteria (gas exchange, ventilatory mechanics, PEEP level, resolution/stabilization of the cause leading to invasive MV and appropriate level of consciousness). Both modes were adjusted to predefined clinical criteria (PSV to reach a respiratory rate about 25 bpm and PAV+ to reach a physiological inspiratory effort INTRODUCTION. Presence of expiratory asynchronies (EA) (ineffective efforts, cough and continued contraction of inspiratory muscles) is a common problem associated with increased duration of mechanical ventilation, longer stay, higher costs and increased mortality. Because of the lack of systems that automatically detect and report EA, their identification is currently done by examining ventilator interface at the bedside or by applying dedicated algorithms in investigational conditions.

Validate the accuracy of linear mathematical algorithms to automatically detect EA built in a new computerized system that grabs and process data from different bedside ICU monitors and mechanical ventilators.

Observational and prospective study in a general ICU of 16 beds. Two beds were equipped with a software (Better Care Ò ), a technological platform responsible for data acquisition and synchronization, processing, storing-as non static and processable DICOM objects-and also for integrating all this data with health information systems. By using the Better Care Ò platform, a total of 2,600,000 breaths from 8 consecutive adult patients were collected with at least 24 h of mechanical ventilation. Algorithm #1: The EA algorithm consisted in a mathematical analysis of the airflow and airway pressure waveform variations during expiration not followed by a mechanical breath. Algorithm #2: Designed to select 1,000 breaths out of the total number. This algorithm sorted and classified the breaths by the percentage of deviation from the expected expiratory curve. The result was 1,000 breaths covering most of the shapes the expiratory curve could have. Five expert attendant physicians independently analyzed the 1,000 selected breaths and classified them as EA or not. The EA algorithm processed the same 1,000 selected breaths and assigned a percentage to each one, according to the variation in the shape and direction of the expiratory airflow and airway pressure curves. The expert criterion against the EA algorithm scores was used to construct a logistic regression model. We calculated sensitivity, specificity, positive predictive value and negative predictive value. The predictive performance of EA algorithm was evaluated using ROC curves.

Optimal sensitivity and specificity were achieved by setting the cut-off point at a EA algorithm score of 42%. A variation in the shape and direction of the expiratory airflow and airway pressure curves [42% compared to the theoretical curve identified an EA with a sensitivity of 91.5%, specificity of 91.7%, a positive predictive power of 80.3% and a negative predictive power of 96.7%. INTRODUCTION. Near-infrared spectroscopy (NIRS) in combination with a vascular occlusion test (VOT) has been proposed to assess and identify metabolic and microcirculatory alterations during sepsis and shock in critically ill patients. However, to automatize repeated measurements at the bedside, this technique can potentially cause discomfort to the patient. Vascular arterial occlusion performed in the finger may be a more attractive method to execute repeated measurements at the bedside because of more tolerability from the patient. We have previously showed in healthy volunteers that NIRS can be used on finger to assess the StO2 response to VOT and that 5 min was an adequate occlusion time to provide the best curve fit for NIRS dynamic variables 1 .

OBJECTIVES. We aimed to investigate whether StO2 response to VOT obtained from the finger could predict conventional StO2 response measurements obtained from the thenar of critically ill patients.

Parameters of StO2 response were measured with an InSpectra Spectrometer model 650 (Hutchinson Technology Inc.) equipped with a 15-mm or a 5-mm probe. The 15mm probe was placed over the thenar eminence and the 5-mm probe was place over the ventral face of the middle finger. We performed in each patient a series of two vascular occlusion tests (VOT): one on the finger (5 min) followed by one on the arm (3 min). The measurements were obtained within 24 h of intensive care admission and every 24 h thereafter until day 3. VOT-derived StO2 traces were analyzed for baseline, ischemic (RdecStO2, %/min) and reperfusion (RincStO2, %/s) parameters.

We performed 63 paired of NIRS measurements in 25 critically ill patients (Age 62 ± 15; 16 M/9F). Although StO2 did not differ significantly between thenar and finger (76% ± 11 vs. 79% ± 12; P = 0.31), RincStO2 and RdecStO2 were statistically lower in the finger (1.5%/s ± 0.7 vs. 2.5%/s ± 0.8, P = 0.001; 3.1%/min ± 1.0 vs. 11%/min ± 3.3; P = 0.001). We performed bivariate linear model with correlated errors in which StO2 outcomes on thenar and on finger were treated as responses. The correlation was significant for StO2 and RincStO2, but not for RdecStO2 (Table 1) . Furthermore, mixed model analysis showed that thenar-StO2 as dependent variable could be significantly predicted by finger-StO2 parameters with estimation coefficient (± SE) of 0.7 ± 0.06 (P = 0.0001), 0.9 ± 0.09 (P = 0.0001) and 0.98 ± 0.3 (P = 0.0061) for StO2, RincStO2 and RdecStO2, respectively. CORRELATION OF STO2 RESPONSE: FINGER VS. THENAR 

A prospective randomized clinical trial performed in ICU's of an university and teaching hospital during a 2.5 year period, involving 120 septic and non-septic patients, randomized (after stratification) to hemodynamic monitoring, by PiCCO TM or PAC with both techniques allowing cardiac output and central/mixed venous O 2 saturation monitoring.

METHODS. Hemodynamic management was guided by extravascular lung water index (EVLWI) and global end-diastolic volume index (GEDVI) in the PiCCO TM group and by the pulmonary capillary wedge pressure (PCWP) in the PAC group for 3 consecutive days. Primary outcome measures were ventilator-free days (VFD), for which the study was powered, and lengths of stay in ICU and hospital. Secondary measures were the course of cardiorespiratory parameters, fluid and vasopressor requirements, lactate levels, organ functions and mortality.

In the study period, 72 septic and 48 non-septic patients were included. 60 patients received a PiCCO TM and 60 a PAC catheter. Monitoring arms were comparable at baseline, although sepsis differed from non-sepsis in hemodynamics and severity of lung injury. Premorbidity was greater in non-septic patients. The fluid infusions and balances did not differ between monitoring arms, except at t = 24 h when the PiCCO TM group had a more positive balance (P = 0.044). Cardiac index and central venous O 2 saturation increased more in the course of time in the PiCCO TM than in the PAC group. The decrease in norepinephrine requirements strongly tended to favor the PiCCO TM group (P = 0.06). The course of lactate levels and organ failure did not differ between monitoring arms. VFD did not differ among monitoring arms. PiCCO TM monitoring was associated with relatively fewer mechanical ventilation and ICU days in sepsis but more in non-sepsis (after day 28). The changes in respiratory parameters, SOFA and number of catheter-related complications did not differ among the arms of the study. Overall, 19 patients (32%) died in the PiCCO TM group before day 28 and 21 (35%) in the PAC group (P = 0.85).

CONCLUSION. Hemodynamic management guided by PiCCO TM monitoring is safe and results in better tissue oxygenation than guidance by PAC, without inducing pulmonary overhydration, in septic and non-septic, critically ill patients. This was associated with fewer mechanical ventilation and ICU days in patients with sepsis but more days in patients with non-sepsis (after day 28), partly attributable to greater cardiovascular premorbidity in the latter. The major primary and secondary endpoints, VFD and mortality, were not affected. INTRODUCTION. Non-invasive evaluation of endothelial function may be easily accomplished by ultrasound assessment of flow-mediated vasodilation (FMD) of the brachial artery, but this technique has not been fully explored in septic patients.

OBJECTIVES. This prospective study aims to investigate the role o FMD analysis on intra hospital prognosis of patients with severe sepsis and septic shock.

Adult patients admitted to the Intensive Care Unit with a diagnosis of severe sepsis or septic shock (\24 h of duration) were consecutively included. FMD of the brachial artery was measured upon admission and after 24 and 72 h using a high-frequency linear transducer (7.5-10 MHz) according to internationally accepted protocols. A group of apparently health subjects paired for gender and age was used as controls for FMD analysis. Patients were followed up to discharge or death.

We studied 42 adult patients 9 mean age 51 ± 19 years, 26 females, 79% on vasopressors with sepsis predominantly of abdominal or respiratory etiology (75%). APACHE II risk score was 23 ± 7 and intra hospital mortality rate was 33%. FMD was similar in patients with or without use of vasopressors at baseline (p = 0.56). FMD in septic patients was significantly lower than in health controls (1.5 ± 7 vs. 6 ± 4%; p \ 0.001). We observed that survivors depicted a gradual improvement on endothelial function, so that 72 h after sepsis onset FMD was significantly lower in nonsurvivors (-3.3 ± 10 vs. 5.2 ± 4%; p \ 0.05; time-group interaction p value = 0.03).

CONCLUSIONS. Brachial FMD is altered in septic patients with hemodynamic instability and its improvement may be an early marker of favorable prognosis. INTRODUCTION. Change in pulse pressure variation (dPP) and respiratory variation of the pulse oxymetry plethysmogram (POP) may predict the hemodynamic effect of PEEP in mechanically ventilated patients [1, 2] . Reported comparisons [3, 4] between POP variations (POPv) and CO or dPP are based on selection of 3-5 consecutive breaths (dPP3b) during a ''stable'' period of POP. Recently, a fully automatic ventilation mode (IntelliVent Ò , Hamilton Medical, Switzerland) that incorporates an automatic and continuous POPv calculation (HLI Ò ) using a dedicated algorithm has been developed. The present study was designed to compare dPP3b, dPP calculated with the algorithm as HLI Ò (dPPalg) and HLI Ò. .

. 11 sedated ICU patients ventilated with Hamilton Medical S1 ventilator (with integrated pulse oxymetry (PO)) were included (age = 66 ± 13 years, SAPS II = 54 ± 24, no arrhythmia, norepinephrine: 1.0 ± 1.2 mg/h in 6 patients, MAP = 74 ± 10 mmHg, Vt = 8.4 ± 0.6 mL/kg). Waveforms of PO from a finger sensor and of blood pressure from a radial catheter were recorded for 1-2 h. From the waveforms, breath by breath (using respiratory flow signal), without pre-selection of stable periods and using known formula [1] dPP3b (averaging 3 breaths without any filtering), dPPalg and HLI Ò were automatically obtained (Matlab Ò ). dPP3b was compared to dPPalg (2600 pairs) using Mann-Whitney t test. 18287 pairs of HLI Ò and dPPalg values (see Fig. 1 below) were compared using linear regression and Bland-Altman method. A dPPalg threshold value of 13% was used to generate HLI Ò ROC curves.

RESULTS. dPP3b and dPPalg were significantly correlated (r 2 = 0.99, p \ 0.001), but standard deviation of dPP3b were higher than the standard deviation of dPPalg (1.2 ± 0.4 vs. 0.2 ± 0.1%, p \ 0.001). dPPalg and HLI Ò were correlated (r 2 = 0.73, p \ 0.001), mean difference was 2 ± 4%. HLI Ò above 13% predicted dPPalg above 13% with a sensitivity of 95% and specificity of 82% (ROC: 0.96).

CONCLUSIONS. dPP3b should be interpreted with caution due to the high variance of this index. In real conditions and during long time monitoring dPPalg and HLI Ò are in acceptable agreement and HLI Ò may help estimating continuously the hemodynamic effects of ventilation. INTRODUCTION. Transthoracic echocardiography (TTE) is supposed not to be useful in ventilated patients (Pt). Echocardiography is usually performed transesophageally in ventilated Pt and is thought to be independent of the examiner's skills. We want to demonstrate that TTE in ventilated Pt could be learned even by medical students with reasonable results and that TTE could add useful informations for interpretation of the hemodynamic status.

OBJECTIVES. In a prospective observational study 42 consecutive patients (pt) were enrolled in a 9-bed medical intensive care unit of a university hospital. Inclusion criteria was septic shock according to actual guidelines.

Transthoracic echocardiography (Acuson CV70, Siemens, Germany) was performed by a medical student in each subject on day 1, day 7 and survival was reported on day 28. TTE-examination was reduced to an apical 4-chamber view for interpretation of left ventricular global function and calculation of left ventricular ejection fraction (EF) with the Simpson method and to a subcostal view in order to examine the diameter of the inferior caval vein (IVC) and to rule out pericardial effusion. Each examination was digitally recorded and was interpreted by an experienced cardiologist. Every single Pt was mechanically ventilated. Cardiac output (CO) was measured with the transpulmonary thermodilutional technique (PiCCO-Catheter, Pulsion, Germany). The insertion of the PiCCO-Catheter took place due to an individual physician's decision. CRP was measured as an parameter of inflammation.

RESULTS. 42 pt, mean age 71 years ± 7.22, 27 male (64%), 23 Pt with known coronary artery disease (55%), 2 Pt with known dilated cardiomyopathy (5%). Mean APACHE II-score 35.9 ± 6.8. 22 pt died within 28 days (52%). PiCCO-Catheter was inserted in 21 Pt (50%). TTE could be successfully performed in 38 Pt (91%). The following values are expressed as mean values ± SD, Student's t test, p \ 0.05 denotes statistic significance. EF on day 1 43.6% ± 16.2, EF on day 7 42.9% ± 15.7, p = 0.04. IVC on day 1 20.3 mm ± 4.5, IVC on day 7 20.2 mm ± 4.5, p = ns. CO on day 1 5.5 l/min ± 3.4, CO on day 7 4.0 l/min ± 3.4, p = 0.05. CRP on day 1 19.82 mg/dl ± 10.4, CRP on day 7 9.91 mg/dl±, p\ 0.0001. Pericardial effusion in no Pt.

In older Pt coronary artery disease is common and EF is at the start of septic shock severely diminished. EF decreased slightly in the early course of septic shock, may be as an expression of septic cardiomyopathy. The IVC diameter did not change and may not be useful as a predictor of preload in ventilated Pt. CO decreased over time as the hyperdynamic circulation in septic shock is getting normalised. TTE adds useful hemodynamic information and should be performed in each ventilated Pt. TTE could be performed in almost each ventilated Pt and is easily learned even by medical students. (1, 2) , which can often be caused by anaemia. In current guidelines the transfusion trigger is haemoglobin (Hb) \ 7 g/dl, but there is no recommendation for ScvO 2 (3).

OBJECTIVES. The aim of this retrospective study was to evaluate the change in ScvO 2 before and after transfusion and to reveal whether CO 2 -gap reflects it.

METHODS. Over a 6 month period Hb, ScvO 2 , CO 2 -gap and O 2 -extraction ratio (O 2 ER) were recorded before and after transfusion. Data are presented as median [interquartile range], for statistical analysis Wilcoxon, Mann-Whitney tests and Pearson correlation were used as appropriate.

RESULTS. Out of 128 transfusion events the ScvO 2 was measured in 50 cases. After transfusion Hb increased significantly: 7.7 [7.1-8.2]-9.0 [7.9-9.7] g/dl, p \ 0.001. The median ScvO 2 was 71 %, therefore two groups were created: ''Low'' (ScvO 2 \ 71 %, n = 27); ''High'' (ScvO 2 C71%; n = 23). Hb increased significantly in both groups (p \ 0.001), but ScvO 2 CONCLUSIONS. In the High-group the low Hb levels did not cause oxygen debt, as after transfusion Hb increased significantly but ScvO 2 did not, and O 2 ER and CO 2 -gap were within the normal range. Our results give further support that not only the Hb level should serve as a transfusion trigger, but measures of oxygen debt such as ScvO 2 and CO 2 -gap should also be considered, hence unnecessary transfusions could be avoided. INTRODUCTION. IntelliVent Ò is a fully closed loop ventilation designed to keep the patient within target ranges of EtCO2 and SpO2. The system includes an automatic adjustment of PEEP and FiO2 following the ARDSnetwork tables [1] . If required PEEP is changed by 1 cmH 2 O every 6 min with a maximal possible value set by the user or depending on an automatic and continuous calculation of the respiratory variations of the plethysmogram from an integrated pulse oxymeter (HLI Ò ), i.e. the higher the HLI Ò the lower the maximal PEEP allowed by the system. The present study was designed to estimate whether changes in PEEP are reflected in HLI Ò changes.

In 17 sedated ICU patients ventilated for 120 min in fully closed loop ventilation with IntelliVent Ò (Hamilton Medical S1 ventilator), 34 episodes of significant changes in PEEP (C 4 cmH 2 O) were selected and HLI Ò values within 5 min before and after PEEP changes were collected. Statistics were done using SigmaStats with p \ 0.05 as significant.

RESULTS. Changes in PEEP and in HLI Ò are shown in the Table 1 15 ± 3 cmH2O 9 ± 4 cmH2O 17 ± 12% 9 ± 6% P \ 0.001

The correlation between change in PEEP and change in HLI Ò is shown on the Fig. 1 .

CONCLUSIONS. Based on these preliminary data changes in PEEP are reflected HLI Ò changes and may help estimating continuously the hemodynamic effects of ventilation. OBJECTIVES. We have tested a 3 axis accelerometer sensor for detection of regional left ventricular ischemia.

In 12 pigs a 3-axis accelerometer was sutured to the left ventricular (LV) apical region in left descending coronary artery (LAD) supply area Accelerometer x-axis measured longitudinal-, y-axis circumferential-and z-axis radial epicardial motions. Epicardial displacements were calculated from the acceleration signals and systolic displacements within 150 ms after peak R on ECG was measured. LAD was occluded for 60 s to induce regional LV dysfunction. Myocardial circumferential strain (shortening) measured by echocardiography in the LV apical anterior region was used to confirm ischemia. The ECG ST-segment in lead II was also monitored. Data are presented as mean ± SE.

Early systolic displacement at baseline was 11 ± 4 mm, 12 ± 2 mm and 3 ± 2 in circumferential, longitudinal and radial directions, respectively. LAD occlusion induced akinesia in circumferential (0 ± 3 mm, P \ 0.001) and radial (0 ± 1 mm, P = 0.01) directions, whereas longitudinal displacement changed less to 10 ± 2 mm (P = 0.280). Ischemia was confirmed by echocardiography strain, showing lengthening in systole (P \ 0.001). No significant changes were observer in the ECG ST-segment during coronary occlusion (P = 0.341). INTRODUCTION. There is increasing evidence to suggest perioperative complications are predictive of long term survival and that reducing them may improve survival rates 1 . Goal directed therapy has been shown to reduce mortality and morbidity perioperatively, with those unable to increase oxygen delivery perioperatively having demonstrably worse outcomes. The advent of non invasive tissue oxygenation monitors using near infrared spectroscopy has allowed further study of oxygen flux during goal directed therapy.

OBJECTIVES. To observe changes in tissue oxygenation during an 8 h oxygen delivery targeted post surgical optimisation program and provide long term mortality followup of a surgical cohort of high risk patients.

METHODS. 40 patients undergoing high risk surgery and postoperative optimisation (targeting of oxygen delivery index of [600 ml/min/m 2 ) on the tensive care unit at a London teaching hospital were enrolled. Each patient underwent a protocolised haemodynamic optimisation protocol as per our standard unit policy for 8 h with consecutive recordings of tissue oxygenation at the thenar eminence using an inspectra 325 monitor. Additional variables relating to global and tissue perfusion were measured concurrently. Patients were followed up for survival status at 3.5 years using routinely available information held within our hospital records.

In hospital mortality was 17.5% (N = 7), whilst at 3.5 years this had increased to 50% (N = 20). There was no significant difference between APII scores 11 (4) versus 9.5 (4), Age 61.95 ± 16.23 versus 65.8 ± 16.35 or operation type for survivors and non-survivors at 3.5 years respectively. Significant differences between groups were found however for admission and mid optimisation protocol (4 h) HR and STO2 (see Table 1 There were no significant differences in measured variables for 30 day mortality.

CONCLUSIONS. There appears to be a statistical and clinical difference in HR and tissue oxygenation between the long term survivors of high risk surgery who undergo monitored postoperative goal directed optimisation. INTRODUCTION. Bronchoscopic bronchoalveolar lavage (B-BAL) is today the gold standard for sampling of inflammatory markers in the distal airways. Nonbronchoscopic bronchoalveolar lavage (N-BAL) by ordinary suction catheter has been investigated as a more easily accessible method for alveolar sampling in the setting of acute respiratory distress syndrome (ARDS). The results, however, were disappointing, probably due to more proximal sampling by the N-BAL.

To investigate wether N-BAL by a catheter with physical properties similar to those of the bronchoscope is comparable to B-BAL.

METHODS. B-BAL and N-BAL by Cook's airway exchange catheter was performed with 3 9 30 ml normal saline on opposite sides 15 min apart at nine different occasions on 5 anesthetized and intubated pigs. The volume of the recovered lavage was noted, after which the fluid was analyzed for albumin, total cell count, viability and differential cell count. Statistical analysis was performed using Wilcoxon's rank-sum test.

RESULTS. N-BAL yielded significantly higher albumin content than B-BAL (20.1 ± 8.7 vs. 11.7 ± 3.4 mg/L, p = 0.027). In all other measurements there were no significant differences between N-BAL and B-BAL (recovered volume 52. OBJECTIVES. We hypothesized that collagen synthesis and degradation are disturbed in acute respiratory failure.

In the FINNALI-study we defined acute respiratory failure as need of noninvasive and/or invasive ventilatory support for more than 6 h (1). After informed consent we collected blood samples for serum procollagen propeptides I and III (PINP, PIIINP) and ICTP levels at study admission, day 2, 7 and 21. Patients with all four blood samples were included in this substudy. Multiple organ dysfunction (MOD) was defined as two or more individual organ SOFA scores of 3-4 at any day during the first week.

RESULTS. The study population comprised 68 of 958 FINNALI patients (1). The mean (SD, range) age was 60 years (15, 20-86) and the majority were male 74%. On admission the mean SAPSII score was 45 (15, 2-83). 16 patients (24%) developed MOD during the first 7 days. Over time PIIINP/PINP-ratio first increased and then decreased to baseline by day 21 while PINP/ICTP-ratio decreased and then decreased to baseline by day 21(P \ 0.001 and P = 0.006, respectively) ( Fig. 1 ). There were no statistical differences in the ratios between patients with or without MOD. CONCLUSIONS. We found that in patients with acute respiratory failure the balance of collagen synthesis was towards degradation of type I collagen and production of collagen type III. 3 ± 43 ng/ml in the ARDS group, and significantly higher than the 53.8 ± 37.0 ng/ml in the ALI (not ARDS) group. The difference in HMGB1 values in the early stage between the group that died up to the by 30th day and the surviving group was not significant, but the HMGB1 values were significantly higher in the group that died until the 60th day and 90th day than in the survival group. It was concluded that differences in HMGB1 values in the early stage after the onset of ALI (not ARDS)/ARDS are useful as outcome determining factors after 30 days of onset. An inverse correlation was observed between the HMGB1 values and lung oxygenation, suggesting the possibility that HMBG1 is involved in the development of respiratory failure.

S. Shibata 1 , G. Takahashi 2 , N. Shioya 3 , S. Endo 3 1 Akita City Hospital, Anesthesiology, Akita, Japan, 2 Iwate Medical University, Emergency Medicine, Morioka, Japan, 3 Iwate Medical University, Critical Care Medicine, Morioka, Japan

Sivelestat sodium hydrate (sivelestat) is a selective polymorphonuclear leukocyte elastase (PMN-E) inhibitor and has also been shown to be effective for pulmonary disorders associated with SIRS in clinical patients. Blood levels of inflammatory cytokines have been shown to be decreased in patients treated with sivelestat. However, since patients with SIRS have already received other drugs, it remains indefinite whether or not sivelestat might suppress the production of cytokines. Moreover, it is difficult to clarify any cells releasing cytokines. In the experiment using cells isolated from the blood, intercellular mutual actions and cytokine networks were blocked and the experiment failed to faithfully reproduce the in-vivo condition.

OBJECTIVES. The possibility of sivelestat suppressing the production of cytokines from granulocytes and monocytes was assessed by intracellular cytokine staining using the whole blood culture method and flow cytometry to faithfully reproduce the in-vivo condition.

METHODS. Blood samples were collected from 9 healthy volunteers. A vehicle (control group), lipopolysaccharide (LPS; LPS group), or LPS + sivelestat (sivelestat group) was added to the whole blood, followed by the addition of a protein transport inhibitor in each group. After incubation, they were subjected to staining of the cytokines retained in the cells by the addition of an anti-interleukin 8 (IL-8) or anti-tumor necrosis factor a (TNF-a) antibody and analysis by flow cytometry. The data were analyzed by the Kolmogorov-Smirnov test. Values obtained [D/s(n)] result from the comparison of the fluorescence histograms of each sample with a control one.

Addition of sivelestat at low concentrations (1 and 10 lg/ml) significantly (P \ 0.01) suppressed the production of IL-8 from granulocytes induced by a low concentration (1 ng/ml) of LPS. On the other hand, the granulocytic production of TNF-a induced by a high concentration of LPS (10 ng/ml) was significantly (P \ 0.01) suppressed by treatment with sivelestat at high concentrations (10 and 100 lg/ml). With regard to the monocytic production of TNF-a and IL-8 induced by LPS, there was no significant suppression of either TNF-a or IL-8 production by sivelestat.

CONCLUSIONS. Sivelestat, a neutrophil elastase inhibitor, suppressed granulocytic production of IL-8 and TNF-a, suggesting the potential usefulness of sivelestat for the treatment of various morbid conditions involving IL-8 and TNF-a in their onset. INTRODUCTION. Coagulation, fibrinolysis and extravascular fibrin deposition are the hallmarks of the pathogenesis of Acute Lung Injury (ALI). PAI-1 has a central role in antagonizing fibrinolysis by decreasing the plasminogen turnover to plasmin. PAI-1 has been suggested as a clinical severity marker of ALI. In previous studies it was associated with higher mortality and morbidity in the critically ill. UPAR is a cell surface receptor activating the serine protease uPA. Increased expression of UPAR is found in various stages, including inflammation, tissue remodelling and malignancies, indicating poor prognosis. PAI-1 antagonizes the proteolytic activities of uPA and plasmin.

OBJECTIVES. We sought to evaluate the prognostic value of sUPAR and PAI-1 for 90-day mortality of patients with Acute Respiratory Failure (ARF).

The FINNALI-study patients needed invasive or non-invasive ventilation for more than 6 h (1). Blood samples were collected from patients at baseline and on day 2 after baseline. Healthy volunteers were also analyzed. Sera were frozen at -80°C until analyses. Concentrations of sUPAR and PAI-1 in blood serum were measured by enzyme linked immunosorbent assay (ELISA). Data are presented as median (IQR). The prognostic value of sUPAR and PAI-1 for 90-day mortality was determined with ROC analysis.

In the critically ill, sUPAR and PAI-1 were 12.7 (10.7-14.7) ng/ml and 2.33 INTRODUCTION. Acute lung injury is a common disease in intensive care, associated to various septic or inflammatory diseases. Inflammation is part of the defense mechanisms of innate immunity, occurring after tissue injury.

OBJECTIVES. The aim of the project was to decipher the transcriptional changes occurring after the onset of an inflammatory injury by intravenous injection of oleic acid.

Experimental study of the lung transcriptome after oleic acid injection. Thirtysix C57Bl/6 J mice, aged of 7 weeks, were sacrificed at 1H, 1H30, 3H, 4H, 18H and 24H after physiological serum or oleic acid injection (20 lL) in the caudal vein. Left and right lung were separated for mRNA extraction and pathological examination. Labelled cDNA were hybridized on cDNA nylon microarray (TAGC, Marseilles, France) and raw data were extracted from scanned images with BZScan2 software. Raw data were normalized with the quantile method, and supervised analysis was conducted with Significance Analysis of Microarray algorithm within the R statistical suite and Bioconductor libraries.

After the administration of oleic acid, the mice were tachypneic and prostrated. All survived during the 24 first hours. The pathological analysis of lung tissue revealed an early inflitration of the lung tissue by polynuclear cells, as well as a pulmonary edema. These alterations were not observed after 4H. The time course analysis of transcriptional lung data identified a thousand genes which expression is modulated after injury. Hierarchical clustering identified 3 major groups of genes. The first one (365 genes) is composed of genes transiently up-regulated between 1H and 3H after oleic acid injection. Th second group (366 genes) is composed of genes expressed between 1H30 and 4H. The third group (366 genes) is composed of genes expressed at the later time points (18H-24H). The functional annotation linked these signatures with keywords related to pro-inflammatory response, vascular endothelium modification and lipid metabolism, respectively. RT-PCR analysis of pro-(TNF, IL6) and anti-inflammatory (IL4, IL10) markers related the pro-inflammatory phase to the earlier time points (1H-1H30) and the anti-inflammatory phase to the late points (after 4H).

CONCLUSIONS. Oleic acid injection in mice induced a transient acute lung injury. This is confirmed by clinical, pathological and transcriptional modifications. The modulation of gene expression after the oleic acid injection revealed an early pro-inflammatory response, followed by an anti-inflammatory response and lipid metabolism modificiations. This model could now be used to describe the specific modulation occuring during pulmonary infection and critical injuries like acute respiratory distress syndrome. INTRODUCTION. Ventilator associated lung injury (VALI) is influenced by tidal volumes, airway pressure and cyclic opening of alveoli during mechanical ventilation. Preserved spontaneous breathing during partial ventilatory support may be protective, but it is not known whether the transpulmonary pressure generated by spontaneous breathing has the same effect on VALI as if generated by the ventilator.

To determine whether hemodynamics, respirtory function and VALI are influenced by the amount of support provided by pressure support ventilation.

After approval from the institutional animal care committee, acute lung injury was induced in 50 anesthetized SD rats by acid aspiration. Ten animals each were then ventilated with positive end-expiratory pressure 5 cmH2O in pressure control (PC), pressureregulated assist control (AC) or pressure support mode with 100% (PS100), 60% (PS60) or 20% (PS20) pressure support of initial distending pressure needed to maintain tidal volume. PC animals were paralyzed. After 4 h animals were killed and VALI determined.

RESULTS. There were no differences in baseline characteristics. Acute lung injury was characterized by a decrease of the P/F ratio from 447 ± 75 to 235 ± 90 mmHg and of the dynamic compliance from 0.5 ± 0.2 to 0.28 ± 0.1 ml/cmH 2 O. CONCLUSIONS. Compared to controlled ventilation, preserved spontaneous breathing activity improved hemodynamic stability, respiratory function and lung edema clearance. The reduction in pressure support did not lead to reduced tidal volume, but transpulmonary pressure was preserved by muscular activity of the chest wall. No difference was observed between full or 60% of pressure support, but further reduction in pressure support resulted in increased wet-dry ratio. OBJECTIVES. We studied the effects of metabolic acidosis on enzymatic and non-enzymatic NO-production in hypoxic and hyperoxic lung regions in a pig model.

Eighteen healthy anesthetized pigs were separately ventilated with hypoxic gas to the left lower lobe (LLL) and hyperoxic gas to the rest of the lung. Six pigs received HCl infusion (HCl group), six pigs received N w -nitro-L-arginine methyl ester (L-NAME) and HCl (L-NAME + HCl group) and six pigs received buffered Ringer's solution (Control group). NO concentration in exhaled air (ENO), NO synthase (NOS) activity in lung tissue, and regional pulmonary blood flow were measured.

RESULTS. Metabolic acidosis, induced by infusion of HCl, decreased the relative perfusion to the hypoxic LLL (Q LLL /Q T ) from 7 (±3) to 3 (±1)% in the HCl group (p \ 0.01), and from 4 (±1) to 1 (±1)% in the L-NAME + HCl group (p \ 0.05), without any measurable significant changes in ENO from hypoxic or hyperoxic lung regions There were no significant differences between the HCl and Control groups for Ca 2+ -dependent or Ca 2+ -independent NOS activity in hypoxic or hyperoxic lung regions.

Metabolic acidosis augmented the hypoxic pulmonary vasoconstriction, without any changes in pulmonary enzymatic or non-enzymatic NO-production. When acidosis was induced during ongoing NOS-blockade, the perfusion of hypoxic lung regions was almost abolished, indicating acidosis-induced pulmonary vasoconstriction was not NO dependent. 

Assessing and monitoring biomarkers in acute lung injury (ALI) may improve knowledge of its pathogenesis, early recognition, and management and predict remote organ injury and multiple organ failure.

OBJECTIVES. Early consents for research are difficult to obtain in patients with or at risk of ALI because of the emotional burden of the severity and sudden onset of the disease. However, study samples may be obtained from left-over clinical blood draws, which are readily available if processed adequately. The aim of this study was to compare fresh and ''waste'' blood samples prospectively in a series of consecutive critically ill patients. The hypothesis is that ''waste'' blood samples if appropriately processed provides accurate and reliable results comparable to the gold-standard, which is immediate collection and processing of fresh blood samples.

Prospective study comparing biomarkers of epithelial injury (sRAGE) and inflammation (20 different cytokines/chemokines) in critically ill patients measured on fresh blood or waste blood, kept at 4 degrees Celsius for 24 h. An automated system performed a daily screening of adults in the ICU with an increased risk for ALI (lung injury prediction score, LIPS) within 12 h of admission and/or on recognition of the diagnosis of ALI, using the American-European Consensus Conference criteria. Risks factors for ALI include pneumonia, sepsis, pancreatitis, shock, aspiration, high risk surgery and high risk trauma. IRB approved the protocol and written consent was obtained from patients or their surrogates. Statistical measurements were performed using the Bland-Altman analysis for correlation between fresh and waste blood sample data.

Between May and December 2009, 30 patients were enrolled. One patient was excluded due to lack of sample. Samples were obtained either at one time point (N = 25) or two, on consecutive days (N = 13). Female/male patient ratio was 12/17. Seven of the 29 patients had ALI. Twenty two patients had risk for ALI with a median LIPS score of 4 (IQR 3.5-4.5). Sepsis was the most common risk factor, present in 23 patients. In-hospital mortality was 28% (8/29). The Bland-Altman plot (mean bias ± SE, limits of agreement) showed good correlation for IL-1ra (-9 ± 14.3 pg/ml, -38.3 to 20.4 pg/ml), IL-6 (0.9 ± 2.9 pg/ml, -5 to 6.8 pg/ml), IL-8 (-0.3 ± 0.3 pg/ml, -0.9 to 0.3 pg/ml), IL-12 (p40) (-0.2 ± 0.3 pg/ml, -0.7 to 0.4 pg/ml), MCP-1(-0.2 ± 2 pg/ml, -4.2 to 3.9 pg/ml) and sRAGE (-24 ± 53 pg/ ml, -130 to 83 pg/ml) between fresh blood and ''waste'' blood samples.

In patients with ALI, properly stored blood, drawn for clinical purposes, can be processed within 24 h for research purposes. However, the stability of each biomarker of interest needs to be individually validated before using stored blood INTRODUCTION. Pulmonary surfactant inactivation following acute lung injury might promote alveolar derecruitment and reduce the airspace available for ventilation, making the lung more prone to ventilation-induced lung injury (VILI).

Our aim was to test the potential for a protective effect of exogenous surfactant treatment in a model of acid aspiration and VILI.

METHODS. 17 male C57/BL6 mice were anesthetized, mechanically ventilated (Vt 15 ml/ kg; RR 130/min; PEEP 2 ± 0.2 cmH 2 O; FiO 2 0.5) and immediately subjected to intrabronchial (right) instillation of 1.5 ml/kg HCl 0.1 M. Mechanical ventilation went on for 420 min. 180 min after the acid instillation, 9 mice were treated with exogenous surfactant (80 mg of phospholipids/ml) given as bolus of 1 ml/kg in the right bronchus (surf group). We measured oxygenation, lung compliance (measured every 60 min throughout the experiment), macrophage inflammatory protein (MIP) 2 in broncho-alveolar lavage (BAL) fluid.

. PaO 2 at the end of the experiment was significantly higher in the surf than in control group (383 ± 163 vs. 198 ± 104 mmHg, p \ 0.05). Although surfactant bolus caused a reduction in lung compliance measured 10 and 60 min after treatment, in the surf group compliance restored to 98 ± 7% of the post injury level, while it decreased in control group to 88 ± 8% (p \ 0.05). There were no differences between groups in the dosage of MIP-2 in BAL neither in right or left lung.

CONCLUSIONS. Exogenous surfactant treatment improved lung function in a murine model of two hit lung injury.

GRANT ACKNOWLEDGMENT. INTRODUCTION. Ventilator induced lung injury significantly contributes to the mortality in patients with acute respiratory distress syndrome, the most severe form of acute lung injury. Understanding the molecular basis for response to cyclic stretch and its derangement during high volume ventilation is of high priority.

OBJECTIVES. To identify specific molecular regulators involved in the development of ventilator induced lung injury.

We undertook a comparative examination of cis-regulatory sequences involved in the coordinated expression of cyclic stretch responsive genes using microarray analysis. Analysis of stretched vs. non-stretched cells identified significant enrichment for genes containing binding sites for the transcription factor ATF3 (Activating Transcription Factor 3).

To determine the role of ATF3 in vivo, we compared the response of ATF3 gene deficient mice to wild type litter mates in an in vivo model of ventilator induced lung injury.

RESULTS. ATF3 deficiency results in increased sensitivity to mechanical ventilation alone or in conjunction with inhaled lipopolysaccharide (10 mg/kg) as determined by assessment of lung and bronchoalveolar lavage cell infiltration and pro-inflammatory mediator release, pulmonary edema and indices of tissue injury. The expression of genes containing an ATF3 cis-regulatory region was significantly altered in gene deficient animals. ATF3 protein expression and nuclear translocation is increased after mechanical ventilation.

CONCLUSIONS. ATF3 deficiency confers increased sensitivity to mechanical ventilation alone or in combination with inhaled endotoxin. In our model, ATF3 acts to ''counterbalance'' cyclic stretch and high volume-induced inflammation, limiting its potential to cause additional lung injury and consequently protecting animals from injurious cyclic stretch. OBJECTIVES. Our aim was to evaluate the role of the alveolar macrophages in a murine model of ALI, by selective depletion of this type of cells from the air space achieved by clodronate administration.

Mice were treated (it) with 100 ll of clodronate (Clo)-or PBS (PBS)-liposomes. After 24 h mice were anesthetized and ventilated (Vt 8-10 ml/kg, RR 140 min -1 , FiO 2 0.21); in order to induce lung injury 2 ml/kg of HCl (0.1 M) or air bolus (sham group) was instilled in the right bronchus. Mice were ventilated for 10 min, and extubated after awakening. 24 h after injury, animals were sacrificed and Broncho-Alveolar Lavage (BAL) and blood gas analysis (FiO 2 = 0.21) were performed.

. 24 h after lung injury animals with alveolar macrophages depletion, showed a better oxygenation versus PBS-treated group. However, recruitment of neutrophils in BAL was not statistically different between clo_HCl and PBS_HCl group. RESULTS. High levels of OC were found in patients treated by 150 mg of OT bid. OC levels ranged from 2,507 to 4,551 ng/mL in these patients. Concentrations of OC were five-to tenfold higher than concentrations reported in healthy volunteers. Lesser levels were found in patients treated by 75 mg of OT bid. Nevertheless, the patient with the moderate renal failure seemed to accumulate OC (levels ranged from 906 to 1470 ng/mL) whereas concentrations reported in the patient with a normal renal clearance were below (152-349 ng/mL).

CONCLUSIONS. ECMO seemed not to have any influence on OC concentrations while renal insufficiency seemed to be the parameter leading to OC accumulation. As IC50 was very low and reached even with usual dosage, increasing OT dose to 150 mg bid appeared to be unnecessary. OBJECTIVES. Aim of our study was to evaluate the effect of nursing care on patients undergoing venous-venous ECMO for acute respiratory distress syndrome (ARDS).

METHODS. we recorded physiological and ECMO parameters (heart rate, arterial blood pressure, mixed venous saturation (SvO 2 ), arterial oxygen saturation (SpO 2 ), body temperature and extracorporeal blood flow (BF)) before and during daily nursing in 5 patients undergoing vv-ECMO for several days (each patient was followed on average for 4.6 days, 23 cases in total). Arterial blood gases were also collected before and after nursing care. Daily nursing was performed following defined steps (sponge bath, oral hygiene, change position of endotracheal tube, elevation with scooping stretcher for sheets replacement and back hygiene, dressing replacement) in agreement with a standard protocol in use in our department. 

(Expressed as mean ± standard deviation). All patients were affected by ARDS H1N1-related. Patients were sedated with propofol (182 ± 68 mg/h) or midazolam (7.9 ± 1.8 mg/h) plus an opioid drug (fentanyl 175 ± 45 mcg/h or remifentanil 0.2 ± 0.01 mcg/kg/min or sufentanil 0.32 ± 0.09 mcg/kg/min). Ramsey score before nursing was 5.4 ± 0.9. In 11 cases patients were paralysed. In Table 1 we summarized the adverse events observed during nursing care, divided into hypertensive or tachycardic episodes, blood oxygen desaturation, reduction in SvO 2 or reduction in BF. Forty-nine sedative bolus were administered during nursing (mean request for each patient: 2.1 ± 1.9), always after an episode of hypertension or tachycardia (most frequently during elevation with scooping stretcher and changing position of endotracheal tube). Although in 13 cases preventive bolus of sedation were administered before nursing, in 8 of those cases (61%), additional bolus were required. We found an inverse correlation between BF and the increase in heart rate, drop in arterial saturation and SvO2. Despite active warming, we observed a drop of 0.31 ± 0.15°C (p \ 0.01) in body temperature.

Nursing care may have a significant impact on physiologic parameters of patients during vv-ECMO. Tachycardia, hypertension and reduction in oxygenation were commonly recorded and were not prevented by pre-nursing bolus of sedation but were attenuated in patients with higher BF. INTRODUCTION. Prone position has been used in cases of ARDS with refractory hypoxemia but some physiological effects are still unknown. Prone position could increase intraabdominal pressure (IAP) and could lead to acute renal failure (ARF). Acute kidney injury in ICU is associated with increased mortality.

OBJECTIVES. The aim of this study was to determine whether prone position could increase intraabdominal pressure and possibly promote ARF.

We studied all adult ARDS patients who were ventilated using the protective strategy defined by ARDS Network criteria and who needed prone position to improve oxygenation. We collected respiratory data (ventilator parameters and gas exchange) and hemodynamic variables (heart rate, systolic, diastolic and mean arterial pressure). IAP was measured using the Abdo-Pressure TM bladder transducer following World Society of Acute Compartment Syndrome recommendations. Abdominal perfusion pressure was calculated as mean arterial pressure minus IAP. Main renal parameters were: filtration gradient (FG), creatinine clearance, fractional excretion of sodium (FENa) and urea (FEUrea). Patients were classified according to RIFLE score after each manoeuvre. All data were recorded in prone and in supine position at least once per day.

RESULTS. The study included 18 patients (14 male) admitted to a medical-surgical ICU over a one-year period. Their mean age was 47.3 ± 17.8 and length of ICU stay was 23 ± 18 days. All 18 patients had primary ARDS and 9 had received nephrotoxics. ICU mortality reached 50%. We recorded at least 3 manoeuvres per patient (A, B, C). Prone positioning improved PaFiO 2 ratio from 101.5 ± 5 to 150 ± 72 (p = 0.027). IAP showed a small increase from 6.1 ± 2.7 to 8.9 ± 2.9 mmHg (A; p = 0.004), from 8.4 ± 3.4 to 11.2 ± 3.7 mmHg (B; p = 0.038) and from 7.6 ± 1.3 to 9.9 ± 0.8 mmHg (C; p = 0.001). There were no statistically significant changes in hemodynamic parameters or abdominal perfusion pressure. Renal function parameters (FG, creatinine clearance, FENa and FEUrea) showed no modification after each prone positioning. In contrast, when patients were classified according to RIFLE score, we observed a trend towards worsening, though this was not statistically significant.

CONCLUSIONS. Prone positioning improved arterial oxygenation in primary ARDS patients and was associated with an increase in IAP. However, creatinine clearance and glomerular filtration remained unchanged. 

Percutaneous Extracorporeal life support system (p-ECLS) including ECMO becomes widely used in medical and surgical emergent situation, such as refractory cardiogenic shock, cardiac arrest and acute respiratory failure. Patients requires highly specialized intensive care and monitoring system.

We reviewed our ECLS experience and tried to analyze the clinical outcomes, factors for survival and frequently faced problems during management for improving weaning and survival rate (medical vs. surgical patients). INTRODUCTION. in spite of the huge efforts spent over the last years, conventional treatment of acute hypoxemic respiratory failure (AHRF) is often inadequate and alternative procedures must be instituted. ICUs skillful in extracorporeal membrane oxygenation (ECMO), as recently shown [1] , may improve survival of these patients. Since 1989 we developed a treatment algorithm for AHRF which encomprises: (1) low flow venous-venous ECMO (LF-ECMO) consisting in a relatively low initial blood flow (BF, 2-2.5 L/min) to maximize extracorporeal CO 2 removal while providing partial oxygenation (if needed, BF can be increased up to 4.5-5 l/min to keep arterial pO2 above 45 mmHg); (2) femoral-femoral percutaneous cannulation with 21-25 Fr cannulas to allow free movements of the neck and increase patient's tolerance; (3) early institution of spontaneous assisted ventilation (SB) and weaning from sedation and mechanical ventilation (MV) while on ECMO.

OBJECTIVES. to review our last 10 years LF-ECMO activity.

METHODS. study period was January 2000-2010. LF-ECMO entry criteria were: potentially reversible acute hypoxemic respiratory failure, LIS C3, no evidence of intracranial bleeding and no absolute contra-indications to heparinization. ECMO was performed with different type of heparin coated hollow-fiber artificial lungs.

. we treated 31 patients (mean ± SD, 37 ± 18.5 years old, 68% males, BMI 27 ± 6, SOFA 7.5 ± 3, OI 38 ± 14). 48% of these patients were placed on ECMO at other hospitals and transported to our ICU by a dedicated ECMO team. Ventilation days before ECMO were 12 ± 16 (range 1-69). Before ECMO Vt/kg was 6 ± 1.5 and RR was 33 ± 12: after ECMO beginning Vt/kg was unchanged while RR decreased to 11 ± 4 (p\ 0.01). ECMO was set at BF 2.5 ± 1 L/min, GF 3.3 ± 1.5 L/min, FiO2 0.98 ± 0. INTRODUCTION. Ventilating patients with acute lung injury (ALI) in supine position potentially leads to an impaired pulmonary gas exchange. Prone position (PP) is an attractive means to improve ventilation-perfusion (V/Q) ratio [1, 2] but has several contraindications and showed no improvement in survival so far [3] . Another therapeutical option is an upright position, which is easy to perform and has theoretical advantages over PP: the upward shift of the abdominal compartment is less pronounced, thus increasing thoracoabdominal compliance [2] . 4 However, to date regimes of an upright position did not tilt patients more than 45° [4] . 4 OBJECTIVES. We hypothesised that a 60°standing position (SP) during mechanical ventilation may improve respiratory function. Furthermore, we aimed to determine the feasibility of a SP for 2 h during mechanical ventilation.

We studied 30 adult patients, receiving mechanical ventilation for more than 48 h in the intensive care unit of an university hospital. After recording baseline data, patients were placed in a 60°SP with the body entirely straight. Further data sets were recorded during 2 h in SP, and after patients position was readjusted to supine position.

Functional residual capacity (FRC) increased immediately after reaching SP (p \ 0.001) and remained elevated after repositioning to supine position. PaO 2 /FiO 2 ratio and compliance decreased initially during SP, but increased (p \ 0.05) after patients were retransferred to supine position. Haemodynamic variables remained stable under a moderate increase of doses of catecholamines during the study period.

CONCLUSIONS. Changes in respiratory function during SP are probably explained by a downward shift of the diaphragm due to gravitational forces 2 leading to an increased FRC but not altering V/Q ratio as demonstrated by the paO 2 /FiO 2 ratio. After reaching the initial supine position the opening of the lung proved by the elevated FRC is the predominant effect now associated with an increase in oxygenation as reflected by the paO 2 /FiO 2 ratio due to an optimised V/Q ratio. Our results are confirmed in a subgroup analysis for 9 patients meeting ALI criteria. Ventilating patients in SP may be a new therapeutical approach to improve respiratory function in patients with ALI. (1) . There are several clinical trials investigating the efficacy of the free radical scavenger N-acetylcysteine (NAC) in ARDS, but its advantage remains uncertain.

OBJECTIVES. Critically appraise and summarize all randomized clinical trials involving intravenous NAC administration in adult patients suffering from ARDS.

We included trials involving participants with ARDS according to the American-European Consensus Conference Criteria (2) regardless of the underlying cause, and where one of the groups was treated with intravenous N-acetylcysteine in bolus intravenous doses or as continuous infusion, or combination of the two, and the other group was given placebo or standard treatment. CONCLUSIONS. The main finding of this meta-analysis is that intravenous NAC is ineffective in reducing mortality, length of stay or duration of mechanical ventilation in ARDS.

We also found that late administration of NAC may be associated with adverse outcome. The mechanism of this potentially deleterious effect remains unclear, but dosing and timing of NAC appear to be critical issues. OBJECTIVE. To evaluate if extubation during ECLS is harmful or beneficial.

A 28-year-old woman was admitted to our intensive care unit (ICU) after removal of a left ventricular assist device. This device was implanted as bridge to recovery for postpartum cardiomyopathy and ventricular function seemed to have recovered sufficiently. However, shortly after ICU admittance she developed massive left and right ventricular failure. Therefore a centrally cannulated veno-arterial ECLS (Maquet Permanent Life Support) was implanted as a bridge to transplant. Four days later she was extubated while on full ECLS support, in order to reduce the risk of ventilator associated pneumonia. While on ECLS, the patient was mobilized, practiced with an ergometer and chatted with her family. Three days later the patient underwent cardiac transplantation. The postoperative period was characterized by temporary pulmonary failure, due to the combination of lung edema and atelectasis. Eventually she made a full recovery.

DISCUSSION. ECLS provides a valuable means as bridge to transplantation, bridge to bridge or bridge to recovery. With the increasing use of ECLS for circulatory failure, debate about the necessity of mechanical ventilation during this treatment ensues. ECLS is usually applied under deep sedation and controlled mechanical ventilation. Discontinuation of sedation possibly prevents intensive care acquired weakness. Extubation during ECLS may provide better pulmonary perfusion due to negative intra-thoracic pressure. Furthermore, the awake and extubated patient is able to mobilize and exercise which may reduce the risk of atelectasis and ventilator associated pneumonia. Our patient however developed pulmonary edema and atelectasis after discontinuation of ECLS. The edema was probably a consequence of reperfusion injury, due to severely decreased pulmonary flow while on ECLS. An absent ventilatory drive while on ECLS may have led to hypoventilation while the patient was extubated, resulting in atelectasis. An extensive Medline search resulted in one other case report describing an extubated patient on ECLS. 1 Intermittent non-invasive positive pressure ventilation was used to prevent atelectasis, but the patient developed pneumonia after 30 days of ECLS. Our patient was successfully extubated while on ECLS. However, we conclude that there is insufficient evidence to recommend or oppose extubation of patients on ECLS for circulatory failure. Severe ARDS and refractory hypoxemia were defined with a PaO2/ fraction of inspired oxygen (FiO2) ratio of B100, or uncompensated hypercapnea with a pH of \7.20 despite receiving optimal conventional treatment. The ECMO can be used as a rescue treatment in these case.

OBJECTIVES. Evaluation of severe ARDS treated with extracorporeal oxygenation (ECMO). All these ARDS were due to bacterial pneumonia or H1N1 influenza.

Over the last year (December 2009-January 2010), the recourse to extracorporeal oxygenation (ECMO) was used in ten patients with severe ARDS and severe hypoxemia. Two groups were defined: bacterial pneumonia with ARDS (BP group, n = 5), and H1N1 influenza with ARDS (H1N1 group, n = 5). All ECMOs were implanted at the bedside to facilitate intra-hospital or inter-hospital transfer, because of severe hypoxemia or hemodynamic instability making impossible patient mobilization before ECMO. RESULTS. Data sets of 44 patients of 47 consecutive patients treated with ECMO were complete and included into analyses. We had no clinical or radiological evidence for thrombosis or clotting within ECMO-circuit with a target-PTT of 35 s. One patient with systemic aspergillosis died because of intracranial hemorrhage. One ECMO circuit had to be replaced due to insufficient oxygenator function after 10 days. Further data are presented in Tables 1 and 2 . CONCLUSIONS. In this retrospective analysis of 44 patients who underwent ECMOtreatment, AC with low-dose heparin (target-PTT of 35 s) was safe and without any observation of macroscopic thrombosis or clotting within the circuit. Transfusion requirements and intracranial hemorrhage were low as compared with previous reports [1, 2] . Therefore our data suggest that it is possible and safe using ECMO-therapy with low-dose Heparin. INTRODUCTION. In response to H1N1 pandemy, Italy and Lombardy created a national and a regional ICU network, respectively, for treatment of ARDS patients. Our hospital Policlinico San Matteo of Pavia participated with a team for inter-hospital ECMO implantation and subsequent patient transport.

OBJECTIVES. Description of the Pavia ECMO team and activity analysis.

METHODS. Our team is composed by a cardiac surgeon, two intensivists, a perfusionist, an ICU nurse, two emergency rescue technicians and a driver. All necessary aids for implantation and intensive care are ranged in three trolleys and three transport bags. Equipments are firmly mounted on a two-level steel bridge connected to a spinal board. A portable ultrasonograph is also available. The ECMO team was alerted by the national call center. Each mission used two ambulances, and in one case the ambulances were embarked on a Hercules C130 J.

From 31 October to 31 December 2009, four patients were implanted and transported, three suffering from H1N1 influenza (including a 190-kg body weight patient) and one from acute mitral valve rupture. All patients, already mechanically ventilated with maximal support, had veno-venous ECMO implanted by femoro-femoral percutaneous cannulation. The median mission duration was of 7.25 h (range 6-17 h). All patients were transported to our ICU, where the median ECMO duration was of 8 days (range 6-10 days).

No major managing issue occurred during the ECMO missions, and patient hospital survival was of 100%.

A multispecialist team with good knowledge of ECMO can provide an effective support in severe respiratory failure, with ECMO implantation in peripheral hospitals and subsequent patient transport, thus realizing a fast and safe continuum between phone call activation and admittance to the reference center. INTRODUCTION. When patients with sever respiratory failure are treated with V-V ECMO the right heart sometimes fails. This is a serious complication with a high mortality. In our unit these patients have been converted to V-A ECMO, although it is not fully agreed upon in the ECMO community due to previously depressing results.

OBJECTIVES. To evaluate the results of conversion to V-A from V-V ECMO in case of right heart failure.

Retrospective analyses of all patients with severe respiratory failure, treated between 1987 and 2009 at the Karolinska ECMO centre. Patients who were converted to V-A ECMO due to right ventricular failure were evaluated.

A total of 214 patients (78 adults, 57 peadiatric, 79 neonatal) were treated on V-V ECMO for severe respiratory failure. Of them 50 (33 adults, 9 peadiatric, 8 neonatal) needed conversion to V-A ECMO due to right ventricular heart failure demonstrated clinically by multiorgan failure and verified by ECHO cardiography. The survival after conversion to V-A ECMO was 10/33 (30%) in the adult age group, 7/9 (78%) in the peadiatric age group and 2/8 (25%) among the neonates.

CONCLUSIONS. Given the high risk of fatality if not treated, conversion to V-A from V-V ECMO should be considered when the right ventricle fails. Patients on V-V ECMO with right ventricle heart failure have very bad prognosis. It is concluded from the present results that conversion to V-A ECMO can save some of these patients.

Cardiac surgery and regional hemodynamics: OBJECTIVES. To test whether TAPSE and right ventricular systolic (Sm) and diastolic (Em and Am) tissue Doppler imaging velocities are related with pulmonary artery systolic pressure (PASP) and length of the weaning process in mechanically ventilated patients with acute heart failure (AHF).

METHODS. RV fractional area change (RVFAC), left ventricular ejection fraction (LVEF), PASP, TAPSE, Sm, Em, Am RV TDI velocities, early diastolic mitral E wave and E 0 maximal TDI velocities of the mitral annulus at the lateral wall were obtained at admission by Doppler echocardiography in a cohort of 32 patients with AHF, presented with pulmonary oedema, who required positive-pressure ventilation for more than 48 h in the Intensive Care Unit (ICU). Echo-derived measures were compared between patients with and without pulmonary hypertension, whereas their association with duration of mechanical ventilation and length of the weaning process was tested with multivariate linear and logistic regression analysis. and increased E/E 0 ratio (10.8 ± 1.21 vs. 7.6 ± 0.54, p \ 0.001) compared with subjects with normal PASP (n = 10). These variables were negatively associated with duration of mechanical ventilation (R 2 = 0.55, beta slope = -0.89 for TAPSE, R 2 = 0.52, beta = -0.57 for Sm, R 2 = 0.45, beta = -0.27 for Em/Am, p \ 0.001) and were proven to successfully discriminate patients with (n = 12) and without (n = 20) prolonged weaning ([7 days of weaning after the first spontaneous breathing trial failure, p \ 0.001 for all comparisons).

CONCLUSIONS. We suggest that in critically ill patients with AHF presented with pulmonary oedema, low TAPSE and RV TDI velocities upon admission are associated with pulmonary hypertension and prolonged length of the weaning process. OBJECTIVES. The aim of the study was to study changes in cerebral blood flow (CBF), as determined by TCD, during the early postoperative course of CVS and to correlate such changes with post-operative NC.

We studied 72 patients undergoing extracorporeal circulation CVS (coronary by-pass, valve replacement or both) between March 2007 and March 2008. CBF was assessed by measuring bilateral MCA flow velocities by TCD before and 10, 24 and 48 h after CVS. Changes C20% between consecutive TCD results were considered significant. Demographic and clinical variables, co morbidities, Euroscore, SOFA, type and duration of surgery and type and severity of NC were also recorded. Patients were assigned to 3 groups according to CBF changes from baseline: A) Changes B20%; B) CBF increases C20%, C) CBF decreases C20%. NC were classified as major (stroke, TIA and coma) and minor (delirium, encephalopathy, transient cognitive impairment). We used descriptive statistics and inference by V 2 , ANOVA and Pearson's correlation.

Of the 72 patients, 20 were excluded (2 early post-operative death and 18 due to technical difficulties or incomplete TCD recordings). Of 52 evaluable patients, 17 (33%) had no CBF changes (group A), 12 (22%) had increases C20% (group B) and 23 (45%) had decreases C20% after CVS (group C). A positive correlation was found between CBF changes and duration of circulatory arrest (p \ 0.02), maximum SOFA score (p \ 0.001), respiratory dysfunction (p \ 0.016) and duration of mechanical ventilation (p \ 0.001). Neurological complications occurred in 16 patients (31%), of which 4 (25%) were major and 12 were minor (75% INTRODUCTION. The sole monitoring of macrohemodynamic variables is not always sufficient in the early detection of tissue hypoperfusion, especially in cardiac surgical patients that frequently present with microcirculatory derangements. Near infrared spectroscopy (NIRS) is an easily applicable non invasive technique that has been used to provide an estimate of tissue oxygenation at the bed side.

OBJECTIVE. The aim of our study was to evaluate the effect on outcome of guiding hemodynamic therapy and specifically inotrope titration in cardiac surgical patients postoperatively with NIRS.

METHODS. Patients operated on with cardiopulmonary bypass were assigned, after stratified randomization (gender, Euroscore-cutoff of 6), to an Intervention (IG) and a Control Group (CG). Postoperatively, following cardiac Intensive Care (cICU) admission, after initial resuscitation according to cICU protocol, StO 2 (%) was measured in patients of the IG in 3 muscle sites: thenar, masseter and deltoid. If it was less than 80% in 2/3 sites, dobutamine was administered in incremental doses (2.5 lg/kg/min), with the StO 2 (%) measured every half hour. The interventional period began upon cICU admission and lasted for 6 h, after which both groups were treated according to cICU protocol. Primary outcome measured was the oxygen consumption rate at the end of the 6 h intervention period as assessed with NIRS vascular occlusion technique.

. 18 patients were included in the study (8 in the intervention group and 10 in the control group). The 2 groups did not differ statistically significantly regarding age, Euroscore, and macrohemodynamic variables postoperatively (with the exception of CVP). Microcirculatory parameters upon admission to the cICU also did not differ, excluding masseter StO 2 (%). The oxygen consumption rate and the reperfusion rate increased in the 6 h study period in both groups, without differing statistically significantly between the groups at any time point (CG oxygen consumption rate 20.1 ± 12.4 upon cICU admission and 32.8 ± 15.5 6 h later, IG 15.4 ± 6.6 and 19.7 ± 7.7 respectively) (CG reperfusion rate 297 ± 210 upon cICU admission and 592 ± 254 6 h later and IG 201 ± 119 and 304 ± 163 respectively). As far as outcome parameters were concerned, the 2 groups did not differ statistically significantly in the total hours and total dose of vasopressors ± inotropes received, in the hours of mechanical ventilation, in the duration of cICU or hospital stay, and in SOFA scores the days following the operation.

CONCLUSION. NIRS guided titration of inotropes did not lead to a greater improvement in the microcirculation 6 h postoperatively, or to a better outcome. The limited power of the study prevents definite conclusions on the role of NIRS in hemodynamic therapy in cardiac surgery patients. OBJECTIVES. To estimate the prevalence of pulmonary embolism among MV patients in ICU and its association to deep vein thrombosis (DVT).

In a monocentric prospective observational study, we included all the patients requiring mechanical ventilation with no previously diagnosed PE, who underwent a thoracoabdominal CT contrast scanner for any medical reason. We used a modified protocol for PE diagnosis with a 64-multidetector row CT scan read by two independent radiologists. The association with a DVT was explored by performing venous compression ultrasound of four limbs. OBJECTIVES. The aim of this animal study was to evaluate the effect of intraabdominal hypertension on left ventricular diastolic function.

After approval by an institutional animal care committee, 7 rabbits were anesthetised before mechanical ventilation. An intraperitoneal infusion of 1.5% glycine solution was used to increase intraabdominal pressure to 20 mmHg. The right common carotid artery was catheterised in the neck in order to introduce a Millar Mikro-Tip catheter (Millar Instruments inc., Houston, USA) into the left ventricle. Heart rate, arterial pressure, central venous pressure, oesophageal pressure and intraabdominal pressure were measured. The s time constant of relaxation which is considered as best index of relaxation was calculated using the derivative method (1). All haemodynamic measurements were registered at baseline and after inducing intraabdominal hypertension. Data are presented as mean (IQR) and were compared using a Wilcoxon rank sum test.

RESULTS. Heart rate (from 202 ± 100 to 166 ± 100 beat/min, p = 0.6), mean arterial pressure (from 60 ± 29 to 50 ± 34 mmHg, p = 0.20) and dP/dt max (from 3,590 ± 255 to 2,111 ± 197 mmHg/s, p = 0.14) were not significantly modified by intraabdominal hypertension. However, the s time constant of relaxation increased significantly (from 16 ± 3 to 43 ± 18 ms; p = 0.05).

CONCLUSIONS. In this animal model, intraabdominal hypertension impairs left ventricular relaxation. These changes in the condition of the microcirculation have been related to the degree of organ dysfunction and thus patient outcome ie hospital length of stay. Near infrared spectroscopy (NIRS) is an easily applicable non invasive technique that has been used to provide an estimate of tissue oxygenation at the bed side.

OBJECTIVES. The aim of our observational study was to examine whether impaired tissue oxygenation as assessed with NIRS immediately postoperatively correlates with hospital length of stay.

Patients undergoing a planned cardiac surgical procedure on CPB were included in the study. Patients' thenar tissue oxygenation (StO 2 %) was assessed with NIRS postoperatively in the cardiac Intensive Care Unit (cICU).

RESULTS. 22 patients undergoing cardiac surgery on CPB (13 male/9 female) (age: 64 ± 14 years, EuroSCORE: 5.7 ± 3; mean ± SD) were enrolled in the study. Patients length of stay was 8.5 (3-26); median(range). The haemodynamic parameters of our patients upon admission to the cICU were: MAP 84 ± 12 mmHg, CVP 9 ± 4 mmHg, PCWP 12 ± 4 mmHg, MPAP 24 ± 5 mmHg, CI 2.6 ± 1.1 L/min/m2, SVR 1407 ± 501 dyne x s/ cm5, PVR 239 ± 106 dyne x s/cm5, HR 97 ± 15 bpm, Hb 11.7 ± 1.7 g/dl, lactate 2.8 ± 1.8 mg/dl; (all variables expressed as mean ± SD). Upon admission to the cICU all patients were mechanical ventilated, under vasopressor ± inotrope support and their central temperature was 36.9 ± 0.68; mean ± SD. The thenar StO 2 % was 80 ± 11; mean ± SD. Thenar StO 2 % correlated statistically significantly with hospital length of stay (r = 0.53, P = 0.11).

DISCUSSION. Tissue oxygenation as assessed with NIRS reflects the balance between regional oxygen delivery in relation to oxygen utilization. An elevated StO 2 in the presence of normal macrohemodynamics may reflect impaired oxygen consumption and thus an impaired microcirculation.

CONCLUSION. Patients with impaired tissue oxygenation immediately postoperatively have a longer hospital length of stay. Further studies are needed to confirm these results and to investigate the potential benefit from incorporating this information regarding tissue oxygenation in the treatment algorithm. OBJECTIVES. The goal of this study was to compare two different sedative agents for implantation of CRT-Ds related to incidence of adverse events and patient's satisfaction.

METHODS. The study included forty-two, ASA III-IV patients, undergoing transvenous implantation of CRT-Ds under local infiltrative anesthesia with 20 to 30 mL of 1% lidocaine. Intraoperative sedation was established with intermittent boluses of midazolam (1-5 mg) to achieve desirable level of sedation. Before the induction of ventricular fibrillation in order to test the defibrillator function of the CRT-D device, patients received an additional bolus of either propofol (1.5-2.5 mg kg -1 , P group, n = 20) or etomidate (0.1-0.25 mg . kg -1 , E group, n = 22) targeting BIS values in the range 50-60. The incidence of apnea, hypotension, nausea, myoclonus, pain at injection site, allergic reactions as well as patient's satisfaction with anesthesia described as feel of well being were registered and compared between groups.

RESULTS. In 23 subjects (56%) no complications were recorded. Myoclonus was registered in 5 patients from E group (23%) and in none from P group (p \ 0.05). No patients receiving etomidate reported pain at injection site compared to 4 patients (20%) receiving propofol (p \ 0.05). There was no significant difference in incidence of apnea between two groups (15 vs. 9%, p = 0.30). Two patients in P group (10%) and 1 in E group (4.5%) became hypotensive after delivering the hypnotic agent (p = 0.36). Also, there was no statistically significant difference between groups considering the frequency of nausea (0% vs.9%, p = 0.26). All the patients whom propofol had been delivered (100%) reported feel of well being and only four of them filed the same after etomidate (18%) (p \ 0.01). No allergic reactions and major adverse events were registered.

CONCLUSIONS. Implantation of CRT-Ds and its testing can be successfully performed with administration of both propofol and etomidate as a safe procedure with low per operative morbidity and shorter complication rates. Still, treating with propofol tends to be more satisfactory for the patients. INTRODUCTION. Ultra-short-acting b 1 selective adrenergic antagonists are now widely used to control tachycardia and tachyarrhythmia perioperatively. Among them, landiolol, a new ultra-short-acting b 1 -blocker, has been reported to exert a more potent negative chronotropic effect with little effect on blood pressure than esmolol (1). However, detailed mechanisms underlying different cardiovascular actions are still unknown.

OBJECTIVES. In this study we evaluated direct effects of landiolol on cardiac performance and single cell electrophysiology in comparison to those of esmolol.

METHODS. The present study composed of two parts. The first part of the study used isolated guinea-pig hearts which were perfused in the Langendorff mode at constant flow with oxygenated Tyrode solution at 37°C. The coronary perfusion pressure (CPP) was continuously monitored throughout the experiment, and intrinsic heart rate (HR) and isovolumetric left ventricular contraction were measured with a thin saline-filled balloon inserted into the left ventricle. The second part of the study was to measure action potentials and ionic currents in ventricular myocytes isolated enzymatically from guinea-pig hearts. Comparison of data was conducted by repeated-measure ANOVA with post hoc test (Bonferroni's correction). CONCLUSIONS. Esmolol had a more potent negative inotropic effect than landiolol. This effect is, at least in part, derived from shortening of APD. In addition, increase of the coronary resistance would facilitate the negative chronotropic action of esmolol in vivo. CONCLUSIONS. NMA moderates HPV in the conscious spontaneously breathing beagle, but not to the same degree as ACZ. As compared to ACZ, the additional methyl-group in NMA may impair its capability in vivo to act on a non-CA ACZ-sensitive cellular receptor or channel or that both, CA-dependent and CA-independent actions of ACZ yield a greater effect. INTRODUCTION. TEE with bubble test is considered as the ''gold standard'' method to detect a PFO with right to left shunt. TCD is a non-invasive method which has been shown to be as accurate as TEE for PFO detection.

We conducted a multicenter trial to estimate the prevalence of PFO, the influence of the size of the heart chambers on the prevalence of PFO and the accuracy of TCD as a non invasive method for PFO detection in mechanically ventilated ICU patients.

One hundred ICU patients (74 M and 26 F) under mechanical ventilation who needed a TEE study for hemodynamic assessment were included in the study. In each patient, the presence of a PFO was detected by TEE and TCD. Three bubble tests with agitated Haemacel Ò were performed by each method, with TEE probe at 608 and 908 rotation and with TCD the gate of pulse wave Doppler (PWD) at the M 1 segment of the middle cerebral artery (MCA). Patients without temporal acoustic window to perform TCD were excluded from the study. The size of PFO was classified as grade I, II and III according to the number of microbubbles passing from the right to the left atrium and the number of HITS (High Intensity Transient Signals) detected with PWD in the MCA (Grade I: \6 microbubbles or HITS, Grage II: [6 and \25 and grade III: more than 25 microbubbles or HITS). For each patient included in the study we measured and correlated the presence of PFO with the tidal volume (V t ), the plateau pressure (P plat ), the compliance of the respiratory system (C rs ) and the size of the right (RV) and left (LV) ventricle.

RESULTS. Mean P a O 2 /FiO 2 was 198 (min 50, max 350), mean C rs was 39 ml/cmH 2 O (min 11, max 72), mean V t was 574 ml (min 260, max 980) and mean P plat was 22 cmH 2 O (min 14, max 37). The prevalence of PFO detected with TEE was 28% and with TCD 48%. There was no PFO detected with TEE and missed by TCD. TCD was more sensitive than TEE in detecting PFO of grade I (7 with TEE, 17 with TCD) and II (6 with TEE, 16 with TCD), while for grade III the two techniques had equal sensitivity (15 with TEE, 15 with TCD). No correlation was found between P plat , C rs , V t and the presence of PFO. On the contrary, a strong correlation was found between RV dilatation and the presence of PFO (p \ 0.001).

CONCLUSIONS. The prevalence of PFO detected by TCD is very high in mechanically ventilated ICU patients and this may have important clinical implications. TCD is more sensitive than TEE in detecting a small PFO. The presence of RV dilatation increases the prevalence of PFO. OBJECTIVES. The aim of our study was to identify in mechanically ventilated patients for ALI/ARDS the prevalence of PFO and to evaluate the factors that may influence the prevalence of PFO.

METHODS. Two groups of MV patients, one with ALI/ARDS and one without respiratory failure (RF), were enrolled in the study. All patients underwent a TEE study for hemodynamic assessment. In each patient three consecutive bubble tests with agitated Haemacel Ò were performed at 60 0 and 90 0 rotation of the TEE probe. The bubble test was performed through a central line in the inferior or superior vena cava (IVC, SVC). A PFO was diagnosed by the presence of microbubbles in the left atrium within five cardiac cycles following the injection. Furthermore, in ALI/ARDS patients in whom a PFO was not detected at baseline MV, three consecutive bubble tests during recruitment maneuver at 45 cmH 2 O for 20 s were performed. The compliance of the respiratory system (C rs ), blood gas exchange and the ventilatory settings (P plat , V t ) were recorded in both groups. O, respectively. The presence of RV dilatation was a strong predictor for the FO opening (p \ 0.01); on the contrary, no statistical significant difference was found between the site of injection (SVC vs. IVC), the C rs , V t , and P plat and the presence or absence of a PFO.

A high prevalence of PFO was found in ALI/ARDS patients. RV dilatation seems to be the reason of this high prevalence. RV dilation may be due to the lower C rs and higher P plat of the ARDS patients. INTRODUCTION. The clinical evaluation of arterial tone is mainly based on the calculation of total systemic vascular resistance (TSVR). However, given the pulsatile nature of arterial flow, this parameter provides an inadequate assessment of vascular tone. Another approach proposed would take account of changes in pulse pressure and blood flow, relationship known as arterial elastance (Ea). So, for a given stroke volume, the blood pressure generated in the circulatory system will depend on Ea (1).

To assess the ability of the dynamic arterial elastance (Ea dyn ), defined as the relationship between pulse pressure variation (PPV) and stroke volume variation (VVS), to predict the hemodynamic response in mean arterial pressure (MAP) to a increase in stroke volume (SV) in hypotensive preload-dependent patients with acute circulatory failure.

We performed a prospective clinical study in a 17-bed multidisciplinary intensive care unit, including 21 patients with controlled mechanical ventilation and monitored with the Vigileo Ò monitor, for whom the decision to give fluids was taken due to the presence of circulatory, including arterial hypotension (MAP B 65 mmHg or systolic arterial pressure \90 mmHg), and preserved preload-responsiveness condition, defined as SVV C10%. Dynamic arterial elastance (VPP/VVS ratio), arterial pulse pressure to SV ratio, MAP/SV ratio, TSVR and MAP were compared to predict a MAP increase C15% after volume expansion (MAP-responders).

RESULTS. At baseline, only Ea dyn was significantly different between MAP-responders and nonresponders. VE-induced increase in MAP was strongly correlated with baseline Ea dyn (r 2 = 0.82, P \ 0.0001) and changes in Ea dyn after VE (r 2 = 0.77; P \ 0.0001). The only predictor of MAP increase was Ea dyn (AUC 0.98 ± 0.03; 95% C.I.: 0.8-1). A baseline Ea dyn value [0.89 predicted an increase C15% in MAP after fluid administration with a sensitivity of 92.9% (95% C.I.: 66.1-99.8%) and a specificity of 100% (95% C.I.: 59-100%).

CONCLUSIONS. Dynamic assessment of arterial elastance by PVV to SVV ratio during controlled mechanical ventilation could be used to predict mean arterial pressure increase after volume loading in hypotensive preload-dependent patients. Severe sepsis is one of the major reasons for intensive care unit (ICU) admission and leading causes of mortality. Some of these score systems have been customized for patients such as APACHE II, APACHE III, SASP II and MODS. This study is to assess the validity of mortality prediction systems in severe septic patients.

OBJECTIVES. The aim of this study was to compare and evaluate four severity scoring systems in intensive care unit (ICU), including APACHE II, APACHE III, SASP II and MODS in severe septic patient.

METHODS. Fifty-six severe septic patients were divided into two groups. One was survival group and the other was non-survival group. Besides general data, the continuous surveillance of APACHE II, APACHE III, SASP II and MODS were recorded by 1st, 3rd and 7th day.

RESULTS. Compared with survival group, MODS was significant difference in non-survival group only in 1st day (7.9 ± 3.1 vs. 10.1 ± 2.9, P \ 0.05) but APACHE II, APACHE III and SASP II were significant difference through 1st, 3rd and 7th day(P \ 0.05). In seven-day comparison, P value of APACHE III in non-survival group was the minimum (P = 0.000) and P value of MODS was the maximum (P = 0.008). In optimal survival evaluation, it seemed that APACHE III was the best (APACHE III [ APACHE II = SASPII [ MODS).

CONCLUSIONS. In order to evaluate the critical condition and prognosis of severe septic patients, APACHE III was the best and APACHE II and SASP II were followed and MODS was the worst. OBJECTIVES. To assess compliance with the CEM standards for management of severe sepsis across three ED sites in the West Midlands.

METHODS. Data was collected retrospectively over 3 months. Patients presenting to the ED within this period were assessed for likelihood of severe sepsis by the diagnostic code given to each patient upon leaving the ED. Data was analysed using a scanned copy of the ED clerking. Patients' notes were assessed for SIRS criteria and signs of new infection. If these criteria were met, and organ dysfunction was present, they were included in the audit.

RESULTS. 255 patients with severe sepsis were identified. Of these 17% were documented as septic by ED staff. The CEM standards of care were received in 41% of patients with a documented diagnosis of severe sepsis in the ED, and 23% of patients overall. 89% of patients received the 'treatment' aspects of care: Oxygen, IV antibiotics (with blood culture) and IV fluids. 71% of severely septic patients had no documented consideration of ICU referral.

CONCLUSIONS. Early recognition of severe sepsis in the ED led to greater performance in meeting the CEM standards. Although 100% of patients received observations and 89% received the treatment interventions, we performed poorly in meeting the remaining CEM standards. The trust has developed a severe sepsis proforma which incorporates the CEM standards to accurately record the completion of each intervention. A sepsis course for staff has been launched trust wide, and a formal referral process to ICU for all severely septic patients is being implemented. OBJECTIVES. To observe association of body temperature (BT) and antipyretic use with mortality in the critically-ill.

A prospective multi-national, multi-center observational study. Consecutive patients whose ICU stay were expected to be more than 48 h were recruited from 15 centers in Japan and 10 centers in Korea. Patient's BT was prospectively recorded every 4 h until patient's death, discharge from the ICU or up to 28 days. Information including patient's clinical characteristics at admission, presence of infection, and use of steroids, extracorporeal circuit, and antipyretics were recorded. (1). While blood culture results take time, treatment for bloodstream infection should be provided swiftly, usually before results are available (2) . Prior treatment with antimicrobials increases the chances of false negative results. Haste, poor technique and alteration in commensal flora may increase the chances of falsely identifying pathogens.

OBJECTIVES. We have investigated the utility of blood culture tests in our general critical care unit over 1 year in terms of results yielded and actions prompted.

METHODS. The indication for blood culture was clinician's discretion. All critical care sourced blood cultures for the period Oct 2008 to Sept 2009 were reviewed from the microbiology laboratory database. Blood culture specimens were collected in BacT/ALERT Ò bottles (Biomerieux, Durham, NC, USA 27704). Notes review was made of the positive blood culture episodes to determine actions after the results were known. Consideration was given to the source of the blood sample: clean stab versus from an intravascular device. Categoric data was analysed using the Chi-squared test and P value of 0.05 was accepted as significant. OBJECTIVES. We hypothesized that in the emergency department of our hospital many patients with sepsis are not recognized as such.

METHODS. In a retrospective design, patients of an age of 18 years and older who were admitted to the emergency department during a period of 3 months between January-April 2008 and diagnosed as having an infection were included. The diagnose infection was made on admission by the emergency department nurse. The included patients were either classified as having sepsis or not having sepsis, according to the SIRS criteria. CONCLUSIONS. H1N1 infection was associated with significant morbidity and mortality. It occurred mainly in young pts with co-morbidities and was associated with severe hypoxemia, a trigger for prolonged mechanical ventilation and frequent use of lung rescue therapies. A significant delay in hospital admission and start of antiviral therapy should also be noted. Admission to administration time difference between cycles was 1.26 h, with a mean reduction of 1.04 h between clinician assessment and prescription time in cycle two.

We identified delays against the standard after both cycles of the audit. We demonstrated that the method of prescription should be taken into consideration when prescribing antibiotics in patients with suspected sepsis. There are a multitude of factors that could contribute to a reduction in the clinician assessment to prescription time, which may be investigated in further audits. CONCLUSIONS. Despite high levels of resistance among PSA and AB from these ICUs, CFR for most carbapenem dosing regimens were above the reported susceptibility. Doripenem provided greater CFR than meropenem, which was superior to imipenem against these isolates. While higher doses combined with prolonged infusions significantly improved CFR against PSA, alternative therapeutic strategies will be required to address these highly resistant AB.

GRANT ACKNOWLEDGMENT. The PASSPORT study is supported by a grant from Janssen-Ortho-McNeil. INTRODUCTION. Drug interactions are common, and the effects of these interactions can range from innocuous to deadly. Critically ill patients often receive a variety of potent drugs, including antimicrobials, making this population extremely susceptible to drug-drug interactions. Therefore, physicians must be familiar not only with the antimicrobial drugs capable of producing adverse drug events, but also their potential drug-drug interactions. There are scarce data about the incidence of these types of drug interactions and the how frequently it might cause adverse events.

OBJECTIVES. The purpose of this study is to evaluate the incidence of potential drug interactions involving antimicrobials and the possibility to cause adverse events.

The clinical pharmacist has prospectively analyzed ICU prescriptions between January 2009 and December 2009 with the purpose to identify potential drug-drug interactions involving antimicrobials. The screening was done with the relief from a software (Epocrates Rx Ò drug reference). The interactions detected were classified in eight groups according to the affected system (neurological, cardiovascular, gastrointestinal, renal, endocrine, hematological, musculoskeletal and others) and through the type of interaction (pharmacokinetic, pharmacodynamic and others). We have identified the most common potential effects, the medications involved and have observed the incidence of adverse drug events.

RESULTS. The ICU admitted 347 patients during the study period. We have analyzed 661 physician orders with 8209 prescribed items. We have identified 871 antimicrobial drug interactions (71 different interactions) which compound 20% of the total drug interactions (n = 4349). The cardiovascular system and the pharmacokinetic interaction were the most potentially affected (38%; 42%). The most common medications involved were: fluconazole (24%), clarithromycin (23%), levofloxacin (12%); linezolid (11%). The clinical pharmacist has made an intervention regarding medication safety in 3% (n = 27) and the acceptance rate by the medical ICU staff was 74%. We have not been able to identify any adverse drug event caused by drug interaction even with our active search and the spontaneous reports. However, sub notification must be taken into consideration.

CONCLUSIONS. Clinicians should be aware of potential drug-drug interactions when making therapy selections for critically ill patients. Antimicrobial drugs are susceptible to interact with other drugs, which may increase the risk of adverse drug events. The clinical pharmacist interventions may improve clinical outcomes by optimizing medication use, monitoring potentially preventable adverse drug events and promoting information about this important issue to the ICU multi-professional team. INTRODUCTION. Cefazolin is one of the most frequently administered antimicrobial agent for prophylaxis in ''clean'' surgery. Its broad spectrum against Gram + micro-organisms and its pharmacological characteristics make it an easy-to-use choice to prevent infections caused by Staphylococcus aureus and coagulase-negative stapylococci.

OBJECTIVES. The aim of this study is the evaluation of the plasma concentrations of cefazolin administered as a prophylactic antimicrobial agent during cardiac surgery with CPB. Adequate cefazolin plasma levels can maintain a tissue concentration high enough to prevent the risk of developing post-operative infections.

After obtaining Ethical Committee approval and personal written consent, two groups of patients were enrolled in this prospective study. The first group, 12 patients, received cefazolin, 2 g, 30-60 min before skin incision and 1 g adjunctive dose after 4 h. Then, three 2 g doses were administered every 8 h. In the second group of 12 patients the adjunctive 1 g cefazolin dose was given at the beginning of the CPB. Blood samples were collected immediately before the first dose and every hour for the whole time of surgery, and, only in the second group, after surgery, at 12th, 18th and 24th hour. Plasma cefazolin concentration was determined with a biological radial diffusion assay.

RESULTS. Plasma cefazolin was constantly higher than the MIC90 of the most involved micro-organisms (according to CLSI). In the first group, cefazolin concentration suddenly decreased after starting CPB. The 1 g adjunctive dose immediately restored it. The earlier administration of this dose in the second group prevented this sudden fall. Plasma cefazolin was maintained at effective inhibitory levels for the whole time of surgery in all patients ([32 mcg/ml). During the postoperative period cefazolin decreased slowly, but inhibitory plasma levels were always maintained. The rate of cefazolin clearance was found equal to the creatinine clearance in all patients.

Perioperative plasma cefazolin concentration CONCLUSIONS. The administration of cefazolin 2 g every 8 h can guarantee effective inhibitory plasma concentrations during surgery and during the first 24 h after surgery. CPB causes a sudden fall in cefazolin plasma levels. This can be avoided administering an adjunctive 1 g dose immediately before starting CPB. OBJECTIVES. Vancomycin dose regimen was adjusted based on trough plasma levels in burn patients that were distributed according to the extension total burn surface area (TBSA); also pharmacokinetics changes were compared.

METHODS. Twenty seven adult burn patients of both sexes, requiring antimicrobial therapy with vancomycin for the control of sepsis were investigated. Pharmacotherapeutic follow up was performed in a serial of periods (112 observations) for all patients investigated by collection of blood samples, 1 mL each from the venous catheter as follows: 1st blood sample collection, 2 h after the beginning of drug 1 h infusion and a 2nd sample blood collection at the trough, immediately before the next dose. If necessary, additional sample blood collections were performed based on the laboratorial data for patients any time, for dose adjustment purpose and optimization of drug therapy. Vancomycin plasma concentrations were determined by highperformance liquid chromatography. Plasma curve decay was plotted, and pharmacokinetics was analyzed by one-compartment open model against the reference data reported.

RESULTS. Burn patients receiving the empiric dose regimen showed trough plasma level lower than the minimum effective concentration, consequently dose adjustment was required. Vancomycin adjusted dose regimen showed statistical significance differences according to TBSA (p \ 0.05) as follows for daily dose normalized to body weight and expressed by mean ± SD: 25.5 ± 11.9 mg/kg/day were required for patients with TBSA below 20%, 32.0 ± 6.4 mg/kg/day for TBSA 20-40% and 34.8 ± 12.2 mg/kg/day were required for TBSA above 40%. Relevant changes on pharmacokinetics were observed by drug plasma clearance increased according the increase of TBSA (p \ 0.05), while the apparent volume of distribution and also the biological half-life remained unchanged. Additionally, a weak correlation was observed between vancomycin plasma clearance and creatinine clearance (R 2 = 0.36; p = 0.0004), probably due to the contribution of the extra-renal clearance on total drug elimination.

On the basis of data obtained in the present study and to prevent therapeutic failure and also to reduce the risk of bacterial resistance, dose adjustment in burn patients is recommendable based on vancomycin plasma monitoring and also on the extension of total burn surface area. INTRODUCTION. The importance of early antibiotic therapy has been recently demonstrated. Regarding a rapidly increasing number of obese patients, appropriate drug dosage in these patients is an important challenge of critical care since it has been shown that not only early start of antibiotics but also correct target concentrations decrease mortality. Vancomycin is administered according to body weight (BW). Nevertheless, little is known about the percentage of obese patients achieving pre-defined target serum levels within 24 h after initiation of vancomycin therapy compared to patients with normal BW.

OBJECTIVES. Therefore, it was the aim of our study to analyze the appropriateness of serum vancomycin levels in patients with a BW between 40 and 150 kg. Vancomycin is almost entirely excreted by the glomerulus and may be responsible for nephrotoxicity [1] . However, there is a lack of definitive evidence linking concentrations to either outcome or toxicity [2] . Few reports exist comparing intermittent dosing and continuous infusion. Ingram [3] suggested that whilst associated with a slower deterioration in renal function, there was no difference in the prevalence of nephrotoxicity. Similarly, Hutschala [4] demonstrated worsening creatinine in 149 patients following cardiac surgery with both intermittent and continuous infusion but infusion tended to be less nephrotoxic despite receiving higher doses. We wish also to report our experiences with vancomycin infusion in critically ill cardiac patients. METHODS. We examined retrospective data from 2,512 patients treated with vancomycin. We perform adjusted and un-adjusted analysis using SOFA on the day of starting vancomycin and total dose received. 

To assess the differences in either an initial pulmonary or non pulmonary presentation.

METHODS. Prospective, observational, multi-center study conducted in intensive care (ICU). We reviewed demographic and clinical data for all Pandemic H1N1 Influenza A infections reported in the ESICM H1N1 registry.

RESULTS. 512 patients were screened from the registry. 330 patients with completed data entry for pulmonary and non pulmonary with outcomes were identified and analysed. All patients had either suspected, probable or confirmed pandemic H1N1 Influenza A infection and were being cared for in an ICU. 53% of the patients were male with a median age of 43 (IQR 32-55) years. The admission mean SAPS3 score was 54 ± 16 and the APACHE II score was 21 ± 9. 33% of the patients subsequently received non invasive ventilation and 72% received invasive mechanical ventilation. The ICU mortality rate was 30%. The Hospital mortality was 32%. 77% of patients presented with a pulmonary presentation. 73% of these were admitted with ARDS and/or bacterial pneumonia and 22% with an acute bronchospastic exacerbation. 23% of patients were admitted to the ICU with a non pulmonary presentation. The main reasons for admission in these patients were: cardiovascular instability (33%), altered level of consciousness (20%), renal failure (11%) and acute coronary syndromes (5%). Patients with a pulmonary presentation were older, had a increased history of asthma or COPD and were more likely to be ventilated. They had a higher mortality rate in the ICU. Non pulmonary presentations were more likely to suffer from chronic renal impairment. 

A total of 512 episodes of pandemic Influenza A (H1N1)v infections in critical care setting were analyzed: 169 with bacterial pneumonia (94 males and 73 females) and 343 with wheezing or viral pneumonia (182 males and 161 females). The mean age was 45 (± 17) years in patients with bacterial pneumonia and 40 (± 17) in patients viral pneumonia. The mean APACHE II score was 25 (± 9) and 20 (± 9), with a corresponding probability of death of 37 (± 25)% and 24 (± 21)%.

Comorbidities were common, but without significant differences between the two groups (only exceptions pregnancy-more prevalent in patients without bacterial pneumonia-and dialysis dependence-more prevalent in patients with bacterial pneumonia). At ICU admission shock and acute renal failure were more common in patients with bacterial pneumonia. In patients without pneumonia; Severe hypoxia and ARDS did not presented significant differences between groups. AIMS. Evaluation if an isocaloric beginning of artificial nutrition in critically ill medical patients is associated with increased nutritional related side effects compared to a hypocaloric start.

METHODS. 110 critically ill medical patients with an expected need for artificial nutrition of [5 days were included into this prospective, randomized clinical study. Artificial nutrition was started either isocalorically right from the beginning (group A; n = 55) or hypocalorically (50% of the energy demands) followed by a stepwise increase over the next 2 days (day 2: 75%); day 3: 100%) (group B; n = 55). Nutrition related side-effects were defined as the occurrence of hyperglycemia, hyperlactatemia, hypertriacylglycerolemia, upper digestive intolerance, cholestasis, or diarrhea as well as disturbances of serum electrolytes and were assessed on a daily basis.

Patients were randomized to receive either an artificial nutrition started isocalorically (group A) or hypocalorically followed by a stepwise increase (group B). Of the 150 patients, 100 patients completed the study (group A: n = 55; group B: n = 55). The calculated, cumulative energy requirements of patients of group A and B were 9,058 ± 1,236 and 9,125 ± 1,569 kcal, respectively (p = ns). Patients of group A received 76 ± 17% and patients of group B 69 ± 12% of the calculated energy requirements (p \ 0.01). The incidence of nutritional related side effects was not different comparing both groups, except for hypophosphatemia, which was more pronounced in group A. Additionally, exogenous phosphate needs were higher in patients of group A. The number of interruptions of the artificial nutrition did not differ between groups.

CONCLUSIONS. An isocaloric start of artificial nutrition provided more energy during the first 5 days of their ICU stay than a hypocaloric beginning. There was no difference in the number of interruptions and in the incidence of nutritional related side effects, except hypophosphatemia suggesting the presence of refeeding syndrome. In studies carried on to demonstrate positive effects of glutamine (Gln) that has innumerable biological features, the main point of discussion isn't whether Gln has positive effects in sepsis but rather the effect difference between different administration routes. Only enteral (en.) or parenteral (pn.) administration was analyzed in this respect and no studies on combined administration were performed. The primary endpoint in this study was to analyze the effects of administration of en. and pn. Gln together or separately on intestinal mucosa + immune system in the experimental sepsis model. For this purpose villus atrophy, bacterial growth in blood and tissue, levels of blood Gln, TNFa and IL10 were examined. The secondary endpoint was to evaluate the different administration models in terms of cost.

Wistar, adult female rats were used. They were fed standard. Sepsis was developed in 4 groups (all rats) by injection of intraperitoneal(ip.) 1 mL (5 9 10 7 CFU/mL) E. coli. Grup C (n = 12):en./pn. isotonic saline (1 mL/day; 2 mL/d); Grup E (en., n = 10):en. Gln (0.5 g kg -1 day -1 ) + pn. saline (2 mL/d); Grup P (pn., n = 11):pn. Gln (0.4 g kg -1 day -1 Gln) + en. saline (1 mL/d); Group EP (en. ± pn., n = 11):pn. Gln (0.3 g kg -1 day -1 ) Ala-Gln = 0.2 g kg -1 day -1 Gln) + en. Gln (0.3 g kg -1 day -1 ); were administered. Feeding of rats began 4 h (h) after administration of ip E. coli. Blood Gln (with spectrophotometer), TNFa and IL10 concentrations(with ELISA) were examined at the start (baseline levels) and at 24-96 h after the experiment started. Samples of tissue from mesenteric lymph node, liver, lung, blood and small intestine were collected. Ala-Gln = 0.27 g kg.

. Rates of reproduction of the strain administered were found lower for Group EP than Group C (p \ 0.05). Rates of villus atrophy in ileum of Group EP, P and E were lower than Group C (p \ 0.05).Plasma Gln levels were found lower in groups EP and P at 24 h, and higher at 96 h than other groups (p \ 0.05). When plasma Gln levels at 96 h were compared with their baseline levels, significant increases were detected in Groups EP and P and significant decreases were detected in Groups C and E (p \ 0.01). Serum TNFa and IL10 levels were found lower for Groups EP and P at 24 and 96 h when compared between groups (p \ 0.05). When Serum TNFa and IL10 levels at 96 h were compared with their baseline levels, more distinctive increases were detected in Groups C and E than other groups (p \ 0.01). Significant positive correlation was determined between TNFa and IL10 levels at 24 h (p \ 0.01) and 96 h (p \ 0.01). Cost of simultaneous administration of en. and pn. Gln was higher than en. administration but close to pn. administration at these doses. METHODS. Medline and EMBASE were searched. Hand citation review of retrieved guidelines and systematic reviews was undertaken and academic and industry experts were contacted. Only methodologically sound randomised controlled trials (RCTs) were eligible for inclusion in the primary analysis. The primary analysis was conducted on clinically meaningful patient oriented outcomes, which included mortality, functional status and quality of life. Secondary analyses considered vomiting/regurgitation, pneumonia, bacteremia, sepsis and multiple organ dysfunction syndrome. Meta-analysis was conducted using the Peto analytic method, which is known to minimize bias in the presence of sparse events. The impact of heterogeneity was assessed using the I 2 metric. RESULTS. 4,217 unique abstracts were identified, resulting in the retrieval of 701 papers for detailed eligibility review. Four RCTs were identified to be on topic however one RCT reported excessive loss to follow-up such that an intention to treat analysis could not be conducted. Analysis based on the three methodologically sound RCTs demonstrated the provision of early EN was associated with a significant reduction in mortality (OR = 0.20, 95% confidence interval 0.04 to 0.91, I 2 = 0). No other outcomes could be pooled. Sensitivity analysis including all four on-topic RCTs (OR = 0.26, P = 0.04, I 2 = 0), and a simulation analysis conducted using a different analytical method. (OR Exact = 0.14, 95% CI 0.01 to 0.97), confirmed the presence of a mortality reduction. CONCLUSIONS. Although the detection of a statistically significant reduction in mortality is promising, overall trial size was small. The results of this meta-analysis should be confirmed by the conduct of a large multi-center trial. REFERENCE(S). RESULTS. The mean IBP was 10.7 ± 4.1 and mean IGP was 11.6 ± 4.1. Correlation between the IBP and IGP was significant however moderate (R2 = 0.51). Analysis according to Bland and Altman showed a bias and precision of 0.8 and 2.7 mmHg respectively, however the limits of agreement (LA) were large and ranged from -4.5 to 6.1 mmHg. The median GRV proto was 80 mL (0-1,050) and median GRV classic was also 80 mL (0-1,250). Correlation between the 2 methods was excellent (R2 = 0.89). Analysis according to Bland and Altman showed a bias and precision of -0.8 and 52.3 mL respectively and the limits of agreement (LA) ranged from -103 to 102 mmHg. The median drainage time and return times were 5 min (0.5-15) and 2.5 min (0-21) for GRV proto compared to 2 min (0.1-10) and 1 min (0-8) for GRV classic. A preliminary cost effectiveness analysis shows that the price of measuring GRV with the classic method ranges from 3.84€ to 24.18€ per day, depending on the GRV size. Price of measuring GRV with the gastro PV system is independent of GRV size and is estimated at 9.49€ per day. The gastro PV system if priced at 8.5€ could become cost effective at GRV of 100 cc and more.

CONCLUSIONS. The interim results of an ongoing multicentre pilot study show that the gastro PV is a good alternative to the standard method for measuring GRV. Because the nurse can perform other tasks during drainage and return of the GRV, and the fact that the system remains closed during measurement, this could be a major step forward in standardisation of GRV measurement. Furthermore it allows screening for intra-abdominal hypertension via IGP estimation.

ACKNOWLEDGMENT. The gastro PV devices were provided by Holtech Medical, free of charge. INTRODUCTION. The importance of early enteral feeding of the critically ill patient has been well documented. It is the more physiological approach, which is associated with lower rates of infectious complications. Early enteral nutrition within 24 h is recommended by the ESPEN Guidelines on Enteral Nutrition. A recent meta-analysis revealed that mortality and the incidence of pneumonia were significantly reduced in patients with enteral nutrition within 24 h. Parenteral nutrition may be associated with higher mortality.

OBJECTIVES. Evaluation of a new technique for the placement of postpyloric feeding tubes by intensive care physicians.

METHODS. Prospective cohort study in 27 critically ill patients subjected to transnasal endoscopy and intubation of the pylorus. Attending intensive care physicians were trained in the handling of the new endoscope for transnasal gastroenteroscopy for 2 days. A jejunal feeding tube was advanced via the instrument channel and the correct position assessed by contrast radiography. Primary outcome measure was successful postpyloric placement of the tube. Secondary outcome measures were time needed for the placement, complications like bleeding and formation of loops and the score of the placement difficulty graded from 1 (easy) to 4 (difficult). Data are given as mean values and standard deviation.

Out of 34 attempted jejunal tube placements, 28 tubes (82%) were placed correctly in the jejunum. The duration of the procedure was 28 ± 12 min. The difficulty of the tube placement was judged as follows: grade 1: 17 patients, grade 2: 8 patients, grade 3: 7 patients, grade 4: 2 patients. In 3 cases, the tube position was incorrect, and in another 3 cases, the procedure had to be aborted. Only in one patient, bleeding occurred that required no further treatment.

CONCLUSIONS. Fast and reliable transnasal insertion of postpyloric feeding tubes can be accomplished by trained intensive care physicians at the bedside using the presented procedure. This new technique may facilitate early initiation of enteral feeding in intensive care patients.

GRANT ACKNOWLEDGMENT. The authors acknowledge the support of Pentax, Hamburg, Germany, who provided the endoscope used in the study and of Fresenius Kabi, Bad Homburg, Germany who provided the feeding tubes. 

A well-nourished condition before prolonged endotoxemia results in a better ability to adapt to endotoxin-induced metabolic deterioration of arginine-nitric oxide metabolism than does reduced caloric intake before endotoxemia (1). The role of individual organs in the arginine-citrulline metabolism during malnutrition and sepsis is unknown and may be key to direct future interventions.

To study the effects of reduced caloric intake and endotoxemia on the citrulline-arginine metabolism in the gut-liver-kidney axis.

Organ arginine-nitric oxide metabolism was measured by using a primedconstant stable-isotope infusion of [15N2]arginine and [13C-2H2]citrulline during 4 conditions; a 7-day reduced caloric intake feed regimen (STARV; n = 9), normal control feed regimen (CO; n = 9), endotoxemia alone (CE) and reduced caloric intake and endotoxemia (RE) in. Catheters for blood sampling were placed in the abdominal aorta, which, in combination with the catheters in the portal, hepatic and renal veins, served for metabolic measurements across the portal-drained viscera, liver and the kidneys, respectively.

RESULTS. Interestingly, RE animals had similar citrulline appearance from the gut (331 ± 36 nmol/kg/min) compared to control and animals during CE, but higher in endotoxemia alone (102 ± 14, P \ 0.01). This was related to a significantly higher NO production from the gut in the RE group (1,644 ± 153 vs. 720 ± 41, P \ 0.01). In the kidney arginine appearance from citrulline decreased significantly during RE compared the control animals (3 ± 7 vs. 224 ± 30 nmol/kg/min, P \ 0.01). In contrast, the liver disposed more arginine in the RE group compared to the other conditions, while NO production was not higher.

CONCLUSIONS. Despite reduced caloric intake prior to endotoxemia, the gut remains capable of increasing release of citrulline, although the capability of the kidney for the de novo production of arginine is severely compromised. Metabolic control of the citrullinearginine metabolism in the gut-liver-kidney axis should focus on increasing de novo arginine production from citrulline. OBJECTIVES. The aim of this study was to measure duodeno-caecal transit times of enteral feed in this patient group using a scintigraphic technique.

A prospective observational study was performed in 16 mechanically ventilated critically ill patients (12 M, age 49 ± 17 yr, BMI 25 ± 5 kg/m 2 , ICU admission day 5 ± 3, APACHE II on study 20 ± 7; mean ± SD) and 6 healthy subjects (3 M, age 24 ± 9 year, BMI 24 ± 45 kg/m 2 ). After a 6 h fast a 100 ml enteral feed (Ensure 1 kcal/min), labelled with 20 MBq 99m Tc-sulphur colloid, was infused into the distal duodenum over 6 min. Dynamic anterior scintigraphic images were recorded in 3 min frames for 240 min and the time of first appearance of activity in the caecum was recorded by two blinded operators (KJ, AR). Data were assessed using Mann Whitney U test and are presented as median (IQR). INTRODUCTION. Erythromycin, a macrolide antibiotic is widely used as a prokinetic agent in intensive care unit (ICU) despite the lack of data supporting its prolonged effectiveness in enteral nutrition (EN) intolerant critically ill patients.

To evaluate impact on clinical outcome of erythromycin prescription as prokinetic agent in ICU.

All patients consecutively admitted from January 2008 through December 2009 mechanically ventilated for more than 5 days and receiving EN were included in an observational cohort study. EN intolerance was defined clinically as a 8-hourly gastric residual volume (GRV) C150 mL or vomiting. Successful EN was defined as a GRV\150 mL with a feeding rate C40 mL/h. Erythromycin prescription was left to practician appreciation. OBJECTIVES. This study aims at evaluating the relationship between diarrhoea and EN in ICU patients.

METHODS. During 1 month, the days with and without diarrhoea (C 3 liquid stools/day) and the characteristics of nutritional support of all patients staying in our ICU were recorded. Patients staying \24 h or presenting an intestinal stomy were excluded. We compared, between days with and without diarrhoea, total energy coverage and energy coverage by EN as % of needs, EN energy intake and EN volume for each patient. Needs were estimated as 25-30 kcal/kg body weight for women and men, respectively. The relationship between antibiotics, laxative treatment and diarrhoea was also analysed. Results are presented as mean ± SD. Comparisons were made by Mann-Whitney test. The risk of diarrhoea with EN was calculated by odds ratio and confidence intervals (CI).

The study included 1133 days of hospitalisation of 171 patients (60 ± 16 years, BMI 26 ± 5 kg/m 2 , SOFA score at admission 6 ± 6). EN was present in 85 days of diarrhoea and 581 days without diarrhoea. Determining the small bowel function is of great concern in ICU patients, because a malfunctioning small bowel may predispose to malnutrition and may increase the risk of SIRS. A recently developed test, the Citrulline Generation Test (CGT), measures the enterocytes' capability to convert glutamine into citrulline. The production of citrulline exclusively takes place in functioning enterocytes, therefore this conversion represents small bowel function. OBJECTIVES. We aimed to define the CGT reference values in 16 'stable' ICU-patients to assess small bowel function. Secondly, we wanted to compare four different CGT methods; enteral and iv administration of Dipeptiven and measurement of citrulline in both arterial and venous samples.

We performed the CGT on 16 stable ICU-patients, defined as having respiratory failure but not dependent on vasopressors. They had a normal renal function and were able to tolerate enteral nutrition. A 5 h fast was followed by administration of 20 g of glutamine-alanine (Dipeptiven Ò ) either intravenously or enterally, randomly determined. The next day the same test was performed by using the other route. After each administration of Dipeptiven, citrulline levels, both arterial and venous, were measured at fixed time points using reverse-phase high performance liquid chromatography (HPLC). RESULTS. Nine females and 7 males were admitted to the ICU with either a medical (11) or a surgical (5) diagnosis. They had a mean (± SD) age and BMI of 60.9 ± 10 years and 26.7 ± 7, kg/m 2 respectively. Their median APACHE II score was 24.5 (IQR = 19.5-26.8).

On the day the CGT was performed their median SOFA score was 4.0 (IQR = 3.0-4.8 Early post-pyloric feeding has been shown to improve clinical outcomes [1] . Commonly used methods for placing a nasojejunal tube (NJT) are blind, endoscopic or fluoroscopic placement. The later two methods are relatively invasive, expensive and can cause delay to feeding, whereas blind placement is often unsuccessful. Electromagnetic sensor guided NJT insertion is a bedside technique able to confirm successful placement without the need for abdominal X-ray. The system incorporates a liquid crystal display and a receiver unit. The receiver is placed over the patient's xiphoid process and picks up the signal from an electromagnetic transmitter located at the tip of the feeding tube. The screen provides a visual aid to enable the operators to trace the route of the tube tip and identify its' location according to anatomical markers. OBJECTIVES. We were interested to determine the suitability of electromagnetic sensor guided NJT insertion especially in relation to success rate and procedure time.

METHODS. Fifty patients were referred for electromagnetic NJT insertion on units at the Leeds Teaching Hospitals. Insertion time was measured from oesophageal visualisation until post-pyloric placement. Various positional manoeuvres were employed along with administration of sedatives, prokinetics and air insufflation when applicable. All insertions were carried out by experienced investigators. All NJT insertions were confirmed by abdominal X-ray. Data collection included patient demographics, hospitalisation and procedural information.

RESULTS. Forty male and 10 female patients, mean age 40 (range 1-81 years), BMI mean 25 (14-33), had attempted electromagnetic NJT placement. Patients had been hospitalised for a median of 12 days (1-180). Indication for NJT insertion was either large aspirate and/or reflux (86%). Seventy six percent of patients had an artificial airway and 50% of patients were receiving sedation. Forty six percent of patients received metoclopramide and 58% air insufflation. Thirty six percent of patients were moved into either left or right lateral position. Successful post-pyloric placement was achieved in 90% of patients confirmed by additional abdominal X-ray. Procedural time varied from 2 to 180 min (mean 23). Two of the placement failures were due to patient intolerance. CONCLUSIONS. Bedside electromagnetic guided NJT placement technique is an acceptable method of placing post pyloric feeding tubes with a high success rate. 

Gastrointestinal failure (GIF) score has been suggested (1). The GIF score defines GI failure as the occurrence of feeding intolerance (FI) and intra-abdominal hypertension (IAH) simultaneously.

To compare the outcome of patients with primary vs. secondary GIF.

METHODS. All consecutive, mechanically ventilated (MV) patients treated for at least 24 h during January 2008 to December 2009 in two ICUs were studied. GIF was defined as GIF score equal or above 3 points according to the GIF score (1). 3 points = FI and IAH simultaneously; 4 points = abdominal compartment syndrome (ACS). FI was defined as the need to stop enteral feeding for any clinical reason (vomiting, high gastric residuals, bowel distension etc). IAH was defined as mean intra-abdominal pressure (IAP) C12 mmHg on any day. ACS was defined as IAP [ 20 mmHg with the new onset organ failure. When GIF developed in a patient with primary pathology in abdomino-pelvic region it was classified as primary GIF, when occurred without previous pathology in abdomino-pelvic region it was taken as secondary. OBJECTIVES. In this study the biochemical quality and prion safety of the pharmaceutically licensed plasma OctaplasLG Ò was evaluated. The prion reduction factor achieved by Western blot was confirmed by animal studies.

Eighteen consecutive batches of OctaplasLG Ò (Octapharma PPGmbH, Vienna, Austria) were tested on global coagulation parameters, fibrinogen levels, activities of coagulation factors and protease inhibitors, activation markers, as well as von Willebrand factor multimers. In parallel studies, plasma pool was spiked with exogenous spike material, derived from brains of hamsters infected with hamster-adapted scrapie 263 K, and a down-scale of the OctaplasLG Ò manufacturing process was performed. The PrP Sc reduction factor for the resin was investigated in both Western blot and hamster bioassay studies.

A reduction factor of C3.1 log10 PrP Sc was found for this process step by Western blotting. The outcome of the hamster bioassay confirmed that the high level of removal PrP Sc seen during OctaplasLG Ò manufacturing was equivalent to a removal of infectivity (3.0 log10). In Octa-plasLG Ò , a parallel reduction of the S/D virus inactivation step led to significantly higher activities of plasmin inhibitor.

Our studies demonstrated that the same amounts of PrP Sc and prion infectivity bind rapidly and with a very high affinity to the chromatography resin. OctaplasLG Ò has the same clinical safety and efficacy profile compared to that demonstrated by Octaplas Ò over the last 18 years, except for the increased safety margin in terms of prion disease transmission and the possible effect of a significantly increased plasmin inhibitor activity. Uniplas Ò is a second generation solvent/detergent (S/D) treated, coagulation active plasma for infusion produced with an implemented prion removal step. It was developed as an alternative to the blood group specific S/D plasma products, OctaplasLG Ò and Octaplas Ò , in order to obtain an universally applicable (i.e. blood group independent) plasma that can be used without taking into account the blood group of the recipient. Due to an initially controlled, optimal mixing of plasma of different blood groups prior to S/D treatment, in Uniplas Ò , the blood group specific antibodies (anti-A and anti-B of both IgM and IgG type) are neutralised and/or removed by free A and/or B substances and red blood cells (RBCs) to a clinical acceptable level with very limited or no complement activation. OBJECTIVES. In this study an extensive biochemical characterisation of the first Uniplas Ò validation batches was performed.

METHODS. Three batches of Uniplas Ò were produced by OctapharmaPPGmbH (Vienna) under production conditions in [2008] [2009] . Uniplas Ò batches were tested on all important coagulation factors, protease inhibitors, activation markers, ADAMTS13 and factor H levels, as well as von Willebrand factor multimers. In addition, anti-A and anti-B titres of IgM-and IgG-type were investigated. Finally, complement activation products, as well as key components of the complement system, were measured. RESULTS. In Uniplas Ò batches, all coagulation factor activities were higher than 0.7 IU/ml and all protease inhibitor activities, including protein S and plasmin inhibitor, were higher than 0.5 IU/ml. Uniplas Ò contained standardised levels of ADAMTS13 and factor H, within the normal ranges for single-donor freshfrozen plasma. There was no activation of FVII obtained during manufacturing, thrombin-antithrombin (TAT)-complex, prothrombin fragments (F1 + 2) and D-dimer levels were within the normal ranges. Anti-A and anti-B titres were within the Uniplas Ò specification, i.e. anti-A IgM and anti-B IgM\1:8 as well as anti-A IgG and anti-B IgG \1:32, respectively. Uniplas Ò did not contain an increased amount of immune complexes and the manufacturing of Uniplas Ò associated with more complement activation than the one seen for OctaplasLG Ò . CONCLUSIONS. The present study confirmed that Uniplas Ò displays the same high quality and clinical efficacy as the S/D treated blood group specific plasma OctaplasLG Ò , but with the additional advantage in being a blood group independent universally applicable plasma. Most pts received more fluids than calculated by Parkland formula (7 ± 7.5 ml/kg 9 %TBSA). Interestingly, nonsurvivors received less (3.9 ± 4.1 vs. 8.3 ± 8.2 ml/kg 9 %TBSA). Gastric decompression, ascites drainage and the implementation of a stool protocol with rectal enemas (18 interventions in 12 pts) was able to remove 2.2 ± 1.3 L of body fluids and this was related to a significant decrease in IAP and CVP and an improvement in oxygenation and urine output ( CONCLUSIONS. PRIS is a difficult condition to diagnose and routine monitoring of the adverse effects of high-dose propofol remains sub-optimal. Hypothermia has been reported to alter propofol pharmacokinetics 3 and we propose that active cooling may increase the risk of developing PRIS. This may be particularly relevant in patients with TBI who are on high doses of propofol to control ICP in addition to concomitantly administered catecholamines to maintain cerebral perfusion pressure. We recommend that further research is required in this area in view of the increasing use of induced hypothermia in ICU. OBJECTIVES. To compare differences in fluid resuscitation based on direct or indirect admissions to the London Burns Unit.

METHODS. 67 admissions to the burns unit with [20% burned surface area (%BSA) were identified over 5 years. 16 were excluded from analysis due to palliation or death within the first 8 h. 12 sets of notes were randomly selected for analysis of fluid balance in the first 24 h period of fluid resuscitation after the burn injury.

RESULTS. Mean (SD) time from burn injury to arrival at the Burns Unit was lower for patients transferred direct to the burns hospital rather than via another hospital (117.9 ± 75.27 vs. 372.5 ± 255.7 min p = 0.032). Mean (SD) error in burn size estimation was lower for patients initially treated by burns specialists versus non-burns specialists (4 ± 3.6 vs. 10.6 ± 4.3%, p = 0.016). All patients were resuscitated according to the Parkland formula calculated at one of 2, 3 or 4 ml/kg/%BSA. The mean (SD) actual fluid volume differed from the target by 30.06% (±40.43%); the lower the calculated fluid target, the greater the error between actual and planned resuscitation volumes; there was no difference in accuracy of fluid resuscitation at 24 h between patients initially managed by burns specialists versus non-burns specialists (44.14 ± 49.04 vs. 10.34 ± 7.70% respectively, p = 0.163).

CONCLUSIONS. Burned patients transferred directly to specialist Burns care receive a faster and more accurate assessment of their burn injury. Despite this, we found no difference in fluid targeting errors at 24 h, though this may reflect corrective fluid management on arrival at the specialist centre. 

Echocardiography is an useful and minimally invasive tool that allows to know the heart filling pressures, also it has proven highly accurate in predicting the response to volume in critically ill patients.

We try to determinate the response to fluid infusion by static variables as CVP or LAP, comparing with the variation of IVC.

METHODS. An observational prospective study with 18 patients undergoing coronary cardiac surgery (7 patients were excluded by a no presenting a good echo views), in the postoperative period under mechanical ventilation (Vt 8 ml/kg, FiO2 50%, PEEP 3). We performed an echocardiography if the patient presented hypotension, just before the habitual fluid load protocol were started (500 ml HES 6% in 30-45 min). We collected data before and after the infusion, and determine the responsiveness to volume if the cardiac output increased more than 10%. Data in the report included invasive CVP and LAP, and echo measures, ratio E/e', diameter and variations of inferior vena cava (IVC) and variations of stroke volume by echocardiography (ØLVOT x VTI LVOT) and with VigileoÒ system.

. The correlation between low values of CVP/LAP and volume response was poor, the relationship between CVP below 6 mmHg with increased cardiac output had a correlation (Pearson correlation -0.06) with a significance (2-tailed) 0.981, and the relationship between LAP \4 mmHg and an increase in cardiac output had a correlation (Pearson correlation 0.229) with a significance of (2-tailed) 0.362. The measurement of the variation of the inferior vena cava, led us to calculated a cutoff point more sensitive to determine which patients were responders to volume. Through the ROC curves (sensitivity/specificity), with the area under the curve of 82.5% (SE = 11.2%) and with a confidence interval of 95% (p significance of 0.002), resulted in a 12% variations of IVC with a sensibility of 100% and specificity of 42% (Younden's index of 42.9%). The same calculation, based on Kraemer's quality indices (QI) gave us a 13% of variation in IVC, with a w = 0.2 specificity rather than sensitivity (QI 0.58), and with a w = 0.8 sensitivity rather than specificity (QI 0.61) OBJECTIVES. To ascertain whether postoperative hypothermia is linked to high or low risk surgical patients.

We conducted a prospective systematic analysis looking at the incidence of postoperative hypothermia in adults who underwent general anaesthesia. Children age \18, pregnant women and patients undergoing regional anaesthesia were excluded from the survey. 

To identify the current level of doctors' knowledge on perioperative fluid management.

METHODS. The survey was conducted at George Eliot Hospital, Nuneaton, UK in May 2009. Questionnaires consisting of ten multiple-choice questions on basic sciences and clinical scenarios were devised by a consultant anaesthetist. These were personally distributed to doctors of all grades working in anaesthetics and the surgical specialties. Doctors were asked to complete the questionnaire within 5 min. Of the 66 questionnaires distributed, 52 were completed.

RESULTS. The mean questionnaire score varied between specialties from 87% in the anaesthetics department to 64% for doctors in surgical specialties. The mean score of registrars and FY1 doctors in surgical specialties was found to be 72 and 65% respectively. The overall mean score was 74%. Of all doctors surveyed, the daily maintenance water requirement was known by only 52%, 62% knew the daily maintenance sodium requirement and 75% knew that of potassium. The electrolyte contents of 0.9% sodium chloride and Hartmann's Solution was answered correctly by 65% and 63% respectively.

There is a significant deficiency in doctors' knowledge on perioperative fluid management. More emphasis on optimal perioperative fluid management is required in undergraduate and postgraduate training. Increased awareness of the British Consensus Guidelines on Intravenous Fluid Therapy for Adult Surgical Patients would aid training. Based on this survey, a regional online survey of junior doctors is planned to further identify gaps in perioperative fluid management training. Optimal fluid management could also help to reduce prolonged hospital stay which can result from fluid-related complications. OBJECTIVES. To evaluate dynamic echocardiographic parameters as predictors of volume responsiveness in surgical patients.

METHODS. 25 patients were included in the study after laparotomy surgery performed on the same day (4 breathing spontaneously and 21 mechanically ventilated in volume controlled mode with tidal volume of 10 ml/kg). A fluid challenge was performed in spontaneously breathing patients by passive leg raising and infusing saline (7 ml/kg). Echocardiographic analysis of respiratory changes of inferior vena cava diameter (DDIVC) and aortic blood flow (DABF) was performed in all patients. A threshold of 18% for DDIVC was used for classifying patients as volume responders or non-responders. Age, sex, gender, BMI, CVP, IAP, MAP, left ventricular ejection fraction, left ventricular systolic and diastolic area, and stroke volume in all patients, as well as ITBVI, CI, PPV and SVV in 16 patients were measured.

A positive correlation with DDIVC was established for ITBVI (r = 0.49, p = 0.05), IAP (r = 0.47, p = 0.01) and EF (r = 0.38, p = 0.05). A positive correlation with DABF was not established for any variable measured. 16 patients (64%) were classified as volume responders and 9 (36%) as non-responders. Responders had overall higher IAP than non-responders (13.45 ± 4.99 mmHg vs. 9.82 ± 3.33 mmHg respectively, p = 0.04).

Respiratory changes of IVC diameter showed positive correlation with ITBVI. So, conclusions about ITBVI could be indirectly made from DDIVC values in patients who are not being invasively monitored. PPV and SVV did not show positive correlation with ITBVI. Surprisingly, we confirmed a positive correlation between DDIVC and IAP. We detected 9 patients with high IAP, while all the volume responders had overall higher IAP. Although further investigations are needed to establish how longer duration of high IAP may influence DDIVC, it seems that DDIVC is a good parameter of volume responsiveness during first 24 h after laparotomy surgery. Unlike from other studies, we could not establish a positive correlation between DABF and any variable measured. These studies were performed in hypovolemic septic patients, so this could be the reason for such different results. More studies are needed in a larger set of patients undergoing laparotomy surgery to evaluate DABF. INTRODUCTION. Fluid optimization after major cardiac surgery was shown to improve patients postoperative outcome significantly. Several hemodynamic parameters were proposed for the guidance of therapy but never compared in a head to head trial.

OBJECTIVES. In this prospective randomized trial patients scheduled for elective cardiac surgery underwent early goal directed fluid therapy guided either by stroke volume variation (SVV) or by oxygen delivery index (DO2i). We hypothesized that while SVV is easier to obtain it will not be inferior to DO2i in outcome parameters.

METHODS. Following ethics committee approval and signing of a written informed consent, 84 patients were randomized in two groups to undergo either fluid optimization guided by Do2i or SVV in the first 8 postoperative hours in the ICU following elective cardiac surgery (CABG). Following a standardized EGT protocol the parameters were collected by using hemodynamic monitoring based on a pulse contour analysis and a transpulmonary lithium dilution (Lidco Plus, Lidco,UK). We compared amount and type of volume infused, need and amount of inotropic or vasopressor substances, time spent on ventilator, LOS in the ICU and postoperative complications. Statistics were evaluated by using a T Test for unpaired samples. Table 1 . Compared to the DO2i group fluid optimization using SVV showed reduced ventilator times (p = 0.043) and less complications (p = 0.004) in the first 80 days after surgery. No differences between the groups were detected concerning the type and amount of volume infused, need for inotropes or vasopressors or the LOS in hospital CONCLUSIONS. While SVV is less invasive, cheaper and easier to be obtained than DO2 outcome was at least not inferior and even showed improvements in postoperative cardiac surgery patients.

23rd ESICM Annual Congress -Barcelona, Spain -9-13 October 2010 S377

INTRODUCTION. Over the years, there have been concerns over incompatibility of transfused blood with various intravenous fluids during blood transfusion, especially related to increased levels of haemolysis. It is often impractical, particularly in an emergency situation, to flush through a giving set with a so-called ''safe'' fluid prior to and after delivering blood. We wanted to investigate whether this is actually necessary and whether the usual fluids used in the perioperative period really do cause any demonstrable alteration in the composition of transfused blood.

OBJECTIVES. The purpose of this study was to expose packed red cells to a variety of different intravenous fluids commonly used during the perioperative period and to measure a number of parameters in the blood following their contact with each different fluid, including a blood film to examine for clumping of cells or haemolysis.

A unit of A positive blood was passed through blood giving sets which were primed with various intravenous fluids. After adequate mixing of blood with fluids, samples were collected for full blood count, urea and electrolytes and blood films. One millilitre of mixed blood was taken in each bottle at a time. The intravenous fluids used in this study were Normal saline, Hartmann's solution, 5% Dextrose, 10% Dextrose, starch and gelatin.

There was no significant rise in blood parameters suggestive of haemolysis. The potassium and LDH levels were not significantly different with various fluids. The haemoglobin and haematocrit levels were also comparable to one another.

There was no demonstrable changes in blood parameters suggestive of haemolysis, nor were there any change in electrolyte values. This suggests that all of the fluids investigated during this study would be suitable to be used via the same giving set before and after the transfusion of pack red cells. OBJECTIVES. To assess the compliance with the national guidelines in avoiding inadvertent peri-operative hypothermia in an acute district general hospital in England.

We prospectively studied our local practice on maintaining normothermia in 165 consecutive adult surgical patients {61 men, mean age 57.4 years, 112 patients with ASA grade 2 (67.7%), 60 emergency surgical patients (36.4%), 49 patients with significant cardiac disease 29.7%}. We used a questionnaire that was filled pre-operatively by anesthetic nurses, intra-operatively by anesthesiologists, and post-operatively by recovery nurses. Patients were recruited from the following surgical subspecialties: General surgery (34%), Gynecology (24%), Trauma (22%), Breast surgery (9%) and Orthopedics (7%). Day surgery patients were excluded. Peri-operative hypothermia was defined as temperature \36°C as per the NICE guidelines.

RESULTS. Less than half of our patients (47.9%, n = 79) had their temperature measured preoperatively, on whom incidence of hypothermia was 25.3% (n = 20). Only one of these patients was warmed prior to induction. Patients requiring emergency surgery and those with ASA Grade 2 had increased incidence of preoperative hypothermia (30.8% and 28.6% respectively, p \ 0.05). Based on NICE guidelines, 164 patients needed intraoperative forced air warming but only 64 (38.8%) patients received it. Intraoperative temperature measurement was made on 111 patients, of whom 28.8% (n = 32) were hypothermic. Incidence of intraoperative hypothermia was high in surgical procedures lasting longer than 30 min (p \ 0.05) but was not affected by the use of regional anesthetic techniques. 161 patients had their temperature measured on arrival to recovery of whom 46 (26.6%) were hypothermic. 38 patients (23.6%) had their temperature measured every 15 min (NICE recommendation) and the mean time interval for temperature measurement in recovery was 26 min. 10 patients were still hypothermic on leaving recovery.

CONCLUSIONS. Majority of our surgical patients did not receive adequate perioperative care on maintaining normothermia. Consequently, the incidence of hypothermia was significant pre-, intra-and post-operatively. We are currently analyzing the data to investigate the effect of hypothermia on duration of recovery stay, length of hospitalization and mortality in our patients. 

We completed a double-blind randomized trial in patients undergoing cardiac surgery in which we compared fluid resuscitation with a hydroxyethyl starch (HES, 10% 250 MW pentastarch) and saline. Use of HES resulted in markedly less use of catecholamines the morning after surgery. An underlying design principle was that assessment of cardiac index (CI) is essential for a proper fluid protocol. In this analysis we examine that supposition.

All subjects had pulmonary artery catheters. Patients were consented preoperatively, but randomized post operatively to receive up to 4 blinded 250 ml boluses for predefined hemodynamic targets; CI \2.2 L/min/m 2 , blood pressure (BP) set by admitting team, CVP \3 mmHg, or urine output \20 mL/h. Hemodynamic measurements were made before and after each bolus. After the 4 study boluses, only saline was used.

RESULTS. 237 patients received fluids, 119 HES and 118 saline. There were 727 study boluses, 348 HES and 379 saline. Of these, 99 boluses (14%) could not be assessed for this hemodynamic analysis (but were still used for the primary outcome) because of protocol violation or missing data. Of the rest, 235 (37%) of boluses were given for a low CI; in 33 BP and 11 CVP were also low so that CI was the only trigger in 30%. A low BP was a trigger in 290 (46%). Low CVP was the trigger in 95 (15%). Only 50 HES and 65 saline patients required the maximum allowed 4 blinded boluses. At the 4th bolus, low CI was the trigger for 8 (16%) of HES but 20 (31%) of saline patients. There were 511 that could be evaluated for hemodynamic response based on four possible outcomes of CVP and CI. OBJECTIVES. The aim of our study was to evaluate the predictive value of CVP with regard to GEDI, and to correlate these parameters to cardiac index (CI). CONCLUSIONS. Volume depletion according to GEDI was found in more than half the patients. The predictive values of CVP with regard to volume depletion were low GEDI and its changes significantly correlated to CI and its changes, which was not observed for CVP.

Therefore, GEDI appears to be more appropriate for volume management during mayor liver resections. INTRODUCTION. Regional anticoagulation with citrate is an effective and established anticoagulation strategy during CRRT in critically ill patients, especially in surgical patients with a high risk of bleeding and in case of a heparin-induced thrombocytopenia (1). However, citrate CRRT could be associated with major metabolic derangements such as metabolic alkalosis, hypocalcemia, hypernatremia and citrate toxicity.

OBJECTIVES. The aim of our study was to investigate efficacy, safety and metabolic stability during citrate CRRT in critically ill patients with acute kidney injury.

METHODS. The retrospective study was performed in a mixed surgical and trauma ICU in a university hospital. Patient charts were reviewed for demographic data, the period and dosage of citrate CRRT and metabolic parameters. Reasons of admission, comorbidities and severity of illness were also evaluated. Citrate CRRT was performed using commercially available equipment and fluid solutions (Multifiltrate Ò with integrated Ci-Ca Ò -system; Fresenius Medical Care; Germany). To maintain stable metabolic and haemodynamic conditions we used an internal standard protocol for citrate CRRT. Statistical analysis was performed using descriptive methods (mean, median and standard deviation) and a Mann-Whitney U test where appropriate. P \ 0.05 was regarded as statistically significant. CONCLUSIONS. Although minor metabolic imbalances were observed, none led to a termination of citrate CRRT and all of them could be managed by adjustments of blood flow and dialysate rates according to a preset protocol. Our findings suggest citrate CRRT to be a safe and effective strategy for CRRT even in patients with hepatic dysfunction. Nevertheless, metabolic parameters need to be monitored regularly to avoid severe metabolic derangements. INTRODUCTION. The liver is central to ammonia metabolism, being the main site of urea cycle enzyme pathways. In acute liver failure (ALF) and decompensated chronic liver disease (CLD) ammonia dysmetabolism results in hyperammonaemia, thought to be of central importance in the pathogenesis of hepatic encephalopathy and, in ALF cerebral oedema [1] . Continuous renal replacement therapy (CRRT), commonly used in critically ill patients may be an effective method of clearing ammonia. Little is known of the efficacy such techniques have on ammonia clearance. OBJECTIVES. To quantify the clearance of ammonia using an Aquarius haemofilter (AHF) using different renal replacement doses and techniques.

METHODS. Patients with a circulating ammonia level[100 lmol/l due to commence CRRT were enrolled. The AHF was programmed to run in either pre-or post-dilution modes at a blood flow rate of 200 ms/min using a 1.2 or 1.9 m 2 filter depending on the CRRT ultrafiltration (UF) dose, which included 35, 60 or 90 ml/kg/h (adjusted for ideal body weight). 2 ml of blood and effluent fluid were collected, on ice into lithium/heparin and serum separation tubes, from pre and post filter access points and effluent tubing to calculate urea and ammonia clearance using the Cordoba formula [2] . Delta whole body ammonia clearance was determined by measuring arterial ammonia at 0 and 60 min. Ammonia measurements were performed using a Pocketchem Ò Blood Ammonia bedside testing machine.

RESULTS. 20 patients (9 ALF and 11 CLD) were recruited (mean age 45 years, SD (14), with mean arterial ammonia 123 lmol/l, SD (49). 60 min whole-body ammonia clearance was -12 lmol/l, p = 0.016, Paired t test). Ammonia and urea clearance were correlated (R = 0.748, p = 0.020); UF rate correlated negatively with filtrate ammonia (R = -0.496, p = 0.007) and positively with ammonia clearance (R = 0.515, p = 0.005). Filter ammonia clearance was not dependent on filter size for the standard blood flow rate. Pre or post dilution modes did not affect ammonia clearance (p = 0.100, Student's t test). A constant filter size and blood flow rate achieved ammonia clearance of 39 ml/min/m 2 for 35 ml/kg/h, 60 ml/min/ m 2 for 60 ml/kg/h and 52 ml/min/m 2 for 90 ml/kg/h (p = 0.008, One way ANOVA).

CONCLUSIONS. 60 ml/kg/h based on ideal body weight appears to be the optimum dose of CRRT for ammonia clearance when using a blood flow rate 200 ml/min and a 1.2 m 2 filter. Filter and delta whole body ammonia clearance may be increased further using the combination of a higher dose (90 ml/kg/h) with a larger filter size and higher blood flow rates. INTRODUCTION. Malnutrition is common in intensive care following the catabolic state induced by critical illness. Patients who progress from enteral nutrition back to oral feeding are usually in an energy deficit. ESPEN guidelines recommend increasing calorie delivery during the recovery period to cover this anabolic phase. Oral nutritional supplements (ONS) are widely used to facilitate calorie delivery within the hospital setting however the effectiveness of this strategy is dependent on patient compliance with the products. Compliance among the elderly ward-based population has been considered (1) however that of intensive care patients has not been reported.

To evaluate compliance to ONS in a mixed medical and surgical adult intensive care unit (ICU) in a District General Hospital.

Prospective observational study was conducted over a 2 month period with data compiled from fluid chart analysis and discussions with nursing staff. All adult ICU patients prescribed, or offered without prescription, an ONS were included until the point they were discharged to the ward. The supplements studied, Resource Ò Energy, 2.0 Fibre, Fruit and Dessert (Nestlé Nutrition), were selected based on their availability within the trust. Patients were offered a choice of flavour.

RESULTS. Data was collected and analysed for 51 patient days. A total of 122 supplements were prescribed. Of the prescribed supplements, 42.6% were offered to patients and 26.6% consumed. 20% were offered the same at nursing discretion based on clinical need and 39.8% were consumed. Resource Ò Energy was the most frequently prescribed and offered product (76.2 and 59.2% respectively). Most common flavours selected by patients were strawberry and vanilla. Resource 2.0 fibre was better tolerated (91.7%) than Resource Energy, Resource Fruit and Resource dessert (73.9, 70.8 and 50.0% respectively). Across all products the best tolerated flavours were apricot, chocolate and coffee (100%). The highest calorie supplement, Resource Ò 2.0 fibre, resulted in the best compliance in both tested flavours.

Compliance with ONS demonstrated here is higher than previous studies (1) partly attributable to one-on-one nursing of ICU patients enabling active encouragement with feeding. Nursing staff discretion had better uptake than routine prescription of ONS. However, difficulties with ONS still remain. Interestingly in our study the highest calorie density supplement was tolerated the best and thus giving the most benefit to the patient. Despite the difficulties associated with ONS uptake we would recommend its regular use on ICU with a drive towards the highest calorie supplements being offered. INTRODUCTION. Cirrhosis is a chronic disease and the patient's quality of life is affected in a negative way due to the problems like ascites, jaundice, nutrition deficiency, fatigue, activity intolerance, itching, pain, insomnia, anxiety, hopelessness, work loss and depression.

OBJECTIVES. The aim of this study is to examine the changes in patient's lives that diagnosed with cirrhosis of the liver disease owing to the symptoms they experienced.

METHODS. This research is a qualitative study that has been carried out with 13 inpatients diagnosed with liver cirrhosis in the gastroenterology clinic of a teaching and research hospital. Average age of patients was 54 (ranging 39-70). Descriptive characteristics form and semi-structured interview form were used in the data collection. Interviews with patients have been performed individually and face to face. The data were evaluated by using Colaizzi's phenomenological data analysis method.

As a result of the data analysis, three categories and six themes were identified. Categories include: (i) problems of symptoms related to the physical limitations (ii) psychosocial issues. Patients suffer mostly from fatigue and malaise (12 Patients), while those in the later stages suffer from, additionally, physical ailments caused by acid. Inability to sleep due to anxiety and increase in tendency to sleep in advanced stages have been identified after being diagnosed. The majority of patients were identified to have undergone an anxiety besides having a fatal disease due to concern for the future, being forced to quit the job and being affected by the experiences of the patients in advanced stages. It also has been discovered that the patients had experienced social isolation because of fatigue and weakness in particular.

As the result of this study it has been determined that patients with cirrhosis have mainly problems of fatigue, weakness, sleep disorders, anxiety and associated problems.

23rd ESICM Annual Congress -Barcelona, Spain -9-13 October 2010 S379 [1] . While in patients with acute liver failure, elevation of arterial ammonia levels has been linked to cerebral complications and increased mortality, the role of arterial ammonia in HH patients is unknown.

Our study aims at evaluating arterial ammonia levels in patients with HH. Furthermore, we wanted to elucidate the potential consequences of high ammonia levels in these patients.

Arterial ammonia levels were measured and documented in 72 HH patients without liver cirrhosis who were admitted to the medical ICU. ICU mortality and overall 28day-survival were documented. COX regression was performed to describe the impact of ammonia levels on mortality. Mann-Whitney Test was used for comparison of metric variables.

RESULTS. Overall median arterial peak ammonia level in our patients was 65 lmol/L (41.9-100.6 lmol/L), whereas median arterial peak ammonia value was significantly higher in ICU non-survivors compared to survivors (77 (53-119.2) vs. 52.9 (36.4-71.1); P \ 0.01). SAPS II and SOFA score were significantly higher in ICU non-survivors (P \ 0.01 and P \ 0.05, respectively). COX regression revealed that arterial peak ammonia levels were significantly associated with higher 28-day-mortality (P \ 0.01), even after adjustment for SAPS II.

Median arterial peak ammonia levels in patients with verified brain edema were significantly higher than in patients without (130.7 lmol/L (92-349.7 lmol/L) vs. 70.6 lmol/L (43.2-81.7 lmol/L); P \ 0.05) after exclusion of patients following cardiopulmonary resuscitation with consecutive hypoxic brain damage.

Our results suggest that increased levels of ammonia are associated with high mortality and can lead to brain edema in patients with HH. 53% of patients had a diagnosis of sepsis and 21% of patients were admitted under the neurosurgical team, the latter of which may have contributed to the relatively low anticoagulant use of 54%. Systemic heparinisation was the sole anticoagulant used, but compliance with local protocols was poor with 48% of APPTs below the therapeutic range and 46% of infusions commenced at the wrong rate. 68% of filter changes were due to clotting and mean filter life was 32 h. Despite this, dose delivery was acceptable, with 84% of prescribed dose delivered. CONCLUSIONS. As previously reported 1 , our demographic data confirm the relatively poor outcome of patients needing CRRT. We have identified areas where care for these patients could be optimised and endeavour to do this locally via improved protocol design and an ongoing educational programme. Many of the components of CRRT could be incorporated into care bundles, but certain aspects of treatment remain controversial 3 which may be a barrier to their adoption. Given the high numbers of neurosurgical patients in our unit, consideration should be given to the use of regional anticoagulation such as citrate.

INTRODUCTION AND OBJECTIVES. Accurate prognostic indicators of patient survival in an intensive care unit (ICU) help guide clinical decision making. Factors known to portend poor prognosis in acutely ill cirrhotics in ICU include the need for mechanical ventilation, development of shock, renal failure and sequential increase in the number of failing organs. While serum lactate is now an established marker of survival and/or the need for transplantation in fulminant liver failure, its impact on critically ill cirrhotics is less well known. METHODS. We retrospectively studied 133 consecutive acutely ill cirrhotics admitted to the ICU between 2005 and 08 at the Royal Free Hospital, a tertiary referral centre in liver diseases and transplantation. Data were collected on demographic variables, aetiology of liver disease, liverspecific prognostic scores [Child-Turcotte-Pugh (CTP), Model for end-stage liver disease (MELD), United Kingdom model for end-stage liver disease (UKELD)], and acute illness scores [acute physiological score and chronic health evaluation (APACHE II), sequential organ failure assessment score (SOFA) ]. In addition, serum lactate levels at 0, 24 and 48 h were also recorded. Multivariable logistic regression analysis was performed, and the discrimination ability of each of the above-mentioned scoring models in predicting ICU and hospital survival of these patients was evaluated using the area under the receiver operating characteristic (ROC) curve. 

CONCLUSIONS. One third of LT recipients present a documented bacterial infection within 1 year after surgery. We found a high prevalence of ciprofloxacin resistance and a low incidence of S.aureus witch was often resistant to methicillin. Non fermentative gram negative bacilli represent 18% of the pathogens and should be taken in account for treatment of the most severe patients. 

Extracorporeal liver support therapy is in its infancy but is valued as a detoxification treatment option for patients with cirrhosis who have rapid worsening of their liver function. We report the use of Prometheus Ò , a new extracorporeal liver support system allowing the removal of protein bound and water soluble toxins by fractionated plasma separation and absorption (FPSA) in a patient with Wilson's disease (WD) who developed rapid worsening of their liver function.

A 26-year-old female patient, diagnosed with WD since the age of 17, was initially treated in an irregular pattern with penicillamine. Therapy was discontinued. Now, 2 years later, she developed acute decompensated liver failure with hepatic encephalopathy with a MELD 29. Liver transplantation (LT) was the treatment option for this patient. But, in this case, the rapid and adverse evolution of the liver failure with renal failure and the unknown waiting time for a emergency liver donor in our country led us to use the Extracorporeal liver support therapy.

After 5 h 30 min of therapy we reduced the amount of bilirrubin for less than a half, we increase the urinary output and next day the patient went to liver transplant, stable, with a renal function improved.

CONCLUSIONS. Acute liver failure due to WD is most of the time fatal without emergency LT. This case report highlights discontinuation of chelants treatment in a patient with WD. As the patient progressed to decompensated liver cirrhosis with encephalopathy, LT was the only treatment option but while we don't get a donnor, we can use, for a short period of time, an Extracorporeal liver support therapy as a very useful bridge. Results from two studies presented at the recent EASL 2010 Congress have shown that treatment with extracorporeal devices may not confer a survival advantage for severe liver failure patients, despite positive dialysis effects. However, results among a small sub-group of patients show promise like severely ill patients with hepatorenal syndrome type 1 or a MELD score over 30. (1) . Metoclopramide is used to stimulate the upper GI tract and seems to have no effect on colonic motility.

OBJECTIVES. The aim of this in vitro study was to compare the prokinetic potency of those substances.

A tissue bath with guinea pig colonic segments fixed on a polyacrylic tray allows the evaluation of the transit time (TT), the time necessary for a wooden pellet to perambulate. A decrease of the TT reflects stimulation, and an increase inhibition of peristalsis.

After stable peristalsis activity the effect of increasing concentrations of prucalopride, neostigmine or metoclopramide on TT were evaluated. Dose response curves were constructed, two way ANOVA (Sigma Stat) was used for statistics, p values B 0.05 were considered to be significant.

Effect of prucalopride and neostigmine on motility RESULTS. Prucalopride stimulates normal peristalsis in vitro only in the highest tested concentration of 3 lM (p \ 0.05). Neostigmine's prokinetic effect was limited to a small concentrations range (0.1 lM, p \ 0.05), the concentration of 0.03 lM had a moderate, but not statistically significant prokinetic effect and the highest tested concentration (1 lM) lead to a complete block of peristalsis (Fig. 1) . Metoclopramide, as expected, was devoid of any effect on colonic motility.

CONCLUSIONS. This experimental setting is a reliable method to evaluate the effect of different substances on colonic motility in vitro. Prucalopride's prokinetic activity is concentrations dependent and limited. Neostigmine is well known to improve colonic motility, but it seems imperative that the drug's effective dose range be use-higher concentrations have inhibitory effect on peristalsis. OBJECTIVES. Robotic radical prostatectomy involves extreme changes in patient position and often associated with a longer operative time than other commonly performed laparoscopic procedures. This review discusses the anesthetic considerations in robotic radical prostatectomy while analyzing potential risk factors related to pulmonary complications.

We retrospectively reviewed the medical records of all the patients who had undergone robotic radical prostatectomy at our institution. Among the total patients of 80, aged 49 to 82 years, 58 patients were capable of spontaneous respiration at the end of surgery (Group I) whereas 22 patients needed assist ventilation (Group II). The demographic characteristics, coexisting diseases, anesthesia and operation time, anesthetic agents, the amounts of blood loss, infused fluid and transfused blood products were compared between the groups.

RESULTS. The mean age of the patients was 67.2 ± 7.3 years. The mean operation times were 445.3 ± 210.3 min (range, 235-1250 min). Age, body mass index (BMI) and ASA status did not differ significantly between the two groups, whereas operation time, the amount of blood loss and the incidence of transfusion were significantly higher in the group II. Although 3 patients with subcutaneous emphysema and atelectasis needed prolonged ventilator care for 48 h, the incidence of atelectasis and subcutaneous emphysema was similar between the groups.

CONCLUSIONS. Prolonged laparoscopic surgery in a steep Trendelenburg position has a high possibility of postoperative respiratory insufficiency and the possible contributing factor is a long operation time. OBJECTIVES. We examined the frequency of postoperative cough reflex and its effect on postoperative clinical outcome retrospectively.

We examined the patients who admitted into the ICU after the esophagectomy with lymphadenectomy during the period from September, 1999, to February, 2010. In addition to usual criteria for extubation we removed their tracheal tube if the cough reflex was identified when one milliliter of half saline was distilled into their trachea. If the cough reflex was absent until 7 days after the operation the patient underwent tracheostomy and after that they weaned from the ventilator.

RESULTS. There were 125 patients (F/M 24/101), and their mean age was 58.5 ± 7.8. Cough reflex were confirmed by seventh postoperative day in 95 patients (76%) but residual 30 patients underwent tracheostomy because of absence of cough reflex ( INTRODUCTION. The technique of laparoscopic cholecystectomy carried with carbon dioxide pneumoperitoneum may lead to adverse events in mechanical, hemodynamic and respiratory systems as a consequence of physiopathological changes such as increased intraabdominal pressure. _ It may cause hypoxemia, hypercapnia, hemodynamic instability and impairment of oxygenation. Decreased functional residual capacity, ventilation/perfusion imbalance and sympathetic stimulation effects of CO2 that is absorbed from peritoneum are basic problems. In perioperative period, application of mechanical ventilation and anesthesia should be reviewed because of these physiopathological mechanisms.

In this study, we aimed to investigate the effects of 5 cmH 2 O PEEP application on ETCO2, minute ventilation and arterial oxygenation during laparoscopic cholecystectomy operations. For this reason, the study included total 40 patients and they were randomly divided into two groups. Same anesthetic protocol was applied in both groups. For general anesthesia induction; 1 mg/kg dose of fentanyl, 2 mg/kg dose of propofol were administered. Following this procedure endotracheal intubation was applied with 0.15 mg/kg dose of cisatracurium. Patients received %50 O2-%50 N2O (mixture with equal amounts) with 0.8-1.2 MAC end-tidal sevoflurane for anesthesia maintenance. Before CO 2 insufflation, respiratory parameters were recorded on the respiratory apparatus adjusting ETCO2 32-36 mmHg, respiration rate 12/min., inspiration/expiration rate 1:2, Vt: 8-10 ml/kg. Patients were ventilated by volume controlled mechanical ventilation. Heart beats, mean arterial blood pressure and peripheric O2 saturation (spO2), ETCO2, minute ventilation(V) and peak airway pressure(P _ IP) values of all patients were recorded just before insufflation (T0). After recording, 5 cmH 2 O PEEP was applied to the first group (Group 1). PEEP wasn't applied to the 2nd group (Group 2). These parameters were repeated in 5 periods such as 5 (T1) and 30 (T2) minutes after insufflation, preexsufflation (T3) and postexsufflation (T4) in both groups. Before insufflation, respiration rate (12/min) and ETCO2 (32-36 mmHg) values were adjusted as planned in both groups and minute ventilation was also adjusted. At the same time, total insufflated amount of CO2 for distending abdomen was recorded. Arterial blood gas analyses were made just before induction (while patients were breathing normal room air, T0), 30 min after induction (T2) and just before the end of the operation (T4).

In our study, we found that minute ventilation to stabilize ETCO2 32-36 mmHg was significantly increased in group 2 in which PEEP was not applied (p \ 0.01). None enhancement was needed in minute ventilation in Group 1 and arterial oxygenation was significantly increased in Group 1 (p \ 0.01). Aside from the cholesterol lowering effects of statins, as a class of drugs they have been shown to exert anti-inflammatory effects and have the potential to be therapeutic in neuroinflammatory disorders 3 .

We tested the hypothesis that atorvastatin improves memory retrieval post unilateral nephrectomy in a murine model.

METHODS. C57/BL6 mice were randomly allocated into 4 groups (n = 8-10/group): control plus placebo, control plus atorvastatin, nephrectomy plus placebo and nephrectomy plus atorvastatin. Animals were given either a placebo (0.4 ml normal saline) or 250 lg in 0.4 ml normal saline of atorvastatin by gavage once a day for 5 days. On day 4 all animals underwent fear conditioning training using a conditional stimulus of a 70 dB tone and an unconditional stimulus of a 0.70 mA electric shock. On day 5 the surgical animals underwent unilateral nephrectomy, whilst the control animals received no surgery. At post-surgical day 3 all animals were tested for hippocampal dependent memory retrieval using the fear conditioning paradigm, with freezing response to the 70 dB tone as a marker of memory retrieval. All animals were then terminated.

RESULTS. Surgery evoked a reduction in hippocampal dependent memory retrieval in the nephrectomy plus placebo group as measured by % freezing time (mean ± SD: 40 ± 16) when compared to the control plus placebo group (70 ± 19; p \ 0.01); a situation mimicking POCD. This change was obviated in the nephrectomy plus atorvastatin group (59 ± 16; p [ 0.05 vs. control plus placebo).

CONCLUSIONS. Our data suggested that atorvastatin has the potential to improve postoperative cognitive performance in a murine model of POCD. The proven safety of the drug along with its already widespread use and cost effectiveness would permit rapid instigation of a human randomized controlled trial to explore efficacy in the clinical setting. 

A. Puxty 1 , R. Docking 1 1 Glasgow Royal Infirmary, Department of Anasethetics, Glasgow, UK

Hypotension in the post-operative period is common but guidelines recommend its prevention/treatment [1] . Epidurals are common practice following major surgery in many institutions and can prevent pulmonary complications [2] but have also been associated with falls in blood pressure when compared to other analgesic techniques [3] . Fluids therapy is a common intervention for hypotension but fluid overload has been associated with worse outcomes in surgical patients [4] . We decided to audit the incidence and management of hypotension in the surgical high dependency unit of a large tertiary referral hospital.

To determine the incidence and management of hypotension in the surgical high dependency unit in pancreatic, upper GI and lower GI patients.

We prospectively looked at 48 patients who underwent major upper GI, lower GI or pancreatic surgery involving epidural analgesia. The first 24 h of care from onset of anaesthesia was closely looked at with regards to fluid management, epidural management and actions taken on episodes of hypotension or severe hypotension (defined as systolic blood pressure of \90 and \80 respectively). Each episode of hypotension was looked at to determine the actions taken at that point.

Of the 48 patients looked at, 17 were major pancreatic, 17 lower GI and 14 upper GI patients. 40 (83%) had at least one episode of hypotension, with 11 (23%) having at least one episode of severe hypotension. Mean Fluid in during the first 24 h was 6523 ml, with a mean fluid balance of 4133 ml. There was no difference between the doses of epidural local anaesthetic in 24 h between the hypotensive and non hypotensive groups (p = 0.275). Management of hypotensive episodes was variable, but the most common intervention at episode one was fluid bolus (80%) and discontinuation of epidural was most common at episode two (39%). Use of vasopressors for hypotension was very low with only two infusions being started altogether.

CONCLUSIONS. Hypotension is very common in our high dependency unit. Fluid balance in our patients was far more positive that we had expected. Management of hypotension was variable. We plan to institute a protocol for hypotension and fluid administration to determine if improvements can be made. OBJECTIVES. To identify predictive factors associated with the need for relaparotomy in patients with SSP.

Adult SSP patients undergoing laparotomy between 2004 and 2009 included within a single-center peritonitis registry (PERIT) were collected. Patients subjected to relaparotomy were studied. We excluded patients with severe peritonitis secondary to appendicitis. APACHE II and SOFA score at ICU admission after the initial laparotomy were recorded. Variables with a p value.1 in a bivariate analysis were included in a multivariate logistic regression for further analysis of predictors for need for re-laparotomy.

RESULTS. Two-hundred forty-seven patients were obtained from PERIT registry. A total of 212 patients with SPP were included in the analysis. Eighty seven patients (41%) required relaparotomy. Median number of re-laparotomies was 3. Most SPP were associated to colon (n = 94, 44.3%), small intestine (n = 83, 39.2%) and biliary tract (n = 33, 15.6%) perforations. Cultures were positive in 74.5% of first laparotomy: gram negative bacteria were isolated in 53.3%, gram positive bacteria in 16.5% and fungi in 4.7%. Hospital mortality was 17% (n = 36). Multivariate analysis is described in the Table 1 . CONCLUSIONS. In obese patients scheduled for surgery, the previous use of CPAP has not shown an improvement in blood gas parameters. The use of CPAP in the hours before and immediately after surgery has not been associated with better postoperative oxygenation. Combined ICU-Surgery Dpt. action in these cases seem to contribute to better patient outcomes. OBJECTIVES. We set out to quantify the intensive care workload and changes to that workload over the first 4 years following the transfer of a Specialist Bariatric service to our hospital.

A prospectively collected bariatric surgical database was cross-referenced to the ITU database (Ward-Watcher) to identify admissions to the 10-bedded Critical Care Unit of all patients who had undergone any bariatric procedure. For each patient identified; demographics, reason for admission, level of support, length of stay and outcome were recorded. Data were grouped into 12-month periods for trend analysis. 

Research in emergency situations and especially in resuscitation field raises important ethical and regulatory issues. The globalization of the resuscitation science through multicentric trials for example highlights the need for a more consistent approach to regulatory aspects to enable the science to grow while protecting human rights.

OBJECTIVES. The purpose of this analysis is to compare the different regulations approaches in emergency research in North America (Canada, USA) and in Europe (European Directive, France). CONCLUSIONS. This analysis emphasizes the lack of international standardization of regulatory measures and ethical decisions. However some countries like the US seem to advance in the democratic process by mandating additional regulatory measures (community consultation, public disclosure to the communities) prior to initiation of clinical investigation; nonetheless, there is little evidence of their effectiveness. Many challenges are raised. Firstly, the variability in regulations, and consequently in local board's assessments, is problematic, pleading for international regulations. Secondly, the current heterogeneous ethical review process and demanding unsubstantiated regulatory measures poses a risk to all when it is not evidence based and it is applied inconsistently between countries, within a country and worse at the level of each individual hospital review board. It puts the investigator at risk for unnecessary criticism and the community at risk as it is unknown if we truly consult or inform our target communities about waiver of consent research through our current ethical and regulatory processes. Globalization and evaluation of the ethical and regulatory processes are urgently needed; regulatory community has to work towards a standardized evidence-based process upon which to base regulatory decisions. INTRODUCTION. In research outside the intensive care field it is known that a high score for the psychological factor ''perceived hopelessness'' experienced by healthy individuals increases risk of death several fold.

OBJECTIVES. The aim of this study was to examine if the score of the psychological factor ''perceived hopelessness'' may predict long term mortality (mean or high perceived hopelessness score) when assessed post ICU care in former ICU patients.

METHODS. Prospective, multicenter study in three mixed ICU's in Sweden. Questionnaires, including the 2-item hopelessness scale, demographic data and previous illnesses, were sent 6 months after discharge to all former adult ICU patients who thereafter were followed for another 3 years. A reference group of 6093 individuals from the uptake area of the hospitals served as controls.

RESULTS. 980 (59%) patients returned the questionnaires. The ICU patients reported significantly higher mean scores in perceived hopelessness score compared with the general population, 2.5 (SD 2.2) compared with 2.1 (SD 2.1) (p \ 0.001), and 40% (n = 392) of the ICU patients perceived a mean or high hopelessness score compared with 36% of the general population (p \ 0.001).

The ICU patients who died during the follow-up period reported a significantly higher perceived hopelessness score (n = 123) 3.7 (SD 2.5) (p \ 0.001) as compared with those who survived up to 3 years after discharge (n = 857) 2.4 (SD 2.1). In a logistic regression model the long term mortality for the ICU group was found to be affected by: pre-existing disease [odds ratio (OR):2.2], age (OR: 1.04) and perceived hopelessness score (OR: 1.18).

The new and interesting finding of this study is that ICU patients score higher on ''perceived hopelessness'' than a control population and this increase is predictive for the post ICU mortality. Furthermore, the size of this effect is significant and only exceeded by pre-existing disease and age. We performed a retrospective observational study to evaluate what proportion of MET calls was associated with LOMT issues.

To estimate the proportion of MET reviews involving patients with a not-forresuscitation (NFR) order and the timing of MET calls in relation to admission and death or discharge from hospital. To compare the patient characteristics and outcome for MET calls associated or not associated with LOMT issues.

We obtained Hospital Research Ethics Committee approval. We performed a retrospective observational study involving five-year (August 2005-April 2010) in a single tertiary Australian hospital. We obtained information on demographics, on the MET review and hospital outcome. LOMT included NFR orders, not for MET orders and palliative care plans. RESULTS. We analysed 4829 MET reviews in 3629 patients. Table 1 and Fig. 1 summarize major findings for overall population and the two subgroups of patients with or without LOMT. Patients with LOMT care plan were older, more likely to have medical diagnoses, were reviewed later during their hospital stay and closer to their hospital discharge or death. Fewer LOMT patients were admitted to ICU. Hospital length of stay was shorter, mortality in LOMT care patients was double that of non-LOMT patients. However, more 50% of patients with LOMT were discharged alive from the hospital.

CONCLUSIONS. More than one third of MET activations deal with LOMT issues. Although the mortality of these patients is high, a large proportion survives to hospital discharge.

Evaluation of the patient experience in Intensive care (ICU) frequently depends on reports from surrogates such as relatives. There is a concern regarding the validity of the surrogate opinion which might not represent the values of the incapacitated patient and treatment decisions therefore maybe biased [1] . Others have found that there is a strong preference within a population for utilizing relatives as surrogate decision-makers in the event of admission to ICU and this attitude is not influenced by ethnicity, religion or education level [2] .

OBJECTIVES. The objective was to measure the ability of the relative to answer on behalf of the patient. A further wish was to determine the validity of their surrogate responses.

A retrospective study, which surveyed relatives of patients who had died within a critical care service during a 2-year period (2005, 2006) . The 31 item questionnaire allowed for the collection of quantitative and qualitative data with respect for each item to overcome the limitations of the quantitative format which may not be sensitive to all the issues which can surround the provision of end-of-life care [3] . For 3 items, relatives were asked specifically to grade their capacity to represent the patient.

RESULTS. Quantitative data from the 3 items designed to test the relatives' perception of their ability to act as surrogates indicates that relatives considered they could respond to these items for 46% (average) of instances. When the relative did answer on the patient's behalf, the level of concordance between the surrogate (relative) and the patient's perceived opinion was 60% suggesting that when the relative is willing to act as surrogate the response is likely to have validity. (Table 1) . 

Results from the qualitative data indicates that the low (46%) level of willingness to answer these questionnaire items reflected a reluctance to answer on behalf of a sedated or ventilated patient, rather than an inherent inability to represent the patient.

CONCLUSIONS. The response rate to the 3 items vindicates concerns regarding the ability of relatives to represent the patient in ICU settings and supports a need for further study.

Where the relative is willing to act as surrogate, concordance does exist. Qualitative data clarified quantitative results and was instrumental in promoting a better understanding of the concerns of relatives who have a family member admitted to ICU. . The majority of patients that died in ICU were provided some kind of therapy restriction. An important conflict strains between clinical practise, bioethical principle and jurisdiction laws; the solution of this conflict is more and more urgent. Therapy restriction has also important economical aspects since the number and cost of available treatments constantly increase. Our survey studied therapy restriction procedures in Hungary for the first time.

In 2007 we performed a survey with questionnaire among intensive care physicians. Questionnaires were sent out electronically to 743 registered members of the Hungarian Society of Anaesthesiology and Intensive Care. Respecting anonymity we have statistically evaluated 103 replies (14%) with t test and ANOVA. We grouped intensive care physicians based on gender, years spent in work, religion and type of department they were working, and we compared data from these groups.

Intensive care physicians generally make their decisions alone, based on the patient's long-term life prospects and physical status (3.75/5 points). They are slightly influenced by the opinion of the patient (2.57), the relatives (2.14) and other medical personnel (2.37). If the physician sees any chance of recovery but the patient or relative requests treatment restriction then 67.3% of physicians that completed the forms would continue therapy against the will of the patient or relative. Only 27.7% would accept the patient's/relative's opinion and autonomy in such a case and would stop therapy. In fact 28.3% of physicians would make their decisions without considering or even against the opinion of patient if they think therapy is useless. If there is no chance of recovery despite medical treatment 3% of physicians stop the treatment, 8.1% would continue it without informing the patient or the relatives, 17.2% informs the relatives but continues useless treatment irrespective of the will of the patient or relative. Having analyzed the groups we found two significant differences. In case of useless treatment physicians working in university hospitals more often choose treatment restriction without informing relatives (p \ 0.001) then those working in non-university hospitals. Physicians who declare themselves as atheist rather choose the continuation of treatment without informing relatives (p = 0.007).

CONCLUSIONS. The Hungarian practise of end of life decisions among intensive care specialists is paternalistic, physicians make their decisions alone, do not consider the requests of the patient or relatives. Our goal is to strengthen patient autonomy and to support their opinion by training ICU physicians. On the other hand it is inevitable to define what useless medical treatment exactly is and to introduce this category in medical ethics and also in jurisdiction practise. OBJECTIVES. To determine the frequency and processes of EOL care at our centre.

Between October 2007 and December 2009, 150/388 (39%) patients staying in the ICU for more than 1 day, underwent some form of EOL care in the ICU. ICU staff notified investigators whenever an EOL decision was made. We recorded demographic details, documentation of the EOL care process in the case notes, and interviewed ICU staff to determine the EOL care processes involved.

RESULTS. 97 patients (65%) were male, 53 (35%) were females. Mean age was 42.4 ± 2 years. ICU stay was 6.7 ± 6.3 days, admission APACHE II score was 18.1 ± 7.8 which increased to 24.9 ± 9.4 on the day of EOL care decision. 33% patients had metastatic cancer. Reasons for initiating EOL care were refractory acute illness in 53%, advanced cancer in 51%, brain death in 11%, and lack of finances in 11%. EOL discussions were initiated by the family in 11%, and by the ICU medical team in 89% patients. 16 families wanted to take the patient home to die. The ICU consultant was involved in all discussions with the family, the primary consultants in 91% and primary team residents in 79%. Nurses were involved in only 3 patients. Agreement on EOLC was reached after 1 discussion in 70%, 2 discussions in 22%, and 3 discussions in 8% of cases. Documentation of the EOL care process was not done in 51% cases. Withholding of life support (WH) was practised in 110/150 patients (73%) and

Withdrawal of life support (WD) in 27%. Intubation was withheld in 14.5% patients, cardiopulmonary resuscitation in 50%, inotropes in 76% and dialysis in 8%. Regarding WD, only 3/40 patients were extubated and the ventilator withdrawn in another 7/40 patients. Inotropes were withdrawn in 20 patients (50%). Reduction of FiO2 0.21 without discontinuing mechanical ventilation was the commonest mode of WD, in 35 patients (88%). All patients received morphine infusions during LOLS/WOLS. Family members were present by the bedside in 61% cases.

CONCLUSIONS. WH is preferred over WD. Documentation of the EOL process does not occur in a significant proportion of cases. Nurses are rarely involved in the EOL care decision making process. Legal issues may be barriers to good EOLC in our ICU, and perhaps in India. OBJECTIVES. To know the point of view of the staff is essential to understand their beliefs, attitudes and decisions.

Brazilian private general ICU with 23 beds. The following items were analyzed: profile of the interviewed; their opinion about end of life questions: fear of death, fear of experience pain before death, the best place to die, advanced directives, decision-making process, therapeutic withhold of mechanical ventilation, nutrition, fluid management, antibiotics, vasoactives drugs, sedation and analgesia in patients which death is imminent and irreversible.

RESULTS. About 84.41% of our ICU team answered the research (n = 65). The mean age is 32.61 years (SD 6.74), 58.50% of female, 58.50% married, 50.85% Protestants and 38.50% Catholics and ICU professional experience of 6.75 years (SD 4.65). Using a visual analog scale (0, no fear to 10, the worst fear possible) the team pointed 5.96 as their fear of death; the fear of suffering pain before death was 8.17. For 44.60% of the responders, the best way to die would be with their lovely ones, no matter if at home or at hospital. Only 9.20% would prefer to die an ICU. The majority of the team (88.25%) would share the EOL decision-making process with the family instead only by the medical staff (52.30%). About 73.10% would leave an advanced directive with their therapeutic preferences like do not resuscitation orders. The ICU team agreed on the withdrawal of vasoactives drugs (41.50%), antibiotics (57.00%), nutrition (18.50%) and mechanical ventilation (1.53%) in patients out of treatment.

Our results showed the staff vision about their own death and their opinion about the end-of-life care issues. In developing country as Brazil there is a still gap between everyday practice and the current legislation. Fortunately, the debate about EOL issues has increased in last years. The end-of-life discussions and decisions should begin by respect to points of view of all involved: patients, family, medical staffs with a legal support of the society's beliefs and expectations.

Prospective observational study conducted in 8 Greek multidisciplinary ICUs. We studied all consecutive ICU patients who died, excluding those who stayed in the ICU \48 h or were diagnosed with brain death. 306 patients comprised the study population [mean age 64 ± 17 (SD) years, mean APACHE II score on admission 21 ± 7].

RESULTS. Of patients studied, 41% received full support including unsuccessful cardiopulmonary resuscitation (CPR). 48% died after withholding of CPR, 8% after withholding of other treatment modalities besides CPR, and 3% after withdrawal of treatment. Patients in whom therapy was limited had a longer hospital (p = 0.01) and ICU (p \ 0.01) stay, a lower admission GCS score (p \ 0.01), a higher APACHE II score 24 h prior to death (p \ 0.01), and were more likely to be admitted with a neurological diagnosis (p \ 0.01). Patients who received full support were more likely to be admitted with either a cardiovascular (p = 0.02) or trauma diagnosis (p = 0.05), and to be surgical rather than medical (p = 0.05). The most important factors affecting the physician's decision to provide full support were reversibility of illness and prognostic uncertainty; the physician's religious beliefs and legal concerns had minimal impact. The main factors guiding the decision to limit therapy were unresponsiveness to treatment already provided, prognosis of underlying chronic disease, and prognosis of acute disorder; old age was not a determinant, while economic cost and lack of ICU beds seem to play no role. Relatives' participation in decision-making occurred in 20% of cases and was more frequent when a decision to provide full support was made (p \ 0.01). The principal reason for not discussing end-of-life dilemmas with relatives was the fact that the family was thought not to understand (60%) Advance directives were rare (1%).

ICUs. However, in a large majority of cases, it involves the withholding of CPR only.

Withholding of other therapies besides CPR and withdrawal of support are infrequent. Physician has a dominant role in decision-making. OBJECTIVES. The primary objective of this study is to determine the prevalence of inappropriate or non-beneficial care in ICU patients as perceived by their ICU healthcare providers, as well as the reasons for this perception. Second, we want to determine which factors are associated with the perception of inappropriate care.

A descriptive survey design is used. A single-day cross-sectional evaluation of perceptions of inappropriate care among 2,327 ICU healthcare providers in 85 ICU centres in 10 European countries will take place on May 11th 2010. Questionnaires will be administered to ICU healthcare providers (nurses, head nurses, junior and senior ICU physicians) providing bedside care to adult ICU patients on that particular day. In this study, inappropriate care is defined as a patient care situation that is similar to one or more of seven scenarios. These scenarios were created based on the literature and a multidisciplinary conference attended by experts in intensive care, geriatrics, and palliative care.

. The cross-sectional study will take place on May 11th 2010. Preliminary results will be given at the ESICM conference.

We have designed a one-day cross-sectional study to record inappropriate or non-beneficial care in European ICU's. Results will be available for the ESICM conference.

GRANT 

Since the introduction of the Mental Capacity Act in the UK in 2007, the impact within research in the intensive care environment has not been elucidated. Since many of the patients are incapacitated and therefore unable to consent, it is now stipulated by the ethics committee that the researcher must make reasonable attempts to identify a consultee, failing this, nominate a person unrelated to the research project to be consulted. In order to comply with the Mental Capacity Act, retrospective consent must be obtained, once the patient regains capacity.

OBJECTIVES. The aim of the study was to highlight the difficulties in obtaining retrospective consent, evaluate the methods used and demonstrate the adaptations made to increase retrospective consents.

METHODS. This explorative analysis investigated the process of obtaining consent in 246 patients enrolled in an observational study on critically ill patients. Consent was obtained on admission if the patient had capacity. Assent from the patient's next of kin or a legal professional representative was obtained before enrolment in patients who lacked capacity. After discharge from ICU, a member of the research team re-visited these patients to explain their involvement in the research, its purpose, procedures, implications and any further participation required by the subject. At this point, the patient could consent or withdraw from the study. If the patient decided to withdraw from the study, all data collected and samples stored were destroyed. The researcher visited the patient for a minimum of two visits; firstly to explain the study; secondly to establish if the patient has retained the information and to gain retrospective consent.

RESULTS. 246 patients were recruited within the time period of which 44 (17.7%) died. In 65 (26.2%), consent was obtained on admission as the patients had capacity, 36 (14.6%) were discharged prior to obtaining retrospective consent, 10 (4%) lacked capacity on the researcher's visits, and 1 patient (0.4%) withdrew from the study. 90 patients (36.3%) were successfully consented retrospectively. Overall, the researchers performed 221 visits to obtain from the 137 patients for whom retrospective consent was required.

CONCLUSIONS. The process of recruiting patients who lack capacity within the intensive care unit is challenging and time consuming. Stipulations set by the ethics committee to seek retrospective consent once the patient has regained capacity, has a major impact on research staff time and finances. Detailed recommendations as well as guidelines how to assess capacity in the post-ICU patient and how the assessment of capacity has to be applied to Intensive Care research are needed to fully comply with ethical and legal requirements. OBJECTIVES. We wanted to know if patients expressed to surrogate decision makers, after ICU discharge, specific resuscitation directives, and we have investigated any factors related to the patients and their illness or care process that might be associated with this.

We reviewed patients admitted in the ICU between December 2008 and May 2009. A random sample of 60 survivor patients has been defined. Seven patients were excluded (3 for language barrier, 2 died, 2 were no more reachable). Fifty three patients took part in semistructured interview at 6-12 month post ICU discharge. The questionnaire discussed in detail the aspects of advance directives. Patients had also completed a quality of life questionnaire (Euroqol 5D), and we calculated the EQ-5D visual analog scale. We reviewed medical records in ICU data base: age, gender, length of stay, SAPS II, BMI, length of ventilator support and central venous catheterization as well as prescription of transfusion, hemodialysis or adrenergic agonist. Multivariate logistic regression was practiced to investigate any factor associated to expression of specific resuscitation directives after ICU discharge.

After ICU discharge, 60% of interviewed survivors expressed specific resuscitation directives to an appropriate identified surrogate (written ''living will'' or oral statement). EQ-5D visual analog scale was 58 ± 19. On multivariate regression analysis, only one studied variable was significantly associated to the post-ICU expression of specific resuscitation directives: age (odds ratio = 0.94, z = -2.1, p = 0.035).

CONCLUSIONS. After ICU discharge, a majority of our patients expressed to surrogate decision makers specific resuscitation directives, especially the younger patients. Our findings suggest that surviving to ICU is an opportunity to specify oral or written directive, and both may help to illuminate future decision making from the patient's perspective. OBJECTIVES. To explore the issues around EOLC provision for cancer patients in a critical care unit through family, professional and patient experiences. To explore how a diagnosis of cancer impacts upon EOLC provision for critically ill patients.

A Heideggerian phenomenological interview approach was undertaken, in order to gain personal experiences. Families of those patients who died after decisions to forgo lifesustaining treatment (DFLSTs) were interviewed. Patients who were seriously critically ill (APACHE II [ 25 or had received CPR) who experienced critical care were also interviewed, since patients' views about EOLC provision are very rarely explored. Doctors and nurses also contribute their vision for, and experiences of, EOLC in a cancer critical care unit. Thirty seven participants were interviewed.

Tensions between treating families versus treating patients impacted on timeliness of EOLC. Achieving a good death was possible through caring activities that made best use of technology to prevent prolonged dying. Decision-making and EOLC could be difficult to separate out which, in turn, affects prospects for EOLC. Three main themes included: Dual Prognostication; The Meaning of Decision-Making; and Care Practices at EOL: Choreographing a Good Death. These themes outlined the essence of moving along a continuum toward patients' deaths and the impact that had on opportunities for care and a good death.

CONCLUSIONS. Cancer affected the trajectory in unexpected ways. The trajectory could be very quick, especially in unexpected death and some newly diagnosed cancers. Even in the face of a life-limiting and serious disease like cancer, death could be unexpected. The rapidity of trajectory related to cancer diagnosis, prognosis, withdrawal and patient demise significantly impacted on the potential for, and timing of, EOLC. A sentiment of moving on from historical practices around critical care for cancer patients, and related poor prognoses, was overwhelmingly agreed on but important caveats in cancer prognostication remains. CONCLUSIONS. These data suggest that oscillation settings of 5 and 10 Hz provided more optimal PEF/PIF ratio ([1.1). Our data also suggests that airway clearance using HFCWO may facilitate improved gas exchange in mechanically ventilated patients. Further study is required to confirm these results GRANT ACKNOWLEDGMENT. Partial funding support in the form of devices was provided by Hill-Rom Inc.

A. Esquinas 1 , M. Folgado 2 , J. Serrano 3 1 Hospital Morales Meseguer, Intensive Care Unit, Murcia, Spain, 2 Hospital Virgen de La Concha, Zamora, Spain, 3 Hospital Reina Sofia Córdoba, Intensive Care Unit, Cordoba, Spain OBJECTIVES: We hypothesized that the use of intrapulmonary percussive ventilation (IPV) could effect hypercapnia/acidosis and airway secretions control during treatment with Noninvasive Mechanical Ventilation (NIPPV) in exacerbations of COPD associated with bronchial secretions.

Prospective multicenter study. The study was performed in the medical ICU of 3 Spanish university hospitals members of the Spanish IPV Working Group. We enrolled COPD exacerbation patients with secretions and the need for NIPPV in ICU. Criteria of exacerbations of COPD are: a respiratory frequency C25/min, a PaO 2 [ 45 mmHg and pH B 7.35. We define two IPV strategies as complementary treatment during NIPPV to evaluate the effects of IPV. Strategy Group I: NIPPV at first line and combination of IPV in early periods without NIPPV in spontaneous breathing and pH C7.35. Strategy Group II: First line of IPV with mouthpiece/face mask and oxygenation previous to the application of NIMV with pH \ 7.35. In both groups daily sessions IPV were applied by for 30 min/3 day by mouthpiece or face mask during stay in ICU. NIPPV was applied with BiPAP Ventilator (Respironics) and face mask with BiPAP mode. Cardiopulmonary monitoring, clinical and arterial blood gases were evaluated. Therapy was considered as successful when patients did not need NIPPV support and clinical and arterial blood gases returned to baseline.

RESULTS. 65 patients with COPD exacerbation were admitted in ICU for NIMV, age 70 ± 12 years, male (90%) 15 were excluded for severe hypoxemia (PaO 2 :FiO 2 B 200) associated with pneumonia (5/15) and cardiac insufficiency (10/15). Fifty patients were enrolled in the study. -UP TILT-TABLE REHABILITATION BETTER THAN SITTING IN A  CHAIR FOR VENTILATED ADULTS IN INTENSIVE CARE IN TERMS OF  IMPROVING LUNG FUNCTION? J. Manners 1 , A. Thomas 1 , S. Boot 1 , G. Mandersloot 1 1 Barts and The London School of Medicine and Dentistry, London, UK

Physiotherapy intervention is a fundamental part of the patient stay in an intensive care unit (ICU) and treatment is often aimed at maintaining/improving respiratory function. Physiotherapists use the upright posture to elicit these improvements and sitting in a chair and standing with a tilt-table are commonly used interventions. To date there are no published reports comparing the efficacy of these interventions in ventilated subjects.

• To compare the effects of these two positioning techniques employed with ICU patients.

• To measure changes in respiratory rate, tidal volume and minute volume during these positioning interventions. • To measure functional residual capacity during positioning interventions.

• To measure the change in metabolic demand during positioning interventions.

METHODS. Convenience sampling of ventilated subjects meeting the inclusion criteria was employed. Subjects acted as their own controls undergoing sitting in a trauma chair and standing on a tilt table at 70 degrees in random order on the same day. Respiratory rate (RR), tidal volume (V T ), ventilation (V E ) and oxygen consumption (VO 2 ) were measured at minute intervals during baseline and intervention for 10 min. Functional residual capacity (FRC) was measured once at rest and following each intervention. Measurements were recorded using the ''E-COVX'' module for the ''GE Carestation ventilator''. RESULTS. 10 subjects were recruited. No adverse events occurred during interventions. Significant increases from baseline RR (p \ 0.0001), V E (p \ 0.0001) and VO 2 (p = 0.009) occurred during the tilt table intervention. There was an increase in FRC during tilting of 0.6 L which failed to reach significance. Significant increases from baseline RR (p \ 0.0001), VO 2 (p = 0.024) and a decrease in V T (p = 0.015) occurred with the chair intervention. CONCLUSIONS. These interventions are safe in a critical care population. Increased muscular activity associated with upright interventions elicited expected elevations in VO 2. The tilt-table produced an increase in V E driven by an increased RR at the expense of V T . V E was not elevated during chair sitting despite an increased VO 2 and was accompanied by an unexpected decrease in V T. INTRODUCTION. UK guidelines about rehabilitation after critical illness highlight the need for outcome measures to determine patient progress and efficacy of treatment [1] . There is no consensus about the most appropriate measures of patient function. The AusTOMs [2] tool was designed by therapists in Australia to measure activity and function across nine scales assessing structural and functional difficulties and ability to perform activities. Scales are split into four domains (impairment, activity limitation, participation restriction and distress/wellbeing) and scored from 0 to 5 with 0.5 intervals allowed. Acceptable inter-rater variation is defined as an absolute difference of 0.5. AusTOMs has not been appraised in patients recovering from critical illness. OBJECTIVES. To prospectively determine the inter-rater reliability of the AusTOMs physiotherapy scales in adult patients who had undergone cardiothoracic surgery and required critical care admission for over 5 days. METHODS. The therapy (physiotherapy and occupational therapy) team underwent a 1 h teaching session using the AusTOMs handbook prior to commencing the trial. AusTOMs was then used over eight consecutive weeks during the weekly therapy goal setting meeting. Each week a patient was selected to be scored using the most appropriate functional scales. The clinical history was presented to the team by the therapist leading the patient's care. Therapists were then asked to independently score patients across the four domains for each scale. Reasons underlying differences in scores were explored by group discussion. The difference between the 10th and 90th centiles of the initial scores was calculated for each domain as a measure of inter-rater variability. RESULTS. 6-8 therapists were present at each meeting. Respiratory function and musculoskeletal movement related function were the most common scales used. The mean difference between 10th and 90th centiles was greater than 0.35 (± 0.21) for all domains.

None of the scales/domains showed consistent inter-rater reliability over the 8 week period. Overall the activity limitation domains of each scale showed the least inter-rater variance of scores. Clinical experience of therapist did not appear to influence scores.

CONCLUSIONS. The AusTOMs outcome measure showed poor inter-rater reliability when evaluated over an 8 week period on our intensive care unit. Further work is ongoing to evaluate the ability of AusTOMs to reveal changes over time when scored by therapists. INTRODUCTION. UK guidelines on the rehabilitation of patients after critical illness highlight the importance of establishing and reviewing individualised rehabilitation goals for all patients that are at risk of developing physical and non-physical morbidity [1] . Our institution's practice is to create objective goals that are SMART-Specific, Measurable, Achievable, Realistic and Timed [2] .

OBJECTIVES. The aim of this audit was to prospectively collect data regarding the setting of rehabilitation goals in a group of patients admitted to a cardiothoracic intensive care unit.

METHODS. All consecutive patients admitted under the Intensive Care team in November 2009 were included. Data regarding the timings of initial physiotherapy assessment, goal setting, and concomitant sedation were collected using a structured questionnaire completed by the treating physiotherapist.

RESULTS. 30 patients were admitted under the critical care team. 29 patients were assessed by a physiotherapist within 24 h of admission. Of these 29, 21 had SMART goals set within a median of 7 days of initial assessment (range 1-29 days). There was a correlation between level of consciousness and the number of days taken to set goals. Patients who were fully conscious or drowsy on initial contact (n = 11) had a SMART goal set in a median of 3 days. By contrast patients who were sedated/paralysed on initial assessment (n = 18) had goals set in a median of 12 days. Initial goal setting did not include other therapists or the family. Goals fell in to 5 categories, range of movement, hoisting out to chair for periods of time, sitting on the edge of the bed, transferring out to the chair by standing and mobility goals-i.e. walking set distances. The maximal interval between reviews of the patients' goals was 7 days.

Most patients had SMART goals defined and regularly reviewed. However, despite physiotherapy assessment within 24 h of admission, there was often a delay in setting these objective goals. The need for continuous sedation acted as a barrier to explicit setting of goals. The results emphasised the need to improve patient and family/carer involvement with initial goal setting in order to be compliant with UK standards. OBJECTIVES. Investigation of EMS effects on muscle strength and exploration of issues in relation to handgrip dynamometry in ICU patients.

One hundred seventy two consecutive patients with APACHE II score C13, were randomly assigned to the EMS (n = 86, age: 62 ± 18 years, APACHE II: 18 ± 5) or the control (n = 86, age: 59 ± 18 years, APACHE II: 18 ± 5) group. EMS sessions applied daily in muscles of both lower extremities. The strength evaluation of various muscle groups of the upper and lower extremities was made clinically upon awakening with the MRC scale, ranging from 0 to 5 (normal strength) for each group. The same scale was also employed in the diagnosis of CIPNM (MRC \ 48/60). A subgroup of these patients also performed handgrip dynamometry.

RESULTS. Fifty seven patients (EMS: 28, control: 29) were finally evaluated. EMS patients scored higher than controls (p B 0.05) in wrist flexion, knee extension, ankle dorsiflexion and right side hip extension, while they tended to perform higher in all other muscle groups (p: 0.20-0.10) ( Table 1) . GRANT ACKNOWLEDGMENT. This project has been co-financed by E.U. and the Greek Ministry of Development. BACKGROUND. Secretion removal is major aim of respiratory physiotherapy in intensive care. Manual hyperinflation provides a tidal volume to the lungs that is greater than baseline. It is effective in secretion clearance and is frequently used [1, 2] . There is a limited evidence that addressed the effects of combining rib-cage compression and suctioning on oxygenation, ventilation, and airway-secretion removal in mechanically ventilated patients [3] .

OBJECTIVES. The aim of this study was to investigate the effects of manual hyperinflation administered in combination with expiratory rib-cage compression on lung compliance, gas exchange, and secretion clearance in mechanically ventilated patients.

METHODS. Twenty-two intubated, mechanically ventilated, and hemodynamically stable patients were studied. The patients received manual hyperinflation, with or without expiratory rib-cage compression, with a minimum 3-h interval between the two interventions. Manual hyperinflation with or without expiratory rib-cage compression was performed for 5 min before endotracheal suctioning. Respiratory mechanics and hemodynamic variables were measured 5 min before (baseline) and then 5 and 20 min after the interventions. Arterial blood gases were determined 5 min before (baseline) and 20 min after the interventions. Secretion clearance was measured as sputum weight. The two measurements were obtained on the same day.

RESULTS. No significant differences were observed in gas exchange and secretion clearance between the two interventions (p [ 0.05). In each case, static lung compliance and tidal volume improved significantly at 5 min post-intervention (p \ 0.01), whereas at 20 min postintervention, only static lung compliance had improved significantly above baseline (p \ 0.01).

Our results suggest that the addition of expiratory rib-cage compression to manual hyperinflation does not improve lung compliance, gas exchange, or secretion clearance in mechanically ventilated critically ill patients. 

Recently, there has been an interest in mobilization of acutely ill patients who are in an intensive care unit (ICU). In the literature, the major safety issues while mobilizing critically ill patients has been outlined. Cardiac reserve [(CR) (% of age predicted maximal heart rate)] and respiratory reserve [(RR), ratio of partial pressure of oxygen in arterial blood to the inspired fraction of oxygen (PaO 2 /FiO 2 )] are the important factors that can affect the ability to tolerate the mobilization. Patient who has RR more than 300 and CR lower than 50% is considered to have sufficient reserve to tolerate mobilization [1, 2] .

OBJECTIVES. The aim of this study was to compare the effects of mobilization on respiratory and hemodynamic parameters in patients with sufficient and insufficient respiratory and/or cardiac reserve.

Mobilization events are divided into two groups (sufficient, insufficient) according to the pre-mobilization CR (sufficient, \50%; unsufficient, [50%) and RR (sufficient, [300; insufficient, \300). Heart rate (HR), systolic/diastolic/mean arterial blood pressure (SBP, DBP, MABP), respiratory rate (RsR) and percutaneous oxygen saturation (SpO 2 ) were recorded from the monitor. Respiratory and hemodynamic parameters were collected just prior to the mobilization, just after the completion of the mobilization when the patient had been returned the supine position and 5 min of the recovery period and compared between the groups.

A total of 87 abdominal surgery patients (47 male, 40 female) received 113 mobilization treatments in ICU. The mean age was 60.1 years, mean body mass index (BMI) was 24.9 kg/m 2 , mean APACHE II score was 20.5 and mean ICU stay was 7.3 days. Mobilization events included 88 (78%) sitting on the edge of the bed, 12 (11%) standing, 13 (11%) walking to chair and sitting in the chair. 85% (96) of mobilization events had insufficient RR and 15% (17) of mobilization events had sufficient RR. 74.3% (84) of mobilization events had insufficient CR and 25.7% (29) of mobilization events had sufficient CR. All respiratory and hemodynamic parameters were found similar in sufficient RR and insufficient RR group at all stages of the mobilizations (p [ 0.05). SpO 2 was higher, while HR and RsR was lower at all stages in sufficient CR group compared to insufficient CR group (p \ 0.05).

Resting HR and CR may affect the safety of mobilization, for this reason it is important to consider respiratory and hemodynamic parameters prior to and while mobilizing the ICU patients. INTRODUCTION. Obesity is a chronic disease and a major health problem. Obesity in critically ill patients is associated with a prolonged duration of mechanical ventilation and intensive care unit (ICU) length of stay [1] .

OBJECTIVES. The aim of this study was to investigate the effects of mobilization on respiratory and hemodynamic parameters in the critically ill obese patients.

[ 30.00 kg/m 2 )] were included as soon as their cardiorespiratory stability allowed mobilization protocol. Mobilization was defined as sitting in the bed, sitting on the edge of the bed, standing, walking to chair and sitting in the chair. Heart rate (HR), systolic/diastolic/mean arterial blood pressure (SBP/DBP/MABP), respiratory rate (RR) and percutaneous oxygen saturation (SpO 2 ) were recorded from the monitor. Respiratory and hemodynamic parameters were collected just prior to the mobilization (supine position), just after the completion of the mobilization when the patient had been returned the supine position and 5 min of the recovery period. All parameters were compared with initial values. The ratio of partial pressure of oxygen in arterial blood to the inspired fraction of oxygen (PaO 2 /FiO 2 ) was calculated from the arterial blood gas samples before and after the mobilization. INTRODUCTION. The use of respiratory therapy for patients with a variety of lung disease is a standard in medical care [1] , including in the intensive care unit (ICU) setting [2] . In this context, it is widely accepted the routine use of physical therapy in several situations in the intensive care, such as the care of critically ill patients not requiring ventilatory support, assistance during the postoperative recovery and the assistance to critically ill patients requiring ventilatory support [3] . At present definitive recommendations cannot be made regarding the use of respiratory physiotherapy for decreasing relevant clinical outcomes in critical ill patients requiring mechanical ventilation.

OBJECTIVES. This study aimed to determine the impact of providing chest physiotherapy on the duration of mechanical ventilation, intensive care length of stay, intensive care and hospital mortality in mechanically ventilated patients.

Single-centre, randomized, controlled trial in a university hospital general intensive care unit (ICU). Were included in the study 139 patients aged more than 18 years, admitted to the ICU needing mechanical ventilation for longer than 48 h. Physiotherapists provide group intervention (P) with the intensity and frequency of therapy they felt appropriate based on their assessment of the likely treatment benefit. Control patients (group C) only received suctioning, decubitus care and general mobilization.

RESULTS. Primary outcomes were ICU and hospital mortality regardless of the cause of death. Secondary outcomes were length of ICU and hospital stay, length of mechanical ventilation, weaning and extubation failure. Patients in the P group more frequently achieved parameters to start weaning, but there were no significant differences between P and C groups on weaning and extubation failure, length of mechanical ventilation and length of ICU stay. There was fewer hospital, but not ICU, mortality in the P group.

CONCLUSIONS. We demonstrated that respiratory physiotherapy decrease hospital mortality and suggest that this effect was, in part, secondary to the effect of the intervention on weaning from mechanical ventilation. INTRODUCTION. Critical illness can cause diverse cerebral dysfunctions ranging from unconsciousness to minor cognitive impairments (MCI). Severe cerebral dysfunction, as delirium, is known to affect outcome after critical illness but it is uncertain whether minor impairments affect mortality or morbidity [1] .

OBJECTIVES. The primary aim of this study was to estimate the incidence of MCI in a group of general ICU survivors immediately after ICU stay and three and 12 months after discharge. Secondary we wanted to explore if type of cerebral dysfunction after ICU discharge affected mortality and morbidity.

METHODS. Patients admitted to our general ICU were included prospectively. 

We included 80 patients. 11/80 (14%) were delerious and 14/80 (18%) were not delerious but had MMSE \ 24 after ICU stay. Of the 55 patients with MMSE C24, 28 were possible to classify as having MCI or not. 18/28 (63%, 95% CI: 45-83%) were found to have a MCI after ICU discharge. On 3 and 12 months these numbers were respectively: 11% (95% CI: 0-20%) and 10% (95% CI: -1 to 21%) There was an increased risk of both death and being institutionalised at both 3 and 12 months regarding delirious patients and patients with MMSE \ 24 compared to patients with MMSE [ 24. No such differences were found regarding patients with or without MCI. (Tables 1 and 2) . CONCLUSIONS. The incidence of MCI after critical illness is high on discharge but drops on 3 and 12 months after. Severe cognitive impairments affect mortality and morbidity, but minor cognitive impairments do not. OBJECTIVES. This study analyzes mid-term survival and risk factors associated with survival of patients undergoing cardiac surgery in Son Dureta hospital.

METHODS. 1938 patients were consecutively operated from November 2002 to December 2007. Patients who were discharged alive from hospital were followed until December 2008. We did Kaplan-Meier survival analysis and logistic regression study of variables associated with mid term mortality.

RESULTS. In-hospital mortality was 1.96% (95% CI: 1.36-2.60%). Information was available on 1,844 (97%) of 1,900 patients who survived until hospital discharge. At the end of the follow-up period, observed mortality was 6.5% (CI 95%: 5.4-7.7%). Survival probability at 1, 3 and 5 years of follow-up was 98, 94 and 90%, respectively. The mean time of follow-up was 3.2 years (range 0.01-6.06). Patients C70 years showed a lower survival rate than patients \70 years of age (log rank \ 0.0001). Age C70 years, history of severe ventricular dysfunction (EF \30%), diabetes mellitus, preoperative anemia and hospital stay were independently associated with mid-term mortality.

CONCLUSIONS. Mid-term survival of patients alive after hospital discharge was very satisfactory. Mid-term mortality varied according to age and several preoperative chronic diseases. 

A closed-ended questionnaire was developed by the Nurse congress commission of the Société de Réanimation de Langue Française (SRLF). An invitation to complete it online was sent by email to caregivers registered on the SRLF push-list. Results were analyzed by ICU or by respondent.

RESULTS. 731 caregivers working in 263 ICUs completed the questionnaire (65% were nurses, 24% were doctors, 9% were nurse's aides, 86% worked in adult ICUs and 14% in pediatric ICUs). 7% of adult ICUs (n = 222) had unrestricted policy but 58% had a visiting time of less than 4 h per day. At the opposite, 59% of pediatric ICUs (n = 41) had unrestricted policies. 63% of the respondents working in ICUs with a visiting time \4 h per day considered very useful or essential to enlarge visiting periods but 29% of them considered this enlargement as unhelpful. At the opposite, 9% of the respondents working in ICUs with unrestricted policy found very useful or essential to reduce visiting periods. 81% of caregivers working in ICUs with unrestricted policy but only 35% of caregivers working in other ICUs thought that an unrestricted policy was able to improve often or systematically the relations with families. Moreover, only 8% of caregivers working in ICUs with unrestricted policy but 46% of caregivers working in other ICUs thought that an unrestricted policy disturbs the organization of care. 96% of respondents found very useful or essential to give information in a dedicated room whereas it was often or systematically done in only 52% of ICUs. Identically, 97% of respondents found very useful or essential to give information to proxies with the patient's nurse whereas it was often or systematically done in only 54% of ICUs. Some cares were often or systematically programmed for family participation in 56% of pediatric ICUs but in only 0.5% of adult ICUs. Indeed, proxies often or systematically participated in nursing in 66% of pediatric ICUs but never in adult ICUs. At the opposite, proxies often or systematically participated in tracheal aspirations in only 5% of pediatric ICUs and in 0.1% of adult ICUs.

CONCLUSIONS. More than half of respondent's adult ICUs are closed but caregivers working in ICUs with unrestricted policy perceive it favorably. Some improvements are also expected by caregivers on the use of dedicated rooms for information and on the participation of nurses in meetings with families. Finally, participation of families to care is not a practice of French adult ICU caregivers. METHODS. Included: Patients with dysfunction of two or more organs in the first 24 h, admitted and discharged from ICU during 2008. Excluded: neurocritical and politrauma patients. Contact 1 year following discharge; questions were asked concerning symptoms related to a period in intensive care that presented following discharge and which were not present prior to admission. In the case that the patient was not contacted, the next of kin was asked.

RESULTS. 154 patients included. General characteristics during admission to ICU: 63% male; age 60.2 ± 17.5 years; SOFA* 7 ± 2.8; APACHE** II 16.8 ± 6.4; APACHE** IV 57 ± 18.8; length of stay in ICU: 5.1 ± 28.4 days; 42.9% were on invasive mechanical ventilation and 17.5% on non-invasive mechanical ventilation. Data collection was carried out over 18 ± 4.4 months, on average 19 months (range: 12-26 months). 18.2% (28 patients) had died at the time of contact. The person contacted was the patient in 33.3% of the cases, the spouse in 13.7% and immediate family (patient's parent/child/sibling) in 43.1% of the cases. 46.6% had difficulty sleeping following discharge from ICU with an average time since discharge of 10.5 ± 8.9 months; 51.1% suffered feelings of sadness and difficulty in finding enjoyment which had persisted for 8.7 ± 8.1 months; 30.3% had experienced difficulty in concentrating over an average of 9.3 ± 8.8 months; 28.2% had suffered some form of memory loss after discharge over an average period of 8.9 ± 8.6 months; 76.5% presented with asthenia over an average of 10.8 ± 8.3 months; 64.7% had arthromyalgia over a period of 12.6 ± 8.4 months; 35.3% had experienced changes in appetite over an average of 6.6 ± 7 months; 43.3% had changes in intestinal habit over an average of 7.78 ± 9 months; of which 64.3% had diarrhoea, 30.4% constipation, and 5.4% both symptoms; 18.8% presented with headache over a period of 12.9 ± 9.2 months; 15.2% had tremors, that had not previously been present, over an average of 12 ± 7.3 months; 12.8% had experienced reduced vision, over an average period of 16.5 ± 7.1 months; 10.7% presented with speech/ language problems, over an average period of 14.6 ± 8.6 months; 14.5% exhibited newly presenting changes in micturition, over 16 ± 7.7 months. Another less frequently occurring symptom was loss of hearing (1.3%).

CONCLUSIONS. Severely ill patients that are admitted to ICU frequently present with ''residual'' symptomatology following discharge, most notably arthromyalgia and asthenia. Many of these conditions persist for months. 

Intensive care unit (ICU) readmission rates range from 3 to 12%, in spite of initial recovery from critical illness. Previous researches report that the Revised Acute Physiology and Chronic Healthy Evaluation (APACHE II) score at either admission or discharge is an important predictor for readmission after ICU discharge. However, there are a few papers concerning the association of discharge APACHE II score with readmission after discharge from surgical intensive care unit.

OBJECTIVE. We compared the ability of the discharge APACHE II score with that of the admission APACHE II score in predicting readmission, especially early readmission within 48 h, after discharge from ICU. CONCLUSION. This study showed that both discharge APACHE II score and admission APACHE II score are useful predictors for readmission after ICU discharge, but discharge APACHE II score is only independent factor in predicting early readmission within 48 h after ICU discharge.

INTRODUCTION. Health related quality of life (HRQoL) is decreased in former ICU patients. In research outside the intensive care field it is well known that the psychosocial factors, coping strategies and perceived hopelessness affect HRQoL. However, the influence of coping and hopelessness on HRQoL after intensive care is unknown.

OBJECTIVE. The aim of this study was to examine how coping strategies and perceived hopelessness among former ICU patients compares to corresponding in a reference group. We also evaluated the effect of coping and hopelessness and ICU related factors on HRQoL.

METHODS. Prospective, multicenter study in three mixed ICU's in Sweden. Patient demographics, length of stay, APACHE II score, reason for admission and time on ventilator were collected for all adult patients. Questionnaires, including the coping instrument Pearling-Schooler Mastery Scale (PMS), the 2-item hopelessness scale, SF-36, demographic data and previous illnesses were sent 6 months after discharge from hospital to the patients. The reference group (n = 6,093) was a random selection of persons from the same catchment area as the study patients.

. 780 (59%) ICU-patients, 18-74 years, returned the questionnaires. The patients reported significantly lower mean scores in coping 13.6 (SD 4.6, p \ 0.001) and higher perceived hopelessness 2. CONCLUSIONS. This study indicates that coping strategies and perceived hopelessness are important for the HRQoL of previous ICU patients. However, the magnitude of these effects are smaller than that of pre-existing diseases. INTRODUCTION. Mortality on a medical intensive care unit (ICU) is estimated to occur in about 25% of patients. Its association with age, severity of illness and comorbidities is well established. For other diseases like coronary artery disease it has been shown that pre-existing depression is a risk factor for worse outcome. The role of depression regarding the outcome of ICU patients has not been investigated so far. We studied the association between pre-existing depression and mortality in medical ICU patients and present preliminary data of this ongoing study.

OBJECTIVES. Assessment of a possible association between mortality of ICU patients and prevalent depressive mood at time of ICU admission. The primary endpoint was 28-day mortality.

METHODS. Prospective cohort study. All patients admitted to a medical 8-bed ICU in a university hospital, older than 18 years, were eligible. Postoperative patients and patients who had an expected length of stay below 48 h (survey) were excluded. Patients whose cognitive function allowed appropriate comprehension and response answered the hospital anxiety and depression scale (HAD). Prevalent depressive mood at admission was defined by a score C8 in the depression dimension. All other patients were assessed by observer rating by next-ofkin. In this case the Hammond scale, a validated instrument for observer rating of depressive mood (cut-off C4), and a modified version of the HAD for observer rating (cut-off C10) were used. In addition APACHE II, SAPS II, SOFA, age, sex, comorbidities, reason for admission, length of ICU stay and ventilator days were recorded.

. By now 173 patients had complete follow up data. Of these patients 53 (31%) were classified to have depressive mood at ICU admission. In total 46 patients had died by day 28 (27%). The 28-day mortality was 40% (21/53) in patients with depressive mood and 21% (25/120) in patients without (p = 0.01). Patients with and without depressive mood did not differ with respect to age, sex, APACHE II, SAPS II or SOFA score at admission. Multiple logistic regression analysis with 28-day mortality as the dependent variable revealed that prevalent depressive mood at the time of ICU admission was an independent risk factor for mortality (Table 1) .

CONCLUSIONS. Pre-existing depressive mood is an independent risk factor for mortality in medical ICU patients. INTRODUCTION. Some classical post-ICU discharge predictors of death are described, such as age, severity of disease and level of nursing care [1] . Besides these factors, some laboratorial data at ICU discharge are potential predictors of post-ICU death.

OBJECTIVES. The aim of this study was to investigate whether standard base excess (SBE), pH, lactate, hemoglobin level, creatinine, platelets, leukocytes and albumin at the ICU discharge as well as the 25% decrease on C-reactive protein concentrations (CRP [ 25%) from the day pre-ICU discharge to the day of ICU discharge may be useful predictors of in-Hospital outcome.

Patients discharged from the ICU after at least 72 h of stay were retrieved from our prospective collected data base. A multivariate analysis was performed using a backward-LR binary logistic model taking in-hospital death as a dependent variable and the cited data as independent variables.

RESULTS. 396 patients were retrieved. The average age was 49 ± 19 years old, mean APACHE II score was 17 ± 8, and the main causes of admission were septic syndromes and respiratory failure. The in-hospital mortality after ICU discharge was 18%. The ICU length of stay was 9 ± 9 days. At the time of ICU discharge pH was 7.41 ± 0.04, SBE was -0.3 ± 4.0 mmol/L, lactate was 1.9 ± 1.0 mmol/L, hemoglobin 9.6 ± 2.0, creatinine was 1.3 ± 1.46 g/dL, albumin was 2.68 ± 0.64 g/dL, platelets was 280,645 ± 166,100/mm 3 , leukocytes was 10,795 ± 8,769 cells/mm 3 and the number of patients who lowered CRP at least 25% were 159 (40% CONCLUSIONS. This study demonstrated that SBE, Lactate, Hemoglobin and Albumin concentrations on the day of ICU discharge are independent predictors of in-hospital mortality. Moreover, the reduction on CRP levels above 25% in the last 24 h of ICU stay is a strong predictor of better in-hospital clinical outcome. We suggest that these variables together with the clinical judgment may be taken into account on the ICU discharge decision process. 

Readmissions to the intensive care unit (ICU) are usually associated with increased morbidity and mortality, and they may evidence the quality of patients' care. The risk for ICU readmission varies across studies, and is generally analyzed just before ICU discharge, leading to deviation of ICU team and patients' daily goals. Early prediction may improve the care for patients in risk for ICU readmission, and help developing mechanisms for its prevention.

OBJECTIVES. To analyse risk factors for readmission in intensive care unit looking at the first 24 h data after unit admission.

METHODS. The first intensive care unit admission of patients was analyzed from January to December 2009 in a medical-surgical unit. Readmission to the unit was considered those during the same hospital stay or within 3 months after intensive care unit discharge. Deaths during the first admission were excluded. Demographic data, acute illness and comorbidity prognostic scores, and use of mechanical ventilation were submitted to uni and multivariate analysis for readmission. Numeric variables were expressed as median or percentage. CONCLUSIONS. Age, medical admission, SOFA score and respiratory-and/or sepsisrelated admission are early associated with increased ICU readmission risk. OBJECTIVES. The aim of this study was to examine patient perceived HRQoL in former ICU patients that die in the period from 6 month up to 3 years after discharge from intensive care unit and the hospital.

METHODS. Prospective, multicenter study in three mixed ICU's in Sweden. Questionnaires, including HRQoL (SF-36), demographic data and previous illnesses, were sent out six, 12, 24 and 36 months after discharge to all former adult ICU patients. Data for this study were only collected among those dying before the 36 months post-ICU follow-up.

Of the 980 patients who returned the questionnaires 123 (12.6%) died, 31 (25.2%) between 6 and 12 months, 55 (44.7%) between 12 and 24 months, and 37 (30.1%) between 24 and 36 months. The most frequent admission diagnoses were respiratory problems n = 36 (29.3%) and gastrointestinal diseases n = 33 (26.8%). Examining HRQoL in the former ICU patients the following observations were made: (see Fig. 1 ). A pronounced and quantitatively large decrease in HRQoL is seen for the surviving patients with pre-existing disease as compare to the previously healthy survivors. Although already at a very low value further decreases in HRQoL for the patients dying before 3 years post ICU is significantly less as compared to the ICU patients with pre-existing disease that survives. The decrease is mainly in physical function, role physical function and role emotional function (marked in the figure).

CONCLUSIONS. Yes, health related quality of life is extensively affected, mainly in the dimensions physical function, role physical function and role emotional function. Importantly, in these two affected physical dimensions a shorter time to death increases such a decrease. The finding further stresses the importance of pre-existing diseases for the final HRQoL outcome of former ICU patients. INTRODUCTION. Despite initial recovery from critical illness requiring ICU admission, many patients remain at risk of subsequent deterioration and death [1] . Recent studies have shown readmission rates ranging between 4 and 7% [2] ; this population had mortality rates six times higher and were eleven times more likely to die in hospital [1] .

1. To calculate the readmission rate in our mixed ICU unit over a 6 months period 2. To identify risk factors associated with readmission into the ICU 3. To study the outcomes of these readmissions METHODS. a retrospective observational study, data was collected from an ICU computer database (Metavision) and analysed manually RESULTS. The total number of admissions in this period was 365, average patient age was 57 ± 17 with 52.4% being males. Readmissions constituted 10.5% of the total admissions with 23.8% of those readmitted within 24 h of their initial discharge. 40% of the initial discharges from the unit were made out of hours i.e. unplanned, presumably due to heavy demand on beds. Readmissions were particularly associated with patients discharged to surgical wards 23.8% and the hepatobiliary HDU 28.6%, the latter might reflect the proportion of that particular patients population received. 52.4% of the readmissions required to stay 72 h or less in ICU. The overall mortality of the patients requiring more than one admission in this 6 months period was 57.1%.

There is an urgent need for expanding ICU services in our hospital, i.e. extra beds, staff, outreach teams, etc in addition to investing in nursing capacity building especially in surgical wards. 

We agree with others studies that compared with the general population, ICU survivors report lower HRQOL. Moreover, a relationship between several factors like sepsis, renal failure, SOFA (first and second day score), critical illness polyneuropathy, mechanical ventilation, sedation time, previous psychiatric history and blood products transfusions were found in our study population. CONCLUSIONS. According to our data, subclavian vein was the most common insertion site used, especially as 2nd and 3rd placement and was related with the lower incidence of becteremia episodes. Although the risk of placing a CVC for inflection complications is against the risk for mechanical complications, we have to improve our CVC policy, preferring the subclavian or the jugular site of insertion, in order to minimize the infection risk for a nontunneled CVC. OBJECTIVES. Objectives for this study were to determinate the frequency and the risk factors associated with BOS. Secondarily, we searched several variables as civil status, age, sex, work seniority as potential risk factors.

Inclusion criteria were to work in Critical Care Unit (CCU) the Hospital Clínico Universidad de Chile (HCUCH). This unit included 2 subunits: intensive care unit (ICU), middle care unit (MCU). The MBI Ò Instrument was applied between April to July of 2009. All staff of CCU were asked to response the instrument. As previously reported, BOS was defined with high EE, high DP and low PA. Risk of BOS was anything of the three dimensions positive for BOS. We gave information on specifics objectives and the schedule of a future intervention programme. For analysis, comparisons were made based on student t test, Chi-square test with Yates corrections or Fisher exact test as corresponded. For all tests we used confidence interval 95% with p \ 0.05.

A total of 111 MBI Ò tests that included all sub-units in CCU. This is a 56% of all personal working in the CCU. BOS was found in 8.26% of cases. Women (75%), unmarried (75%), with an average of age 31.9 years old. (23-47 years old) and with a work seniority younger than 5 years (50%). EE is high (31.5%), for nurse and paramedical personals. DP was 22.5 and 31.5% to middle level, for nurse and medical doctor, and low PA in 34% for paramedical personal, with longer work seniority (more than 10 years). Risk factors were female gender, unmarried status, childless, middle aged (23-47 years old) and recent start in the job (stay younger than 5 years). INTRODUCTION. Burnout is a prolonged response to chronic emotional and interpersonal stressors on the job, and is defined by three dimensions: exhaustion, cynism (depersonalization), and inefficacy. ICU physicians are exposed to several stress factors and are particularly predisposed to this syndrome.

To describe the prevalence of burnout syndrome among intensivists and its relation to their quality of life.

METHODS. An epidemiological cross-sectional survey conducted to evaluate all adult ICU physicians in Salvador, BA (Brazil), from October to December 2006. The quality of life and burnout syndrome were evaluated respectively by the WHOQOL-Bref instrument and the Maslach Burnout Inventory (MBI). Burnout was classified into low, moderate and high levels for the three studied dimensions, according the MBI classification, and it was defined by the presence of a high level in at least one dimension. The quality of life was evaluated in four domains: physical, psychological, social relationships and environment, graduated from 0 to 100, with higher scores denoting higher quality-of-life. [1] ) has been successfully used to measure nursing workload on an Intensive Care Unit over a 24-h period. In contrast to Intensive Care, the nursing care workload on MC is not evenly spread over a twenty four period, but tends to vary between shifts.

OBJECTIVES. The aims of this pilot study were (1) To assess the fitness of NAS as an accurate reflection of nursing workload on an MC unit. (2) To determine the nursing work load, per patient, per 8 h shift.

Prior to the commencement of the study all thirty one nurses taking part received instruction in the content and registration of NAS. At the end of each 8 h shift, each nurse retrospectively scored their patient(s) using NAS. This consists of a check list containing twenty three items giving a possible score between 0 and 177, where 177 equates to 1.77 Full Time Equivalent (FTE) Intensive Care nurse. The NAS were entered in to a database and the average scores, per patient, per shift were calculated. Three hundred patients were retrospectively scored over a 2-month period in October and November 2009. Not all patients were scored on all three shifts as some patients had been transferred out of the unit before shift end. In addition any incorrectly completed forms were discarded and excluded from the study. METHODS. This multicenter pilot study included 14 doctors working at (pediatric) intensive care units (ICU). Subjects were randomly assigned to two groups: one was first tested during day, then during night, while the other was tested in reverse order. The d2 Test of Attention [3] was used to assess attentional performance. Total performance (Tn-F) score, standardized for age and level of education, was used to express attentional performance. Subjective, 1-to-10 scores were gathered in two questionnaires.

RESULTS. Figure 1 displays standardized total performance scores of 14 doctors. Measured attentional performance showed high intra-and interpersonal variability and did not differ between both shifts (p [ 0.9). In contrast, doctors expected alertness to be decreased (7.5 ± 1.7 and 6.1 ± 1.7 (mean ± SD) on subjective 1-to-10 scale during day and night shifts, respectively; p \ 0.02) and the chance of making errors to increase (from 3.3 ± 1.7 to 6.2 ± 1.3 (mean ± SD); p \ 0.01) during night shifts.

CONCLUSIONS. Physicians working at ICU are aware of the risk of making errors during night shifts. However, we showed that doctors perform equally during night and daytime when confronted with a short-time challenging task. Consequently, a discrepancy between measured attentional performance and expected alertness was observed. These results suggest nocturnal alertness might be comparable to daytime during short-lasting tasks that elicit a high level of stress and motivation (e.g. testing, medical emergency). Further research is needed to elucidate if longlasting (routine) tasks reflect decreased sustained attention and contribute to medical errors. 

We studied 297 physicians, the majority of whom were male (70%). Mean age and time since graduation were 34.2 and 9 years, respectively. High levels of emotional exhaustion, depersonalization, and reduced personal accomplishment were found in 47.5, 24.6, and 28.3%, respectively. Prevalence of burnout syndrome, defined as a high score in at least one dimension, was 63.3%, while prevalence was 7.4% for all three dimensions. In conclusion, burnout syndrome was common in this sample of ICU physicians. AIMS. Our goal was to assess the physician's opinion about potential competencies of a triage nurse.

A representative cross sectional study design was applied with self-fill-in questionnaire about physician's attitude related to skills of triage nurses. The questionnaires were distributed between September and November 2007 in 24 (out of 31) EDs. In this survey 159 physicians' questionnaires were processed. Chi-square and Student-t test was used for comparison of variables. P values less than 0.05 were considered statistically significant.

RESULTS. 79.2% of physician would support the special training of triage nurse. 55.3% of physician suggests that the nurses use the patient's physical examination regularly in EDs. The full time (FT) emergency physician significantly would reduce the basic competencies of nursing (e.g. dressing, feeding of patient, P = 0.008, and P \ 0.001, respectively) than parttime (PT) emergency physicians. Significantly greater part of the FT physician would widen the competency of triage nurses in the field of physical examination of nervous system (P \ 0.001) and cardiovascular system (p = 0.005) than the PT physician.

CONCLUSION. Hungarian emergency physician would widen the competency of triage nurse, but only half of physician would like to that nurses apply physical patient examination in practice. The full time physician would give more competencies for triage nurse than part time ones, but the final field of competency will be depended on other factors.

Healthcare-associated infections (HCAI) are estimated to affect 1.4 million people worldwide, causing longer hospital stay, increasing hospital costs and excess mortality [1] . Hand hygiene represents the single most effective way to prevent healthcareassociated infections. Compliance with hand hygiene amongst healthcare workers (HCW) has been demonstrated to be quite low at 40% [2] .

To quantify the degree of compliance to hand hygiene norms in the ICU and to assess the short term success of strategies to improve hand hygiene compliance.

SETTING. 34 bedded Medical-Surgical ICU in a Tertiary care centre.

DESIGN. Prospective observational.

METHOD. Unobtrusive observer (single person). Observed over sessions of 1 h. The compliance was calculated as :Number of times the staff performed hand hygiene/Number of hand hygiene opportunities. The number of hand hygiene opportunities was based on the WHO tools [3] : before touching a patient, before clean/aseptic procedures, after body fluid exposure risk, after touching a patient and after touching patient surroundings. INTRODUCTION. ICU delirium represents a form of brain dysfunction that in many cohorts has been diagnosed in 60-85% of patients receiving mechanical ventilation. Delirium is a common but complex clinical syndrome characterized by disturbed consciousness, cognitive function or perception, which has an acute onset and fluctuating course and is associated with poor outcomes. And yet, it can be diagnosed and treated. In the UK, reporting of delirium is generally considered to be poor. In light of updated NICE guidelines on delirium due out this year, specialist clinical assessment will soon become gold standard as a means of diagnosing and reducing the prevalence of this condition in the ICU setting. NICE recommends that CAM-ICU (Confusion Assessment Method) be used by healthcare professionals who are trained and competent in the diagnosis of delirium. On our 16-bed unit, we are currently implementing CAM-ICU assessments to be performed twice daily (at the commencement of each nursing shift) as well as RASS (Richmond Agitation and Sedation Scale) scoring on an hourly basis for all patients. OBJECTIVES. To implement training of all our ICU nursing staff in the use of CAM-ICU and RASS scoring. To periodically validate and reinforce earlier training, so as to improve assessment and reporting of delirium. METHODS. Our 'Delirium Group' comprising both nursing and medical staff, taught CAM-ICU and RASS to 48 staff members using multimedia presentations in small groups and/or individual teaching sessions over 6 weeks. ScoringofCAM-ICU andRASSwassubsequentlyauditedon3occasions post training. Discrepancies were discussed and post-audit retraining provided where necessary. RESULTS. The following audit and validation data were generated on our unit as documented in Table 1 . No statistical analysis was undertaken. We anticipate focusing on the challenges encountered and strategies used in managing this change in our ICU practice. METHODS. The factors causing resistance to change based on multisource data. Qualitative technical methods were used: brainstorming and focal groups. The data collection elaboration was created by the collaboration of ICU Nurse, Quality Department Nurse and external reviewers. Finally, the main factors were classified in different categories. Each category was scored by 1 to 5 according to gravity and prevention possibility. Finally, priority was given to more serious and easier prevention problems. RESULTS. The most serious problems for ICU professional was the historical factors. The easier solution problems were ''the lack of information'' and all evaluators were agree with it. We arranged the factors in order to the next classification (Tables 1 and 2 ). DISCUSSION. All investigators were agreed with the low importance of problems with payments and low prevention probability of low organisational flexibility, so they were agreed on not to work about them. The ICU professionals were more pessimistic and have lower confidence in prevention possibilities but they showed more confidence about the capacity to learn new skills. They weren't worried about resistance to do experimental things. Probably, historic factors play an important role in this pessimistic attitude. On the other side, Quality and Safety Experts have more experience in prevention programs and they put all their trust in its. After doing the analysis, we chose the ''lack of information problem'' to plan prevention activities. We consider it is a serious and real problem but at the same time, easy of prevent. CONCLUSIONS. The implementation of the Patient Safety Program in the ICU means a real cultural change. The priority analysis could help to plan strategies in order to avoid the program failure. OBJECTIVES. We concerned about whether medical personnel could recognize management of the cuff of artificial airway or not.

We asked to doctors and nurses working in intensive care unit of Konyang University Hospital, Daejeon, Republic of Korea. We asked 12 questions with 18 contents of questionnaire that was composed of methods of set initially, maintenance and appropriate pressure of cuff.

RESULTS. 96 of medical personnel replied to us. Most of them had worked in intensive care unit, so they had placed of artificial airway. 86.4% of them used manometer to adjust the cuff. We could find that nurses had more cognition compared to doctors for it (94 vs. 76%). Only 28.5% of doctors described pressure of the cuff in medical record. 89 of medical personnel replied that they knew the appropriate range of cuff pressure. 37% (32/89) of them replied that the range of cuff pressure was kept with 25-30 mmHg and 34% (30/89) was 20-25 mmHg. 70% of nurses in the ICU knew that range of cuff pressure was 15-25 mmHg.

Most of them knew complications of high and low pressure of the cuff. 69.8% of medical personnel monitored the cuff balloon during receiving mechanical ventilation and they used manometer to adjust it. 78% of nurses knew that the cuff should be adjusted continuously, but 58% of doctors did. Interval measuring the cuff pressure was 38% of once a day, 38% of three times a day, 17% of more than four times a day CONCLUSIONS. Most of the medical personnel knew to keep appropriate cuff balloon to prevent various complications of artificial airway. They had insufficient cognition about maintaining the cuff balloon and appropriate level of cuff pressure. That was more prominent in doctors than nurses

Interhospital transfer is occasionally required as a consequence of limited therapeutic options or because of a need for a higher intensity of medical care that cannot be given in rural Intensive Care Units. Along with the potential benefit for the to be transferred patient, transport may also lead to hemodynamic and pulmonary deterioration. In order to minimize additional risk of interhospital transport of critically ill patients, a Mobile Intensive Care Unit with a specialized retrieval team was established in our university hospital-based intensive care unit. From March 2009, transport of the critically ill patients in our adherence region are performed by MICU.

OBJECTIVES. In this prospective audit adverse events and patient stability during MICU transfers were assessed and compared to our previous data on transfers performed by standard ambulance [1] . RESULTS. 74 interhospital transfers over a 10-month period were evaluated. Systolic blood pressure, glucose and haemoglobin were significantly different at arrival compared to departure, although never significant values for major deterioration were reached. An increase of total number of variables beyond threshold at arrival was found in 38% of patients, 32 percent exhibited a decrease of one or more variables beyond threshold and thirty percent showed an equal number of trespassed thresholds. There was no correlation between the patients status at arrival and the duration of transfer or severity of disease. ICU mortality was 28%.

Compared to standard ambulance transfers of ICU patients performed in 2005, there were far less adverse events: 12.5 vs. 34%, which in the current study were merely caused by technical (and not medical) problems. Although mean APACHE II score was significantly higher, patients transferred by MICU showed less deterioration in pulmonary parameters during transfer than patients transferred by standard ambulance.

CONCLUSION. Transfer by MICU imposes less risk to critically ill patients compared to transfer performed by standard ambulance and has therefore resulted in an improvement of quality of interhospital transport of ICU patients. INTRODUCTION. Previous studies in adult intensive care units (ICUs) reported rates of pre-mortem to post-mortem discrepancies ranging between 7 and 32% depending on the population studied. And, most of them were retrospective studies, which included small number of patients.

To compare clinical and pathological diagnoses and to determine the types of errors in a large and multidisciplinary ICU-patient population.

We conducted a prospective study of all consecutive autopsies performed on patients who died in the ICU of the Hospital Universitario de Getafe, Madrid, Spain, between January 1982 and December 2007. The diagnostic errors were classified in two categories: class I errors that were major misdiagnoses with direct impact on therapy, and class II diagnostic errors which comprised major unexpected findings that probably would not have changed therapy. CONCLUSIONS. This study found significant discrepancies between clinical diagnoses before death and post-mortem findings. This reinforces the importance of the post-mortem examination in detecting otherwise unexpected diagnoses and improving the quality of care of critically ill patients. INTRODUCTION. Unplanned extubation is associated to a high risk of reintubation end correlates with increased risk of nosocomial pneumonia. On the other hand, reintubation significantly increases morbidity and mortality in critical ill patients, increasing the incidence of ventilator associated pneumonia (VAP) rate and makes the airway management risky.

OBJECTIVES. The aim of our study was to test the rate of unplanned extubation as well as the reintubation rate in our ICU, in order to evaluate the efficiency of our airway and weaning time protocols.

METHODS. During a nearly 4 year's period, 216 patients admitted to the ICU, mean age: 61.8 years, mean APACHE II score: 18.5, mean LOS: 13.5 days, with predicted and actual mortality: 31.1 and 22.47% respectively. From these, 191 were intubated and included retrospectively in our study. 109 patients were extubated, while the others either underwent bedside percutaneous tracheostomy or died. We concerned that the number of days of mechanical ventilation were about equal to the number of days of intubation. Reintubation was defined as the need to reintubate during the first 48 h after extubation.

We recorded four episodes of unplanned extubation. Three of them caused by malfunction of the tube due to secretions and airway obstruction and one of them was undesired extubation caused by the patient himself. The total number of days of intubation was 2,111, mean ± SD: 11.06 ± 11.87, min: 1, max: 51 days. Therefore the rate of unplanned extubation was 1.897%, while the standard limit is below 15 %. The total number of reintubations was 19, while the total number of scheduled extubations was 109. Therefore, the reintubation rate was 17.43%, while the standard limit is below 12%.

CONCLUSIONS.The recorded rate of unplanned extubation was low in our ICU patients, below the acceptable limit, assuming that our sedation and airway management policy is effective. On the other hand, the recorded rate of reintubation was high in our study, above the acceptable limit. Although a low rate of reintubation might indicate excessively long mechanical ventilation times, this did not recorded to our study. Nevertheless, our data suggest that we have to improve further our weaning time protocols, making the extubation procedure safer, and avoiding risk factors for VAP. . PVs such as inappropriate enrollment of patients with a contraindication to the study treatment may lead to excess harm in the active intervention group [1] and failure to deliver the study intervention according to the study protocol may underestimate true treatment efficacy [2] . Full reporting of PVs may aid in the interpretation of RCT results however there are no published reviews on this topic [3] .

OBJECTIVES. To determine reporting rates for key types of PVs and to investigate study characteristics that may be related to reporting. Publications were excluded because they were subgroup or economic analyses of a previously published RCT [6] , not a RCT [5] , not published in the target journal [4] , systematic reviews [2] , or other reason [5] . Median trial size was 700 participants (range: 20 to 162,387). 13/80 (16%) of RCTs were single centre, 58/80 (72%) were Industry funded and 21/80 (26%) reported negative findings. Overall 71/80 (89%) of RCTs reported some form of PV, these included: 51/80 (63%) patient compliance; 40/80 (50%) discontinuation of study intervention due to safety; 32/80 (40%) study intervention-related researcher error; 16/80 (20%) inappropriate enrollment and; 8/80 (10%) technical errors in randomisation. Multi-centre RCTs may be more likely to report study intervention-related researcher errors (44% of multi-centre trials vs. 15% of single centre trials, P = 0.06). Academic trials were less likely to report discontinuation of study intervention due to patient safety (38% of Academic trials vs. 80% of Industry trials, P = 0.002) and were less likely to report technical errors in randomization (5% of Academic trials vs. 23% of Industry trials, P = 0.03).

CONCLUSIONS. Multi-centre trials are accepted to be organizationally complex. On-site education may be required to reduce errors in study intervention delivery attributable to the research team. It is possible the apparent excess harm attributable to Industry trials is a reporting artifact however, if it is real, it must be addressed. Additional research is required to investigate patient safety-related PVs and technical randomization errors, which may be lower in Academic trials. 

To determine the occurrence of harmless incidents and AE related to physician's competences in ICUs, disclosing their potential risk factors. CONCLUSIONS. This prospective study was essential to identify the proportion of our ICU admissions affected by MD-INC and MD-AE, disclosing their nature. Our MD-AE rates, affecting more than 40% of admissions, were higher than those described in prior general studies, including not only ICUs. Among the detected MD-AE, hypoglycemic episodes not related to insulin administration predominated, indicating important deficiencies regarding nutritional support. Severity on admission and length of stay were important risk factors for the occurrence of at least one MD-AE. 

A systematic measurement and analysis of unintended events (UE) have been recommended for patient safety and improvement of quality of care in critically ill patients. However, a spontaneous reporting system may be inefficacious in intensive care unit (ICU) because of a poor data collection, particularly by physicians staff.

OBJECTIVES. The aim of this study was to evaluate the reliability of a staff spontaneous event report by comparison with events collected by an external observer in a surgical intensive care unit (ICU).

To facilitate the reporting and the analysis, we identified a series of events with a serial number and a colour code related to their for each of the following 5 macro-phases: ICU bed booking, admission procedures, patient stay, discharge and emergency procedures. A specific structured form including UE's code and colour, date and hour of the event and type of patient has been prepared and proposed to staff 1-week for each month after a proper phase of education. The report was voluntary and anonymous and the data collected during the morning shift from September to December 2009 have been compared to those collected from an external observer.

In the studied period, healthcare staff reported 53 UEs: 51% collected by nurses, 47% occurred during the morning shifts and 67% were classified as moderate or severe. The rate of UE in the morning shift was 12 UEs per 100 patient days. The external observer identified 55 events in 8 morning shifts with an incidence of 76 UEs per 100 patient days. The violation of isolation rules for patient with multi-drug resistant bacteria infection both by ICU staff and surgical consultant was the UE observed more frequently by the staff (26%) and by the external observer (35%).

CONCLUSIONS. The above data indicated that: 1. in our ICU the incidence of UE is very high, particularly for compliance to isolation of infected patients and 2. the spontaneous reporting system under-estimated largely the real incidence of UEs. INTRODUCTION. Importance of renal assessing in intensive care unit (ICU) patients is unquestionable for a correct drug dosing, fluid requirements or decisions for renal replacement therapies. Serum creatinine (SC) is a very common biochemical parameter in clinical practice for assessment of renal function. Many equations have been designed to estimate creatinine clearance based on SC, but their capacities for providing a correct estimate of glomerular filtration rate (GFR) are suboptimal. This is even worse in critically ill patients due to malnutrition and/or immobilization. In clinical practice, despite its limitations, 24 h-urine creatinine clearance (CrCl 24h ) is used as a reference method to determine GFR. Data show that Cystatin-C could be promising as an endogenous filtration marker in ICU settings.

OBJECTIVES. To assess in a medical ICU population whether the Arnal-Dade formula of cystatin-C clearance (CC) developed from serum cystatin-C (SCC) shows better predictive performance of GFR than SC-based formulae, as regards to patients' renal function: CrCl 24h C60 mL/min 1.73 m 2 or CrCl 24h \60 mL/min 1.73 m 2 . RESULTS. All formulae showed notable bias from the reference method. Interestingly, all equations based on SC-values clearly overestimated CrCl 24h (CG: 19.6%; MDRD: 1.1%; fv-MDRD: 27.5%), whereas CC showed underestimation of these CrCl 24h (CC: -17.5%). In the CrCl 24h C60 mL/min 1.73 m 2 group (n C60 = 146; 89 patients), CC showed the best correlation indexes (CC-CrCl 24h ; r 2 = 0.296, r = 0.544), the second most biased (-16.7%) and the worst precision (24.9%). In this group, MDRD was the least biased (-2.4%) and the most precise (21.3%). In the CrCl 24h \60 mL/min 1.73 m 2 group (n \ 60 = 95; 54 patients), CC was the worst correlated with CrCl 24h (r 2 = 0.214, r = 0.462), in contrast to MDRD (r 2 = 0.299, r = 0.547). In terms of precision, MDRD showed again better results than CC: 33.5% vs. 38.9%, respectively.

CONCLUSIONS. In our ICU population, CC did not demonstrate a clear improvement on the remainder SC-based formulae in either of the two groups according to CrCl 24h . However, in a patient with high MDRD values and suspicion of low GFR, CC could be useful as guidance before obtaining the definitive confirmation by CrCl 24h . INTRODUCTION. There are well established and robust techniques for measuring and categorizing renal function in people with chronic kidney disease (CKD). A number of rapid bedside estimates of renal function have been devised incorporating routine daily measurements, such as serum creatinine, in combination with demographic data (e.g. Cockroft-Gault, the MDRD series). The addition of serum cystatin C measurements to some equations may also improve accuracy of estimation. The current and accepted categorical classification of Acute Kidney Injury (AKI: AKIN/RIFLE) has been useful epidemiologically but does not provide a continuously variable measure of severity of AKI which would be valuable for both clinical management and research.

OBJECTIVES. Previously published abstracts have suggested a role for eGFR in describing renal function in the critically ill but a more comprehensive analysis was needed.

METHODS. 37 (21 male) (mean age 72 range 38-90) critically ill patients with AKI were recruited. A 4 h creatinine clearance ( 4 CrCl) (previously validated as a measure of renal function in critically ill patients) was measured and simultaneous blood sampling was done for creatinine, urea, albumin and cystatin C. Various equations used to estimate GFR were compared to 4 CrCl with regression and Bland-Altman analysis. All patients had a 4 CrCl of\60 mL min per 1.73 m 2 INTRODUCTION. Epithelial-mesenchymal transition (EMT), a key process in tissue development and repair, has also been identified as a major mechanism in fibrogenesis. The cytokine TGFb has been shown to induce transformation of epithelial cells into matrixforming and smooth muscle actin (SMA)-expressing myofibroblast (MF) via EMT. The other prerequisite is an injury-induced loss of intercellular contact, including adherens junctions (AJs). The classical experimental method to induce AJ disruption is the uncoupling of E-cadherin-mediated contacts by Low Calcium Medium (LCM). This concept has been termed as the two-hit model of EMT (1). b-catenin, a scaffold protein of the AJ, released by cell contact injury, can act as a transcription factor and has been shown to facilitate EMT. However, the mechanism whereby cell contact injury promotes EMT is not understood. Our recent studies have shown that Smad3, one of the main signal transducers of the TGFb pathway is a strong inhibitor of epithelial SMA expression, by interfering with myocardinrelated transcription factor (MRTF) [2] . The latter is the main driver of the SMA promoter, through it association with serum response factor (SRF). Intriguingly, b-catenin can bind to Smad3.

To clarify the mechanisms whereby AJ injury promotes SMA expression.

METHODS. AJs were manipulated in kidney tubular cells, either by siRNA-mediated downregulation of E-cadherin, b-catenin or through chemical uncoupling of AJs by LCM. Protein expression was detected by Western blotting and immunofluorescence microscopy, proteinprotein interactions were monitored by co-immunoprecipitation, and the activity of the SMA promoter was determined by luciferase reporter assays.

Knockdown of E-cadherin promoted b-catenin translocation to the nucleus and induced a threefold rise in the TGFb-triggered SMA expression. Conversely, silencing of b-catenin strongly suppressed the two-hit (TGFb + LCM)-induced activation of the SMA promoter, and inhibited SMA protein and mRNA expression by 50%. The same stimuli induced strong association of b-catenin with Smad3. Transfection of cells with a b-catenin expression vector dose-dependently prevented the inhibitory action of Smad3 on the MRTFinduced activation of SMA promoter. Moreover the active (myogenic) MRTF-SRF complex was restored, as b-catenin preempted Smad3's inhibitory effect on the complex.

These studies define a novel mechanism whereby epithelial injury activates the myogenic program, a central process in organ fibrosis. Our results imply that b-catenin, liberated from the injured AJs, facilitates the activation of the myogenic program by preventing or mitigating the inhibitory action of Smad3 on MRTF. These hitherto unknown interactions among Smad3, b-catenin and MRTF represent novel targets to lessen fibrogenesis. INTRODUCTION. In intensive care unit (ICU) patients, kidney function is monitored by the creatinine clearance (crcl). It can be measured by two methods. Urinary crcl (Ucrcl) is directly measured, using the urinary and serum creatinine. But commonly crcl is estimated from serum creatinine (Scr) alone, as estimated glomerular filtration rate (eGFR); using equations validated in chronic kidney diseases. There is paucity of literature on validation and comparison of these methods in ICU (Hoste) .

OBJECTIVES. We compared 4-h timed Ucrcl and eGFR in the newly admitted critically ill. We also sought to ascertain the incidence of high crcl and the agreement between 2 methods in this subgroup. CONCLUSION. The use of RIFLE criteria gives a high incidence of AKI in the ICU setting.

In this unselected population of critically ill pts, CysC seems to be superior to Cre in predicting pts who will develop AKI and will need RRT during their hospitalization in the ICU. Early identification of high risk patients may allow potentially beneficial therapies to be initiated early in the disease process, before irreversible injury occurs. INTRODUCTION. The contrast-induced nephropathy (CIN) is consider to be the most frecuence reason of acute renal failure in hospitalized patients. They are defined by a fixed increase (0.5 mg/dl) o a 25% rise serum creatinine level after to be exposed 24 h to the contrast. The main complications are kidney and cardiac problems and this will lead to longer hospitalization and increased mortality.

OBJECTIVES. To compare CIN occurrence after a injecting a iso-osmolar contrast (IOC, Idixanol) or a low-osmolar contrast (LOC, Iohexol) to a group of patients submitted to coronary angiography, with o without percutaneous coronary intervention (PCI). To establish unrelated CIN markers and to evaluate the efficiency of the kidney protection protocol used in our hospital. CONCLUSIONS. The LOC was associated to a greater number of CIN than IOC. Patients who developed CIN were significantly longer hospitalized. The use of point giving system that includes CIN's predictors like DM, hematocrit \39%, AMI, and treatment with diuretics helps us to classify CIN risk and use a correct kidney protection protocol.

INTRODUCTION. The incidence of Acute Renal Failure in the intensive care unit (ICU) is around 50% of cases and is related to increase in mortality in patients who required dialysis as far as 80%. Early detection of acute kidney injury (AKI), after damage is not on set could be crucial to develop therapeutic strategies to modify the course of injury. Blood and urinary concentrations of NGAL are early biomarkers of AKI 1 ; to date, little information exists regarding NGAL usefulness in critically ill patients.

OBJECTIVES. To analyze: 1. the capacity of urine NGAL (uNGAL) to predict AKIevaluated by RIFLE score-in critically ill patients and, 2. the uNGAL values in patients with SIRS, sepsis or septic shock.

METHODS. NGAL was measured in urine sample by an automatic analyzer device (ARCHITECT ci4100 Ò ; Abbott Diagnostics) at admission and 48 h later in patients admitted to a general ICU. Patients were classified both by RIFLE score at admission and 24 and 48 h later and by uNGAL concentrations at admission. To the later classification, the cut-point for AKI prediction was obtained by ROC curve analysis. uNGAL values at admission were compared in patients with SIRS, severe sepsis or septic shock. Clinicians were blinded to NGAL results.

The study included 76 consecutively-admitted patients (29 female) with mean age 59.6 ± 17.6 years, and length of ICU stay of 10.4 ± 10 days. Fifty-four SIRS, 6 severe sepsis and 16 septic shock. Thirteen patients developed RIFLE F score, 9 of them at ICU admission; extracorporeal renal therapies were required in 6 cases. When patients were classified according to their RIFLE score at 24 h of admission, uNGAL values at admission were: 26 (10-81) ng/mL in 48 patients with RIFLE 0, 163 (49-998) ng/mL in 5 with RIFLE R, 357 (151-596) ng/mL in 8 with RIFLE I and 384 (31-539) ng/mL in 10 with RIFLE F (p = 0.02). Five patients were excluded, three died before 24 h with uNGAL 816 (17-1,029) ng/mL and two were discharged before 24 h with uNGAL 182 (13-352) ng/mL. The area under ROC curve of uNGAL at admission for AKI prediction was 0.84 (95% confidence interval 0.74-0.92, p \ 0.0001), with an optimal cutoff value of 114 ng/mL with 69% sensitivity and 84% specificity. Forty-seven patients have Ungal B 114 ng/mL. uNGAL concentrations at admission were 30 (10-132) ng/mL in 54 patients with SIRS, 36 (10-235) ng/mL in 6 patients with severe sepsis and 400 (126-592) ng/mL in 16 patients with septic shock (p = 0.001).

CONCLUSIONS. Urine NGAL concentrations measured at ICU admission appeared as a useful predictor of AKI in critically ill patients; in addition, uNGAL concentrations showed an increasing pattern from SIRS to severe sepsis and septic shock. 23rd ESICM Annual Congress -Barcelona, Spain -9-13 October 2010 S403

INTRODUCTION. Two previous studies using the RIFLE criteria in intensive care patients have found the incidence of acute kidney injury (AKI) to be 36 and 67%. However, these studies used calculated basal value of creatinine in a considerable proportion of their patients, which is a possible source of error.

OBJECTIVES. The aim of this study was to investigate the incidence and severity of acute kidney injury in intensive care patients using true baseline creatinine values. OBJECTIVES. The aim of this study was to define the status of Hcy and B vitamins at admission and 7 days of ICU stay in critically ill patients, and to evaluate its relationship between them.

A prospective study was done on critically ill consecutive patients with inclusion criteria: C18 years old, SIRS and APACHE II [15. Hcy, B12 and folic plasma levels were measured by enzymoimmunoassay and enzymatic method. for B1, B2 and B6 in erythrocyte. Permission was obtained from an institutional ethical committee and written informed consent was asked.

RESULTS. At 0 and 7 days of ICU stay 50 and 53% of patients were B 1 deficient, respectively. 12 and 20% were B 12 deficient on both times, respectively. Folic levels show significant differences between 0 and 7 days of ICU stay. We found association between B 2 vitamin and Hcy at admission and 7 days. No differences were found between 0 and 7 days Hcy values. INTRODUCTION. Cytochrome P450 3A (CYP3A), the most abundantly expressed cytochrome P450 enzymes in liver, are responsible for the metabolism of over 50% of drugs used across several therapeutic classes. In adults, CYP3A is represented primarily by the major isoform, CYP3A4, and a polymorphically expressed isoform, CYP3A5. Individuals with at least one wild-type CYP3A5*1 allele synthesise functionally active enzyme while homozygotes for the *3 allele are functional non-expressers of the enzyme. The presence of functional CYP3A5 increases the hepatic metabolism of CYP3A substrates such as tacrolimus. CKD is known to reduce the hepatic metabolism of drugs via the CYP3A enzyme system and we have shown, recently, that AKI has a similar effect and that the length of time with AKI is the most important variable. We hypothesise that expression of functional CYP3A5 may reduce the impact of AKI on hepatic drug metabolism as has been shown to be the case for drug interactions with the imidazole antifungals.

METHODS. 72 (45 male) (mean age 72 range 23-90) critically ill patients with no AKI and varying degrees of severity of AKI were recruited. Midazolam concentration was measured 4 h after intravenous administration as a probe-drug for hepatic CYP3A 4/5 enzyme activity (T 4 [midazolam] ). This is a validated method for testing CYP3A activity in critically ill patients. Patients were excluded if they were on any known CYP3A4/5 inhibitors.

RESULTS. Two patients with severe AKI had unexpectedly high T 4 [midazolam] . Figure 1 demonstrates the following: Without a CYP3A5*1 allele, the rate of midazolam metabolism increased with duration of AKI (r 2 = 0.24; p \ 0.0001) (solid line). Patients who had at least one *1 allele (dashed line) were protected from the inhibitory effect that AKI has on hepatic drug metabolism (significant difference between the correlation lines p = 0.011). If the two major outliers are removed (dotted grey line) from the *3/*3 group (r 2 = 0.23; p \ 0.001), the correlation lines remain statistically different (p = 0.048).

CONCLUSIONS. The presence of an allele which codes for functional CYP3A5 protects critically ill patients from the inhibitory effect of AKI on the hepatic metabolism of midazolam. 

Thyroxine replacement therapy has become commonplace in the management of organ donors to reverse hemodynamic instability and homeostasis, yet the pharmacokinetics of thyroxine are unknown in this patient population [1, 2] . Since T3 is only available in oral form, we studied the pharmacokinetics of oral versus intravenous T4 to determine if oral administration is suitable.

OBJECTIVES. (1) to study the pharmacokinetics of oral versus iv T4 therapy;

(2) to determine if oral thryoxine therapy is suitable.

With ethics approval and signed consent from the substitute decision maker, 34 patients who were determined to be neurologically dead and consented for organ donation, were randomized to receive either an oral or intravenous dose of T4 (2 mcg/kg). All patients received an oral and iv preparation; one of which was a placebo. This study was also double blinded and randomization occurred in blocks of 4-6. Free serum levels of T4 and T3 were measured hourly until the time of organ procurement. The area under the curves (AUC) were determined and compared using.

RESULTS. There were 15 patients (13 males) in the oral versus 17 patients (8 males) in the iv group, with an average age of 61 ± 14 vs. 53 ± 17, respectively. There was no significant difference at baseline or 6 h between groups for hemodynamic variables, free T4, free T3 or TSH levels. The only exception was MAP where it was higher at baseline in the oral group and there was a significant increase at 6 h in the iv but not the oral group (77-96 vs. 91-89 in the oral). The AUC for T4 was greater for the iv group (301 pmol/l/12 h) compared to the oral group (274 pmol/l/12 h). There was no statistically significant difference in any of the levels from 0 to 12 h between the oral and iv groups. Oral bioavailability of T4 was 91%.

CONCLUSIONS. Administration of iv T4 resulted in a slightly greater AUC compared to oral administration. However, oral bioavailability of T4 in our population was very high, at 91%. T3 is currently the recommended thyroid replacement in neurologically dead organ donors. However, intravenous T3 is unavailable in many jurisdictions. IV T4 has been used as a substitute. Our study shows that in this select population, oral bioavailability is high suggesting that oral T4 may be a reasonable alternative. Further work is needed to determine whether there was a difference in the number and rate of organ retrieval in the oral versus intravenous groups. INTRODUCTION. Specific characteristics of metabolic derangements occurring in critical illness is domination of developing catabolic state particularly in acute necrotizing pancreatitis. As a result, we faced such a problem as developing a clinically apparent protein-calorie deficiency which is resistant to standard nutritional support. The treatment of acute necrotizing pancreatitis in chronic abuse patients is difficult to handle for the clinician and should include sufficient energoplastic supply.

OBJECTIVES. In our research we aimed to assess the efficacy of adding of ornithineaspartate complex in carbohydrate metabolism in chronic abuse patients with acute necrotizing pancreatitis.

METHODS. 56 comparable chronic abuse patients with acute necrotizing pancreatitis (control group N = 28, mean age 45.1 ± 8.5; Ornithine group N = 28, mean age 44.0 ± 8.1) received early parenteral nutrition from the moment of admission to hospital with universal system ''three-in-one''. Ornithine group also received ornithine-aspartate complex by parenteral administration (40 g/day). On the second day the patients were admitted parenteral nutrition and tube feeding 24 h/day. The volume of parenteral nutrition was gradually decreasing. Biochemical and metabolic endpoints were measured at baseline and on 6th day (nitrogen balance, amino acids spectrum, plasma whole protein, transferring concentrations, glucose and insulin levels) at the Clinical Laboratory

In all patients metabolic disturbances with protein status and carbohydrate metabolism shifts were revealed. Dynamic of the whole protein, albumin/protein ratio and nitrous balance in both group showed similar tendency of metabolic improvement. Dynamic of essential and nonessential amino acids concentration remained normal showing adequate energoplastic supply in both groups. Glutamine concentration in Ornithine group remained stable and even increased by the 6th day of nutritional support, while in Control group glutamine concentration was decreasing, and by the 6th day of nutritional support it was below normal values. In Ornithine group higher levels of endogenous insulin at normal values of glucose and faster Fisher index improvement were detected.

CONCLUSIONS. Administration ornithine-aspartate complex in therapy of acute necrotizing pancreatitis in chronic abuse patients, probably, may influence on disease outcome.

In Ornithine group duration of delirium tremens causes was 4 ± 1 days versus Control group (6 ± 1 days). Restoration of metabolic activities confirms adequate nutritional support in both groups but ornithine-aspartate complex adding provides faster improvement of protein and carbohydrate metabolism. OBJECTIVES. This study was designed to evaluate the nutrition indexes including serum prealbumin level as prognostic indicators of patient recovery in critically ill patients with comparing severity scoring systems.

We selected 59 patients over 16 years old, supplied with total parenteral nutrition (TPN) for more than 7 days in surgical intensive care unit, Ajou University Hospital, Suwon, Korea. The serum prealbumin, albumin levels and total lymphocyte count were measured at the first, 3rd, 5, 7, 10, 14th days of nutrition support care by TPN. We checked APACHE (acute physiology and chronic health evaluation) II score, SAPS (simplified acute physiology score), MODS (multiple organ dysfunction score) and SOFA (sequential organ failure assessment) score of patients.

RESULTS. There were 43 male patients and 16 female patients with mean age 54.3 years. The mean day of SICU staying was 13.2 days. We compared two groups; survivor group (N = 49) and non-survivor group (N = 10). There were significant statistical differences in ICU staying days (p = 0.021), APACHE II score (p \ 0.001), SAPS (p \ 0.001), MODS (p = 0.007) and SOFA score (p = 0.005) between two groups. However, serum prealbumin level (p = 0.370), albumin level (p = 0.444) and total lymphocyte level (p = 0.584) did not showed significant difference between two groups. Receiver Operating Characteristic curve showed low accuracy of serum prealbumin level as a prognostic factor (area = 0.559). Prealbumin level showed correlation with albumin (r = 0.561), however did not show correlation with APACHE II (r = -0.151), SAPS (r = -0.056), SOFA (r = -0.056) and MODS (r = -0.076).

CONCLUSIONS. Nutrition indexes including prealbumin did not correlated with clinical outcome of critically ill patients. INTRODUCTION. Physical function is impaired following critical illness [1] . Anaemia is a common complication of critical illness and has the potential to influence physical function [2] . It is not known whether anaemia affects the physical components of quality of life, the ability to carry out the activities of daily living (AODL) or the actual physical function of patients during recovery from critical illness.

To determine the physical quality of life, ability to perform activities of daily living and actual physical function in a cohort of ICU survivors dichotomised on the presence of anaemia at 3 months following ICU discharge.

one other organ failure were recruited from a general ICU population. Patients with a preexisting haematological condition were excluded. Baseline and characteristics of ICU stay were recorded. The patients were assessed with the SF-36 quality of life questionnaire (PCS), The Frenchay Activities Index (FAI) of AODL recalled for pre-morbid status and at 3 and 6 months, and the 6 min walk test (6 MWT) for actual physical function at 3 and 6 months following discharge from ICU. Organotopic measures of haemaglobin, creatinine, serum C-reactive protein and albumin concentration were also recorded. The results were dichotomised on the presence of anaemia at 3 months for statistical analysis. Baseline characteristics were compared with Student's t test. A 2 way ANOVA was performed on the PCS and FAI score as well as comparisons with t test between each time-point. The distance walked as part of the 6 MWT was compared with Mann-Whitney U test.

Patients who remained anaemic at 3 months were older, had a longer ICU stay and had a greater requirement for inotropes during their ICU stay. The PCS score of quality of life and the FAI score was significantly impaired in both groups during follow up, but there was no effect of anaemia. The results of the t tests showed that there was a significant difference between the groups at 6 months for PCS but not for FAI scores. The distances walked were severely impaired compared to the normal population (170 and 228 m at 3 and 6 months for anaemic group and 240 and 358 m for non-anaemic) in both groups was not significantly different between the two groups. The non-anaemic group did increase the distance walked significantly from 3 to 6 months. There was no difference between albumin, CRP and creatinine concentrations between the groups. METHODS. This experiment was divided into two procedures. The first procedure is to choose two kinds of cell strains, including Jurkat cell strain (comes from leukemia) and CCRF-CEM cell strain (comes from acute lymphocyte leukemia).we cultivate this two kinds of cell strains to mature stage, then inoculate every kind of cell strain into four culture dishes, two culture dishes was stimulated by 1 lg/ml lipopolysaccharide(study group), and the other two culture dishes serve as blank control(not stimulated by 1 lg/ml lipopolysaccharide). Eight hours later, we extracted the microRNA in each culture dish. The second procedure is to use the technique of gene microarray to analysis the difference expressions of microRNA. In the context of a high altitude expedition human subjects can safely be submitted to prolonged hypoxia and the resulting changes in mitochondrial function can be explored in a controlled fashion. The effect of hypoxia on immune cells-key players in the pathophysiology of sepsis-is of particular interest.

To measure mitochondrial function of monocytes during prolonged hypobaric hypoxia.

METHODS. Serial blood samples were collected and oxygen saturation was measured in twelve climbers before and throughout a high altitude climbing expedition to Pik Lenin (7,134 m). Measurements were performed at 440 m (baseline) and at the altitudes of 3,600 m (day 1), 4500 m (day 4) and 5,400 m (day 9) above sea level. Pure monocytes were isolated by the use of an antibody-antigen mediated immunomagnetic cell isolation procedure and lysed for determination of activities of mitochondrial enzymes cytochrome c oxidase and citrate synthase. Repeated measurements ANOVA followed by least significant difference (LSD) post hoc test were used to compare results on different altitudes.

Mean oxygen saturation was 91 ± 3% on 3,600 m, and decreased to 86 ± 3% on 4,500 m and 81 ± 5% on 5,400 m (p = 0.001). We observed an increase in citrate synthase activity on all altitudes compared to baseline levels (p = 0.003). Compared to the baseline, prolonged hypobaric hypoxia induced an increase in the mitochondrial respiratory chain enzyme cytochrome c oxidase enzymatic activity only at 4,500 m (p = 0.003). Normalization of cytochrome c oxidase enzymatic activity by citrate synthase activity (relative enzymatic activity) yielded a decrease in relative cytochrome c oxidase enzymatic activity during hypoxia on 3,600 and 5,400 m (Fig. 1) . Expressing cytochrome c oxidase enzymatic activities as a ratio to citrate synthase is intended to act as a safeguard for potential differences in mitochondrial enrichment.

CONCLUSIONS. The data demonstrates that prolonged hypobaric hypoxia leads to a decrease in relative cytochrome c oxidase activity. This is due to an increase in citrate synthase activity as a marker enzyme for the mitochondrial matrix representing mass and/or number of mitochondria which is not counterbalanced by a corresponding increase of cytochrome c oxidase activity. RESULTS. Glycocalyx degradation was increased in the LPS-treated animals (0.3 lm, p \ 0.05) compared to controls. Intracellular tissue NO concentrations were two-to threefold higher in the LPS-treated mice compared to controls (liver, kidney, heart, gut). The number of infiltrating MPO-positive cells increased significantly during endotoxemia. Levels of both plasma ARG and CIT were significantly lower in LPS-challenged mice than in controls, whereas plasma ornithine levels were significantly higher.

CONCLUSIONS. In this new developed murine sepsis model, the prolonged infusion of LPS resulted in increased glycocalyx degradation and associate endothelial leakage. The enhanced NO levels correlated with decreased plasma levels of ARG and CIT. Our murine model with prolonged infusion appears applicable as a model for the human clinical situation, enabling adequate investigation of the influences of the ARG-NO metabolism on endothelial dysfunction in sepsis. 

Critical illness polyneuromyopathy is a muscular weakness occurring in intensive care unit. One of the major risk factor is sepsis. An early decrease in membrane excitability was described [1] but corresponding mechanisms are imperfectly known. TNFa is released in the first time of sepsis and could be involved in the physiopathology.

OBJECTIVES. The aim of our study was to investigate TNFa effects on muscular voltage gated sodium channels (Nav) in an in vitro model.

Early effects of TNFa on Nav were analysed by macro-patch clamp on muscular fibers isolated from rat peroneus longus. Measurements were performed on control fibers and after addition of TNFa at concentrations ranging from 2.5 to 100 ng.mL -1 . The effects of chelerythrine, a specific inhibitor of protein-kinase C (PKC), were also tested. Experimentations were realised in a laboratory with permission of experimental research on animals and under the supervision of an authorized person (no 29-028).

TNFa produced a concentration-dependant inhibition of Nav currents (Fig. 1) . Maximal inhibition (75% of control current) was observed with concentrations from 50 ng mL -1 and above. This decrease was fast: 40% of maximum inhibition was observed in less than 5 min. Moreover, chelerythrine inhibited TNFa action on Nav.

CONCLUSIONS. In our experimental model, TNFa induce a rapid and concentration dependant decrease of muscular Nav currents like observed in chronic sepsis [2] . As this effect is too quick to be a transcriptional one, and as it is blocked by chelerythrine, it can be assumed that TNFa action is mediated by a Nav phosphorylation secondary to PKC activation. In conclusion we evidenced that TNFa reduce muscle excitability in the early stages of sepsis. Further studies are needed to obtain a precise description of TNFa mechanisms. may also contribute to cell signaling and regulation of the immune response. NAD(P)H oxidase in leukocytes and the vascular wall is a major regulated source of O 2 . We hypothesized that mice deficient in the p47phox (KO) component of NAD(P)H oxidase would have less pulmonary inflammation than wild type (Wt).

We treated Wt or KO mice with IV saline or LPS and assessed lung injury by: 1. wet-dry-weight ratio; 2. leak of Evans Blue (EB) labeled albumin; and 3. histological score for edema.

We used myeloperoxidase activity to indicate neutrophil (PMN) accumulation in lungs, and measured accumulation of macrophages and neutrophils in bronchial alveolar lavage (BAL). Apoptosis was assessed by tunnel staining. We also expression of ICAM-1, an adhesion molecule, and nitric oxide synthase (NOS) enzymes, eNOS and iNOS (western and northern analysis) as well as nitrotyrosine formation.

RESULTS. Lung injury was increased in both groups. Surprisingly there was greater EB leak in KO than Wt at 2 h and a greater edema score at 4 and 8 h. PMN and macrophage accumulation in BAL were the same in both groups at 2 h but greater in KO mice at 8 h. Myeloperoxidase activity was similar at 2 h post LPS in KO and Wt indicating that similar accumulation of PMN in the lungs. Apoptosis was increased in both groups at 4 h, but resolved in Wt at 24 h and persisted in KO. Nitrotyrosine was increased in both groups but appeared higher in KO. Expression of eNOS and iNOS increased in both groups but was greater in KO than Wt.

CONCLUSIONS. In contrast to our prediction, lung injury was greater in p47phox KO mice which indicates that this complex is not essential for lung injury. However, the injury was more severe and prolonged in KO mice indicating that O 2 may regulate the inflammatory response. INTRODUCTION. Septic shock remains the main cause of mortality in the ICU, thus a persistent challenge. Recently, DNA and mRNA analysis by microchip and gene expression by real time PCR highlighted 2 proteins S100A8, S100A9 and their complex, known as the calgranulins, as potential key prognostic markers for this disease: those two proteins, whose expression seems to be restrained to phagocytes cells are newly recognized components in sepsis-induced inflammation. Moreover, they were shown to be at significantly higher concentrations in the plasma of septic shock patients that were going to die. In the contrary, those who were to survive saw their plasmatic concentration decrease, all severity scores in between the 2 population being the same.

OBJECTIVES. The aim of this study was to determine the repartition of these proteins in immune cells, their intracellular variation, at baseline and after cell activation and finally to understand the relation between their intracellular and extracellular expression.

We used an in vitro model close to the immuno-inflammatory aggression that is septic shock. We stimulated in vitro for 1, 3 and 6 h whole blood from healthy volunteers using agonists found in the inflammatory storm that is septic shock (LPS, fMLP, GMCSF, IFNg). We also induced death cell, either using an apoptotic agonist, or by necrosis technics. We then analysed the intracellular variation of the calgranulins using flow cytometry technics. The extracellular quantification was made using ELISA methods. All the statistic analysis were made using a Mann-Whitney test.

We showed in this work for the first time that the intracellular repartition of the calgranulins is different depending on the type of cell: the complexe is the main form in the monocyte cytoplasma, whereas S100A9 is the main intracellular form of the PMN. This repartition remains after cell activation. We also checked the absence of calgranulins in lymphocytes. After cell activation we showed that intracellular S100A8, S100A9 and S100A8A9 increased, but at different levels depending on the cell and the agonist used. Extracellular S100A8 also raised after cell stimulation, but the concentration found were very low compare to those found in the plasma of septic shock patients.

CONCLUSIONS. Together, these results suggest a different regulation depending on the form of the protein and of the cell and thus of proper distinct function of each monomer and of the complex. In the limits of our model the increased concentrations found in the plasma of patients with a septic shock can't be explained by immune cell activation. OBJECTIVES. Although there is no specific antidote for these potent toxins, drugs like penicillin G and silibinin have been used with conflicting evidence. We successfully managed two patients with mushroom poisoning by using silibinin and NAC.

METHODS. Two members of a family, a mother 38 years old, and her son 15 years old were admitted to our ICU 36 h after the ingestion of wild mushrooms. They presented with abdominal cramps, vomiting, profuse diarrhea ([20/day), myalgias, confusion and agitation.

The clinical examination showed severe dehydration, tachycardia, oliguria with grade I-II hepatic encephalopathy. Laboratory exams revealed elevation of liver enzymes SGPT: 1106/880 U/L, SGOT: 1884/930 U/L. Coagulation parameters were as following: Prothrombin time 15.5 00 /15.1 00 , factor V \55%/60%, factor VII \50/55%. High ammonia levels were noted, reaching 360 and 400 ng/dl, respectively. Metabolic acidosis was also present with mild renal dysfunction. The ultrasound performed in both patients showed hepatosplenomegaly. Aggressive fluid and electrolyte replacement started upon admission. Silibinin was given at a dose of 20 mg/kg/day intravenously, in four divided doses, for three consecutive days, while NAC was given as a continuous infusion at a dose of 150 mg/kg for the first hour, 50 mg/kg for the next 4 h, and thereafter 100 mg/kg/day for the following four days. Hepatic encephalopathy, mild jaundice and renal dysfunction resolved within 48 h, and liver function tests returned to normal within 4 days. The patients recovered fully and were discharged to a medical ward.

Recent experimental and clinical studies have shown a strong protective and antioxidant effect against hepatic cell injury in Amanita toxicity by the administration of NAC and silibinin, either as monotherapy or as a combination therapy. Although further clinical research is required to confirm their efficacy in reducing mortality and transplantation rate, NAC has been used in our ICU in hepatic dysfunction of different etiologies with promising results. 

We have recently shown that in patients with lactic acidosis due to metformin intoxication (serum drug level = 61 ± 25 lg/ml; therapeutic level is B4 lg/ml) systemic oxygen consumption (VO 2 ) can be abnormally low despite a preserved global oxygen delivery (DO 2 ) (1). The study, however, suffered from being retrospective.

OBJECTIVES. To prospectively clarify whether metformin primarily impairs VO 2 .

METHODS. Eight sedated, paralyzed and mechanically ventilated pigs received a continuous i.v. infusion of metformin, at a rate of 1.6 g/h. The amount of metformin administered to each animal ranged from 4 and 8 g. The experiment always finished 9 h after the initiation of drug infusion. Use of sedative and neuromuscular blocking drugs, as well as ventilatory setting, were always kept constant. Serum metformin concentration was measured at the end of the experiment, using High Performance Liquid Chromatography (HPLC). Arterial pH, lactatemia, VO 2 (indirect calorimetry) and DO 2 (computed from cardiac output measured by pulmonary artery thermodilution) were recorded hourly. Data are presented as mean ± SD. Statistical testing was performed using the one-way repeated measure ANOVA and the linear regression analysis.

Metformin infusion produced toxic serum drug levels (77 ± 30 lg/ml; n = 5). Arterial pH drop from 7.63 ± 0.05 (prior to infusion) to 6.96 ± 0.26 (end of the experiment) (n = 8; p \ 0.001) and lactatemia rose from 1 ± 0 to 17 ± 7 mmol/L (n = 8, p \ 0.001). VO 2 progressively decreased (from 117 ± 37 to 66 ± 30 ml/min; n = 8, p \ 0.001) while DO 2 did not significantly change over time (from 274 ± 74 to 239 ± 66 ml/min; n = 7, p = 0.39). The decrease in VO 2 was proportional to the dose of metformin administered (R 2 0.62; n = 8, p = 0.02) and to the serum drug level reached by the end of the experiment (R 2 0.69; n = 5, p = 0.08).

CONCLUSIONS. Lactic acidosis develops during metformin intoxication in the presence of a diminished VO 2 but in the absence of any clear evidence of inadequate DO 2 . This finding suggests that impaired oxygen utilization, rather than availability, may have a role in the pathogenesis of metformin-induced lactic acidosis. : 180 min) . Death was consequent to multiorgan failure, anoxic encephalopathy or capillary leak syndrome if ECLS was performed under cardiac massage. Four patients presented with documented brain death, allowing organ donation in 2 cases. Among these patients, the heart of one flecainide-poisoned patient was successfully transplanted, after normalization of ECG and myocardial function as well as toxicant elimination under ECLS. Prognostic factors in ECLS-treated poisoned patients were as follows: QRS enlargement on admission (p = 0.009), SAPS II score on admission (p = 0.005), ECLS performance under massage (p = 0.008), arterial pH (p \ 0.001), lactate concentration (10.7 [6.6-19.6] versus 15.0 mmol/l [6.2-29.5], p = 0.003), as well as red cell (p = 0.008), fresh plasma (p = 0.003), and platelet (p = 0.03) transfusions within the first 24 h.

CONCLUSIONS. To our knowledge, this is the larger series of ECLS-treated poisoned patients ever reported. ECLS appears to be an efficient salvage technique in case of refractory toxic cardiac failure or arrest, with a 28% survival rate. Our series clearly demonstrate that toxic refractory cardiac failure remains the best indication with a 46% survival rate. OBJECTIVES. Aim of the study was to investigate the incidence of infections in patients treated with hypothermia while receiving SDD.

In this retrospective case control study 35 patients treated with prolonged hypothermia (cases) were identified and 169 patients with severe brain injury were included (controls). Propensity score matching was performed to correct for differences in baseline characteristics and clinical parameters. Primary outcome was the incidence of infection. The secondary endpoints were the micro-organisms isolated from surveillance cultures and during infection.

The demographic and clinical data indicated that the cases and controls were well matched. The length of stay in the ICU and duration of mechanical ventilation were comparable between the groups. The overall risk of infection during ICU stay was 20% in the hypothermia groups versus 34.4% in the normothermia group (p = 0.388). Pneumonia was diagnosed in 11.4% of patients in both groups (p = 1.000). The incidence of meningitis, wound infection, bacteremia, and urinary tract infection was low and comparable between the groups. Staphylococcus aureus was most frequently identified as the causative infectious microorganism in both the hypothermia (14.3%) and normothermia (36.3%) group (p = 0.375), followed by coagulase negative staphylococci (14.3% in the hypothermia and 27.3% in the normothermia patients, p = 0.625) Gram-negative bacteria were isolated from the surveillance cultures in 51.4% of patients treated with hypothermia and 31.4% of patients in the control group (p = 0.143). Colonization of the rectum with gram-negative bacteria was significantly more frequent in patients treated with hypothermia compared with normothermia (48.6 vs. 20.0% respectively, p = 0.041). In contrast, colonization of the upper gastrointestinal tract and sputum was comparable between the groups with an incidence of 14.3% in the hypothermia patients versus 11.4% in the normothermia patients (p = 1.0).

Use of SDD mitigates the increased risk of infection in patients treated with hypothermia. Based on the surveillance cultures, it seems that oropharyngeal decontamination is the most effective part of the SDD regimen in the prevention of pneumonia.

INTRODUCTION. Prognostic scores specific for critical patients were developed in order to predict mortality based on physiologic and laboratorial variables. On the other hand, specific scores for burn patients are calculated taking into consideration inhalation injury, age and total burned surface area (TBSA), among others. However, scores utilized in general ICU have not been evaluated in burn patients.

OBJECTIVES. Therefore, the aim of the present work was to validate APACHE II, SAPS 3 as well as initial SOFA in a population of patients with massive burn. These scores were compared to some specific burn patient scores, including ABSI (Abbreviated Burn Severity Index) and Estimates of the Probability of Death.

Retrospective study employing data collected prospectively from May 2005 to February 2010 (58 months) at an ICU specialized in burn patients at a teaching hospital which is considered a reference centre in trauma care. All patients admitted during this period were included.

One hundred and fifty-four consecutive patients were studied (male: 73%; female: 27%), with averaged age of 37.7 ± 15.9 years and a hospital stay of 33.8 ± 28.4 days. Mortality rate of our sample was 43.5%. Incidence of inhalation injury was 45% and total burn surface area (TBSA) was the following: 35.5% of patients had 20% or less; 36.1% had 20-40% of TBSA whereas 28.4% showed 40% or more.

Area under curve of Receiver Operating Characteristic (ROC) of evaluated indexes is displayed on Table 1 . 

Computerized head tomography is routinely performed as a diagnostic tool after the occurrence of neurologic deterioration in the ICU adult patients. However, the CT findings in this setting are rarely reported. We hypothesized that the analysis of a series of cranial CTs would help to understand the neurologic conditions of the critically ill patients and improve their management.

OBJECTIVES. To analyze, over a three-month period, the head CT scans performed in the Adult ICU in the Albert Einstein Hospital in São Paulo, Brazil.

METHODS. All cranial CTs performed in the ICU patients during the studied period were analyzed by two radiologists from the Albert Einstein Hospital staff From May 1st to August 31st, 2009, according to a pre-established protocol: 1. presence of acute cerebral ischemia; 2. presence of previous cerebral ischemia; 3 . presence of acute cerebral hemorrhage; 4. presence of cerebral edema; 5. cerebral aneurisms; 6. cerebral tumors and 7. normal cerebral tomography.

We studied 69 CT scans from 34 (49.2%) males and 35 (50.8%) females, mean age 64.43 ± 21.7 years. The head CT findings were the following: (1) presence of acute cerebral ischemia = 10 (14.4%); (2) presence of previous cerebral ischemia = 17 (24.6%);

(3) presence of acute cerebral hemorrhage = 27 (39.1%); (4) presence of cerebral edema = 7 (10.1%); (5) cerebral aneurisms = 2 (2.8%); (6) cerebral tumors = 2 (2.8%) and (7) C16 years C 18, ABC (assessment blood consumption) CP: C2 and ETS (emergency transfusion score) CP: C3, C4 years C6. These scales handle the following combinations of variables for calculation: age, sex, type of admission, mechanism, blood pressure, focussed assessment for the sonography of trauma, hemoglobin, orthopedic or pelvic trauma, heart rate. MT was defined as the transfusion of 10 units or more of packed red blood cells in the first 24 h. We study the sensitivity (S), specificity (SP), positive and negative predictive value (PPV, NPV), likelihood ratios positive and negative (LHR+ , LHR-) and area under the receiver operating characteristic curve (AUROC) of different scales for the predictive power of TM validated in the literature. 568 patients were available for analysis (77.6% men, ISS 30 ± 13, blunt trauma 93.8% OBJECTIVES. We measured patient-reported outcome following surgical management with DC using a quality of life instrument. METHODS. Survivors discharged between 3 and 18 months after severe TAT were contacted after obtaining approval by our institutional IRB. We excluded patients with neurotrauma. We applied self-response version EuroQoL questionnaire (EQ-5D) and visual analog scale (EQ-VAS: 0 (worst health)-100 (best health). EuroQoL it is based on a descriptive system that defines health in terms of 5 dimensions: mobility, self-care, usualactivities, pain/discomfort and anxiety/depression. Each dimension has 3 levels of response: no problems (level 1), some problems (level 2) severe problems (level 3). RESULTS. Thirty four patients were contacted. Mean ± SD age was 31.8 ± 11.6 yrs, Male were 88.2% and penetrating trauma occurred in 79.4%. Mean ± SD in severity scores were: ATI 24.4 ± 9.6, ISS 28.1 ± 8.5 and APACHE II 20 ± 6. The median time from discharge was 12 months (IQR 6-15 months). The EQ-5D dimensions in which the largest proportion of patients reported severe problems were usual-activities (work, study) and pain/discomfort 14.7% and 5.9% respectively as shown in the CONCLUSIONS. Survivors of severe trauma and DC, reported acceptable quality of life with minimal limitations with social functioning. A prospective study should assess Quality of Life in these patients from hospital discharge and systematically over time.

INTRODUCTION. Brain tumors surgery is one of the main causes of admittance to the NICU. It is important to know the risk factors associated to hospital mortality of patients admitted to NICU due to this reason.

To identify perioperative factors associated to higher hospital mortality in a series of patients admitted to NICU immediately after a BT elective resection.

METHODS. Data of 101 patients operated for BT elective resection and consecutively admitted to NICU at IMSS UMAE 1 Bajío were prospectively obtained. NICU BT database includes 269 perioperative items. We divided the series in two groups: surviving and deceased patients. Then, we analyzed the perioperative behavior differences between both groups. Either Student's t test or Chi-square test was used, as it corresponded, for the analysis of differences observed between both groups. Values of p lower than 0.05 were considered significant.

RESULTS. The hospital mortality observed in this series of 101 patients was 8.9% (9/101). Data of the nine variables showing significant differences between surviving and deceased patients groups are shown in Table 1 . 

Even if hypoxic brain injury has been reported as the strongest factor affecting the poor outcome of near-drowning patients, little has been known about prognostic factors affecting the outcomes of those patients receiving mechanical ventilation.

To define prognostic factors affecting the outcomes of patients mechanically ventilated after near-drowning. , white blood cell counts (OR, 1.00; 95% CI, 1.00-1.00; P = 0.001), serum creatinine (OR, 0.19; 95% CI, 0.06-0.67; P = 0.010), and serum lactic acid (OR, 0.95; 95% CI, 0.93-0.98, P = 0.002) were associated with favorable outcomes, respectively. However, only higher body temperature as a clinical parameter and the level of serum lactic acid as a laboratory parameter were significant predictors of favorable outcomes in multivariate analyses; the OR were 2.87 (95% CI, 1.01-8.14; P = 0.048) and 0.96 (95% CI, 0.93-0.99; P = 0.011), respectively.

CONCLUSIONS. Initial body temperature and the level of serum lactic acid were two most important clinical and laboratory prognostic factor in nearly drowned patients. The outcomes were not affected by the degree of initial hypoxemia. 

To determine the use of automated external defibrillators (AED) and manual defibrillators deployed in the various hospital wards (unmonitored areas) in a university hospital.

A prospective study was performed according to Utstein style of all cardiac arrests occurred in the hospital during the first 6 months after the implantation of a new protocol of care for hospital cardiac arrest. Because of this plan automated external defibrillators were located for hospital wards and common service areas (radiology areas, outpatients, …) where one would expect a lower incidence of cardiac arrests, according to the risk map elaborated previously. In areas of greatest risk manual defibrillators previously existed. All resuscitation attempts in these areas were analyzed, excluding the emergency department because of a separate protocol against the rest of the hospital. Special attention was given to the use of AEDs by wards staff before the arrival of resuscitation team. Also a comprehensive volunteer training program was designed, but it began after the analyzed period was finished.

RESULTS. During the first 6 months we collected a total of 20 PCR in hospital wards and public areas, with a median age of 71 years and predominantly male (12 patients). The most common origin was respiratory (13 patients) followed by cardiac (4 patients). The most frequent rhythm detected was non-shockable (17 patients), only in 2 was shockable and unknown in 1. Before the resuscitation team arrival only two patients had been manually defibrillated and were never used the new AEDs.

CONCLUSIONS. The AEDs provided in the hospital were completely useless in the first months after placement, probably due to the lack of a comprehensive training plan associated to the population goal. METHODS. descriptive longitudinal Study. 208 patients were studied by encephalic Death, as potential donors of organs, alerted to the Network of regional Transplant 5 (CDTOT), by units of intensive care, for 36 months, in Barranquilla's city. It was applied qualifying each of the variables in agreement to the vital opposing signs and biochemical tests brought in this moment.

. 78.3% of the subjects were male; the average of age was 37.34 years (±sd: 16.24). The values of blood sugar, sodium, osmolaridad, tonicidad, PO2, FC, PAM, and Glasgow, determined a score of 22, qualification that there had patient with encephalic death with the scale MBCM, as a test of certainty of the scale to diagnose encephalic death in total absence of reflections of stem.

CONCLUSIONS. there is recommended the application of MBCM's scale to every neurological patient by diagnosis of encephalic death in proof of certainty, in absence of others. By the high specificity of the already demonstrated scale there is recommended that scores lower than 19 they should restate the qualification. A score of 22 is an encephalic death in absence of reflections of stem.

GRANT ACKNOWLEDGMENT. Clínica General del Norte-CDTOT INTRODUCTION. Prospective analysis of tracheostomies performed in patients admitted to a neurotrauma ICU, the reasons for its implementation, and intraoperative complications in the first week.

METHODS. All patients admitted to the ICU of neurotrauma, which underwent a tracheostomy after admission. Data were collected: affiliation, Cause of admission, average stay, cause for realization of tracheostomy, tracheostomy time delay from its indication, place of performance of the procedure (ICU or operating room), perioperative complications (event at transfer to operating room or during surgery: hypoxia, hypotension, arrhythmia, bleeding, premature extubation, false cannulation, cardiac arrest, pneumothorax or death), and postoperative complications in the first week (bleeding, difficulty in changing cannula, stomal infection, pneumothorax, death). INTRODUCTION. The S-100B protein is a brain-specific protein release from astroglial cells into the circulation after Traumatic Brain Injury (TBI). Researches indicate that the S-100B serum level could be a useful indicator of TBI severity, however there is not evidence enough about the role of S-100B in nonsevere Head Trauma. The hypothesis that S-100B is a useful screening tool to detect brain injury in patients with a normal level of consciousness after a Head Trauma was tested.

A total of 115 patients with the diagnosis of Mild TBI without decrease of consciousness (according to the GCS) with at least one neurological symptom or finding like amnesia, headache, dizziness, convulsion and vomits, were prospectively included. We recorded the clinical data on admission and a blood sample before 6 h after TBI, for S-100B inmunoluminescence analysis. A routine cranial computed tomography scan (CT) was obtained within 24 h after the injury (categorized in normal or pathological). The diagnostic properties of S-100B serum levels.105 lg/L, for prediction of intracranial lesions revealed by CT were tested with receiver operating characteristic (ROC) analysis.

Seventy of the patients (60.9%) were men, with a mean (SD) age of 49.07 (20.65) years (range, 14-92 years). A total of 11 patients (9.6%) had intracranial lesions. Serum S-100b levels were significantly higher in patients with intracranial lesions than in the remaining patients. The average value of the protein in patients without intracranial lesion was 0.315 lg/L with a CI 95% (0.265-0.365 lg/L), and in those with pathological findings in CT was 0.601 lg/L with a CI 95% (0.403-0.798 lg/L). Significant differences were found between levels of S100b protein and the presence of pathological findings in the CT (p = 0.001) (Fig. 1) . The ROC curve analysis showed that S100b protein is a useful tool to discriminate the presence of intracranial injury in CT (AUC, 0.76, 95% CI, 0.685-0.907, p \ 0.001). S100B analyses with a cut-off level of 0.105 lg/L showed a sensitivity 100% but a specificity 19.2%. We evaluated different cut off values and in our series, the best cut off of the S100b protein is at 0.253 lg/L with a sensitivity of 100% and specificity 58%. (Fig. 2) CONCLUSION. Determination of serum protein S-100b is a useful biochemical indicator of brain damage in head trauma. Our results show that an increase in the cut-off point of S-100b to 0.253 lg/L increases its accuracy in the prediction of the existence of macroscopical lesions.

KEY WORDS. Protein S-100B, brain injury, minor head trauma, cranial computed tomography.

Critically ill patients with systemic inflammatory response syndrome frequently suffer muscle weakness due to critical illness myopathy (CIM) and polyneuropathy (CIP). Several in vitro studies have shown that the cause of muscle weakness is a loss of membrane excitability accompanied by membrane depolarization [1] .

OBJECTIVES. We investigated membrane polarization and excitability parameters in muscle and motor nerve in vivo within the first week after intensive care unit (ICU) admission.

METHODS. The study was approved by our local ethics committee. Patients with SOFA scores C8 on 3 consecutive days underwent nerve conduction studies including direct muscle stimulation to categorize patients as ICU-control, CIM-(dmCMAP \3 mV) and/or CIPpatients (reduced SNAP amplitude) within the first 8 days after ICU admission. To assess excitability parameters we recorded stimulus-response behaviour, threshold electrotonus, current-threshold relationship and recovery cycle from abductor pollicis brevis muscle following stimulation of the median nerve [3] . Data are shown as median and 25%/75% percentile. CONCLUSIONS. We describe for the first time that critically ill patients in general show muscle-and nerve membrane depolarization, whereas patients later suffering from muscle weakness due to CIM or CIM/CIP feature additionally reduced membrane excitability. This suggests that membrane depolarization in critically ill patients is caused by energy failure leading to dysfunction of the Na-K pump, the motor of membrane repolarisation-whereas reduced membrane excitability in CIM or CIM/CIP needs an additional dysfunction of voltage gated sodium channels for example occurring in the presence of endotoxins [2] .

In intensive care patients with central nervous system (CNS) disease, the systemic inflammatory response syndrome (SIRS) criteria are often unreliable as a basis for identifying the inflammatory process. Even with the presence of some infection they could be signs of the diencephalons-catabolic syndrome. Diencephalons-catabolic syndrome like SIRS constitutes of hyperthermia over 38°C, tachypnea of over 20 per minute, tachycardia, and arterial hypertension. Thus, SIRS symptoms may occur after antibacterial treatment even if there is no infection or inflammation. We suggest a more precise method which could help to avoid the excessive antibacterial therapy and to control it in patients with CNS disease-a procalcitonin test.

OBJECTIVES. Reduce the use of wide specter antibiotics makes the control over antibacterial therapy in patients with CNS diseases more precise; reduce the number of complications related to unnecessarily long antibacterial treatment.

After obtaining the informed concern, in our investigation we included 30 patients with different neurological disorders, who had recently transferred neurosurgical operations. All of them demonstrated SIRS symptoms on different postoperative terms. When SIRS symptoms occurred, we checked the level of procalcitonin in the patient's serum by a semi quantitative method on a disposable BRAHMS PCT-Q system. The procalcitonin level was determined against a color scale. Procalcitonin level over 0.5 ng/ml (16 patients) considered a sign of infection and in such cases we prescribed antibacterial treatment 2,000 mg of Selenase for 5-7 days. If the test result was negative (14 patients) we repeated it in 24 h and in cases with the same results, no antibacterial treatment was administered even if there were SIRS symptoms. If PCT-Q test was negative patients were sedated (fentanyl 0.8-1.5 lg/kg/h and clonidine 0.2-0.5 lg/kg/h) to achieve autonomic stability and attenuate clinical manifestation of SIRS.

We had not observed any cases of sepsis in both groups of patients. By mince of PCT, we had managed to reduce the quantity of wide specter antibiotics, used in neurosurgical patients for 34.7%.

CONCLUSIONS. Procalcitonin test in neurosurgical clinic let us determine the necessity of antibacterial treatment reduce the use of wide specter antibiotics, medical costs and prevent the forming of polyresistant infection.

L. Combe 1 , R. Appleton 1 , C. Gilhooly 1 , J. Kinsella 1 1 University of Glasgow, Department of Anaesthesia and Critical Care, Glasgow, UK

Intensive Care Unit-Acquired Weakness (ICUAW) is increasingly recognised as a common complication of critical illness with potentially prolonged debilitating sequelae. The estimated incidence is 46% in patients with sepsis, multi-organ failure or prolonged mechanical ventilation [1] and suggested risk factors include: the systemic inflammatory response syndrome (SIRS), sepsis, higher severity of illness, hyperglycaemia, renal replacement therapy and parenteral nutrition.

OBJECTIVES. The aims of this study were to determine the incidence, risk factors and outcomes for patients diagnosed with ICUAW in Glasgow Royal Infirmary's (GRI) ICU.

The study was undertaken in two parts, firstly as a case-control study [matched for age (within 5 years), sex and admission APACHE II score (within 5 points)] and secondly by comparing identified cases of ICUAW to a 3-month cross-sectional sample (1/10/09-31/12/09, 112 patients) of GRI's ICU patients. Data for both parts of the study was obtained from two electronic databases, WardWatcher and CareVue. CareVue was searched to identify patients with ICUAW and WardWatcher was used to identify the controls. Data collected included: patient and illness characteristics, severity of illness scoring, organ support and treatments provided, laboratory results and outcomes. Minitab software was used for statistical analysis. CONCLUSIONS. The incidence of ICUAW was very low, we hypothesise this to be explained by the absence of systematic evaluation of patients for ICUAW. The risk factors and outcomes for ICUAW were consistent with some of the published literature. Prospective study is now planned to systematically evaluate this condition. With increasing age, comorbidity, and socioeconomic deprivation being associated with higher risk pregnancies, there comes a potential higher risk of complications. Neurological and neurosurgical complications, which can be particularly devastating during the peripartum period, include those due to medical conditions of pregnancy (hypertensive disease, sepsis, thromboembolic disease, hypoxic-ischaemic brain injury), iatrogenic complications secondary to anaesthetic or obstetric interventions, incidental illness or injury (pharmacological alterations, trauma, tumour), and deliberate self-harm and violence.

OBJECTIVES. To ascertain the frequency of neurocritical care admissions in the West of Scotland, the nature of the admission diagnoses, the impact they have on our service (length of stay), and maternal and foetal outcome.

METHODS. Using the Scottish Intensive Care Society Audit Group WardWatcher patient database, female patients aged 14-50 years old who were admitted to the neurocritical care unit were identified (January 2008-December 2009). We manually reviewed the electronic admission note for each of these women in order to gain diagnoses; a targeted case note analysis ensued.

Within the 24 month study period there were a total of 915 admissions to neurocritical care, of whom 164 fulfilled the age and gender criteria; 6 admissions (0.66% of total) were for neurological complications in the peripartum period.

The age range was 18 to 45 years (median 31 years). Three women (50%) were intrapartum (20-36 weeks gestation) at the time of their admission, and three were postpartum (1 day-7 months). Half of admissions were due to incidental illness or injury, a third to pregnancyrelated medical complications, and one case was iatrogenic in nature. Length of stay in ICU was 1 to 12 days (median 4.3 days). One patient sustained a residual facial nerve weakness and deafness.

CONCLUSIONS. This survey provided insight into the incidence and nature of pregnancyrelated pathology requiring acute referral to a regional neurosciences centre. As highlighted in other surveys, there may be many more peripartum patients with neurological complications who are cared for in general critical care units, and do not require admission to a tertiary referral centre [1] . Further work is underway to ascertain the true numbers of neurological complications of pregnancy countrywide. Our approach represents a paradigm for the continuing audit of pregnancy-related critical care resource use in Scotland. INTRODUCTION. Hypertonic saline has an osmotic effect on the brain because of its high tonicity and ability to effectively remain outside the blood-brain barrier. There may be a minimal benefit in restoring cerebral blood flow, which is thought to be mitigated through local effects of hypertonic saline on cerebral microvasculature. Most comparisons with mannitol suggest almost equal efficacy in reducing ICP but not compared their effects on EEG.

OBJECTIVES. We aimed to compare the effects of 20% mannitol, 3% or 7% hypertonic saline on hemodynamic parameters, intracranial pressure and electroencephalography in experimental head trauma.

Bilateral craniotomy were carried out in the parietal region and head trauma was applied for all rabbits. The rabbits were randomly divided into four groups. In group I rabbits were only observed. In group II: 20% mannitol, in group III: 3% hypertonic saline and in group IV: 7% hypertonic saline was administered intravenously to achieve similar osmolar load. Electroencephalography, mean arterial pressure, heart rate, intracranial pressure were recorded before trauma and 5 and 60 min after trauma.

RESULTS. Increased intracranial pressure was significantly decreased by mannitol, 3 and 7% hypertonic saline solutions at the end of study (p \ 0.05). But intracranial pressure values of mannitol and 7% hypertonic saline groups were lower than the other groups (p \ 0.05). The electroencephalography scores decreased after trauma in all groups (p \ 0.05). At end of the study, 3 and 7% hypertonic saline groups had similar electroencephalography scores with pretrauma scores (p [ 0.05). The mean arterial pressure and heart rates increased after trauma in all groups (p \ 0.05). Mean arterial pressure values were found lower only in mannitol group at end of the study (p \ 0.05).

Our study showed that when used in intracranial hypertension treatment, 7% hypertonic saline solution is as effective as mannitol, and preserves hemodynamic parameters, and normalizes traumatic electroencephalography abnormalities better than mannitol. OBJECTIVES. To identify the causes of new onset seizures in patient admitted in medical ICU.

METHODS. All the patient admitted in ICU and who had new onset seizures were evaluated. The patients were evaluated for metabolic profile. Imaging (CT/MRI) was done whenever needed. Patients with preexisting seizure history were excluded from study.

. 44 (29 males, 15 females) patients, who had first seizure during hospitalization in ICU were included. 43 patients had generalised and one had focal seizures. 28 patients had metabolic abnormalities. 13 (46.42%) had evidence of hepatic encephalopathy. 3 (11.36%) had only hepatic encephalopathy while rest had associated uremia, hyponatraemia, hypophosphatemia and hypomagnesemia. Out of 10 patients, who had renal failure, 4 had evidence of uremia while rest had associated hyponatraemia or hypophosphatemia. Only one patient had evidence of hypocalcemia. Imaging was done in 16 patients. 14 (30%) had abnormal CT scan results. 4 (9.09%) had intracranial hemorrhage, 4 (9.09%) had infarct, 2 (4.55%) had brain metastasis, 2 had evidence of hydrocephalus and one each had evidence of extradural hemorrhage and tuberculoma. CSF analysis was done in 7 (15.9%) patients. 3 (6.8%) had evidence of tuberculosis and 1 (2.27%) had evidence of pyogenic infection. To study the role of various investigations and CT in evaluating these patients.

All patients admitted with new onset seizures within 72 h prior to presentation were included. All the patients were questioned and an attempt was made to assign an electroclinical syndrome to seizure. Patients were evaluated for metabolic profile, neuroimaging. CSF examination was done in those who had persistently altered mental status, infectious symptoms and fever.

RESULTS. 110 patients were admitted (1.03% of total patients who came to emergency) with history of new onset seizures. 71.8% patients were diagnosed to have acute symptomatic seizures and were placed in ILAE category 4.13 and three patients were placed in ILAE category of remote symptomatic seizures. The cause of seizures was established in 82 (74.5%) patients and remained unestablished in 28 (25.5%) patients. 14 (12.7%) patients were diagnosed to have neurocysticercosis. Other important causes were acute infarct, uremia, hyponatremia, hypernatremia, viral encephalitis, post partum eclampsia, pyogenic and tubercular meningitis. Alcohol withdrawal seizures were seen in 3.6% patients. Metabolic derangements were seen in 19 (17.3%) patients. Computed tomography was done in 88 patients and 50% had abnormal findings. MRI was done in 13 patients and 12 had abnormalities.

CONCLUSIONS. Neurocysticercosis was found to be most common cause of seizure activity in our part of country. Though Metabolic derangement can cause significant proportion of new onset seizure patients routine imaging of brain should be performed in patients with new onset seizures.

Work environment and organisational issues: 1335-1347   1335   SUBJECTIVE AND OBJECTIVE RESEARCH INTO THE WORKING CONDITIONS  AND THEIR EFFECT ON THE HEALTH AND SAFETY OF PEOPLE WORKING IN  ICU, FOCUSING MAINLY ON THE NATURAL FACTORS OF TEMPERATURE,  HUMIDITY, VENTILATION, LIGHTING AND NOISE (PART 1) N. Karachalios 1 , E.C. Katsilaki 1 , D. Sfyras 1 1 General Hospital of Lamia, ICU, Lamia, Greece

The aim of the project is the subjective and objective investigation of the conditions of work and the relation repercussions on the health and safety of people working in the ICU, focusing mainly on the natural factors that are likely to cause the Sick Building Syndrome. For this purpose a protocol of research in two phases has been planned. The first included objective measurements, with the use of suitable equipment, of the natural factors of temperature, humidity, ventilation, lighting and noise. The second phase included the subjective estimation of the working people about their own health and conditions of their work, in the particular area of the Hospital with the use of substantiated anonymous questionnaire. After the subjective and objective study and analysis of 32 questionnaires and 55 measurements of natural factors, we found that the medium temperature of the ICU was 24°C. The mean relative humidity of the ICU was 38% (highest 38.3% and lowest 37.8%). The mean ventilation rate of the ICU was 17 m 3 /h (highest 28.8 and lowest \7.2 m 3 /h). The mean sound pressure was 64.8 dB (highest 75 and lowest 63.5 dB). The average lighting was 331.5 Lux (160 Lux lowest and 598 Lux highest). The objective data seem to keep pace with the subjective opinions of the working people, as they were impressed in the questionnaires of subjective estimate. The objective data were compared with the subjective. The results of the research were also compared with data from the existing bibliography and current legislation, leading to a line of conclusions. (1) Insufficient and bad quality ventilation. (2) The existing temperature of the environment contributes to the appearance of Sick Building Syndrome. (3) The working environment is noisy. (4) The environment of work has problematic or insufficient lighting. (5) The ICU under study is a building area which can be characterized as ''sick'' if immediate action is not taken.

BACKGROUND. Up to 75% of critical care nurses test positive for (symptoms of) Post Traumatic Stress Disorder (PTSD) [1, 2, 3] . It is assumed that these symptoms are caused by professional involvement in life-threatening events [3] . In a sample of intensive care nurses, we investigated which work related incidents were perceived as most distressing.

METHOD. In interviews, 12 nurses (75% female) were asked to memorize and tell about their most traumatic work related event. All interviews were recorded. After verbatim transcription, the 'most critical events' were extracted and categorized bij two independent psychologists.

. None of the nurses reported major life-threatening events such as trauma-related injuries, massive bleeding or seeing patients die as their 'most critical incident'.

CONCLUSION. Not the major life-threatening events but relatively 'normal work related events' under unusual circumstances are mentioned as most critical by nurses. In contrast to major life-threatening events, these 'normal events' are usually underestimated by colleagues, and thus potentially compromise peer-support. 

A Care Bundle refers to evidence based interventions and information grouped together to improve outcomes and consistency of provided care [1, 2] . At the ICUs charge nurses and intensivists as shift leaders are responsible for daily management of unit activities. Several immediately made decisions by shift leaders are made under time pressure and high information load with inadequate information. Though we have evidence of structure and process based factors such as material and human resources, admission and discharge decisions or bed utilization, the support for information transfer and integration is poor in organizational decision-making concerning these factors.

OBJECTIVES. To identify immediate information needs of charge nurses and intensivists during the management of daily activities at the ICU and evaluate how necessary this information is for their decision-making.

From September 2009 to November 2009, all charge nurses (n = 515) and intensivists (n = 223) of 17 university affiliated ICUs providing comprehensive care in Finland were surveyed with an on-line questionnaire using 122 statements. The questionnaire was developed based on our previous observation study and statements of our survey regarded information needs related to the ICU care activities. A rating scale from 0 to 10 (completely unnecessary-absolutely necessary) was used to assess the necessity of the information. For each statement, a response with mean 7 or over was regarded as necessary information for immediate decisions.

RESULTS. The response rate was 47.97% (charge nurses 50.1%, intensivists 43.1%). The working experience varied from 0 to 35 years (Mean 13.6, SD 7.9). Over 50% of respondents worked as a shift leader once a week or more often. 72 statements of 122 were valued as a necessary (Mean [7 or more) for immediate decision-making. Absolutely necessary information (Mean [9 or more) for immediate decision-making were assessed related to the 11 statements. These statements concerned isolations, mechanical ventilation, admissions and discharges, special treatments, patient's condition, and scheduled dates or times for surgery or other procedures.

CONCLUSIONS. Both ICU charge nurses and intensivists identified several information needs that are crucial for immediate decision-making during the whole ICU care process. Information needs of the shift leaders differed and they were strongly connected to the needs of one's professional requirements. An integrated overview and summarization of immediately needed information-a Care Bundle for organizational decision-making-at the ICUs is highly needed for ICU shift leaders. The common interests of both professionals, charge nurses and intensivists, should be emphasized when new technology-based systems are developed. BACKGROUND. The nursing shortage is an international problem that is expected to worsen in the coming years. Studies show that one of the main reasons nurses leave the profession is their dissatisfaction with their work environment. Structural empowerment and nurse-physician collaboration are two elements of the nurses' work environment that are potentially related to one another according to Kanter's theory (1977) . In addition, a nurse's clinical specialization has been found to influence perceptions related to these two concepts.

To examine the level of perceived structural empowerment, the perceptions of nurse-physician collaboration and the relationship between these two variables, among Intensive Care Unit (ICU) nurses and general ward nurses in Israel, and to compare the groups.

A descriptive, correlational, comparative study design was used on a sample of 84 ICU nurses and 88 nurses from Internal Medicine and General Surgery wards in a large university hospital in Israel (response rate 79%). A three section, self administered questionnaire was used to measure the study variables: the Condition of Work Effectiveness Scale-II (CWEQ-II), the Collaboration with Medical Staff Scale (CMSS) and demographic-professional background.

Results: Perceived structural empowerment was found to be moderate (M = 19.41, SD = 3.66, range = 6-30). Nurses tended to agree that there was nurse-physician collaboration (M = 2.69, SD = 0.55, range = 1-4, 1 = strongly disagree, 4 = strongly agree). A correlation was found between structural empowerment and the nurse-physician collaboration (r = .52, p \ 0.01). A significant difference was found between ICU nurses and general ward nurses on their perceptions of nurse-physician collaboration (t (170) = -3.39, p \ 0.01; General wards: M = 2.83, ICU: M = 2.55). No significant differences were found between nurse specialization on perceived level of structural empowerment.

CONCLUSION. Nurses in this study tended to agree that there was nurse-physician collaboration on their unit/ward. Nurses who perceived themselves as having a higher level of structural empowerment, felt that there was a higher level of nurse-physician collaboration. General ward nurses had more positive perceptions about nurse-physician collaboration on their ward as compared to ICU nurses. No difference was found between the two groups on the level of structural empowerment.

RECOMMENDATION. The findings of this study can be used as the basis for the design of interventions, aimed at enhancing structural empowerment and nurse-physician collaboration, in order to improve nurses' work environment, as one of strategy to decrease the nursing shortage. Further study of additional hospitals in the country is also recommended. Teams have expanded and in some hospitals 24 h cover has been instituted. Researchers are questioning the validity of Outreach services and its impact on patient outcomes. As CCO has been viewed as the panacea to all problems, data collection and analysis is fundamental in proving its financial and clinical benefits.

OBJECTIVES. This comparative study aims to evaluate retrospective data from 1 month in 2006 and 1 month in 2010. Data does not encapsulate patient outcomes; it will compare frequency of referrals and interventions. This data provides an indication to the extent CCO has participated in the care of the acutely ill over a given time period.

METHODS. Data was collected from the 4D Medicus database collating intervention data. Analysis occurred using 13 key interventions using Excel software CONCLUSIONS. Whilst the validity of services has been questioned, the data itself indicates that more patients are referred and frequency of interventions has increased. Various system changes occurred during this time period such as a change of Mews trigger scores, the advent of 24 h CCO and courses such as ALERT and Survive Sepsis were introduced into the basic training of staff. It must be noted that the intention in the UK for CCO was a service that empowered staff through education to undertake this care themselves; therefore the increase in interventions could indicate that the educational approach hasn't made progress. Although the study compares 13 interventions, an increase in the type of interventions was also noted such as ward based CCO supervised CPAP and establishing a PICC line service. Therefore this highlights the changing application of interventions. Further analysis is required to look at the appropriate skills required for the delivery of safe care to the acutely ill in the ward environment. Whilst ward staff are increasingly under resourced, both in skills and manpower, CCO do provide the skills, knowledge and time to meet the shortfall in safe timely care. INTRODUCTION. Working as a critical care nurse involves situations where teamwork is essential and rapid, effective communication is of importance [1] . The education to become a specialist ICU nurse gives skills and knowledge to manage patients who are critically ill with rapidly changing conditions [2] . Experimental research is one way of contributing to the acquisition of such knowledge.

To describe how ICU nurses may contribute and perform in the experimental research process, an environment usually unfamiliar to them. We describe our experiences with regard to clinical contribution and our subjective evaluation of involvement in animal experimental research.

METHOD. Three ICU nurses in a Swedish hospital were asked to participate in a research project investigating myocardial metabolism in porcine models of shock. The tasks were anaesthesia and pain management, assisting with catheter insertion and haemodynamic monitoring the pigs during the process RESULTS. Although the situation was new, the nursing role and function in the team were at once similar and different to the daily work situation in the ICU. One major skill learnt was the rigour of experimental measurements and sources of error, which is sometimes neglected in clinical care. Being able to observe changes due to shock in a controlled setting, we improved our ability to critically 'think ahead' in anticipation of clinical deterioration [3] . Our first-hand experiences at the animal experimental laboratory allayed many anxieties and misconceptions with this type of research.

CONCLUSIONS. The critical care environment demands skills such as the ability to accurately define and change priorities rapidly, good communication and teamwork [4] . We believe that the experimental research setting is one way of enhancing this ability. In these units patients condition may change rapidly and they may need close inspection as well as emergency response. Early Warning Scoring (EWS) system may make early recognition of and response to bad condition possible by observation based on systematic parameters. EWS was developed as a simple scoring system to be used at ward level utilizing routine observations taken by nursing staff. EWS is based on five physiological parameters; systolic blood pressure, pulse rate, respiratory rate, temperature and AVPU score (Alert; reacts to Voice; reacts to Pain; Unresponsive).

OBJECTIVES. The aim of this study was to evaluate EWS among patients admitted to PACU.

METHODS. EWS parameters were recorded four times from 167 patients after their admission to PACU. The First record was taken during the first admission to PACU (EWS 1), the second (EWS 2) after 15 min, the third after 30 (EWS 3) and the fourth record after 60 min. The correlation between variables like differences of four EWS, patients age, the ASA score, duration of operation were statistically examined. 

Early treatment and recognition of sepsis is a stated aim of the surviving sepsis campaign [1] but in busy clinical environments the delivery of antibiotics and fluids can often be delayed. We describe the implementation of an audit proforma, based on the survivesepsis.org [2] resuscitation bundle, as a tool to deliver six aspects of management within 1 h of recognition sepsis.

1. Improve the early recognition and treatment of sepsis in acute medical patients. 2. Provide a sustainable change in the management of septic patients 3. Improve mortality and length of hospital stay METHODS. The proforma consist of six treatment management steps, based on the survivesepsis.org ''septic six'': 1 Oxygen, 2 Blood Cultures, 3 Antibiotics, 4 Lactate, 5 IV fluids, 6 Strict fluid management. It is triggered by patients satisfying two or more of the systemic inflammatory response syndrome criteria. All management steps should be implemented within 1 h of the trigger time stated on the form. The forms are collected and analysed every month and the results are displayed for staff working on the medical admissions unit and accident and emergency.

A total of 239 forms have been collected, 30% diagnosed with severe sepsis. The progress on all six parameters is shown below.

Over the initial seven month period we have demonstrated a sustained improvement in the rapid delivery of all six of the management parameters. INTRODUCTION. The early goal-directed resuscitation has been shown to improve survival in patients presenting with septic shock. A recent systematic review demonstrated the inability of central venous pressure (CVP) to predict the hemodynamic response to fluids infusion, and it should not be used to make clinical decisions regarding fluid management in critical patients. The clinical implication of this fact in septic shock is not well-known.

OBJECTIVES. The aim of this study is to determine if the resuscitation with fluids guided by CVP has clinical implications in patients with septic shock.

Post-hoc analysis of a patients' cohort with septic shock admitted in the Medical Intensive Care Unit since June 2007 to June 2009. All of them were treated on basis of a bundle for severe sepsis management. Chi-square analysis was used to compare categorical data. Continuous data were compared using Student's t test. We used multiple logistic regression model to assess the association between the independent variable and mortality, after adjustment for possible confusing factors (we considered variable to be confounding if the estimate of the coefficient changed by more than 10%).

Eighty-five patients were studied. 66% were male. Their average age was 57 ± 17 and 40% had previous chronic diseases. Severity scores: APACHE II 24 ± 8, SOFA 10 ± 4 and 76% of patients had multiorganic dysfunction. Infectious focus was respiratory in 48%. CVP mean was 12 ± 4 mmHg, ScvO2 73 ± 8% and the mean amount of fluids provided was 2840 ± 844 cc. 71% of patients needed mechanical ventilation. Hospital-stay middle was 28 days (1-160) and 10 days in ICU CONCLUSIONS. In our patients' cohort with septic shock treated under the basis of the early goal-directed resuscitation, the volume of fluids infused was associated independently with mortality. A lower fluid administration in the resuscitation probably could be caused by the early reach of a high central venous pressure. 

Blinding of study interventions is necessary to prevent bias in randomized controlled trials (RCT). Since normal saline and 5% albumin are packaged in bags and bottles, respectively and they have different color and texture, a blinding procedure is necessary to ensure the fluids appear identical for comparative RCTs.

OBJECTIVES. To describe the blinding procedure and evaluate sterility and stability involved in the transfer and storage of study fluids in the PRECISE pilot RCT.

A standard operating procedure for concealment, meeting pharmacy guidelines and good manufacturing practices was developed by the manufacturing pharmacist at the coordinating centre and used by all participating sites. Fluids were transferred with aseptic technique into identical 500 ml bottles under a sterile hood by the pharmacy or transfusion medicine technician then covered with an opaque wrapping. Average time to transfer of study fluids from their original packaging was recorded to understand labor involved with creating each study fluid package. Yellow intravenous tubing was manufactured to also conceal the fluid color. Six blinded bottles of normal saline and 5% albumin from the participating centers were stored at room temperature for at least 3 months. Cultures of the fluids using blood culture media and/or endotoxin levels (measured by commercial assay) were obtained to document sterility of the study fluids. Protein electrophoresis was used to assess albumin stability.

RESULTS. Transfer of the study fluids was the responsibility of the research pharmacist/ technician and blood bank at 3 and 3 sites, respectively. Average time to transfer 20 containers of normal saline and 5% albumin into bottles was 65 ± 29 and 77 ± 19 min, respectively. Sterility (culture negative and/or endotoxin undetectable) of study fluids was confirmed from all bottles of normal saline and albumin that underwent testing. Protein electrophoresis of albumin samples showed a single band suggesting no degradation of albumin during transfer and storage.

CONCLUSIONS. The standardized blinding procedure developed for transfer of study fluids in this pilot RCT confirmed sterility and stability of our study fluids for 3 months. These data are important when considering the length of allowable storage time for these study fluids. Due to the resources and time involved with the transfer of these fluids for individual sites, this transfer method needs to be incorporated into budgeting and may not be feasible in the context of a large RCT.

GRANT ACKNOWLEDGMENT. The PRECISE pilot RCT was funded by a grant from Canadian Blood Services.

Covidien, Singapore, Singapore, 18 Yong Loo Lin School of Medicine, National University of Singapore, Biostatistics Unit, Singapore, Singapore INTRODUCTION. The Surviving Sepsis Campaign recommends a 6-h resuscitation bundle and a 24-h management bundle to improve outcomes in severe sepsis. Compliance with and relevance of these recommendations to Asian intensive care units (ICUs) are unknown. OBJECTIVES. The primary objective of the present study was to assess the compliance of Asian ICUs and hospitals to these bundles. The secondary objectives were to evaluate the impact of compliance on mortality, and the organisational characteristics of Asian hospitals which are associated with higher compliance.

METHODS. This was a prospective observational study of patients with severe sepsis who were admitted to the participating ICUs in July 2009. We recorded the organisational characteristics of participating centres, the patients' baseline characteristics, and the achievement of targets within the resuscitation and management bundles. RESULTS. Sixteen countries and 150 ICUs participated, enrolling 1285 patients. Hospital mortality was 44.9%. Achievement rates for the bundle targets were: lactate measurement, 39.8%; blood cultures, 62.5%; broad-spectrum antibiotics, 64.0%; fluids ± vasopressors, 81.4%; central venous pressure, 39.7%; central or mixed venous oxygen saturation, 10.8%; low-dose steroids, 55.8%; drotrecogin alfa, 3.2%; glucose control, 27.1%; lung-protective ventilation, 11.7%. Compliance rates for the entire resuscitation and management bundles were 7.6 and 3.5% respectively. On logistic regression analysis, achievement of the targets for blood cultures, antibiotics, and central venous pressure independently predicted decreased mortality. High-income countries, university hospitals, ICUs with an accredited fellowship programme, and surgical ICUs were more likely to be compliant to the resuscitation bundle. CONCLUSIONS. Compliance to the resuscitation and management bundles is generally poor across Asia. Given the resource limitations in Asia, the most appropriate strategy to improve outcomes in severe sepsis may be to concentrate on ensuring early administration of antibiotics after blood cultures, and appropriate fluid therapy.

Cerebral oxygen desaturation predicts cognitive decline and longer hospital stay after cardiac surgery

Monitoring brain oxygen saturation during coronary bypass surgery: A randomized, prospective study

The work is supported by departmental sources. 1032 CLINICAL FEATURES AND PROGNOSIS OF ORGANIZING PNEUMONIA PRE-SENTING AS ACUTE RESPIRATORY FAILURE IN ICU

REFERENCE(S). 1. Webster NR. Ventilation in the prone position

Prone position in acute respiratory distress syndrome

Effect of prone positioning on the survival of patients with acute respiratory failure

Acute effects of upright position on gas exchange in patients with acute respiratory distress syndrome

This study was funded by Arjo International AG, Florenzstrasse 1d

Metabolic acidosis and fatal myocardial failure after propofol infusion in children: five case reports

Longterm propofol infusion and cardiac failure in adult head-injured patients

Mild hypothermia alters propofol pharmacokinetics and increases the duration of action of atracurium

Intermittent haemodialysis versus CRRT for ARF in the Intensive Care Unit

Dialysis dose in acute kidney injury: no time for therapeutic nihilism

CIRRHOTICS ADMITTED TO ICU, AND WHEN ADDED TO THE LIVER-SPECIFIC SCORES OF MELD OR UKELD, IMPROVES THEIR RESPECTIVE PREDICTIVE VALUE

Intensive Care, London, UK, 3 Royal Free Hospital

Epidural anesthesia, hypotension and changes in intravascular volume

Intraoperative fluid restriction improves outcome after major elective gastrointestinal surgery

SURROGATE DESIGNATION: CAN WE TRUST OUR RELATIVES?

Does chest physical therapy work?

Physiotherapy in Intensive Care: Towards an Evidence-Based Practice

Fisioterapia no paciente sob ventilação Mecânica

This research was supported by grants from the following Brazilian funding agencies/programs: CNPq, CAPES, FAPESC and UNESC

Readmission to surgical intensive care increases severity-adjusted patient mortality

Physiological scoring systems and audit

Predicting death and readmission after intensive care discharge

A case-control study of patients readmitted to the intensive care unit

Severity of illness and risk of readmission to intensive care: a meta-analysis

A comparison of admission and worst 24-h Acute Physiology and Chronic Health Evaluation II scores in predicting hospital mortality: a retrospective cohort study

Learning from the past to inform the future-a survey of consultant nurses in emergency care

Assessing emergency nursing competence

Post-traumatic stress among Swedish ambulance personel

Levels of mental health problems among UK emergency ambulance workers

Partial and full PTSD in Brazilian ambulance workers: prevalence and impact on health and on quality of life

Ambulance personnel and critical incidents Impact of accident and emergency work on mental health and emotional well being

3 Artemis Health Institute, Director, Critical Care, Pulmonology and Sleep Medicine, Gurgaon, India, 4 Artemis Health Institute, Nursing, Gurgaon, India REFERENCE(S)

The Australian Incident Monitoring Study in Intensive Care: AIMS-ICU. The development and evaluation of an incident reporting system in intensive care

Adverse events in critical ill patients

Ministry of Health and Social Policy 1254 COMMUNICATION: A KEY FACTOR IN THE PATIENT SAFETY?

Anemia of the critically ill: acute anemia of chronic disease

Impact of allogenic packed red blood cell transfusion on nosocomial infection rates in the critically ill patient

High dose recombinant human erythropoietin stimulates reticulocyte production in patients with multiple organ dysfunction syndrome: The Journal of Trauma: Injury, Infection and Critical Ca

To the staff of the critical care department, faculty of medicine

Injury severity and quality of life: whose perspective is important?

Quality of life and persisting symptoms in intensive care unit survivors: implications for care after discharge

Variations in health-related quality of life in critical patients

Funded in part by Fogarty International Center NIH Grant No. 1 D43 TW007560-01 and Clinical Research Institute-Fundacion Valle del Lili

Glasgow coma score, use of mechanical ventilation and vasoactive agents, and the occurrence of severe sepsis (according to Bone's criteria-1992). The causes of admission were divided as: ischemic stroke, hemorrhagic stroke, subarachnoid hemorrhage, status epilepticus, traumatic brain injury, elective neurosurgeries, and miscellanea. The foci of infection, microbiological data and bacteremia were analyzed from septic patients. Numeric data were expressed as median and interquartiles, while categorical data were calculated as percentage. Univariate and multivariate (logistic regression) analysis was carried out to point factors associated with hospital mortality. RESULTS. We included 661 patients, with median age 65 years (IQ range 50-78) and 54% were male

30%) patients, while it occurred during ICU stay on 110 (70%) patients. Hospital mortality was associated with age, the admission cause (higher for hemorrhagic stroke, traumatic brain injury and status epilepticus), APACHE II score, Glasgow coma score and severe sepsis on the univariate analysis

CNPq 1317 PERIOPERATIVE FACTORS ASSOCIATED TO HIGHER MORTALITY IN PATIENTS ADMITTED TO THE NEUROLOGICAL INTENSIVE CARE UNIT (NICU) IMMEDIATELY AFTER BRAIN TUMOR (BT) RESECTION

Saldívar 2 1 UMAE 1 (High-Specialty Medical Unit No 1) El Bajío, IMSS and NICU, HRAEB (High-Specialty Regional Hospital of El Bajío)

Anaesthesiology and intensiv care medizin

Anaesthesiology and Intensive Care Unit

Charité Universitätsmedizin-Berlin, Department for Anesthesiology and Intensive Care Medicine

After approval of the local ethics committee, the PDR ICG was measured within 24 h post injury (day 0) using the non-invasive LiMON system (Pulsion Medical Systems

of PDR ICG to supranormal values

Higher SOFA scores were indirectly associated with lower PDR ICG values, particularly for SOFA scores[8. When patients were grouped by ICU length of stay (\11, C11 days, corresponding to the mean ICU LOS of the German trauma registry), logistic regression analysis identified PDR ICG

Consumables were provided by Pulsion Medical Systems

Influence of APOE polymorphism on cognitive and behavioural outcome in moderate and severe traumatic brain injury

Genetic variation of the APOE promoter and outcome after head injury

Effects of apolipoprotein E genotype on outcome after ischaemic stroke, intracerebral haemorrhage and subarachnoid haemorrhage

The association between APOE « 4, age and outcome after head injury: a prospective cohort study

Decreased cerebrospinal fluid Apolipoprotein E after subarachnoid hemorrhage correlation with injury severity and clinical outcome

« Association of ventilation rates and CO2 concentrations with health and other responses in commercial and industrial buildings

« Sensitivity to noise, personality hardiness, and noise-induced stress in critical care nurses

Recommended lighting level for offices » The Chartered Institution of

« Sick building syndrome, sensation of dryness and thermal comfort in relation to room temperature in an office building: Need for individual control of temperature

SILENT MISERY: MOST SEVERE CRITICAL INCIDENTS

Post traumatic stress disorder in the emergency room: exploration of a cognitive model

Trauma exposure and post-traumatic stress disorder in intensive care unit personnel

Increased Prevalence of Post-traumatic Stress Disorder Symptoms in Critical Care Nurses

Drivers of quality in health services: different worldviews of clinicians and policy managers revealed

Systems Thinking, System Dynamics

The Fifth Discipline: The Art and Practice of the Learning Organisation

The development of system dynamics as a methodology for system description and qualitative analysis

Finnish Funding Agency for Technology and Innovation

Nursing activities score

Tradução para o português e validação de um instrumento de medida de carga de trabalho de enfermagem em unidads de terapia intensiva: Nursing activities score (NAS)

Nursing Activities Score in the intensive care unit: Analysis of the related factors

1341 THE SELF-PERCEIVED HEALTH BETWEEN MEDICAL-SURGICAL AND CRIT-ICAL CARE NURSES IN HUNGARY

Deutsch 1 , I. Boncz 1 , A. Sebestyen 3 , A. Olah 1 1 University of Pecs Faculty of Health Sciences

A longitudinal study design was used to explore the self perceived health of inhospital nurses in acute care settings (surgery, casualty, internal medicine, intensive, coronary care, emergency room) in two Hungarian

Factors predicting team climate, and its relationship with quality of care in general practice

Nurse working conditions, organizational climate, and intent to leave in ICUs: an instrumental variable approach

Critical care nurses' work environments: a baseline status report

Quality of practice in an intensive care unit (ICU): A mini-ethnographic case study

VASPS/INTV 2010). Medicinska Fakulteten, Lunds Universitet

Critical Thinking and Clinical Decision Making In Critical Care Nursing

Assessing and developing critical-thinking skills in the intensive care unit

Gulhane Military Medical Academy, Haydarpasa Training Hospital, Istanbul, Turkey, 3 Gulhane Military Medical Academy

Technology as a Catalyst to Transforming Nursing Care

Devices and Desire: Gender, Technology and American Nursing

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

Division of Pulmonary and Critical Care Medicine, Seoul, Republic of Korea, 3 Peking Union Medical College Hospital, Department of Critical Care Medicine

Mai Hospital, Intensive Care Department, Hanoi, Viet Nam, 10 King Saud Bin Abdulaziz University for Health Sciences, King Abdulaziz Medical City, Intensive Care Department

12 Dr Soetomo General Hospital, Department of Intensive Care

Republic of China, 14 RIPAS Hospital, Intensive Care Unit

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

The Surviving Sepsis Campaign: results of an international guideline-based performance improvement program targeting severe sepsis

GREAT DIFFERENCES IN COMPLI-ANCE WITH SURVIVING SEPSIS CAMPAIGN BUNDLES

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

Delayed diagnosis is associated with increased morbidity, mortality and cost in the ICU. As the mortality rate of severe sepsis remains unacceptably high, a group of international expert developed guidelines in 2004, termed the Surviving sepsis campaign (SSC). The SSC group has introduced the ''sepsis care bundles

Surviving Sepsis Campaign guidelines for severe sepsis and septic shock

Implementation of a bundle of quality indicators for the early management of severe sepsis and septic shock is associated with decreased mortality

Improving Outcomes for Severe Sepsis and Septic Shock: Tools for Early Identification of At-Risk Patients and Treatment Protocol Implementation

Observational, prospective follow-up. Patients who were admitted into the Intensive Care Unit in University Hospital Complex A Coruña (CHUAC) during the months of

Hospital mortality was 30

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

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

associated with decreased mortality

Translating Research to Clinical Practice: A 1-Year experience with implementing early goal-directed therapy for septic shock in the emergency department

Improvement in Process of Care and Outcome After a Multicenter Severe Sepsis Educational Program in Spain

Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock

DELTA CO2 (PVCO2-PACO2) AS A PROGNOSTIC FACTOR IN SEPTIC SHOCK

SEPTIC SHOCK USING THE NEW DEVICE INSPECTRA 650: RELATION TO MACRO-AND MICROHEMODYNAMIC AND OUTCOME C. Luengo 1,2 , F. Vallée 1 , C. Damoisel 1 , M. Resche-Rigon

Among the techniques assessing microperfusion, near infrared spectroscopy (NIRS) gained interest. More than baseline StO 2 values, the reperfusion slope after a vascular occlusion test (VOT) NIRS parameters, especially the reperfusion slope

ScvO 2 or SvO 2 ); metabolic (pH, Base Excess and lactate) parameters were collected. Microperfusion data consisted in: NIRS (baseline StO 2 , occlusion and reperfusion slopes (%/s), automated software); skin laser Doppler microflow (baseline flow (TPU), peak flow (TPU) and slope during reperfusion (TPU/s), measured during and after a 3 min VOT. Survivors (S) and non-survivors

] differed between S and NS at day 1. Macro-hemodynamic and metabolic data did not differ between S and NS

Plan quadriennal EA3509

SVO 2 DOES NOT PREDICT FLUID RESPONSIVENESS IN CRITICALLY ILL SEPTIC PATIENTS

Supported by MSM 0021620819 research grant: Replacement of and support to some vital organs

years) were studied. APACHE II and SOFA score at study entry were 21 (range: 5-32) and 8 (range: 1-16) respectively. The septic syndrome was due to sepsis (n = 12), severe sepsis (n = 3) or septic shock (n = 39). Sites of infection included the lung

REFERENCE(S). 1. Ungerstedt U: Microdialysis: principles and applications for studies in animals and man

The pathophysiology and treatment of sepsis

Management of sepsis

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

Relation between muscle Na + K + ATPase activity and raised lactate concentrations in septic shock: A prospective study

Long-term continuous glucose monitoring with microdialysis in ambulatory insulin-dependent diabetic patients

Whether it is worth to correct acidemia by infusion of alkaline solutions is a matter of discussion. There are a number of evidences against the use of alkalinization therapy with respect to the benefits of reversing pH and the side effects of sodium bicarbonate infusion [1]. Nonetheless, as recently shown by means of an on line survey, 67% of critical care physicians administer base to patients with lactic acidosis

mmol/L), animals were randomized to 60 min of: A) sustained lactic acid infusion, A + B) sustained infusion + sodium bicarbonate, O) transient infusion, B) transient infusion + sodium bicarbonate. In the transient infusion (group O and B), at randomization lactic acid was replaced with normal saline. Acid-base status and lactate levels were measured over time. In a number of animals phosphofructokinase (PFK) enzyme's activity was also measured. RESULTS. Following lactic acid infusion blood lactate rose

Unnecessary use of alkali perturb acid-base status and lactate metabolism potentially overcoming metabolic adaptive strategies. REFERENCE(S). 1. Boyd JH, Walley KR. Is there a role for sodium bicarbonatein treting lactic acidosis from shock?

Use of base in the treatment of acute severe organic acidosis by nephrologists and critical care physicians: results of an online survey

Strong Ions Gap (SIG) quantifies unmeasured blood anions and it is calculated by the difference between strong cations and strong anions (all of them, dissociated in blood plasma)

Retrospective, observational study of all patients with septic shock as defined by the American-European consensus, admitted to the ICU from

arterial blood gases, albumin, lactate and electrolytes were obtained at admittance and 24 h later; APACHE and SOFA score, central venous saturation and lactate

Comparison of acid base models for prediction of hospital mortality following trauma

Forty-five sepsis patients [median age, 62 (IQR, 47-37) years; admission SAPS II, 45 (33-53) pts; severest Multiple Organ Dysfunction Syndrome Score

Interaction of vasopressin infusion, corticosteroid treatment, and mortality of septic shock

Comparing two different arginine vasopressin doses in advanced vasodilatory shock: a randomized, controlled, open-label trial

Lambert 1 1 University of Leicester, Division of Anaesthesia, Leicester, UK blood samples were taken: at induction of anaesthesia, at 3 and 18-24 h post-CPB. Neutrophils were isolated, mRNA extracted, DNA cleaned and reverse transcribed

Supported by a grant from the Association of Anaesthetists of Great Britain and Ireland, and the British Journal of Anaesthesia/Royal College of Anaesthetists

Secretoneurin (SN), a neuropeptide, is specifically expressed in endocrine

Elevated nucleosome levels in systemic inflammation and sepsis

Extracellular histones are major mediators of death in sepsis

23rd ESICM

Asymmetric and Symmetric dimethylarginines (ADMA, SDMA) are protein-breakdown markers; both compete with arginine for cellular transport and are excreted in urine. Moreover

ADMA, SDMA, their ratio (marker of ADMA catabolism), arginine, Interleukin-6(IL-6), Tumor-Necrosis-Factor-a (TNF-a), C-Reactive-Protein(CRP) on day 1, 3, 6, 9, 12 and at discharge in 72 consecutive severely-septic patients were measured

SDMA were higher than normal, ADMA/SDMA ratio was halved, arginine was low. ADMA was related to total SOFA and arginine, inversely related to IL-6 and CRP; SDMA was related to SAPS II, SOFA, blood urea, creatinine, arginine. ADMA/SDMA ratio was inversely

In 14 non-survivors, creatinine, IL-6, TNF-a, CRP and ADMA were stable, SDMA increased, ADMA/SDMA ratio remained low

Figure: time course of ADMA and SDMA blood levels (mean ± standard error) during ICU stay and the last ICU day

PROTEIN-HMGB-1 LEVELS AS PREDICTORS OF OUTCOME IN PATIENTS WITH SEPSIS AND SEPTIC SHOCK

HMGB1 as a predictor of organ dysfunction and outcome in patients with severe sepsis

Early low DCs counts may be correlated to disease severity and could predict fatal outcome. However, little is known about DC number in other shock than septic. OBJECTIVES. To evaluate and compare the circulating DCs number in patients with severe sepsis, septic or cardiogenic shock. METHODS. In a prospective multicentric study (3 ICU), consecutive immunocompetent patients with severe sepsis (SS), septic shock, cardiogenic shock were included. Peripheral blood DC counts, measured by flow cytometry, were evaluated and compared between the three populations at admission and 24 h later. Correlation to disease severity evaluated by clinical scores and day 28 mortality was studied. RESULTS. 100 patients were included (age 61 ± 14 years, 56 male, SOFA D0 10.5 ± 3.9, SAPS II 61 ± 18): 49 septic shock, 16 severe sepsis and 22 cardiogenic shock. Mortality at D28 was respectively 43, 14 and 59%. 13 patients presented a sepsis associated to cardiogenic shock. At baseline and at day 1, a dramatic diminution in the numbers of total DCs either myeloid (MDCs) or plasmacytoid (PDCs), was observed in sepsis (severe sepsis or septic shock) compared to cardiogenic shock patients. No difference was seen between severe sepsis and septic shock patients (Fig. 1). We did not observe any correlation between the number of total DCs at admission or at day 1 and severity of illness scores

DC reduced number is a valuable marker of severe sepsis in shock and is not affected by hemodynamic changes. It could not be used as a prognostic marker in severe septic patients. 1382 PRELIMINARY RESULTS FROM A PROSPECTIVE STUDY ASSESSING THE RELATIONSHIP BETWEEN STANDARD LABORATORY COAGULATION AND GLOBAL TESTS OF CLOT-FORMATION USING THROMBOELASTOGRAPHY IN PATIENTS WITH FULMINANT HEPATIC FAILURE V

The routine use of International normalized ratio (INR) to establish the coagulation status in patients with fulminant hepatic failure (FHF) may be misleading. Anecdotally, FHF patients, despite a significantly deranged INR, may display a normal or even hypercoagulable state, as recently shown, albeit in an extracorporeal setting, with frequently clotted circuits, despite raised PT

We prospectively studied coagulation, demographic, survival and outcome measures of FHF patients (defined by de-novo liver failure, coagulopathy-INR [1.5, and encephalopathy) admitted to the Royal Free Hospital liver and/or intensive care unit(s) (ICU), a tertiary referral centre in liver diseases and transplantation

We present the standard clotting tests and TEG results from 10 (of a required 20) patients currently enrolled, demonstrating variable degrees of encephalopathy and coagulopathy

Effect of norepinephrine on cardiac output and preload in septic shock patients

Apparent heterogeneity in splanchnic vascular response to norepinephrine during sepsis

aggressive use of high-dose norepinephrine in the treatment of septic shock

Norepinephrine requirement is not an independent variable to predict outcome in severe septic shock patients

AIM. The aim of this study was to measure the level of PTSD among Hungarian ambulance workers, and explore factors which can influence it.SAMPLE AND METHODS. 411 Hungarian ambulance workers were involved to this crosssectional study (147 ambulance drivers, 173 ambulance nurses, and 91 ambulance team leaders: medical doctors and ambulance officers). Self filling questionnaire were used for data collection, including Briere's Trauma Symptom Checklist, and socio-demographic questions. Chi square test, independent T test and variance analysis were used for comparison of variables.RESULTS. The average PTSD-points of ambulance workers was 20. There was significant association between level of PTSD and gender: women's average 23, men's average 18 PTSD-points (p = 0.049). There were no correlations between level of PTSD and type of settlement, location of ambulance station and level of education. Those who would need psychological support (p = 0.01), and those who had psychologically traumatic experiences in the last 2 years have significantly higher PTSD-points (p = 0.01).CONCLUSIONS. Hungarian ambulance workers are exposed with many effects which can lead PTSD. Professional psychological support is needed in order to cope with PTSD successfully.The results were presented and discussed in our weekly meeting on patient safety and healthcare for all ICU personnel. By the end of this year all the recommendations will be implemented in our ICU.CONCLUSIONS. We improved the safety and quality of in hospital transportation of ICU patients by performing a prospective risk analysis. Bow-tie is a good instrument to identify health care risks. 

To determine the incidence of phrenic neuropathy associated with the catheterization of internal jugular and subclavian veins, without ultrasound support, in patients admitted to an ICU.

A prospective study was performed by following patients admitted in the ICU between October 2008 and May 2009. A normal neurography of both right and left phrenic nerves at the moment of their admission was the main inclusion criteria. After this baseline study, a new neurography was repeated weekly (Chen and Resman method, Sinergy Medelec), during their stay and at the moment of being discharge from ICU. Simultaneously, all vascular subclavian and internal jugular vein catheterization were registered. A final neurography and a fluoroscopy study were performed after being discharged from hospital. RESULTS. 40 patients were included and two hundred and ten neurographies of both right and left phrenic nerves were performed. 14 patients did not receive any vascular punctures in the cervical region during the follow up period, acting as control group. 26 patients underwent a total of 48 vascular catheterization, 43 in subclavian vein (89.6%) and 5 in internal jugular vein (10.4%). A phrenic neuropathy was diagnosed in 4 patients. This represented an incidence of 15% (4/26) of phrenic neuropathy per patient and 8% (4/48) related to subclavian and internal jugular vein catheterization. In relation to patients without phrenic nerve injury who underwent subclavian and internal jugular vein catheterization, patients affected of phrenic neuropathy had longer mechanical ventilation time (33 ± 36 days vs. 19 ± 13, p = 0.3) and longer average stay time in ICU (49.7 ± 35 days vs. 25 ± 14, p = 0.08), although these differences have not statistical significance. We did not find significantly differences related to age (53 ± 9 vs. 55 ± 18, p = 0.2) and APACHE II index (16 ± 3.5 vs. 15.8 ± 5.8, p = 0.4) between both groups (Wilcoxon Two-Sample Test). We performed a control neurography of 2 case patients after being discharged from hospital. We checked the CMAP phrenic nerve reappearance after 3 weeks and 6 months of being diagnosed its neuropathy, respectively. CONCLUSIONS. We found an incidence of phrenic neuropathy of 15% per patient and 8% related to subclavian and internal jugular vein catheterization, during the follow-up period. The time of reappearance of phrenic CMAPs after being detected its neuropathy points to a neuroapraxia or partial axonotmesis as pathogenic type of injury.DISCUSSION. Phrenic neuropathy has to be considered in cases of difficult weaning of unclear etiology. The catheterization of subclavian and internal jugular veins should be recommended employing ultrasound support.

P. Merino 1 , M.C. Martin-Delgado 2 , J. Alvarez 3 , I. Gutiérrez-Cía 4 , Á . Alonso-Ovies 3 , SYREC 1 Hospital Can Misses, ICU, Ibiza, Spain, 2 ISDE, Á rea de Salud, Madrid, Spain, 3 Hospital de Fuenlabrada, ICU, Madrid, Spain, 4 Hospital Clínico Universitario, ICU, Zaragoza, Spain INTRODUCTION. SYREC project aims to improve ICU patient safety. The project includes an epidemiological study. We present the main results.OBJECTIVES. To estimate the near miss (NM) and adverse events (AE) rate in Spanish Intensive Care Units (ICUs). We study the incidence and nature. Finally, we classify and analyze its severity.METHODS. Multicenter prospective observational cohort study. Inclusion criteria: patients admitted to the 79 participant ICUs during the 24-h observation period. During this period, NM and AE detected and reported inside and outside ICU were included. Only outside ICUs were considered when its were the reason for admission. We evaluate the kind of incident, severity and preventability. Data collection studied under the distribution of frequencies.RESULTS. 1,017 patients were included. 1,424 incidents were reported in 591 patients, 943 were NM and 481 AE. Risk: The median risk of NM was 73% versus AE 40%. 22.1 incidents per patient admitted. Incidence rate: the incident rate median was 5.89 per 100 patients per hour ICU stay, the NM of 3.47 per 100 patients per hour ICU stay and that of AE, 2.04 per 100 patients time of stay in ICU. The 66% of the incidents reported have been NM and 34% AE. This incidents causing temporary damage in the 29.50% of occasions and in the 4.28% permanent damage, compromised the patient's life or contributed to death. Classification of incidents (Table 1) . CONCLUSIONS. Our study shows a high individual risk. Our ICUs services present a highrisk environment. Therefore we have to go into the developement of epidemiological studies depth, in order to create further strategies supporting patient safety. RESTORE CARDIOVASCULAR PERFORMANCE IN SEVERE LACTIC ACIDOTIC RATS A. Kimmoun 1 , N. Sennoun 1 , N. Ducrocq 1 , B. Levy 1,2 1 INSERM U961, Groupe CHOC, Vandoeuvre-lès-Nancy, France, 2 CHU Nancy Brabois, Intensive Care Unit, Vandoeuvre-lès-Nancy, France INTRODUCTION. Lactic acidosis during shock is responsible for myocardial failure, vascular hyporesponsiveness and a decrease in sensitivity to vasopressor agents. Sodium bicarbonate is a proposed treatment to correct acidosis, although with deleterious cardiovascular effects. Indeed, hypocalcemia and hypercapnia, both powerful myocardial depressants, are the main side effects of the administration of this therapy [1] . OBJECTIVES. Already studied in experimental models of isolated lactic acidosis, the cardiovascular effects of sodium bicarbonate administration have never been explored after correction for hypocalcemia and hypercapnia. METHODS. We therefore compared, in a rat model of severe lactic acidosis (pH \ 7.2, hyperlactatemia[3 mmol/l) induced by a state of controlled hemorrhagic shock, the cardiovascular effects of: (1) standard resuscitation plus administration of sodium bicarbonate with correction for calcemia and PaCO 2 (''ADAPT'' group, n = 5); (2) standard resuscitation plus administration of sodium bicarbonate without correction for PaCO 2 and calcium (''NONADPAT'' group, n = 5);(3) standard resuscitation; (''STAND'' group, n = 5); (4) standard resuscitation plus calcium administration (''CALC'' group, n = 5). Evaluation at steady and shock state, 60 min and 120 min was focused in vivo on arterial gas and myocardial contractility (Emax) by conductance catheter. Ex vivo vasoreactivity was tested on mesenteric arteries (300 lm) by myography. Sodium intakes were equivalent between groups. RESULTS. Our model displayed a profound acidosis from 7.31 to 7.18 ± 0.01 (p = 0.02) and hyperlactatemia from 2.8 ± 0.4 to 9.2 ± 0.6 mmol/l (p \ 0.01). Emax decreased from 2.5 ± 0.33 to 0.54 ± 0.06 mmHg/ll p = 0.01. In the ADAPT group, at 60 min, pH was normalized at 7.36 ± 0.01 (p = 0.04). Furthermore, Emax was enhanced at 332 ± 43% (p \ 0.001) (STAND: 145 ± 26%, NONADAPT: 94 ± 18%, CALC: 234 ± 52%). The cumulative dose of infused norepinephrine was significantly lower in the ADAPT group 145 ± 26 lg/kg compared to other groups (STAND: 344 ± 10 lg/kg, NONADAPT: 438 ± 10 lg/kg, CALC: 423 ± 103 lg/kg, p = 0.02). Ex vivo mesenteric vasoreactivity in the ADAPT group was normalized (graph 1).Mesenteric vasoreactivity to phenylephrine CONCLUSIONS. In severe lactic acidosis, infusion of sodium bicarbonate after correction of its side effects improves myocardial function and vasoreactivity. [1] . The prevalence and significance of 25-hydroxyvitamin D deficiency in the intensive care unit have not been fully determined. A recent study of an unselected group of ITU patients [2] has suggested low ITU admission 25-hydroxyvitamin D levels are common. OBJECTIVES/HYPOTHESES TO BE TESTED. Royal Free Hospital Intensive Care Unit patients exhibit low circulating levels of 25-hydroxyvitamin D. Circulating levels of 25-hydroxyvitamin D decrease further during the course of hospital admission. Admission circulating levels of 25-hydroxyvitamin D affect ITU morbidity and mortality METHODS. All ITU admissions were assessed within 72 h of presentation and patients who were deemed to have the potential to require admission for at least 1 week were included. Demographic and clinical data were obtained in a prospective manner. Results were recorded from samples obtained at admission, 7 days and 28 days. Standard ITU nutrition protocols were used. No interventions were performed. RESULTS. Clinical and outcome data were obtained for 40 patients. No significant differences between APACHE 2, SAPS 2 or APACHE 3 scores for survivor and non-survivor groups at either ITU or Hospital discharge were noted. 60 further patients await complete data analysis. 0% (0 of 39 for whom results were available) achieved an adequate ([75 nmol/l) circulating 25hydroxyvitamin D level. 3 patients (7.7%) demonstrated levels within the insufficient range (50-75 nmol/l). 2 patients (5.1%) did not have any detectable 25-hydroxyvitamin D. The remaining 34 patients (87.2%) were either in the deficient (41.0%, 25-50 nmol/l) or severely deficient (46.2%, 10-25 nmol/l) ranges. Admission 25-hydroxyvitamin D levels in survivors and non-survivors were compared at ITU and Hospital discharge. No significant differences between the four groups (P [ 0.05, ANOVA) were observed, indicating that in this data set, admission 25-hydroxyvitamin D levels do not appear to alter or determine clinical course. Mean 25-hydroxyvitamin D levels were compared at admission, Day 7 and at Day 28. No significant differences between the three groups (P [ 0.05, ANOVA) were identified. No significant differences between the mean 25-hydroxyvitamin D levels of the survivors and non-survivors at Day 7 or Day 28 were apparent (small numbers). ADMISSION [1, 2] and patients undergoing surgical procedures [3] . Patients with neurological illness can receive significant quantities of NS, chosen primarily for its iso-osmolar properties. OBJECTIVES. NS is commonly used as maintenance and resuscitation fluid by the anaesthetist, and as intravascular flushes by the radiologist during prolonged interventional neuroradiological (INR) procedures. This pilot feasibility study aimed to ascertain the effect of NS infusion on acid-base measurements in patients undergoing INR procedures under propofol-remifentanil anaesthesia. METHODS. We collated routine electrolyte, albumin and acid-base data of 5 patients who underwent coil/glue embolisations of intracranial aneurysms and vascular malformations, both before and after the procedure. Base excess (BE) was partitioned into the effects of sodium chloride difference (Na-Cl), albumin, lactate and unmeasured anions (UMA), using the Stewart-Fencl-Story approach [4] . All values are reported as medians (ranges OBJECTIVES. To investigate the erythropoietic response to hight dose of a weekly schedule of recombinant human erythropoietin (rHuEPO) in critically ill anaemic septic patients.

A total of 60 patients admitted to the intensive care unite (ICU) were enrolled in this study, patients were randomized to receive either rHuEPO or not, 30 patient did to form the rHuEPO group, 30 did not to form the control group.RESULTS. The EPO treated group of patients showed significant increase in reticulocyte count compared with baseline p \ 0.001, as well as with the control group p \ 0.006. The EPO treated group exhibited also a significant increases in Hb concentration compared with baseline p \ 0.001 as well as the control group .0 3. All patients in the control group received RBC blood transfusion 100%, while only 83.33% of the EPO group did. The EPO treated group showed significant decreases in their APACHE II score during the study period compared with baseline p \ 0.001 as well as with the control group p \ 0.05. The EPO treated group showed no significant difference in their SOFA score compared with baseline p \ 0.923, however the control group exhibited continuous and significant increase in their SOFA score throughout the study period compared with their baseline p \ 0.003, there was no significant difference in the final outcome recovery, mortality or morbidity p .337, p \ 0.286 respectively.CONCLUSIONS. The administration of rHuEPO to critically ill anaemic septic patients is effective in raising their reticulocytic counts, Hb concentrations and in reducing the total number of units of RBCs they require. In addition there was a trend toward better in hospital clinical course, increased recovery and decreased mortality in the rHuEPO group.CONCLUSIONS. Anaemia is common following critical illness but does not appear to affect the physical aspects of recovery during medium term rehabilitation. This may be due to an overwhelming degree of symptom burden from other complications of critical illness impairing physical function to such a degree that the effects of anaemia are negligible in the medium term. Although decreases in number and function has mainly been described in skeletal muscle, also other organs seem to be affected and it has been hypothesized that mitochondrial dysfunction might be involved in the development of organ failure.

To study the effect of plasma of patients with septic shock on mitochondrial function in vitro to potentially later on identify a central factor affecting mitochondria in all tissues during sepsis and leading to multiple organ failure.METHODS. After sacrificing 6-8 week old Sprague-Dawley rats, mitochondria from soleus muscle were isolated through homogenization and a series of centrifugations. Mitochondrial function was assessed by measuring of oxygen consumption, using an oxygraph containing a Clarke-electrode, after addition of ADP. Before these measurements, mitochondria were incubated with plasma from septic patients or healthy volunteers, respectively, for 30 min. In our second series, the mitochondria were incubated with different concentrations of IL-6, TNF-a or buffer. Respiration rates were measured in the presence of ADP (state 3; a measure for the oxidative capacity to produce ATP) and without the presence of ADP (state 4; a measure for the amount of uncoupling). Respiratory control ratio (RCR; a measure for the respiratory efficiency of the mitochondria) was calculated by dividing state 3 by state 4 activity. All measurements were related to citrate synthase activity to compensate for the amount of mitochondria. Statistical differences between the groups were analyzed using a Student's t test.RESULTS. ADP dependent (state 3) respiration was 98% higher and RCR 165% higher in the mitochondria incubated with plasma from the septic patients compared to those incubated with plasma from healthy volunteers (Table) . There were no significant differences between the groups incubated with preservation buffer or the different cytokines (Table) . INTRODUCTION. Microvascular fluid loss from the intravascular to the interstitial space generates tissue edema and is one of the major challenges in emergency and intensive care medicine. Isolation of interstitial fluid (IF) from skin makes it possible to study the microcirculation and proteins in this environment both during normal as well as pathophysiological conditions such as acute inflammation.OBJECTIVES. By studying bio-markers from proteomic analysis by mass spectrometry in an inflammation model, we wanted to find proteomes that could be important in explaining inflammation.

We have applied a recently described centrifugation method in a porcine model and compared it with implanted wicks. In nine anesthesized piglets we compared the methods and evaluated the IF, by overhydrating the pigs with 3000 ml of acetated Ringer's solution for 1 h, and thereafter continuously supplemented for 1 h according to fluid losses. IF was isolated from implanted dry wicks, wet wicks and by centrifugation of excised skin. The methods were evaluated by the ability to reflect overhydration and to show the expected composition of plasma proteins in IF by use of HPLC. The IF was also processed further with mass spectrometry to find possible tissue degradation or inflammation due to overhydration. Statistics: by SPSS v 13.0 and GraphPad Instat (version 3.05). Significance level: p = 0.05.

Colloid osmotic pressure in IF was significantly lowered after overhydration for all the tree methods. Wet wicks p = 0.05, dry wicks p = 0.001, skin samples p = 0.05. HPLC of IF collected with centrifugation after overhydration, identified peaks representing molecules smaller than albumin. Mass spectrometry of the same IF identified several proteins associated with inflammation: alpha-1-antichymotrypsin and Lumican, the latter a protein identified as a modulator of inflammation.

We have introduced a new centrifugation method for isolation of IF from the skin of pigs. By further analysis of IF isolated by centrifugation we were able to distinguish proteins found only in the IF of the pigs overhydrated with Ringer's acetate. These proteins could be associated with an inflammatory condition in the skin caused by massive overhydration, again causing tissue degradation. Identification and validation of proteomic biomarkers can be a useful tool in future treatment of inflammation in general, and in sepsis in special. OBJECTIVES. To define the pattern of change in metabolites by MRS in experimental sepsis.

Male SD rats (weight 325-375 g) underwent cecal ligation and puncture or sham procedure (n = 8 per group), and 24 h after surgery were euthanized. Pulmonary tissue was extracted for magic angle MRS (HR-MAS) and processing by the R Metabonomic package. A supervised statistical analysis of main components (MC) was performed on the processed spectra.RESULTS. The MC analysis discriminated both group (septic and nonseptic) indicating a different metabolite profile. In addition, the analysis of MC loading revealed 4 displacement positions in the discrimination between groups with a variation in the signal intensity of 50%.CONCLUSIONS. Metabolomic analysis of pulmonary tissue by MRS is a potentially useful technique for the detection of biomarkers in sepsis.GRANT ACKNOWLEDGMENT. INTRODUCTION. CD14 + CD16 + neutrophils are a key subset of phagocytes associated with severe bacterial sepsis [1] . Their characteristics, and potential neuro-immunomodulation, have not been explored in humans neutrophils exposed to septic plasma from ICU patients.

To assess the effect of adrenergic/cholinergic neurotransmitter molecules on human neutrophil adhesion and activation markers following exposure to human septic plasma.

With IRB approval, neutrophils were isolated from 5 healthy volunteers (Ficoll density gradient separation) and incubated for 24 h with either plasma from healthy volunteers or septic patients plus pathophysiological concentrations of epinephrine (E), norepinephrine (NE) or acetylcholine (ACh) and nicotine (Nic) to assess potential parasympathetic-related neuro-immunomodulation. Flow cytometry (DAKO Cyan) measured expression on neutrophils of CD11, CD14, CD16 antibody markers and viability. Median values are shown; analyzed by ANOVA.RESULTS. Neutrophils were unaffected by NE, E, ACh or Nic after incubation with plasma from healthy volunteers. After incubation with septic plasma, marked neutrophil activation occurred (p = 0.02). However, Nic reduced CD14 + CD16 + activation (* Fig. A ) by 95% (median (35-97%; 25th-75th centiles); p = 0.01). Nic also attenuated CD11 expression, suggesting reduced neutrophil adhesion (* Fig. B) . Neutrophil viability was similar across drug and plasma treatments.

CONCLUSIONS. These preliminary data suggest that nicotine attenuates both the activation and adhesion of human neutrophils exposed to human septic plasma, but does not affect viability. OBJECTIVES. The aim of this study was to evaluate the potential impact of lag between sepsis initiation and start of treatment on mitochondrial respiration. METHODS. 24 animals [40.1 ± 4.1 kg] were randomized (n = 6/group) to a control group (Group I) and three groups resuscitated at 6 (Group II), 12 (Group III), and 24 (Group IV) hours, respectively, after fecal peritonitis induction. Fecal peritonitis was induced with instillation of 2.0 g/kg of autologous feces via intra-peritoneal drain. Resuscitation was performed according to the SSC and ESICM sepsis guidelines for 48 h. Respiration of permeabilized skeletal muscle fibers and their isolated mitochondria was assessed at baseline and after 6, 12, 24, 48 and 72 h, when applicable, or before death occurred, if earlier. At the end of the experiment, also isolated brain, hepatic and myocardial mitochondrial respiration was measured using high resolution respirometry (Oxygraph-2 k, Oroboros Instruments, Innsbruck, Austria). RESULTS. Mortality (33%, each) and organ dysfunction was highest in groups III and IV. In these two groups, different pattern of changes of skeletal muscle mitochondrial complex I-dependent respiratory control ratio (RCR) were observed (Table 1) . No significant differences between groups were observed for complex I-and II-dependent RCR values of hepatic, myocardial and brain mitochondrial respiration (Fig. 1 ). There were no significant differences between the groups for any of the complexes in permeabilized skeletal muscle fibers mitochondrial respiration (data not shown). CONCLUSIONS. Despite the high mortality observed in groups resuscitated at later time points after induction of sepsis, end organ mitochondrial function assessed using physiological substrates was preserved. Despite significant changes in skeletal muscle mitochondrial respiration efficiency in the two groups with the highest mortality, our findings do not support the view that mitochondrial dysfunction plays a major role in the pathogenesis of multiorgan dysfunction in experimental sepsis. GRANT ACKNOWLEDGMENT. Swiss National Fund, nr: 3200-061988; Stiftung für die Forschung in Anästhesiologie und Intensivmedizin. Adipose tissue is an endocrine organ which produces signalling proteins involved in inflammation and glucose homeostasis [1] . One of these proteins, adiponectin, promotes glucose utilisation and fatty acid oxidation and thus improves insulin sensitivity via its two receptors, AdipoR1 and AdipoR2 [2] . Adiponectin expression has been shown to be reduced in type II diabetes, obesity and endotoxaemia [2, 3] . Adiponectin also exhibits antiinflammatory properties [4] . In this study, we have examined whether adiponectin and its receptor gene expression changes in murine adipocytes stimulated by LPS. METHODS. 3T3--L1 adipocytes were grown in culture media (DMEM with 10% fetal calf serum) until confluent. Pre adipocytes were differentiated with the addition of 10 mg/ml insulin, 1 mM dexamethasone and 100 mM IBMX. Media was changed every 48 h. Cells were treated on day 12 with 100 ng/ml, 1 or 10 mcg/ml LPS (Escherichia coli, Sigma-Aldrich). Cells were harvested at 4 and 24 h. mRNA levels were determined by RT PCR in a 12.5 ll reaction volume consisting of 12.5 ng of reverse transcribed cDNA mixed with optimal concentrations of primers and probe and qPCR TM Core kit (Eurogentec, UK) in 96-well plates on a Mx3005P detector. RESULTS. Cell response to LPS was confirmed using IL6 as a reference gene. Expression of adiponectin mRNA was significantly reduced in cells treated with 10 lg/ml LPS harvested at 4 h (7.1 fold p = 0.002). There were no changes in cells treated with lower concentrations of LPS. There were no changes at 24 h. R1 gene expression was significantly reduced following treatment with 100 ng/ml LPS at 4 h (1.5 fold p = 0.04), but treatment with higher concentrations did not change expression. There were no changes at 24 h. R2 expression levels were significantly reduced at 4 h in the 1 and the 10 mcg/ml groups (2.6 fold p = 0.02 and 4 fold p = 0.01) respectively. There were no changes at 24 h. DISCUSSION. Our results add to the evidence that changes occur in the adiponectin system during inflammation. In this model, we observed rapid reduction (at 4 h) in adiponectin at high dose LPS, R1 at low dose LPS and R2 at medium and high doses. There were no changes in expression levels at 24 h. This suggests that a rapid change in the adiponectin system may occur in response to LPS but this change is not maintained at 24 h. In a previous study, our group has shown reduced adiponectin gene expression in adipose tissue depots in LPS induced endotoxaemia [3] . It is interesting that different concentrations of LPS induce different changes within the adiponectin system. Further studies are needed to elucidate whether reductions in both adiponectin and its receptor may contribute to the inflammatory changes and hyperglycaemia commonly observed during sepsis including all CO poisoned patients treated with hyperbaric oxygen. Following parameters were seized: age, sex, date of admission, SOFA, the source of the intoxication, the gravity CO score, the initial clinical examination (realized by first aid), biology, the rate of HbCO, the Murray score and the rate of complication.

RESULTS. 405 patients were included in the study. The sex ratio was 61%, the mean age was 42 ± 18 years and the global mortality was 5, 9%. Among the 405 patients 29% were poisoned by smoke (S group), 60% by pure CO (C group) and 11% by exhaust fumes. More than 30% of the exhaust fumes victims were suicide origin. This characteristic is associated with neurological impairment induce by ingested drugs. Then, their neurological status is impossible to link to the CO poisoning. We have therefore decided to exclude this group. The SOFA score was higher in the S group compared with the C group (0.38-2.17; p \ 0.0001). A CO score equal to 4 was present in 33 versus 1% respectively in S versus C group (p \ 0.001). In the under group of patients having a CO score at 3, 0% (0/101) of CO poisoned patients versus 18.4% (7/38) of smoke poisoned patients were ventilated (p \ 0.001). These patients were intubated either during transport or in the intensive care and none of them received hydroxycobalamine during the first aid (before intubation). The laboratory data showed in the S group a higher lactates level (2.2 vs. 3.9 mmol/l; p = 0.005) and lower initial PaO 2 /F I O 2 ratio (365 vs. 306; p = 0.002). Nine percent of the S group present a Murray score at 4 versus 0% for C group (p \ 0.0001). Pneumonia, shock and death were significantly more frequent in the S group (respectively 8.6 vs. 0.8%, p \ 0.0005; 11.2 vs. 0.8%, p \ 0.0001; and 17.9 vs. 1.2%, p \ 0.0001)CONCLUSIONS. As expected the smoke poisoned group has a higher mortality than pure CO group (mortality 17% vs. overall mortality 5.9%). At equivalent CO gravity score, mortality and complications are always more frequent in the smoke poisoned group. The smoke poisoned group has a high risk of degradation. Those patients require specific monitoring and support and probably early administration of hydroxycobalamine. HYPOTHESIS. AT administration and maintaining higher plasma levels of AT can reduce the need for inotropes in burn shock patients.

We performed a retrospective cohort study of burn shock patients admitted to a single tertiary care center over 7 years period. Patients were eligible for inclusion if they were received fluid resuscitation with Ringer's solution and colloid according to clinical guidelines. Data were abstracted including demographic, burn injury characteristics, resuscitation fluid volume, the type of colloid and the average of plasma AT levels within 72 h after burn injury. Administration of fresh frozen plasma and/or recombinant human AT was defined as AT administration. The decisions of AT administration and inotropic support (dopamine or dobutamine) were made by the attending intensivists. Primary outcome measure was the need for inotropes within 72 h after burn injury. Cox regression model was used to estimate the risk reduction by AT administration and average of AT levels. [1] . Argon, another member of the noble gas family has been reported previously to have a neuroprotective property [2] . The aim of this study was to investigate whether it attenuates neuronal injury in a rat model of neonatal asphyxia. METHODS. Seven-day-old postnatal SD rats underwent right common carotid artery ligation and then recover with their dim for 1 h. Thereafter, they were exposed to 8% O 2 balanced with nitrogen for 90 min. After 2 h, they were treated with 70% argon or 70% nitrogen (positive control group) for 90 min. The cohort pups without intervention served as naïve control. They were perfused 7 days later and their brains were sectioned and stained with 0.5% cresyl violet. Microphotographs were taken from CA area of the hippocampus near -3.6 bregma relative to adult brain at 40 9 magnification. Healthy cells were counted in a blind manner and their mean value was used for data analysis.

RESULTS. The thickness of healthy layers in the right CA area of the positive control group was remarkably reduced compared with other groups (Fig. 1 ). Quantitative analysis revealed that argon treatment significantly increased healthy cell numbers in the right CA area of hippocampus from 36.53 ± 1.201 in the positive controls to 54.37 ± 1.162 (P \ 0.01) (Fig. 2 ). GRANT ACKNOWLEDGMENT. This study was supported by a grant from Action Medical Research, UK. OBJECTIVES. Our objective was to study the mechanisms of death following high-dose citalopram administration in rats.

Experimental study in Sprague Dawley rats with intraperitoneal (IP) citalopram administration; determination of the median lethal dose (MLD)using the Dixon and Bruce upand-down method; clinical descriptive study of citalopram-induced features and measurement of alterations in respiratory pattern (arterial blood gases and plethysmography) and biological parameters including blood lactate (Scout Ò , EKF diagnostic), plasma and platelet serotonin concentrations (high-liquid performance chromatography-fluorometry); determination of the preventive activity on seizures and death of diazepam, cyproheptadine, and propranolol pretreatments with the determination of their minimal effective dose; comparisons using ANOVA for repeated measurements followed by Bonferroni post-test.RESULTS. Citalopram IP-MLD was determined as 102 mg/kg in rats. Seizures were significantly increased in rats receiving 80 and 120% of citalopram MLD versus controls (p \ 0.01 and p \ 0.05, respectively), while death rate was only significantly increased in rats treated with 120% of citalopram MLD (p \ 0.001). Significant decrease in body temperature was observed after 90 min in rats treated with doses[60% MLD in comparison to controls (p \ 0.05). Occurrence of serotonin behavioural syndrome was comparable in all groups. Citalopram administration did not result in significant hypoxemia, hypercapnia, and lactate elevation, thus not supporting the hypothesis of the occurence of any significant deleterious cardiovascular effect in citalopraminduced toxicity. However, a significant moderate increase in the inspiratory time (p \ 0.05) accompanied with an expiratory braking was observed. A significant decrease in platelet serotonin and increase in plasma serotonin concentrations were measured (p \ 0.05). Pre-treatment with diazepam (1.77 mg/kg) and cyproheptadine (17.1 mg/kg) of rats receiving a lethal citalopram dose prevented seizures and death, while propranolol was ineffective.CONCLUSIONS. Citalopram respiratory toxicity remains mild, while deaths result from seizures probably related to serotonin toxicity. Our observations may be helpful to better understand and manage human citalopram poisonings. OBJECTIVES. To define the population pharmacokinetics (PK) of phenytoin in the critically ill, in addition to risk factors for sub-therapeutic dosing.METHODS. Free and total PTN concentrations were measured in serum by means of high performance liquid chromatography following microfiltration, two to three times in the first 24 h after a loading dose. Population PK modelling, including intra and interindividual variability, were determined using NONMEM (R) . In the Netherlands the use of diazepam is advised as first line treatment although evidence is not established and mainly provided through case-reports [1] .

To compare the effect of diazepam on mortality in (hydroxy) chloroquine intoxication to standard therapy.

We performed an extensive MEDLINE search (1950-April 2010) with a manual reference search of identified papers. (Hydroxy) chloroquine intoxication studies and case reports in English, Dutch or French were evaluated. Patients older than 15 years with severe intoxications, based on measured concentrations or life-threatening symptoms, were included. Pooled relative risk (RR) for mortality with corresponding 95% confidence interval (CI) were calculated by means of a Fisher exact test. Our results were compared with two retrospective and one prospective study.RESULTS. There were 32 case reports identified from which 25 case reports met our inclusion criteria. Thirteen patients received diazepam of whom two died, compared to twelve patients who did not get diazepam of whom one died. Statistical analysis demonstrated that treatment with diazepam was not associated with a lower mortality rate (RR: 1.1 CI 0.8-1.5; p = 1.0).Although pooling of case reports is debatable, these results were comparable to the retrospective and prospective studies that didn't show any benefit from diazepam in chloroquine intoxication [2, 3, 4] . The positive effect of diazepam may have been underestimated, due to the fact that it has been given only as rescue therapy.CONCLUSIONS. Based on our analysis there is a lack of evidence concerning any antidotal effect of diazepam. Good supportive treatment is pivotal. If the clinical manifestations of (hydroxy) chloroquine intoxications require sedation or treatment of seizures, diazepam is a good choice based on its pharmacological profile. A prospective study which compares diazepam to sedativa with similar pharmacokinetic and dynamic profile is required to prove that diazepam has any antidotal effect. INTRODUCTION. Brain is one of the first organs affected in sepsis and evaluation of brain function is difficult since patients are under sedation. It has been shown that mitochondrial dysfunction may play a significant role in the pathogenesis of septic encephalopathy. Here we investigated inflammatory and metabolic parameters in a model of polymicrobial sepsis in mouse.

METHODS. Sepsis was induced by intraperitoneal injection of feces. Animal received imipenem 6 h after the procedure. Control animals received intraperitoneal saline and imipenem after 6 h. Blood Cytokines and serum lactate were measured. The animals were sacrificed by cervical dislocation. Brain slices of 400 mcm were used to measure oxygen consumption and glucose uptake.RESULTS. Interleukin 6, MIP 1a and Interleukin 1b significantly raised in the first 6 h after sepsis induction (p = 0.001; p = 0.0017; p = 0.0001 respectively). In 24 h only MIP 1a was significant higher (p = 0.0001). Lactate was elevated 6 and 24 h after sepsis induction (p \ 0.01 and p \ 0.001 respectively). Oxygen consumption increased after 6 h of sepsis and drops under control values 24 h after the induction of sepsis. Glucose uptake, measured by the NBDG fluorescence, was higher after 6 h (p = 0.0475) and 24 h after sepsis induction.CONCLUSION. In a murine model of abdominal sepsis, inflammatory markers, lactate production, and brain glucose uptake increased and were parallel to alterations in the mitochondrial oxygen metabolism. INTRODUCTION. The Royal Bournemouth Hospital has one of the highest out-of-hospital cardiac arrest admission rates in the UK. In 2005, following ILCOR/AHA guidelines [1] , a cooling protocol was developed for patients with return of spontaneous circulation after advanced life support for ventricular fibrillation or pulseless ventricular tachycardia. In preparation for potential new ILCOR/AHA guidelines in 2010, the prospective database of outcomes for these patients was analysed.OBJECTIVES. To evaluate the outcomes of therapeutic hypothermia for patients with return of spontaneous circulation following cardiac arrest.

Outcome data from our prospective registry of cooled patients are summarised.RESULTS. Sixty-three patients were cooled in 4 years (median age 68 years; mode 76; range 19-88 years). 84% survived to ITU discharge and 70% to hospital discharge. 98% of these were discharged home (14% to a rehabilitation hospital before home and one patient to a long term care facility). Ninety-five percent of survivors were alive at 6 months and 80% alive at 1 year with seven status results still pending. Median ITU length of stay was 3.5 days (range 1-13). Six patients required temporary percutaneous tracheostomies for airway protection and weaning from ventilation. Median duration from ITU to hospital discharge was 16 days (range 0-62).CONCLUSIONS. This series is large by comparison to other UK centres. Survival to hospital discharge, at 6 months and 1 year were better than other published results. Although neurological outcomes were not formally assessed, we believe that the capacity to discharge home is a desirable patient outcome and represents the beneficial neurological effect of our cooling protocol. Selection bias will have undoubtedly affected our results. However the age of our patients was higher than in published trials and in other reports is considered an adverse outcome predictor. Our data would not support restricting induced hypothermia on the basis of age alone. We consider the ITU and hospital lengths of stay required to discharge these patients to be long. These data were not reported in original trials. Discharges may obviously be delayed for non-clinical reasons. This aside, neurological recovery progresses for months after cardiac arrest and discharge home may still prove possible if time is allowed. However, post-ITU resource implications should be considered when introducing a cooling protocol. INTRODUCTION. Acute ischaemic stroke (AIS) is the third largest cause of mortality and the leading cause of chronic disability in the industrialized world. In some parts of Europe and the United States 26-43% of patients with AIS may be admitted to a neurological intensive care unit (ICU) for supportive therapy with 8-10% receiving mechanical ventilation [1, 2] . There are currently no agreed UK criteria for the admission of AIS patients to critical care.OBJECTIVES. To review the incidence and outcome of AIS in our tertiary ICU over the last five years.

November 2005 and November 2009. AIS was classified as thrombo-occlusive or embolic. Subarachnoid haemorrhage and primary intracerebral haemorrhage were excluded. Demographic and outcome data were recorded and compared against a mean value of all ICU admissions.RESULTS. AIS comprised 0.8% of ICU admissions during the study period. Demographic data is presented in Table 1 as mean ± standard deviation or median (interquartile range) as appropriate. In 2008 4% (11/272) of hospital AIS admissions were admitted to ICU. 14 patients had surgical procedures including 8 decompressive craniectomies. 27% of survivors had a discharge GCS of 15/15. Mortality for unselected medical admissions over the study period was 36%. There are differences of significance in the mortality according to the age, classified by age groups with an age cut off of 80 years (\80 years 38.6 vs. C80 years 40%, p \ 0.01). Apart from the GCS, the rest of the variables analyzed in the ICH score are not of significance; supra and infratentorial, presence of intraventricular blood neither on the divided volume over or under 30 cc although, in the latter, a p \ 0.1 can be observed and if we only analyze the supraventricular, it comes out as significant. Other analyzed data are the time of the surgery, which is not significant, the need for mechanical ventilation, which is (42.9 vs. 15.8%, p \ 0.05), and the days of ventilation with a mortality clearly higher on those patients with\5 days of ventilation (83%) and on those of shorter stay (lesser then 5 days 79%).CONCLUSIONS. Let be remarked that the samples have been taken from patients admitted in the intensive care unit, losing a possible sample of less serious patients, and with a higher level of consciousness, what might explain why supra or infratentorial location and the volume don't come out as forecasting factors, since its likely that there are many small infratentorial outside the intensive care unit. We highlight also that the high mortality in the first few days can be caused by those patients who are admitted as donors, developing an encephalic death in the first days, conditioning also the data regarding the mortality on fewer days with ventilation. The finish up, we have to point out the fact that the presence of previous hypertension during the treatment might be a bad forecasting factor that should be deeper studied. 

To determine whether a delay exists between the time of diagnosis of intracranial haemorrhage and the time of reversal of anticoagulation, in patients presenting within our region.

Following approval by all audit and haematology departments a 6 month retrospective analysis was performed. We reviewed consecutive patients who received reversal of anticoagulation with PCC and Vitamin K having presenting with intracranial haemorrhage whilst on warfarin. Time of diagnosis was obtained from the time of scan and time of PCC issue was obtained from the blood bank database. Case note analysis was performed to obtain further information.RESULTS. 47 patients were identified, 12 in the Neurosurgical centre and 35 in peripheral hospitals. The median time from scan to issue of PCC was 84 min. 14 patients were reversed within 60 min and 14 patients waited longer than 120 min to have PCC issued. No adverse thromboembolic events were encountered.CONCLUSIONS. Avoidable delay exists between ICH diagnosis and PCC issue. PCC could be stored in the emergency department and a stat dose administered immediately after diagnosis facilitating rapid correction of INR. Repeat audit will be required to assess safety and efficacy. OBJECTIVES. The aim of this study was to compare the functional ability and muscle strength between these two groups of patients.

Twenty-nine patients were evaluated (M:24, F:5) (age: 55 ± 17 years).The diagnosis of critical illness polyneuromyopathy was based on muscle strength measurement according to the Medical Research Council (MRC) of muscle strength methodology. Nine patients were diagnosed with critical illness polyneuromyopathy during their ICU stay (MRC \48/60).The patients were evaluated with MRC and Hand-Grip dynamometry (HGD) every 10 days until their discharge from the hospital. The FIM scale (Functional Independence Measure) was used to evaluate the functional ability (18-126).The first evaluation was done at the discharge from the hospital and the second one 8 ± 1 months afterwards.RESULTS. The patients who developed critical illness polyneuromyopathy had statistically significantly lower MRC (39 ± 12 vs. 57 ± 2, p \ 0.001) and HGD at ICU discharge (left 8 ± 5 kg vs. 23 ± 8, and right 9 ± 7 kg vs. 27 ± 9, p \ 0.001) compared to those who did not. The muscle strength as assessed with the MRC 10 days after ICU discharge had statistically significantly lower (40 ± 11 vs. 56 ± 3, p \ 0.001), just as the second HGD evaluation (left 11 ± 7 kg vs. 25 ± 8 and right 9 ± 7 vs. 21 ± 8 kg, p \ 0.05).Compared to those who did not develop critical illness polyneuromyopathy, the patients who did, had statistically lower FIM values during their discharge from the hospital (49 ± 26 vs. 77 ± 30, p \ 0.05)and 8 months afterwards (90 ± 27 vs 125 ± 1, p \ 0.05).CONCLUSIONS. The patients who developed critical illness polyneuromyopathy had significantly inferior muscle strength at their discharge from the ICU. These patients also had lower functional ability. This functional ability remained defected even 8 months after their discharge from the hospital. These initial findings are suggestive that the appearance of critical illness polyneuromyopathy affects the patients mobility after their discharge either from the ICU or from the hospital and persists for several months after ICU discharge. Further studies are needed to evaluate the effect of this impairment on the quality of life of these patients and also to evaluate therapeutic tools for critical illness polyneuromyopathy. INTRODUCTION. This poster presents a qualitative system dynamics (SD) analysis of the factors which influence the care of acutely unwell ward patients in New Zealand. This systems thinking approach is commonly used in organisational research and offers a way to make sense of complex relationships between variables. This approach has previously been used in health care to demonstrate differences in mental models between policy makers and clinicians (Cavana et al., 1999) . Since the factors which influence the care of acutely unwell ward patients are complex and multi faceted the qualitative SD method becomes an ideal analytic approach (e.g. see Wolstenholme and Coyle, 1983; Senge, 1990; Vennix, 1996; or Maani and Cavana, 2007) .OBJECTIVES. The aim of this study was to examine the factors which influence the care of acutely unwell ward patients from an organisational perspective. Key objectives were to determine the enablers and barriers to care from a nursing, medical and managerial (at ward and executive level) perspective.METHODS. Using a multiple case study approach in four wards in two New Zealand hospitals, focus groups and one to one interviews were conducted with key stakeholders identified as nurses, doctors and managers. Initial coding of the data generated themes. These themes were then clustered to provide variables which were mapped to generate separate causal loop diagrams (CLDs) for each of the stakeholder groups to provide the basis for analysis. The CLDs were compared for characteristics and world views.

Preliminary results demonstrate a difference between clinical and managerial staff in characteristics and world view regarding the factors which affect the care of acutely unwell ward patients.CONCLUSIONS. The qualitative SD approach has offered a novel and helpful way to make some sense of the complexity associated with caring for acutely unwell ward patients. Organizational responses that may improve care delivery to these patients should be based on frank and open discussions between staff at all levels to ensure a shared mental model as the basis for change. OBJECTIVES. The aim of the study is to explain the nursing in the technologicallyadvanced intensive care units.

In this phenomenologically-designed study, a face-to-face in-depth interview was performed with 20 nurses, who were experienced for 1-22 years in the intensive care unit of cardiovascular surgery clinics. During the interviews, a semi-structured form was used. Data were analysed using Colaizzi's method of data analysis. The study was approved by the Ethics Committee of the institution.RESULTS. According to the nurses, nursing in technologically-advanced environment has three stages. These stages constituted three themes of the study: technology shock (first stage), understanding the technology-supported care (second stage), competency in technological environment (last stage). In the first stage, the nurses focus on themselves and technology; perceive the environment as frightening and complex. In the second stage, nurses gain control on technology, feel themselves safe and recognize their responsibility. In the last stage, the nurses experience anxiety related to their accountability. This anxiety may be motivating but also may be wearisome.CONCLUSIONS. The nurses passes through three stages in a technologically-advanced environment. Helping nurses to pass through these three stages appropriately will increase the contribution of technology to the patient care, more utilization of technology by nurses and more job satisfaction. 

Unexpectedly, the compliance rate with the recommendations was significantly better over night. Although the number of nurses is constant in the 24 h, the number of doctors is lower and less differentiated in the night shift. In an attempt to find an explanation for these findings we looked at the patient flow and time span until the first medical observation in the different time periods and we found that over night admissions (between 00:00 a.m. and 08:00 a.m.) corresponded only to 10% of all admissions and were seen sooner, which might explain our findings.

A. OBJECTIVES. The purpose of the study was to assess whether the completion of the sepsis resuscitation bundle within the first 6 h after ICU admission, but beyond the specific time limit of the various bundle interventions, is related to an improvement in survival in patients with severe sepsis/septic shock.

This was a single-center prospective observational study of patients admitted to the medical-surgical ICU of an urban tertiary care teaching hospital with severe sepsis/septic shock. Patients were recruited from June 2005 to November 2009. We assessed the compliance with the different tasks included in the 6-h resuscitation bundle. Furthermore, we ascertained within the first 6 h after ICU admission the compliance with those tasks not carried out within their specific time limits; we have called this variable ''bundle improvement at the ICU''. Results were stratified by the number of tasks of the bundle completed before admission at the ICU, and the lag time between the beginning of severe sepsis and admission to the ICU. These late completed tasks at the ICU were related to hospital mortality by a Cox regression model. OBJECTIVES. The aims of this study were to assess the compliance rate with 6 h bundle as defined in the Surviving the Sepsis Campaign guidelines 2008 in patients diagnosed with sepsis regardless of severity and whether compliance affects the rate of mortality and/or hospital stay.

We conducted a prospective observational study. We randomly recruited 145 adult patients from acute admissions unit and intensive care in an acute district general hospital in England who met the diagnostic criteria for sepsis. For each patient, compliance with sepsis care bundle was obtained from medical notes. The following components of the 6 h sepsis bundle were assessed: Obtaining blood cultures, initiating antibiotic therapy, measuring serum lactate and in the event of septic shock administration of fluid therapy. CONCLUSIONS. Long and unacceptable delays in admission to ICCU were identified despite evidence of significant organ dysfunction in many of these patients. With all bundle elements being met for only 1 patient it is apparent that evidence based endpoints aimed at reducing mortality from severe sepsis are not being met despite all the bundle elements being practically deliverable. Poor compliance with taking blood cultures prior to antibiotic administration and lack of ScvO 2 measurement are areas requiring particular attention. Further work is recommended to identify potential contributing factors to non-compliance. INTRODUCTION. International guidelines recommend that cardiac output measurement is required in addition to arterial pressure monitoring in patients with persistent shock after initial therapy [1] . Nevertheless, these recommendations are not supported by any comparison of arterial pressure and cardiac output for monitoring the effects of the most current treatments like fluid therapy.

OBJECTIVES. To evaluate in which extent monitoring the haemodynamic effects of a standardized fluid challenge with the sole arterial pressure could help for detecting the fluidinduced changes in cardiac index (CI).

In 228 critically ill patients with acute circulatory failure deemed at receiving a 500-mL saline infusion over 20 min, we measured the systolic (SAP), diastolic (DAP), mean (MAP) and pulse (PP) arterial pressure and transpulmonary thermodilution CI before and after volume expansion.RESULTS. Volume expansion significantly increased CI, SAP, DAP, MAP and PP by 24 ± 25%, 15 ± 19%, 9 ± 16%, 13 ± 17% and 21 ± 29%, respectively. The fluid-induced changes in PP, SAP and MAP were significantly correlated with the fluid-induced changes in CI (r = 0.56, 0.55 and 0.37, respectively). The changes (in %) in PP were significantly related to the changes (in %) in stroke volume for all quartiles but with different coefficients of correlation: r = 0.39 for the 1st quartile (36-53 years), r = 0.44 for the 2nd quartile CONCLUSIONS. PP and SAP were the best arterial pressure values for detecting the fluidinduced changes in CI. Using the sole PP for assessing fluid responsiveness led to a non negligible proportion of false negative cases. This supports the recommendation that when a precise monitoring of fluid resuscitation is required, like in refractory shock, a direct assessment of cardiac output is required. OBJECTIVES. Aim of our study is to show that it is possible to reduce high catecholamines in previous improper volume resuscitated patients by forced volume resuscitation combined with active dose reduction and generate the hypothesis of an avoidable catecholamine induced circulation injury. INTRODUCTION. The sialic acid content of the red blood cell (RBC) membrane decreases early in sepsis [1] , and this alters the RBC shape and metabolism [2] . An increased ratio of the RBC proteins Band 3/alpha spectrin was observed in a mouse model of septic shock, suggesting a possible alteration of the RBC membrane integral/peripheral proteins ratio [3] . As there are interspecies differences in membrane composition, these observations need confirmation in humans.

We studied RBCs from patients with (n = 15) and without (n = 9) sepsis at ICU admission and on day 3 in the septic patients. Exclusion criteria were recent RBC transfusion, hematologic diseases, cirrhosis and diabetes mellitus. Procedures included screening for RBC membrane protein alterations by cryohemolysis test and separation of the RBC membrane and skeletal proteins by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate [4] . Comparison between groups was made by the Student's t test or the Mann-Whitney test. A p value .05 was considered as statistically significant.RESULTS. The hemogram, including reticulocyte count was similar in septic and non-septic patients at ICU admission. No significant difference was observed for cryohemolysis test results and the amount of the RBC proteins (Table 1) . OBJECTIVES. Our purpose was to compare a new method (Patrol FR 08-02110) with the reference method (Randox TM ) during CBP.

Patients scheduled for coronary artery bypass (CB) and aortic valve replacement (AVR) under CBP were enrolled after written informed consent in this protocol approved by local ethics committee. Anesthesia protocol was standardized with systematic use of tranexamic acid. Three blood samples were harvested: T1 = induction; T2 = 30 min. after cross aorta clamping; T3 = 24 h after induction. The Patrol method was performed after serum exposition to a photosensibilizer agent then to a laser irradiation leading to the formation of free radicals. Oxidation by those free radicals of a fluorometric sensor allowed an indirect measure of TAS. This measurement in arbitrary unit (AU) corresponded to area under curve compared to a control value from a pool serum. A value higher than 1 indicated a lower capacity for the given serum to neutralize free radicals whereas a lower value indicated a higher capacity. The same sample allowed TAS determination (lmol/L) with Randox TM method. Results were expressed as absolute numbers, mean ± SD. TAS were compared with ANOVA test; p \ 0.05 was significant.RESULTS. The seven patients (5 male, 2 female; 71 ± 10 years old) enrolled underwent cardiac surgery (4 CB and 3 AVR) without any problem. There was no variation in TAS determination with the Randox TM method: T1: 1.39 ± 0.13; T2: 1.41 ± 0.26; T3: 1.33 ± 0.15 lmol/L. Conversely a two fold significant increase was measured during CPB with the Patrol method: Ti: 1.12 ± 0.31; T2: 2.21 ± 0.60*; T3: 0.99 ± 0.34 AU. *p \ 0.001 versus T1.CONCLUSIONS. Oxidative stress due to overwhelming release of reactive nitrogen/oxygen species (RN/OS) is held largely responsible for sepsis-induced organ failure and mortality [2] . Up-front and/or ongoing distortion of the pro-oxidant/anti-oxidant balance is likely to play an important role in this situation and in ischemia-reperfusion. Therefore the Patrol test which appeared to be more sensible than the Randox TM method could a good tool in these cases and for evaluation of new anti-oxidant treatments in critical care medicine. These results have to be confirmed in a larger population. INTRODUCTION. Sepsis is the leading cause of death in critically ill patients. Despite attempts to improve standardized strategies in resuscitation and treatment of sepsis, the morbidity and mortality remain unacceptably high. Early diagnosis and stratification of the severity of sepsis is the key to start timely the appropriate treatment. Sepsis is the systemic inflammatory response syndrome to infection; it can lead to hypoperfusion and organ dysfunction and at the cellular level to aerobic mitochondrial dysfunction. Lactate is the product of anaerobic metabolism and thus may serve as a prognostic factor in this subset of patients.OBJECTIVES. The authors propose to test the association of the first serum lactate at hospital admission with shock and ICU mortality in patients with community-acquired severe sepsis. 

During the study period 1,223 patients were admitted in the unit, of those 300 (25%) had severe community-acquired severe sepsis (CASS). Crude ICU mortality rate among CASS was 33%. Considering the model previously described in methods and when the variables were adjusted only gender, age, SAPS II, severity of sepsis and serum lactate were retained in the final model for ICU mortality and SAPS II nad serum lactate for shock (see Table 1 ). 

A first blood lactate level was independently associated with shock and ICU mortality in patients community-acquired severe sepsis admitted in intensive care. OBJECTIVES. The objective of this study was to test whether SvO 2 can predict fluid responsiveness in these patients.

We studied 65 patients who were monitored with a pulmonary artery catheter for severe sepsis and septic shock. Hemodynamic measurements were obtained before (baseline values) and after a fluid challenge with colloids or crystalloids. Responders were defined as those with a[10% increase in cardiac index (CI). No additional interventions were performed during the test. Student's t test and linear correlation were used for the statistical analysis.RESULTS. Mean patient age was 70 ± 2 years and the mean SOFA score 10 ± 2. Mean arterial pressure was 71 ± 9 mmHg, cardiac index 3.0 ± 0.8 L/min/m 2 , pulmonary artery balloon-occluded pressure 13 ± 3 mmHg, and heart rate 104 ± 19 bpm. Thirty-four patients (52%) responded to the fluid challenge. Responders and non-responders had similar baseline SvO 2 (67 ± 9 vs. 67 ± 8%, p = 0.98). Baseline SvO 2 was[70% in 13 responders (38%) and in 11 non-responders (35%). There was no correlation between changes in CI (%DCI) and the baseline SvO 2 (Fig. 1) . 

Sepsis is a disorder of microcirculation [1, 2] . Although the pathogenesis of microvascular dysfunction in sepsis is extremely complex, neutrophil activation and their interaction with endothelial cells are considered central features of sepsis-induced microcirculatory alterations. To our knowledge, however, no study evaluated the microvascular pattern of septic patients with chemotherapy-induced severe leukocytes depletion.OBJECTIVES. To assess early microcirculatory response to sepsis in patients with and without drug-induced neutropenia.METHODS. Demographic and hemodynamic variables together with sublingual microcirculation recording (OPS-SDF videomicroscopy) were collected in four groups of subjects: septic shock (SS, N = 9), septic shock in neutropenic patients (NSS, N = 8), neutropenia without inflammation (NEUTR, N = 7) and healthy controls (CRTL, N = 13). Except for controls, all measurements were repeated after complete resolution of septic shock and/or neutropenia (TP2). Collected video-files were processed using appropriate software tool and semi-quantitatively evaluated (functional capillary density, FCD (cm/cm 2 ); mean flow index, MFI [1] ) [3] . CONCLUSIONS. Microvascular derangements in sepsis did not differ between non-neutropenic and neutropenic patients. Surprisingly, neutropenia per se without measurable systemic inflammation was also associated with alterations of the sublingual microcirculation. Although we cannot exclude the role of residual neutrophils, our data could indicate that leukocytes are not the only and exclusive modulators of septic microvascular dysfunction. In addition, the role and mechanisms of microvascular changes associated with chemotherapyinduced neutropenia warrants further investigation. 

Multiple organ failure is a leading cause of death in critically ill patients. Improvements in outcome will most rely on our capacity to measure rapidly accessible biomarkers.OBJECTIVES. To investigate if the time sequence of Reactive Oxygen Metabolites (ROMs) production with SOFA score could be prognostic for outcome.

The study included 33 critically ill patients (from September to December 2009) who had ROMs measured (hydroperoxides) during ICU stay, when the diagnostic criteria for sepsis (observed n = 14), severe sepsis (observed n = 11) and septic shock (observed n = 68) were present, 4-5 days and 3 weeks after the diagnosis (samples n = 93); on the same days, the SOFA score was calculated. The plasma ROMs values were assayed by a Diacron-Italia kit, applied to an automatic instrument (Olimpus AU 640). Statistical analysis was performed used Mann-Whitney test and the linear regression analysis.

The ROMs values and SOFA score were inversely correlated (r 2 = 0.64 for sepsis; r 2 = 0.42 for severe sepsis; r 2 = 0.37 for septic shock). The DROMs (the difference between the first and the last measurement of ROMs levels in each individual patient) was significantly different between survivors and non-survivors. Clinical characteristics of the patients are presented in Table 1 . Values are presented as median and interquartile rangers. A p value .05 was considered as statistically significant.CONCLUSIONS. The plasma ROMs values decreased when the critically conditions rapidly evolved towards organ failures with higher SOFA. 

To explore: (a) stress neuropeptides (ACTH, cortisol, prolactin, neuropeptide Y (NPY) and substance P (SP)) in critically ill subjects and controls, (b) potential association between levels of stress neuropeptides, disease severity and pain.

A prospective correlational study, with repeated measurements and cross-sectional comparisons. Fifty-three critically ill patients with diverse primary diagnoses and 53-age and gender-matched healthy controls were studied for 8 days. Serum neuropeptides were quantified by ELISA (NPY, SP) and chemiluminescence immunoassays (ACTH, cortisol, prolactin). Pain levels were assessed by Payen and Puntillo scales. Clinical severity was quantified by Multiorgan Failure Scoring System (MOF) and the Multiple Organ Dysfunction Score (MODS). RESULTS. We observed: (a) statistically significant differences between critically ill and control subjects in regard with cortisol (p \ 0.0001), NPY (p \ 0.0001) and SP (p \ 0.05) levels throughout the study. Specifically, cortisol levels were higher and NPY and SP levels were lower in patients compared to controls, (b) significant bivariate associations between stress neuropeptides (p \ 0.04), (c) statistically significant associations between ACTH and pain intensity levels assessed by Payen (r = 0.391, p = 0.018) and Puntillo (r = 0.509, p = 0.002) scales. There was also a constant but not statistically significant (p = 0.1) trend for lower SP levels in patients receiving opioids than in controls. Moreover, NPY levels were significantly lower in patients receiving analgesia (p = 0.042), (d) lower ACTH and cortisol levels in survivors (p \ 0.05) (e) at the day of least severity, a significant association between SP levels and MOF was observed (r = 0.29, p \ 0.05).CONCLUSIONS. (a) Despite the fact that NPY and SP are stress neuropeptides, their levels appear to be decreased in MODS patients. It is worth-exploring whether critical illness may be a state of suppressed activity of some neuropeptides, (b) the observed association between stress neuropeptide levels and survival in critical illness needs to be explore further, (c) bedside measurement of selected neuropeptides in the future may provide an estimation of pain in uncommunicative patients.Hence, the study of stress neuropeptides may provide new insight for the management of the critically ill. OBJECTIVES. The objective of this study was to compare septic and non-septic inflammatory process in critically ill patients with respect to paraoxonase 1 activity, lipid profile and lipid peroxidation markers.

METHODS. Analyzed were serum paraoxonase 1 activity, lipid profile, oxidized low density lipoproteins and conjugated dienes in critically patients with sepsis n = 30), age/sex/AP-ACHEII matched critically ill controls with non-septic SIRS (n = 15) and age/sex matched outpatient controls without inflammation (n = 30).RESULTS. The activity of PON 1 was lower in septic patients (96.7 ± 47.4 U/ml) as well as in patients with non-septic SIRS (107.8 ± 39.8 U/ml) compared to healthy controls (158 ± 40.5 U/ml). The decrease in paraoxonase 1 activity, high density lipoprotein cholesterol and apolipoprotein A-1 concentrations was closely followed by the counter increase of serum amyloid A in both groups of patients. There was no difference in paraoxonase 1 activity between septic and non-septic critically ill patients. The concentration oxidized low density lipoproteins and conjugated dienes as markers of lipid peroxidation, were raised in both septic and non-septic SIRS critically ill patients as compared with healthy controls. However there was no difference between both critically ill patient groups.CONCLUSIONS. The decreased activity of paraoxonase 1 in negative correlation with lipid peroxidation markers offers a potentially useful nonspecific marker of inflammation in critically ill patients.GRANT ACKNOWLEDGMENT. OBJECTIVES. In the present study, we studied the short-term and direct effects of IVIG with sepsis.METHODS. 16 patients was investigated. Following the administration of 5 g of IVIG for 1 h, we took blood samples immediately following IVIG treatment and at 1 h after IVIG treatment. Blood samples taken at 1 h and just prior to IVIG administration were used as controls.

While there was no difference between 1 h before and just prior to IVIG treatment, statistically significant decreases were observed in the levels of IL-6 after the administration of IVIG. No significant changes were observed in the levels of tumor necrosis factor-a and high mobility group box-1.Changes in serum TNFa, IL-6, HMGB1

We confirmed the results of previous animal studies. While we reported that the administration of IVIG directly reduces the levels of IL-6 in patients with sepsis, a further prospective study of the ant-cytokine effects following IVIG treatment will be conducted in the near future. OBJECTIVES. To investigate the levels of nucleosome in septic patients and to determine whether nucleosome could serve as a biomarker for sepsis.

Sixty-four consecutive patients who were newly admitted in surgical intensive care unit at two university hospitals were enrolled in this study. Whole blood samples were drawn within 12 h of admission and on the third, fifth and seventh days. A last blood sample was drawn after recovery at ICU discharge in survivors or at imminent death in the cases of non-survivors. Plasma levels of nucleosome as well as cytokines IL-6 and IL-10 were detected by means of enzyme linked immunosorbent assay.

. Fifty patients were diagnosed as sepsis and the other fourteen patients were classified as controls. Plasma levels of nucleosome were significantly higher in septic patients than in controls (two-way ANOVA, p \ 0.05), while the levels of IL-6 and IL-10 were comparable between septic patients and controls. The septic patients presented the highest levels of nucleosome on the admission day, which was significantly different from the admission levels of nucleosome in controls (4.16 ± 3.48 vs. 1.61 ± 1.36, p \ 0.05). The plasma levels of nucleosome between survivors and nonsurvivors showed no statistical significance.CONCLUSIONS. Plasma levels of nucleosome may serve as a valuable biomarker for sepsis.INTRODUCTION. High Mobility Group Box protein 1 (HMGB-1) is a cytokine that can mediate inflammatory response in different conditions included rheumatoid arthritis, infections, sepsis and septic shock. HMGB-1 released by activated macrophages/monocytes acts as a late mediator of sepsis. Studies have shown that serum HMGB-1 concentrations were elevated in patients with severe sepsis.OBJECTIVES. In the present study, we evaluated the role of the HMGB-1 levels at the time of admission at the Intensive Care Unit (ICU) as predictor of outcome in patients with sepsis and septic shock.METHODS. Forty-four Patients admitted to the ICU with sepsis and septic shock was recruited. Serum samples were obtained at the time of admission for the determination of HMGB-1 levels. The results were correlated with the origin of sepsis, severity, organ dysfunction, requirements of mechanical ventilation and vasoactives, days at the ICU, comorbidities and mortality at the ICU and 28 days after admission.

Twenty-six patients were male (68.2 vs. 31.8%). Septic shock was present in 28 patients (63.6%). The mortality rate at the ICU was 31.8% (n = 14) and 38.6% (n = 17) at day 28th. HMGB-1 levels were 34.7 ng/mL ± 39.02 (1.34-230.2 ng/mL). HMGB-1 levels were significantly higher in non-survivors at the ICU than in survivors (56.81 ng/mL ± 58.69 vs. 24.38 ± 19.18, p \ 0.05). Higher levels of HMGB-1 in serum at the admission were correlated with a higher mortality rate in the ICU (p \ 0.05) but not at day 28th (p = 0.07). These levels were not correlated with days at the ICU, requirements of vasoactives, mechanical ventilation, and APACHE score.CONCLUSIONS. The determination of HMGB-1 levels at admission at the ICU in patients with sepsis and septic shock is a good predictor of worse outcome and lethality.INTRODUCTION. Recent experimental and clinical data (1, 2) support the hypothesis that costimulatory molecules, such as CD40, play an essential role in the innate immune response during sepsis. Expression of CD40 on the surface of monocytes could represent an important pathway in the modulation of the production of several key inflammatory mediators.OBJECTIVES. To investigate whether the expression of CD40 molecule on the surface of plasma monocytes differs among the various stages of sepsis.

A total of 32 participants (12 ICU patients with sepsis, 14 ICU patients with septic shock and 6 healthy controls) were included in the study (male patients 73.1%, mean age 46.6 ± 16.8 years). Inclusion criteria: ICU patients on mechanical ventilation with first episode of sepsis or septic shock during current hospitalization. Exclusion criteria: immunosuppression, neoplasia, autoimmune disease, cardiovascular disease. Age, gender and comorbid conditions were recorded. A blood sample for quantification of CD40 expression was obtained at the time of enrollment (day 1), and on the fifth day after the onset of sepsis; measurement was made on the same day. CD40 expression on the surface of plasma monocytes (on days 1 and 5) was assessed by flow cytometric analysis. Statistical analysis: Kruskal-Wallis test to identify difference of CD40 expression among the 3 groups was performed. Post-hoc analysis was made by Mann-Whitney U test between independent groups, using Bonferroni correction for multiple comparisons. ROC curve analysis was used to determine the accuracy of CD40 in identifying patients with sepsis or septic shock.

Patients with sepsis had significantly higher levels of CD40 (day 1) compared with healthy controls subjects (0.86 ± 0.41 vs. 0.34 ± 0.07, p B 0.006). On the contrary, patients with septic shock did not show any significant difference compared with controls. A ROC curve analysis for CD40 (day 1) (AUC = 0.72, p B 0.05), revealed that a cut-off value of 0.53 could predict patients with sepsis with a sensitivity of 91% and a specificity of 57%.CONCLUSIONS. Upregulation of CD40 expression may reflect a protective phenomenon during sepsis. On the contrary, low CD40 expression could represent impaired immune function associated with more severe disease. 

In order to increase the cardiac output in the septic shock patients, according to surviving sepsis campaign team, norepinephrine (NE) or dopamine administration was recommended. The both agents increase the sympathetic tone which antagonize against parasympathetic activity used for gastrointestinal motility (involved gastric emptying). Then, it is raised a question whether NE delayed the gastric emptying or not.OBJECTIVES. This study was aimed to evaluate the gastric emptying in the septic shock patients with norepinephrine.

A prospective observational study involved 54 adult septic shock patients, who received NE continuously in ICU Sardjito General Hospital (Yogyakarta, Indonesia). Patients with any head pathologies (trauma, surgical procedures for tumor or bleeding), any gastrointestinal or abdominal pathologies (diarrhea, trauma, surgical procedures for cancer, peritonitis, ileus etc.), and administrations of metochlopramide or alinamin were excluded. Nutrition fluids (100 ml) was given passively via nasogastric tube, then after 60 min the tube was aspirated. The volumes of aspirates were recorded in % as a gastric residue. Once measurement was done with time randomly for every patient. At the measurement time were recorded the dose of NE and the vital signs.RESULTS. The gastric residues were 41.76 ± 28.11% (17 patients), 45.46 ± 25.27% (13 patients) and 13.57 ± 8.52% (7 patients) for the doses of NE of 0.1, 0.2 and 0.3 lg/kg b.w./ min respectively. At the NE doses of 0.4, 0.5 and 0.6 lg/kg b.w./min, all of the gastric residues were zero (17 patients). The correlation between the NE doses and the gastric rescues was statistically significant (p: 0.01). The mean arterial pressures (MAP) were 84.16 ± 18.01 mmHg (ranges from 58 to 120 mmHg. There was no significantly correlation between MAP and the gastric residues.CONCLUSIONS. The gastric emptying in the septic shock patients was not disturbed by administration of NE. INTRODUCTION. Anemia is a frequently encountered problem on the intensive care unit. Several factor lead to anemia, among which are traumatic blood loss and the drawing of blood for routine laboratory tests. It's not known how this may affect innate immunity. Hepcidin is a central regulator of iron homeostasis. It is induced in response to iron and inflammation and reduced in response to anemia and hypoxia. The suppression of hepcidin leads to the internalization and degradation of the iron exporter ferroportin on intestinal cells and macrophages, leading to the uptake of iron from the gut and the release of iron from the macrophages from the reticulo-endothelial system (RES). These cells are central to the innate immune response and the altered iron status of these cells due to suppression of hepcidin may affect the inflammatory response of these cells. We tested the hypothesis that phlebotomy in human volunteers would lead to a suppression of the innate immune response. This abstract provides data of a pilot study carried out in 3 subjects.

To investigate the effect of phlebotomy on the innate immune response of whole blood in human volunteers.METHODS. Three volunteers were subjected to the letting of 400 ml of blood by phlebotomy. Blood for the determination of hemoglobin and iron parameters, leucocyte count and differential, and hepcidin-25 was drawn at day 0, 4 and 7 after phlebotomy. Further whole blood stimulation was carried out at each time point by adding 0.5 ml heparin anticoagulated whole blood to a prepared tube containing endotoxin, pam3cis or RPMI as a control. Final concentrations of LPS and P3C were 10 ng/ml and 1 lg/ml respectively. These tubes were incubated at 37°C for 24 h and centrifuged for 10 min at 4,000 g. The supernatant was frozen at -80 until the measurement of TNF-alfa and IL-6 by ELISA. Cytokine production was corrected for the number of monocytes present. Data are expressed as mean ± SEM.

Hemoglobine decreased from 8.5 ± 0.3 mmol/l at baseline to 8.0 ± 06 at day 4. It returned to normal at day 7. There were no apparent changes in serum iron levels. There was however a clear decrease in serum ferritin levels from 162 ± 48 at baseline to 109 ? 46 at day 7. Leucocyte count and differentiation did not show any significant changes. Hepcidin was clearly suppressed from 4 to day 7 after phlebotomy (from 6 ± 1 to 2 ± 1). TNF-alfa production dropped from 81 to 47 ng/10 7 monocytes at day 7. IL-6 production dropped from 327 to 195 ng/10 7 monocytes. Hepcidin levels correlated well with cytokine production (r 2 0.53 for TNF-alfa, r 2 0.51 for IL-6).CONCLUSIONS. Phlebotomy leads to suppression of the innate immune response in whole blood. This could be a result of the intracellular decrease of iron in immune cells due to the systemic suppression of hepcidin. These findings are relevant to critical care patients that are subject to the repeated drawing of blood while their immune system is often compromised. INTRODUCTION. Hypothermia and hyperthermia occur in many pathological states presenting to the emergency department. Both these processes are known to significantly impair coagulation pathways but as yet there is little evidence to show what affect they have on the evolving clot structure. Previous studies have attempted to determine the effect of temperature on whole blood coagulation using techniques such as thromboelastometry (TEG) but its ability to provide meaningful outcomes in terms of clot quality and structure remains elusive. Recent studies have highlighted the potential of a new technique, Gel Point (GP) and Fractal Dimension (d f ), as a functional biomarker in haemostasis.

To explore both the changes in coagulation pathways and their associated effect on clot structure and quality based on the new biomarkers, GP and df.

Following full ethical approval, 30 healthy whole blood samples were obtained from individuals and tested at temperatures of 27°C (n -6), 31°C (n -6), 35°C (n -6), 37°C (n -6), 41°C (n -6). An oscillatory shear technique [1] using an AR-G2 Instrument (TA Instruments) was applied to each sample. The GP, which indicates the formation of the fibrin network, was obtained for each sample using the Chambon-Winter Gel Point criterion [2] . This method provides the basis from which d f can be determined [3] to interpret the structural properties of the clot network. The results were compared with the standard TEG analysis.

Firstly, results showed a significant progressive change in the clot structure by this new biomarker across the whole temperature range (27-43°C). Secondly, it also highlighted a significant and meaningful correlation between coagulation pathway change (Time to GP, TGP) and the eventual clot outcome (fractal dimension). The TGP of the incipient clot was prolonged and the corresponding d f decreased with reduced temperature values. Although, the changes in the coagulation pathway of the TEG (R time) and the rheometer (TGP) correlated, the new biomarker, d f , provided additional structural data on the fibrin network formed and highlighted the relationship between coagulation pathway changes and the eventual fibrin clot structure.CONCLUSIONS. In this study, we describe and quantify for the first time how temperature affects the coagulation pathways and how this impacts on the fibrin clot network, morphology and strength by using the new biomarkers, GP and d f . The potential of these new biomarkers in determining the effects of temperature change in critical illness and injury needs to be evaluated clinically.