17th Annual Congress – Berlin, Germany – 10–13 October 2004

S79

294

296

PULMONARY ENDOTHELIAL PARAMETERS AS SURVIVAL PREDICTORS IN SEPTIC ALI PATIENTS TREATED WITH rhAPC

DIFFERENCES OF HEART RATE VARIABILITY TIME KINETICS IN SURVIVORS AND NON-SURVIVORS OF MODS

Kaziani K1, Athanasiou C1, Korovesi I1, Douka E2, Armaganidis A2, Roussos C1, Kotanidou A1, Orfanos S E2 1 1st Department of Critical Care, Evangelismos Hospital and M.Simou Lab, 22nd Department of Critical Care, Attikon Hospital and M.Simou Lab, University of Athens Medical School, Athens, Greece

Schmidt H1, Tymiec P1, Hoyer D2, Müller-Werdan U1, Werdan K1 Dept. of Medicine III, University Halle, Halle, 2Institute of Pathophysiology, University Jena, Jena, Germany

INTRODUCTION: Pulmonary capillary endothelium-bound (PCEB) angiotensin converting enzyme (ACE) activity, a quantifiable index of pulmonary endothelial function, decreases early during acute lung injury (ALI) and the acute respiratory distress syndrome (ARDS), and correlates with syndrome severity. Protein C pathway is a major regulatory mechanism between microvascular coagulation and inflammation that is compromised in severe sepsis. Increased procoagulant activity and reduced fibrinolysis are present in the lung microvascular bed in ALI/ARDS, injure pulmonary endothelium thus contributing to the syndrome pathophysiology, and constitute the rationale for treating septic ALI/ARDS with recombinant human activated protein C (rhAPC). In this study we investigated if baseline PCEB-ACE activity in septic ALI/ARDS, prior to rhAPC initiation, correlates with patient survival. METHODS: Applying indicator-dilution type techniques, we estimated PCEB-ACE activity in sixteen (16) severely septic patients suffering from ALI/ARDS, just before the initiation of the 96 h rhAPC (24\mug/kg/h) IV infusion (baseline measurement). PCEB-ACE activity was estimated as the single pass transpulmonary hydrolysis (v) and percent metabolism (%M) of the synthetic ACE substrate 3H-benzoyl-Phe-Ala-Pro (3H-BPAP), and as the modified kinetic parameter Amax/Km (an index of functional capillary surface area-FCSA). Additional recorded or estimated measures were: age, lung injury score (LIS), PaO2/FIO2, chest X-ray score, septic stage, APACHE II score, and SOFA score. Mortality was assessed post 28 days: 9 patients survived (Group 1), and 7 patients had died (Group 2). RESULTS: Both groups exhibited decreased PCEB-ACE activity parameters as compared to previously reported normal human values. Group 2 patients revealed significantly decreased substrate v and %M values as compared to Group 1 (0.24±0.03 vs 0.41±0.06, and 20.9±2.5% vs 32.7±3.7, respectively; means±SEM, p<0.05 for both parameters by t-test). No significant differences were observed between the two groups in any other parameter. A multiple stepwise logistic regression analysis that included BPAP %M and v, LIS, APACHE II, SOFA and septic staging profile was performed; BPAP % M was found to be the only parameter related to 28-day survival (OR=1.25, p=0.08). CONCLUSION: These preliminary data suggest that in our population baseline substrate %M (a reflection of PCEB-ACE activity per capillary) was the only significant predictor of survival in septic ALI/ARDS patients treated with rhAPC, probably implying that patients with less altered pulmonary endothelial function benefit most from rhAPC administration.

Oral Presentations Evaluating the risk of septic patients: What’s new? – 295-297 295 MORTALITY, IPPV DURATION AND INOTROPES COMPARED WITH ADMISSION MICROALBUMINURIA IN 431 ICU PATIENTS Gosling P1, Czyz J2, Manji M2. 1Clinical Biochemistry, 2Critical Care, University Hospital Birmingham NHS Trust, Birmingham, United Kingdom INTRODUCTION: Recent studies showing urine albumin creatinine ratio (ACR) and predictive value for outcome in ICU patients did not examine the effects of diagnostic group or collection time. This study compared bedside ACR measurement at two time points within 6 hours of ICU admission with diagnostic group, mortality, duration of IPPV and inotrope requirements in medical and surgical patients. METHODS: Consecutive patients admitted to a 17 bed adult general intensive care unit in a university teaching hospital were recruited. The only exclusion criteria were anuria, failure to collect urine samples, or inability to measure urine albumin due to hyper pigmentation of the urine specimen. Bedside ACR was measured within 15 minutes of ICU admission (ACR1) and after 4-6 hours (ACR2) by nursing staff using a Bayer DCA 2000 bench top analyser. Reference range <2.3 mg/mmol. RESULTS: ACR1 was above the upper limit of normal in 386/431 (89.6%) of patients and ACR2 in 334/431 (77.5%). For all patient subgroups, median ACR fell significantly after 4-6 hours (table). ACR1 and ACR2 were positively associated with mortality and duration of inotrope therapy (ACR1 rs=0.22 p<0.0001: ACR2 rs=0.21 p<0.0001) and with duration of mechanical ventilation (ACR1 rs = 0.14 p=0.0031: ACR2 rs=0.11 p=0.0197).

1

INTRODUCTION: We have recently shown that the heart rate variability (HRV) parameter VLF (“very low frequency”) is a potent predictor of mortality in multiple organ dysfunction syndrome (MODS) when assessed at admission to ICU. In the present study we aimed to characterize the VLF in the course of ICU stay and hypothesized that there is a difference between survivors (S) and non-survivors (NOS). METHODS: We enrolled 40 patients into this study and studied the HRV at four sequential time points (admission [A]= T1, 4 days after A = T2, 7 days after A = T3 and 14 days after A to ICU = T4). The HRV was calculated according to the international standards (summary in [1]). VLF contains information on parasympathetic modulations of heart rate. All patients were followed up for in-hospital mortality. APACHE II, SOFA and SAPS II scores were calculated as described earlier. We also aimed to assess whether VLF increases or decreases during ICU stay and calculated the difference of VLF at last time point prior discharge (S) or fatal outcome (NOS) and VLF at admission (T1, in ms2). RESULTS: Demographic data of the MODS patients (29 S / 11 NOS, p of ANOVA): age 58.5±13.8 / 61.6±12.0 years, p=0.5, height 173.1±6.5 / 172.4±8.4 cm, p=0.8, weight 76.3±12.5 / 72.8±15.7 kg, p=0.5. Severity of illness at admission: APACHE II 28.7±8.1 / 34.1±7.6, p=0.1, SOFA score 11.7±3.8 / 12.8±4.2, p=0.2, SAPS II score 61.9±17.9 / 72.1±17.4, p=0.2. The VLF data at the observed four time points were as follows (ms2): T1 93.4±102.5 (S) / 143.9±320.5 (NOS), p=0.4, T2 187.2±387.8 / 75.3±96.9, p=0.7, T3 141.0±156.4 / 42.0±68.3, p=0.2, T4 243.4.2±390.5 / 28.7±20.8, p=0.5. The differences of VLF (ms2) at last time point before discharge (S) or fatal outcome (NOS) minus VLF at T1 were 114.7±252.8 (S) and -105.3±283.6 (NOS), p=0.02, which means a slight increase of VLF in S and a significant decrease of VLF in NOS in the course of ICU stay. CONCLUSION: The HRV parameter VLF is a powerful predictor of mortality when assessed at ICU admission. Moreover, this parameter seems to have different time kinetics in the course of the ICU stay: in survivors it increases slightly but in non-survivors the blunting of HRV is even more pronounced. REFERENCE: (1) Schmidt et al. Intensive Care Medicine (2004) DOI 10.1007/s00134-003211131-2 Grant acknowledgement: HS, DH, UMW: DFG SCHM 1398/3-1,3-2; HOY 1634/8-1,8-2. HS and KW: DFG SFB 598/A7

297 HYPOTHALAMUS ENDOCRINE FUNCTION IN SEPSIS Borkowski J1, Siemiatkowski A1, Wolczynski S2, Czaban S L1, Jedynak M1 Department of Anaesthesiology and Intensive Care, 2Department of Gynaecological Endocrinology, Medical University of Bialystok, Bialystok, Poland

1

INTRODUCTION: Dysfunctions of the central nervous system in sepsis are diagnosed in clinical practice, but we hardly know the reason of these disturbances. The hypothalamus endocrine function is incorrect as well. Unfortunately clinical value of this changes is unknown. The hypothalamus is a source of two very important hormones - oxytocin and antidiuretic hormone (ADH) that are markers of its function. Both of them are necessary for the human body homeostasis, especially cardiovascular system. ADH is also used as a potent vasoconstrictor in septic shock. The aim of our study was to qualify the prognostic value of the oxytocin and antidiuretic hormone serum concentrations changes in patients with sepsis. METHODS: The study was carried out in 28 septic patients, survivors (n=14) and nonsurvivors (n=14). Sepsis was defined according to AACP/SCCM criteria. 16 healthy volunteers served as controls. To quantify organ dysfunction, APACHE II (The Acute Physiology and Chronic Health Evaluation Score II), APACHE III (The Acute Physiology and Chronic Health Evaluation Score III), MODS (Multiple Organ Dysfunction score), SOFA (Sequential Organ Failure Assessment) and LIS (Lung Injury Score) scores were used, and blood assays of the oxytocin and ADH concentrations in serum were done at the moment of sepsis recognition and on the 1st, 2nd, 5th and 10th day of the observation between 8.00 a.m. and 9.00 a.m . RESULTS: Significant differences were not observed between septic patients and controls in ADH (16,8±8,84 pmol/L and 31,5±26,85 pmol/L respectively) and oxytocin serum concentrations (3,2±3,20 pg/ml and 2,3±1,89 pg/ml respectively). At the same time oxytocin serum level was significantly higher in nonsurvivors (4,2±3,71 pg/ml) compared with survivors (2,2±2,22 pg/ml). ADH serum concentration between nonsurvivors and survivors did not differ. In septic patients we found significant correlation between the oxytocin serum concentration and APACHE II, APACHE III, MODS, SOFA and LIS scores. We did not show such correlation in the ADH serum level. CONCLUSION: Sepsis induces dysfunction of central nervous system which are reflecting in hypothalamus hormones concentrations. Therefore the oxytocin serum level might be used as an early prognostic factor of poor outcome in this disease.

CONCLUSION: ACR1 and ACR2 were associated with duration of mechanical ventilation, requirement and duration of inotrope therapy and ICU mortality. Median ACR1 and ACR2 were higher in medical patients than surgical patients and fell by half within 6 hours. Any comparison of microalbuminuria with other predictors of ICU outcome should ensure these variables are controlled. Early ACR assessment may be a useful adjunct to clinical assessment of patient risk. REFERENCES: 1. Abid O et al Predictive value of microalbuminuria in medical ICU patients: results of a pilot study. Chest. 2001 ;120:1984-8. 2. Gosling P et al Mortality prediction at admission to intensive care: a comparison of microalbuminuria with acute physiology scores after 24 hours. Crit Care Med. 2003;31:98-103. Grant acknowledgement: JC received a grant from Bayer Diagnostics Ltd

S80

17th Annual Congress – Berlin, Germany – 10–13 October 2004

Oral Presentations Acute myocardial infarction – 298-300 298 BLOOD LACTATE AS AN EARLY BIOCHEMICAL MARKER OF MYOCARDIAL INFARCTION Radonic R, Gasparovic V, Ivanovic D, Merkler M, Gubarev N, Zlopasa O, Gornik I. Dept. of Internal Medicine ICU, University Hospital Centre Zagreb, Zagreb, Croatia INTRODUCTION: There is a lack of good biochemical markers of myocardial ishaemia in the first hours of acute coronary syndrome. Myocardial ischaemia could result in increase in lactate production, with repercussions on blood lactate level. This could be helpful in the triage of patients with chest pain (1). The aim of this double blind prospective clinical study was to validate the blood lactate as an early biochemical marker of myocardial ischaemia in context of clinical picture. METHODS: Blood lactate was measured in patients with chest pain, admitted in our ED. The attending physician and reviewer were blind for blood lactate values. Patients with shock and with other conditions which could potentially lead to elevated blood lactate were excluded. Patients were divided in groups: AP = stabile angina, UAP = unstable angina, NSTEMI= acute myocardial infarction without ST segment elevation, STEMI = acute myocardial infarction with ST segment elevation, PM = pectoral myalgia, Oth = other or unclear. Patients with pectoral myalgia served as controls. RESULTS: In 159 patients all criteria were met. Mean blood lactate level was 1.94 mmol/L in 24 AP pts, 2.12 mmol/L in 26 UAP pts, 2.22 mmol/L in 17 NSTEMI pts, 3.19 mmol/L in 37 STEMI pts, 1.57 mmol/L in 44 PM pts and 2.20 mmol/L in 11 other pts. Blood lactate levels were significantly higher in patients with STEMI then in other groups, even in those presenting in the first hours and in those in whom initial troponin I was <1 ug/L.

300 TREATMENT OF ACUTE MYOCARDIAL INFARCTION COMPLICATED BY SUSTAINED VENTRICULAR ARRHYTHMIAS Lazarov L1, Petrovski B2, Boskov V3, Georgievski A1, Trajkov I3, Dimitrovski M3, Tosev S4, Georgievska B1 1 Postintensive care, 2Angio Lab, 3CCU, 4Angiology, Institute for Heart Diseases,Clinical center,Medical faculty, Skopje, The former Yugoslav Republic of Macedonia INTRODUCTION: Primary ventricular fibrillation (VF) or rapid polymorphic ventricular tachycardia (VT) are still the main cause of high mortality in the early stage of acute myocardial infarction (AMI). The aim of the study was to investigate the in-hospital incidence, most appropriate treatment and mortality of patients (pts) with AMI complicated by initial sustained ventricular arrhythmias (VF/VT). The main question after initiative resuscitation (CPR) was what should be the most appropriate treatment of AMI: primary angioplasty (PCI) or thrombolysis. METHODS: Total of 6.684 pts with AMI admitted in the CCU were studied (413 pts with VF/VT). The study period was January 1994 to January 2003. We have analyzed separately the data of subgroup of pts with AMI+VF/VT treated in the period 1999-2003 (3253 pts with AMI, 201 with VF/VT) because of predominant use of PCI in the last few years. RESULTS: In-hospital incidence of VF/VT in total group was 6.3% with mortality of 44.5%. In the subgroup the incidence of VF/VT was 6.2% with mortality rate of 43.8%. Independent predictors of in-hospital VF/VT were anterior wall AMI (62% pts), lower SBP (<100mmHg), higher Killip class and shorter time to PCI/thrombolysis (<3 hours). After initial cardioversion and CPR, thrombolytics (Streptokinase or TPA) were used in 22% pts and PCI was performed in 23.9% pts. The 3rd group of pts (44%) was treated with low molecular weight Heparin (LMWH)+Lidocaine/Amiodarone/beta blockers. Acute heart failure occurs in 25% pts. The mortality rate in PCI group was 18.7%, in thrombolytic group 33.3% and in the 3rd group 59.8%. Most of deaths in all groups happened during the first hours of admission. CONCLUSION: Despite the considerable improvement in the treatment of AMI in recent years, the onset of VF/VT is associated with a poor prognosis. It is accompanied usually by extensive necrosis w/wo acute heart failure. Primary PCI and thrombolysis have significant effect on lowering the mortality rate of AMI+VT/VF, but PCI should be the first choice after initial CPR.

CONCLUSION: Blood lactate is significantly elevated in patients with STEMI, even in the very early phase, prior than troponin I and can contribute in the stratification of patients with chest pain as an early biochemical marker of myocardial ishaemia. REFERENCE(S): 1) Schmiechen NJ, Han C, Milzman DP: ED use of rapid lactate to evaluate patients with acute chest pain. Ann Emerg Med 1997;30(5):571-7

299 CLINICAL PROFILE AND DETERMINANTS OF REINFARCTION COMPLICATING ACUTE MYOCARDIAL INFARCTION Cebrián J1, Colomina F1, Serralta M1, Ahumada M1, Valencia J1, Cardona J1, Perez M1, Cabadés A1 1 PRIMVAC, INSVACOR, Valencia, Spain

Oral Presentations Ventilator-associated pneumonia – 301-303 301 NOSOCOMIAL PNEUMONIA PROPHYLAXIS IN THE NEUROSURGICAL ICU BY CONTINUOUS SUBGLOTTIC ASPIRATION

INTRODUCTION: The aim of the study is to determine the global incidence and mortality of reinfarction (REAMI) complicating acute myocardial infarction (AMI) admitted in intensive coronary care unit (ICCU) and also to identify its predictive factors.

Petrikov S S1, Tsarenko S V1, Krylov V V1, Gukova E V1, Titova J V1 1 Neurosurgery ICU, Scientific institute of emergency medicine named after N.V.Sklifosovskiy, Moscow, Russian Federation

METHODS: We have studied all patients with AMI enrolled in the PRIMVAC registry (1) in the period 1995-2000. Differential characteristics of patients with REAMI were determined by bivariate analysis (Student t test and Pearson chi square). Adjusted Odds ratios (OR) were estimated by logistical regression to study the association between the presence of REAMI and the following variables: age, gender, high blood pressure, hypercholesterolaemia, smoking status, diabetes mellitus, previous myocardial infarction, Q wave in electrocardiogram and thrombolysis.

INTRODUCTION: Aspiration of glottis and gastric content is the main risk factor of nosocomial pneumonia in neurosurgical ICU patients. Therefore, continuous subglottic aspiration is a perspective method of nosocomial pneumonia prevention. We investigated the role of continuous subglottis aspiration on nosocomial pneumonia development in the 12-bed neurosurgical ICU.

RESULTS: There were 12, 071 patients. Mean age was 65.5 years old and woman percentage was 23.8%. The REAMI incidence was 2.8%. REAMI group was significantly older (69.5 years; DE 10.4) than no REAMI group (65.4 years; DE 12.1). Female sex was more common in the REAMI group (34.3% vs. 23.5%; p<0.001). Mortality was significantly higher in the REAMI group (37.8% vs. 12.6%; p<0.001). After multivariate analysis, only age (OR=1.2), diabetes mellitus (OR= 1.49), previous myocardial infarction (OR= 1.38) and the presence of Q wave in the electrocardiogram (OR=1.36) were independently associated with the presence of REAMI. CONCLUSION: In the Spanish Valencian Community REAMI in the ICCU implies a high mortality. Some simple clinical factors presents at admission can help to identify patients at risk. REFERENCE(S): 1) Cabadés A, Echanove I, Cebrián J, Cardona J, Valls F, Parra V et al. Caracteristicas, tratamiento y pronostico del infarto agudo de miocardio en la Comunidad Valenciana en 1995: resultados del registro PRIMVAC. Rev Esp Cardiol. 1999 Feb;52(2):123-33.

METHODS: 32 patients of the neurosurgical ICU were enrolled in the study. All the patients were mechanically ventilated starting from the first day of the hospitalization. Tracheotomy was developed on the 2-3 day after admission. 16 patients (group 1) received continuous subglottic aspiration through a special channel of tracheostomy tubes (Blue Line Ultra (Portex, England)). The patients of the second group (n=16) were treated with common tracheostomy tubes without the subglottic aspiration channel. The patients were randomized according to their age, sex, type of the disease and protocol of intensive care. 93,7% of patients in the 1-st group and 93,7% of patients in the 2-nd group had Glasgo Coma Scale lower than 9 on admission to the ICU. Mortality was 50% (n=8) in the 1-st group and 56,2% (n=9) in the 2-nd group (P>0,05). RESULTS: Despite the fact that the average time of mechanical ventilation was 18±9,6 days in the first group and 11,7±7,8 days in the second group (p<0,01) nosocomial pneumonia rates were: 48,9 per 1000 days of mechanical ventilation in the 1-st group and 83,3 per 1000 days of mechanical ventilation in the second group (p<0,05). We registered the first signs of pneumonia on the (M±?) 8±1,7 day after admission in the first group and on the 4,1±1,1 day after admission in the second group (p<0,0001). The average volume of subglottic secretions was 102,5±66,9 ml per day (from 30 to 279 ml per day). CONCLUSION: Continuous subglottic aspiration is an effective method of nosocomial pneumonia prophylaxis in neurosurgical ICU patients. It decreased the rate and prolonged the period of mechanical ventilation without clinical and X-rays signs of pneumonia.

17th Annual Congress – Berlin, Germany – 10–13 October 2004

302 PATIENTS REQUIRING INVASIVE MECHANICAL VENTILATION AND VAP: ANALYSIS OF MORTALITY Borges M1, Diaz E2, Alvarez-Lerma F3, Palomar M4, Jordá R5, Friedmann G6, Rello J2, Grupo FR en NAV G7. 1ICU, Hospitl Son Llatzer, Palma de Mallorca, 2ICU, H. Joan XXIII, Tarragona, 3ICU, H. Mar, 4ICUt, H. Vall´Hebron, Barcelona, 5ICU, C. Rodger, Palma de Mallorca, Spain, 6ICU, H. Clínicas, Porto Alegre, Brazil, 7ICU, GTEI, Spain.

S81

Oral Presentations Neurological intensive care and emergency medicine – 304-306 304

CONCLUSION: We identify many variables of mortality in a MV patients in an univariable analysis, but in a logistic regression only five factors: presence of VAP, age more 65 years, APACHE >15, trauma patients and neutropenia. Much „classical“ factors related with VAP mortality were not significant in this prospective and multicenter study.

INTENSIVE CARE DELIRIUM CHECKLIST: ANALYSIS OF A SHORTER CHECKLIST’S PERFORMANCE Ouimet S1, Dumont M2, Skrobik Y1 1Critical Care, 2Corev, Maisonneuve Rosemont Hospital, Montréal, Canada INTRODUCTION: Delirium in the critical care setting is morbid(1, 2). Simple, reliable screening tools permit detection of clinical fluctuations over 24 hours. Timely recognition and early interventions may benefit the patient.We previously demonstrated an 8-item intensive care delirium checklist’s reliability when compared to a physician’s clinical diagnosis of delirium (3). In clinical use of the scale, some checklist items were present simultaneously. We tested whether some items were redundant and if an abbreviated checklist could prove as accurate as the 8-item scale. METHODS: We evaluated 500 patients sequentially admitted to a medical surgical ICU. Delirium diagnosis by the intensivist was considered the gold standard, as we had previously shown this was equivalent to a psychiatrist’s diagnosis (1, 3). All patients had an 8-item delirium checklist evaluation by the bedside nurse each 8 hour shift. Checklist data was compared to the clinical diagnosis of delirium for 2514 patient- days. Analysis: Stepwise logistic regression was used to determine the relative contribution of each of the checklist’s 8 items to the risk of delirium diagnosis. Redundant items without predictive ability were then sequentially removed, and the specificity and sensitivity of the “shortened” checklist re-assessed in comparison to the gold standard, the clinical diagnosis. RESULTS: Of 500 patients, 176 (35%) developed delirium.. Using all patient-days, the elimination of Psychomotor agitation or retardation and hallucinations-psychosis items, while maintaining altered level of consciousness, inattention, disorientation, inappropriate speech or mood, sleep-wake cycle disturbance as well as fluctuation permitted excellent prediction of delirium (Sn=96.9%, Sp=98.5%). Shortening the list further resulted in unacceptable loss of specificity and sensitivity (to 91.1%) Applying the 6-point abbreviated scale to the 1460 patient-days obtained from patients that had been clinically labeled as having delirium at least once during their stay showed a Sn=96.9% and Sp=97.7%. CONCLUSION: The abbreviated, 6 point delirium screening tool described above is sensitive and specific in detecting clinical delirium in the critical care setting. REFERENCE(S): 1) Dubois MJ Bergeron N Dumont M Dial S Skrobik Y Delirium in an intensive Care Unit: a study of risk factors; Intensive Care medicine (2001)27: 1297-1304 2) E. W. Ely, S. Gautam, R. Margolin, J. Francis, L. May, T. Speroff, B. Truman, R. Dittus, G. R. Bernard, S. K. Inouye. The impact of delirium in the intensive care unit on hospital length of stay .Intensive Care Med (2001) 27: 1892-1900 3) Bergeron, N. Dubois M.J. Skrobik, Y. Intensive care Delirium Checklist: evaluation of a new screening tool. Intensive care medicine, Vol.27 no.5 Mai 2001 pp.859-864 Grant acknowledgement: Fonds de Recherche en Santé du Québec

303

305

IMPACT OF ARDS IN RESOLUTION OF VAP

ADVANCED PREHOSPITAL CARE IN CARDIAC EMERGENCIES – PRECISION OF 112-DISPATCH BY THE POLICE

INTRODUCTION: Mechanical Ventilation (MV) is a very important treatment in ICU, but is a great risk factor for VAP. It had been studied in much reports, but with different results and conclusions about mortality. Many points are still controversial. The objective of this study was to make an analysis of mortality in patients with MV and special attention to patients with episodes of VAP. METHODS: Prospective, multicenter cohort study carried out in 17 ICU (15 in Spain and 2 in Brazil) during 12 months to investigate VAP risk factors. All patients ventilated > 12 hours were included. We collected 150 variables and analysed mortality in patients with MV. RESULTS: We included 1704 patients, male sex 65,5%,, mean age 57,3 (18) years, and ICU LOS 17,7 days. Recognized 403 episodes of VAP in 353 patients. The incidence of VAP was 22 episodes each 1000 days with MV. 89,5% of the 1704 patients death in ICU and 10,5% in a ward. The crude mortality of patients with MV were 38,2% (without VAP 36% and with VAP 46,6%, p<0,05). The crude mortality of 1704 patients in 28 and 60 days were 43,5% and 46,1% (NS), respectively: there was not differences between VAP and no-VAP groups in 28 days, but there was in 60 days (6,2% versus 12,5%, p<0,05). We identify 26 variables in univariable analysis associated with mortality: age, LOS, APACHE II, GCS, underlying condition, COPD, chronic renal failure, malignancy, neutropenia, chemotherapy, inmuno-deficiency, malnutrition, cardio-respiratoy arrest, ARDS, blood transfusion, sepsis, shock, previous pneumonia or infection and VAP. In a bivariable analysis was not significance differences in mortality according types or number of pathogens, type of treatment (monotherapy versus combined), early or late VAP, delayed of empiric treatment. Causes of mortality in MV - Variables MODS % Shock % Brain death % Hypoxemia %

Without VAP 44,4 22,3 22,3 10,9

Mortality of MV: logistic regression - Variables VAP Age > 65 years APACHE II > 15 Trauma patients Neutropenia

OR 1,78 1,37 3,12 0,37 6,49

VAP 62,3 15,1 11,9 10,7

p <0,05 <0,05 <0,05 NS CI 95% OR 1,36-2,32 1,08-1,73 2,40-4,05 0,25-0,56 2,13-19,6

Gualis B1, Vidaur L1, Rodriguez A H1, Ramirez R2, Rello J1 1 Critical Care, 2Internal Medicine, Joan XXIII University Hospital, Tarragona, Spain

Andersen M S1, Christensen E F1 Department of anaesthesiology, Aarhus University Hospital, Aarhus sygehus, Aarhus, Denmark

1

INTRODUCTION: Outcome measures in VAP enclose mortality rates, length of stay and costs. However, limited information(1-3)is available on the temporal patterns of resolution. Our objective was to know the pattern of resolution in patients with VAP, depending on the presence of ARDS.A secondary objective was to identify those variables which can be used to monitor the response. METHODS: Prospective observational study in a ICU. Inclusion criteria: intubated patients with VAP. Clinical resolution was defined when at least two of the following parameters improved: Fever<38, PaO2/FiO2>250, white blood cell count (WBC)<10000, clearance of orotracheal secretions and resolution of chest x-ray infiltrates. Analysis was performed by CIA statistical package. RESULTS: 107 patients with VAP were included.68 episodes of VAP were identified as controls and compared with 23 episodes with VAP plus ARDS. Resolution of fever,PaO2/FiO2 and WBC in episodes of VAP was present in 71.8%,75.6% and 52.6% of patients under three days of therapy. Indeed, more than 50% of episodes presented resolution of fever and PaO2/FiO2 within the first day of therapy. In contrast, radiologic opacities and clearance of secretions (median of 13 and 6 days of resolution)were late events. In patients with ARDS, resolution of fever remained the most early parameter, but like PaO2/FiO2 and WBC , after three days of therapy shows a significantly worse pattern:39.1%,17.4% and 26%respectively.Radiologic resolution was an extremely poor indicator and it was only present in 13% of ARDS patients after 15 days of follow up. Failure to improve after 48 hours of therapy was documented in 87% of ARDS patients and 43.6% of controls (p<0.05). CONCLUSION: Hypoxemia resolution and defervescence occurs very early in patients with VAP and both variables would serve as useful indicators of resolution. The subset of patients with ARDS have delayed patterns of pneumonia resolution , being defervescence the earliest parameter. REFERENCES: 1-Ioanas M, Ferrer M, Cavalcanti M,et al :Causes and predictors of nonresponse to treatment of intensive care unit acquired pneumonia. Crit Care Med 2004; 32:938-945. 2-Luna CM, Blanzaco D, Niederman MS et al: Resolution of ventilator-associated pneumonia : Prospective evaluation of the clinical pulmonary infection score as an early clinical predictor of outcome. Crit Care Med 2003; 31:676-682. 3-Dennesen PJ, Van der Ven AJ, Kessels AG, et al: Resolution of infectious parameters after antimicrobial therapy in patiens with ventilatorassociated pneumonia. Am J respir Crit Care Med 2001; 163:1371-1375 Grant acknowledgement: Supported in part by CIRIT SGR 2001/414, Distinció Recerca Universitaria (JR) and Red Respira-ISCiii-RTIC-03/11

INTRODUCTION: In the major part of Denmark, the dispatch of emergency ambulances is by the police. The dispatcher has the choice between dispatching a regular ambulance alone, or a medical emergency mobile unit (MECU) staffed with an anaesthesiologist to severe cases. Round the clock there is one MECU on duty in the city, with approx. 330.000 inhabitants. Patients with myocardial infarction (MI) and cardiac arrest are in need of immediate assistance, a service provided by all ambulances. Lower mortality has been shown in MI patients after MECU compared to a regular ambulance (1). Optimal utilization of the MECU requires proper dispatch. The purpose of the study was to determine the precision of the dispatch of the MECU towards cardiac emergencies. The recordings of 112 phone calls were studied to describe key information from the caller. METHODS: All MECU charts regarding MI, angina pectoris or cardiac arrest (defined as cardiac emergencies) were collected consecutively during a six month period Recordings of these patients’ phone calls to the police 112-central were retrieved and analysed for information about the keywords: „Chest pain“, „Pain radiating to the left arm“, „breathing“ and „level of consciousness“. RESULTS: A total of 393 patients were included in the study. The sensitivity of the police dispatch of the MECU towards cardiac emergencies was 75 %, the specificity 90 % and the positive predictive value 45 %. Among patients where the diagnosis was confirmed by the anaesthesiologist, information about chest pain was present in the 112-call in 51 % of cases. Information about pain in the left arm was present in 24 % of cases, about breathing in 48 % of cases and about level of consciousness in 66 % of cases. Among patients without confirmation of the diagnosis information about chest pain was present in the 112-call in 26 % of cases. Information about pain in the left arm was present in 14 % of cases, about breathing in 43 % of cases and about level of consciousness in 60 % of cases CONCLUSION: The level of sensitivity and specificity was found acceptable and the positive predictive value of 45 % was found to be of moderate precision. The precision of the dispatch performed by the police was not markedly different from results of medically staffed dispatch centres. Training the police in retrieving systematic information about specific keywords from all calls concerning suspected heart attacks would be one way of improving the dispatch. REFERENCE(S): 1: Christensen E F, Melchiorsen H, Kilsmark J, Foldspang A, Søgaard J: Anaesthesiologists in prehospital care make a difference to certain groups of patients. Acta Amaesthesiol Scand 2003; 47: 146-152

S82

17th Annual Congress – Berlin, Germany – 10–13 October 2004

306

308

FACTORS ASSOCIATED WITH FATAL OUTCOME OF POLYTRAUMATIZED PATIENTS WITH SEVERE BRAIN INJURY

DO SWISS PATIENTS WISH TO BE INVOLVED IN DNAR DECISIONS?

Henzler D1, Cooper D2, Mason K2, Tremayne A B3, Rossaint R1 1 Ansthesiology, University Hospital RWTH Aachen, Aachen, Germany, 2Intensive Care, 3Consultative Committee on Road Traffic Fatalities Victoria, The Alfred Hospital, Melbourne, Australia INTRODUCTION: Traumatized patients with brain injury have a high mortality and morbidity. In a pilot study (1) we identified non-survival to be associated with less volume resuscitation, higher body temperature and longer time to treatment, however results were biased by a lower Glasgow Coma Scale (GCS) at scene despite correct matching for Injury Severity Score (ISS) and head Abbreviated Injury Score (AIS). In the following study GCS was included in the matching criteria. METHODS: Deceased multi-trauma patients with head AIS 4-5 and ISS>15 were matched for GCS, age and Revised Trauma Score (RTS) with surviving patients 1:2. Respiratory and haemodynamic variables, fluids, temperature and key intervention times were assessed on arrival, at 4 and 24 hours after admission (mean±SD). RESULTS: 25 non-survivors(NS, age 38.0±19.5) were evenly matched with 50 survivors(S, age 31.4±8.5). GCS was 5.1±3.1(S) and 4.6±3.1(NS), ISS was 32.9±12.0(S) and 31.4±8.5(NS). There were no differences in amount of fluid resuscitation, blood products substituted or blood gases. Nonsurvivors had persistent lower temperatures, lower mean arterial pressures (MAP) during the first hours after admission and later insertion of an intracranial pressure (ICP) monitoring (391±271min.(NS) vs. 316±228min.(S), p=0.032)(Table 1).

Survivors Non-survivors

total fluids [l] 2.8±2.2 3.2±1.9

PaO2

PaCO2

[mmHg]

[mmHg]

301±178 291±160

42±9 44±13

Temp hemoglobin MAP [°C] [mg/dl] [mmHg] 36.2±1.5 107±44 80±15 35.1±1.4* 102±37 71±16*

*p<0.05 Non-survivors vs. Survivors CONCLUSION: In contrast to previous findings mild hypothermia was associated with worse outcome in severely brain injured trauma patients. Survival was not associated with the amount of fluid resuscitation, but early ICP monitoring and cerebral perfusion pressure (CPP) targeted therapy, which was probably <60mmHg in the presence of increased ICP in non-survivors during the early treatment phase. REFERENCE(S): 1. Henzler D, Cooper DJ, and Mason K. Factors Contributing to Fatal Outcome of Traumatic Brain Injury Patients in Victoria:A pilot case control study. Crit Care Resuscitation 2001;3:153-157.

Zürcher Zenklusen R M1, Kaesermann C1, Busato A2 1 Soins Intensifs, Hôpital des Cadolles, Neuchâtel, 2Maurice E Müller Institute, University of Bern, Bern, Switzerland INTRODUCTION: There are no legal regulations in Switzerland, neither in several other European countries, on patient involvement in the decision to perform or to forego cardiopulmonary resuscitation (CPR). Many “Do Not Attempt Resuscitation” (DNAR) orders are made without patient involvement, often based on the assumption that patients might be troubled, shocked or become anxious if confronted with a hypothetical cardiopulmonary arrest (CPA) on their own. In Switzerland, there are currently no data available on whether and under what circumstances patients would like to be involved in this kind of decision. A prospective pilot study with hospitalised patients was therefore performed in order to elaborate guidelines on DNAR order prescriptions respecting patients’ wishes. METHODS: A questionnaire with 13 key questions on patient involvement in DNAR decisions was elaborated by the authors. A study nurse distributed the questionnaire to all eligible patients admitted to the departments of internal medicine and of general surgery at the Hôpital Les Cadolles in Neuchâtel, Switzerland, between January and March 2004. Exclusion criteria were admission for end-of-life care, severely altered mental state or poor general health condition precluding active participation and patients illiterate in the French language. RESULTS: 205 out of 429 patients (47.8%) hospitalised during the two-month study period agreed to participate and to complete the questionnaire. 51 patients refused to participate, 85 could not be included for medical and 88 for logistic reasons. The median age of the participating patients was 63.5 years, range 18-93 years. 82% considered the opportunity to talk about what should be done in case of a CPA as important to very important. 70% wished to take the decision to undergo or forego CPR by themselves, either alone , together with family members (40%) or with a physician (27%). 78% proposed that the discussion on CPR/DNAR should be initiated by the physician, either systematically at hospital admission (45%) or after changes of health status (44%). Only a very small minority (<15%) declared to have had the opportunity to discuss CPR/DNAR orders with a physician so far. 33% of participating patients stated that they would not wish to undergo resuscitation in case of an unexpected CPA. This attitude was associated with age, 50% of patients over the age of 75y refusing CPR. DNAR orders had been written by hospital physicians for 8% of included patients. CONCLUSION: The great majority of patients wish to be involved in the decision to undergo or forego CPR. This contrasts distinctly with the small minority of patients having the opportunity to do so. As up to a third of patients refuse CPR, great efforts have to be accomplished to respect major ethical principles of patient autonomy, beneficience, non-maleficience and justice with regard to CPR.

Oral Presentations End-of-life care – 307-309 307

309

END OF LIFE DECISIONS IN INTENSIVE CARE: A GLOBAL PERSPECTIVE

1

Gunn S R1, Whetstine L M2, Crippen D W1 1 Critical Care Medicine, University of Pittsburgh Medical Center, 2Health Care Ethics, Duquesne University, Pittsburgh, United States

INTRODUCTION: Until 2001, patients often died in our ICU without being accompagnied by their families, due to usual restricted visiting hours in such a setting. Doctors and nurses had the feeling not to meet family wishes. Therefore, we decided to change our policy about visiting hours in end-of-life setting and to evaluate this change on families and nursing team opinion.

INTRODUCTION: The International Critical Care Internet Group, CCM-L, has been instrumental in clarifying ICU decision analysis among multinational providers. METHODS: We presented a survey regarding a difficult ICU end of life case scenario on-line to CCM-L members. The case involved surrogates demanding life sustaining medical therapy (LSMT) that although not technically futile, would prolong discomfort and eventual death. We analyzed the data using the Chi square test on contingency tables of responses by location of practice (North America, Europe, Australia, or other) and job description (physician, nurse, or ancillary healthcare provider). Significant results (p<0.05) were investigated with interaction plots. RESULTS: 261 of 832 (31.4%) list members responded. 177 of 261 (67.8%) were physicians, 38 (14.6%) were nurses, and 46 (17.6%) ancillary healthcare providers. Overall, 92.7% believed surrogates should not have final authority to guide LSMT. However, North American respondents were more likely to allow the surrogates final authority, than were other respondents (p<0.05). Overall, most respondents (79.9%) felt that the ICU physician should have final authority to withdraw LSMT. However, physicians (p<0.05) and those practicing outside of North America (p<0.05) were most likely to agree. Physicians, as compared to other healthcare providers, were more likely to feel justified in omitting discussion of possible therapies that would only prolong the dying process (p<0.05). 52.9% of respondents would involve an ethics committee in discussions with the surrogates. However, nurses and ancillary healthcare providers (p<0.01) as well as North American respondents (p<0.01) were more likely to seek and ethics consultation. North American respondents were more likely to consider the possibility of future legal action by the surrogates when making their decisions (p<0.05). CONCLUSION: Globally, MDs were likely to justify withdrawing LSMT or omit some options in discussion with families. North American physicians are more likely than their multinational counterparts to refer this case to an ethics committee. On-line Internet interest groups are fertile ground to compare and contrast multinational provider preferences.

END-OF-LIFE CARE : EVALUATION OF OPENING VISITING HOURS FOR FAMILIES Boles J1, L’Her E1, Prat G1, Goetghebeur D1, Renault A1 Service de Réanimation médicale, Hôpital de la Cavale Blanche - CHU, Brest, France

METHODS: Starting January 1st 2001, we allowed families to stay unrestrictedly with their relative once a decision of life-prolonging therapy withholding or withdrawing had been reached. We evaluated this decision by means of a questionnaire sent in 2002 to family members of patients who died in our ICU in 1999 and 2000 (period „before“, group A) and during year 2001 (period „after“, group B). A satisfaction questionnaire was also given to all nurses and caregivers working in the ICU. RESULTS: A total 70 answers (35,7%) out of 196 questionnaires sent. Group A : 29 answers (41,4%) : 16 people (55%) would have liked to stay all the time with their relative ; 22 (75,8%) would have liked to be present at moment of death ; 25 (87,5%) complained persistantly for not having been present at moment of death, hampering bereavement up to 3 years after patient’s death ; only 14 (37,5%) were satisfied of help from team at patient’s death. Group B : 35 answers (50%) : 27 people (77%) were present before and at moment of patient’s death, 3 were not but would have liked to and 2 were not present because they did not want to ; communication with the team was jugded satisfactory by 76,5% and overall satisfaction of help from team at moment of death was 95,6%. Nursing team : 41 out of 77 answered (71,9%) : 75,6% recognized a better communication with families, 82,9% did spend more time with families and 84,2% modified accordingly their work organisation without problem. CONCLUSION: The results of our study demonstrate the necessity of allowing family members to stay with their relative when at end-of-life, without any time restriction, to help them cope with their relative’s death. Such a policy does not bother nurses’ working organisation and does enhance relationship with families. We have pursued this policy since, to the obvious satisfaction of families and nursing team.

17th Annual Congress – Berlin, Germany – 10–13 October 2004

Oral Presentations Physiotherapy – 310-312 310

312

PHYSIOTHERAPY AND VENTILATOR-ASSOCIATED PNEUMONIA IN PATIENTS WITH ACQUIRED BRAIN INJURY

1

Patman S M1, Jenkins S C2, Stiller K3 1 Physiotherapy, Royal Perth Hospital, 2School of Physiotherapy, Curtin University of Technology, Perth, 3Physiotherapy, Royal Adelaide Hospital, Adelaide, Australia INTRODUCTION: Respiratory physiotherapy is often provided to prevent and/or treat ventilator associated pneumonia (VAP) in acquired brain injury (ABI) patients due to the increased incidence of VAP in this population and the associated morbidity and mortality. However there are limited data on the efficacy of physiotherapy in ICU(1) (2). The aim of this prospective randomised controlled trial was to investigate the effect of respiratory physiotherapy on the incidence of VAP, duration of ventilatory support, and length of ICU stay in adults with ABI. METHODS: ABI patients admitted with a Glasgow Coma Scale < nine, requiring intracranial pressure monitoring, and invasive ventilatory support for >24 hours, were randomised to a treatment group (six respiratory physiotherapy interventions in each 24-hour period whilst ventilated), or a control group (routine medical and nursing care only). Respiratory physiotherapy comprised a regimen of positioning, manual hyperinflation and suctioning. Consent was obtained for 134 subjects (of 183 patients fulfilling inclusion criteria), with 67 randomised for treatment. Twenty eight subjects were excluded due to unstable neurological, cardiac or respiratory status, five due to early limiting of active management, and consent was declined in five patients. RESULTS: Groups were comparable on demographics variables, with the exception of body mass index (mean±SD: 27.7±5.5 vs. 25.2±6.4; p=0.02). Using multivariate analysis of variance with intention to treat philosophy and analysis by treatment principle, there were no significant differences for VAP incidence [treatment group 14/67 vs. control 18/67; p=0.42], duration of ventilatory support (hr) [178.1±122.5 vs. 215.8±178.0); p=0.16], or length of ICU stay (hr) [224.6±123.0 vs. 264.0±189.3; p=0.16]. Sixteen subjects were withdrawn (five from treatment group), seven due to cessation of active management, six became medically unstable and three received physiotherapy beyond that described in the treatment protocol; 10 withdrawn subjects died. CONCLUSION: Respiratory physiotherapy, in addition to routine medical and nursing care, does not appear to prevent VAP, reduce length of ventilation or ICU stay in adults with ABI. REFERENCES: (1) Stiller K. Physiotherapy in intensive care - towards an evidence based practice. Chest 2000; 118: 1801-1813. (2) Ntoumenopoulos G et al. Chest physiotherapy for the prevention of ventilator-associated pneumonia. Intensive Care Medicine 2002; 28: 850-856. Grant acknowledgement: This study was supported by a 2001 Australian Physiotherapy Association Physiotherapy Research Foundation Seeding Grant.

311 DEEP BREATHING EXERCISE: AN INTERVENTION TO PROMOTE COMFORT DURING ENDOTRACHEAL SUCTIONING Rena R R1 1 Medical Intensive Care Unit, Philippine Heart Center, Manila, Philippines INTRODUCTION: Discomfort has always been an issue during endotracheal suctioning (ETS) particularly in a fully awake patient. In this study, the efficacy of deep breathing exercise (DBE) as an intervention in the promotion of comfort among intubated and mechanically ventilated patients during ETS was investigated. METHODS: A prospective single-blind block randomized clinical trial was utilized for this study. It was conducted from October to November 2002 at the Medical Intensive Care Unit of the Philippine Heart Center involving 30 awake, intubated and mechanically ventilated patients, who have met the inclusion and exclusion criteria set by the researcher. The samples were randomly assigned either to control or to experimental group. Determinants of patients’ level of comfort were physiological response, behavioural manifestations and patients’ comfort rating. These were assessed using an observation checklist and a comfort rating scale. The data were analyzed by means of t-test, Wilcoxon and Mann-Whitney-U test RESULTS: Of the 30 samples, 67% are ages 51 and above (mean age of 64 for the control group and 57 for the experimental group) 53% are male and 47% are female. Majority in the control group were able to reach college level (40%) as compared to only elementary level (33%) in the experimental group but the two groups have almost the same number of samples for other subcategories. In terms of duration of intubation, 80% had been intubated for 1-3 days, 17% for 46 days and 3% for >7 days. 70% had no experience of previous intubation and 30% had an experience of previous intubation. The samples are comparable in all these characteristics. It was found that samples in the experimental group are more comfortable in all determinants of the level of comfort. Physiological parameters (BP, HR and RR are higher in the control group than in the experimental group by as much as 15-30% indicating a decrease in the level of comfort of the patients. In terms of behavioural response, the control group has a comfort behaviour score of 58.93 as compared to the experimental group score of 40.67, which is qualitatively described as very low level of comfort and high level of comfort, respectively. Experimental group’s mean comfort rating was 8.79 while that of the control group was only 0.20. The patients’ level of comfort during ETS from both groups is not significantly related to age, educational attainment and duration of intubation but is significantly related to previous experience of intubation (p = 0.003 for control, p = 0.023 for experimental). Gender affects the level of comfort of only the control group (p = 0.016) but not the experimental group (p = 0.074). CONCLUSION: The result demonstrated that deep breathing exercise could promote high-level of comfort during ETS. It is therefore recommended that this technique be adopted as part of the protocol of routine ETS.

S83

EFFECT OF ABDOMINAL MASSAGE ON GASTRIC EMPTYING IN CRITICALLY ILL PATIENTS Bastin R1, Kulikowski B1, Piagnerelli M1, Thirry P2, Vincent J1 Intensive care unit, 2Chemestry department, Hôpital Erasme, Brussel, Belgium INTRODUCTION: Enteral feeding is important to support nutritional needs and immune and gastrointestinal functions in critically ill patients. Motility disorders have been associated with stomach and trachea colonization and often lead to the interruption of feeding (1). Abdominal massage therapy is a common practice to relieve constipation and has been shown to shorten the duration of postoperative ileus (2). We therefore evaluated, using a paracetamol absorption test (3), whether an abdominal massage would increase gastric emptying in critically ill patients. METHODS: After approval by the Ethics Committee of our institution, 21 patients (> 18 years) who received enteral feeding by nasogastric tube (NG) for a period > 24h were included. Two patients were excluded on day 2 (1 death and 1 gastrointestinal bleed). After aspiration of gastric contents, 2 g of paracetamol dissolved in 150 ml of water was administered via the NG on two consecutive days. Plasma samples were drawn at T0, 5, 10, 15, 30, 60, 90, 180 min, for measurement of paracetamol concentrations. The area under the curve was calculated at 60 min (AUC60) on the two days. An abdominal massage was performed on day 2 in 11 patients, selected randomly. Statistical analysis included an analysis of variance for repeated measures (ANOVA) followed by a modified T-test. Data are presented as mean±SEM. RESULTS: For the 19 patients studied, there was no significant difference in AUC60 on the two days in the massage group (n=11) (644±188 vs 820±172 mmol/min/l) or in the control group (n=8) (942±275 vs 899±235 mmol/min/l). The AUC60 was significantly lower in patients with an initial volume of gastric aspirate > 150 ml (n=10) than in the other patients (411±167 vs 1144±210 mmol/min/l, p<0.01). In patients with a gastric aspirate>150ml who were in the massage group (n=6), the AUC60 was higher on day 2 than on day 1 (510±143 vs 199±40 mmol/min/l, p=0.06). CONCLUSION: Patients with a gastric aspirate volume >150 ml have impaired gastro-intestinal motility as shown by a very low AUC60. In these patients, abdominal massage may represent a simple option to facilitate gastric emptying. REFERENCE(S): (1) : S Adamet al. Intensive Care Med (1997) 23 : 261-66 (2) I Le Blanc-Louvry et al. J Gastrointestinal Surg (2002) 6(1):43-49 (3) : MM Tarling et al. Intensive Care Med (1997) 23 : 256-60