Neurologic complications in cardiac surgery: can risk scores be applied?

Neurologic complications (NC) in cardiac surgery are not rare (5–15%). Their etiopathogeny is multifactorial, and the risk factors are numerous. Neurologic complications result in high morbidity and mortality rates, and high hospital costs. Most risk scores assess mortality, and the risk for stroke assessed by the AHA/ACC score refers only to patients with coronary disease. One may thus question whether risk scores for NC can be applied in a general population.

Introduction Neutrophils have been involved in sepsis-induced organ damage. Neutrophils could be directly activated by TLR binding ligands including LPS. IRAK-1 is one of many intracellular proteins that are activated upon stimulation of TL receptors. This triggers a series of events that results in the migration of NF-κB into the nucleus and the activation NF-κB-dependent genes. Objectives To identify a single nucleotide polymorphism at position 532 (coding SNP) in volunteers and patients with sepsis. To determine whether IRAK-1 SNP532 results in a decrease in neutrophil NF-κB activation in volunteers and patients with sepsis. To evaluate neutrophil gene expression patterns in IRAK-1 SNP532 and wildtype patients with sepsis. Methods Thirty severe sepsis patients and 34 healthy volunteers were enrolled in this study. Peripheral blood was obtained and neutrophils were isolated by plasma-percoll gradients after dextran sedimentation of erythrocytes. Neutrophils from volunteers were resuspended in RPMI and cultured with or without 100 ng/ml LPS for 60 min. The electrophoretic mobility shift assay technique was used to measure the NF-κB activation. Real-time PCR allelic discrimination assays were developed by the assay-by-design service offered by Applied Biosystems (Foster City, CA, USA). Probe and primer combinations were designed at the single nucleotide polymorphism 532. PCR reactions were performed according to the manufacturer's manual using the Applied Biosystems 7500 Real-Time PCR system. Microarray analysis was used to evaluate the neutrophil gene expression in unstimulated neutrophils and after LPS stimulus. Results The median AUC for NF-κB activation was higher in wildtype genotyped neutrophils as compared with IRAK-1 SNP532 genotyped neutrophils (85.2 vs 100.5, P = 0.05) (Fig. 1). In terms of kinetics pattern, we found some differences on nuclear levels of NF-κB in neutrophils from volunteers cultured with LPS. At 30 min after LPS, the culture nuclear translocation of NK-κB was significantly greater in wildtype genotyped neutrophils than in IRAK-1 SNP532 genotyped neutrophils. Even after 60 min, the NF-κB translocation remained high in wildtype genotyped neutrophils, while in IRAK-1 SNP532 genotyped neutrophils the NF-κB translocation was similar to baseline (Fig. 2). In unstimulated neutrophils from septic patients, the NF-κB translocation was significantly lower in IRAK-1 SNP532 genotyped neutrophils than in wildtype genotyped neutrophils (1.20 vs 2.10, P = 0.05) (Fig. 3). Finally, the expression of some inflammatory related genes (IL-8, IL1β, MIP-2, COX-2, and SOD2) was decreased in IRAK-1 SNP532 genotyped neutrophils. Conclusion IRAK-1 SNP532 genotyped neutrophils from volunteers (after LPS ex vivo challenge) and from septic patients are associated with lower NF-κB activation and lower expression of some IRAK1-related genes. These results demonstrate that IRAK1    Objective To create an experimental animal model of cardiogenic shock for learning and to test new therapeutic strategies.
Methods Adult white pigs (70 kg) received both intravenous anesthesia (acepromazine 0.3 mg/kg, midazolam 0.2 mg/kg, fentanyl 250 µg/kg, thiopental sodium 12.5 mg/kg and pancuronium 0.4 mg/kg) and inhaled anesthesia (halothane 1%), and were intubated and mechanically ventilated. An arterial line was obtained through dissection and puncture of the common femoral artery. A continuous cardiac output catheter (Edwards Lifescience, USA) was introduced through the dissected internal jugular vein and was positioned using the arterial pulmonary pressure curve, allowing monitoring of the right atrial pressure, pulmonary artery pressure, pulmonary wedge pressure (PAop) and SvO 2 . Through median sternotomy, the pericardium was opened longitudinally and the heart was exposed. The baseline ECG and hemodynamic data were recorded and after a 6-0 polypropylene suture was passed under the proximal anterior descending coronary artery that was snared for up to 10-15 min. An ECG was then obtained to show typical ischemic alterations, and a regional myocardium color change and regional myocardial hypocontractility were observed. The presence of cardiogenic shock was defined by cardiac output index <1.8 l/min/m 2 , PAop >20 mmHg and mean arterial pressure <50 mmHg. The carotid artery and external jugular vein were cannulated and ECMO support was used (flow 100-150 ml/kg/min) after induced cardiogenic shock. Results The model was tested in eight animals. Four animals died immediately after coronary occlusion because of ventricular fibrillation, and cardiogenic shock was reproduced in the other four animals and these animals were kept alive for 4 hours with supportive interventions (inotropic drugs and ECMO).
Conclusions The experimental animal model created by ischemic myocardial infarction induced cardiogenic shock and can be used to study and test new therapeutic strategies. Background Mitral stenosis is frequently associated with increased pulmonary vascular resistance (PVR), pulmonary hypertension and right ventricular dysfunction that persist even after surgery. Inhaled nitric oxide (NO) has been shown to selectively reduce PVR in patients with pulmonary hypertension. We tested the hypothesis that NO would improve the hemodynamic effects and short-term clinical outcomes of patients with mitral stenosis and severe pulmonary hypertension undergoing cardiac surgery. Methods Twenty-seven patients (three male, 24 female) with a mean age of 46.9 ± 12.9 years with mitral stenosis and elevated pulmonary artery systolic pressure (PASP) were randomly allocated to receive continuously inhaled NO at 10 parts per million (NO group) or oxygen therapy (control) for 48 hours immediately after surgery. The hemodynamic data, the number and doses of vasoactive drugs, the duration of stay in the ICU and short-term complications (infections, respiratory and/or renal failure, and death) were assessed. Results The mean mitral valve area, gradient and PASP were 0.88 ± 0.20 cm 2 , 15.7 ± 5.0 mmHg and 70.9 ± 10.3 mmHg, res-pectively, for all patients. After 48 hours, patients receiving NO showed an increased cardiac index compared with patients receiving oxygen therapy, with a reduction in the number of vasoactive drugs used. There was a significant reduction in PASP in both groups compared with preoperative levels but no differences were observed between the groups. A tendency towards a reduction in pulmonary vascular resistance, ICU stay and acute complications was observed in the NO group but did not reach statistical significance. Conclusions Use of inhaled NO immediately after surgery in patients with mitral stenosis and severe pulmonary hypertension improves cardiac hemodynamics and may have clinical benefits in short-term outcomes. Objective To study a group of patients included in a fast-track program after cardiovascular surgery concerning the medical, economical, psychological and dynamic conditions of the protocol in the ICU. Materials and methods Seventy patients operated on from August to December 2000 were included. Inclusion criteria were: age, no operation events, hemodynamic stability and no comorbidity. Early extubation was achieved using bendiazepan antagonist (Flumazenil) and respiratory physiotherapy with noninvasive ventilation (CPAP or BIPAP). ICU discharge was on the first postoperative day. Results Among the 70 patients, 57% were male with a mean age of 56.2 years. With regard to the type of operation, 74% were submitted to coronary bypass surgery, 17.1% to valve surgery, and 8.9% to another type of operation. The average extubation time was 153 min; 22% had hypertension and 2.8% were reintubated. From the psychological point of view, 95% of patients considered the shorter ICU stay satisfactory. With regard to the dynamics of the ICU, there was a 50% decrease in duration of ICU stay, and an increase of 30% in patient admission and a reduction of 40% in cost. No patient had significant clinical complication and no one was readmitted. Conclusion A reduction of ICU stay was possible in selected patients with satisfactory medical and psychological conditions, as well as cost containment and greater availability of beds.

Gap care in diagnostic and prognostic evaluation of chest pain in the elderly
Introduction Despite the greater prevalence of coronary disease, aortic pathology and pulmonary thrombolysis in elderly patients, some studies have shown under-utilization of diagnostic and therapeutic resources in this age group. Methods A total of 541 patients (220 [46%] female) attended the Hospital Pró-Cardíaco Chest Pain Center, Rio de Janeiro, Brazil, from January to December 2004. The patients were divided into four age groups: I: <65 years, n = 264 (48.7%); II: between 65 and 74 years, n = 131 (24.2%); III: between 75 and 84 years, n = 104 (19.2%); and IV: >85 years, n = 42 (7.7%). Diagnostic and/or risk stratification tests (treadmill stress test, myocardial scintigraphy, pulmonary scintigraphy, stress echocardiogram, angio-tomography, angio-magnetic resonance, transesophageal echocardiography, coronariography) were analyzed and patients were divided into two groups: DIAG (patients with at least one test done) and NO DIAG (patients without any test done). The intrahospital mortality (MORT) rate was also analyzed and compared between the age groups. Results Table 1 shows the diagnostic test evaluation and the intrahospital mortality rate according to age group. Conclusion Elderly patients, especially the 'oldest old', that come to the emergency room with chest pain have a greater likelihood of discharge without any diagnostic and/risk stratification test being performed, compared with younger patients. The intrahospital mortality rate increased with age. These findings show a gap in care of the elderly with chest pain, with in turn may be associated with a worse prognosis in that population.

Treatment of acute coronary syndrome without ST-segment elevations in the elderly
Introduction Extracorporeal circulation may trigger an extensive inflammatory response and the release of cytokines, which can induce myocardial ischemia.
Objective The goal of this study was to evaluate the relationship between inflammatory markers and biochemical evidence of myocardial cell injury in patients who underwent coronary bypass grafting (CABG) with (on-pump) or without (off-pump) extracorporeal circulation. Methods All patients had myocardial infarction with multivessel coronary artery disease with preserved ventricular function and without renal failure or other cardiac diseases. The mean age was 46.6 ± 12.5 years. Cytokines were measured 6 and 24 hours postoperatively by ELISA and immunoassay, and were correlated to the occurrence of the following clinical complications: fever, atrial fibrillation, significant pericardial effusion, pulmonary complications, and release of CK-MB and troponin I.

Results
Of the 724 CABGs performed in a single-center tertiary hospital, 218 were off-pump and 506 were on-pump. The mean age was 46.6 ± 12.5 years and the mean time of extracorporeal circulation was 72 ± 23 min. Clinical complications were more frequent among on-pump patients (Table 1). This was associated with higher levels of C-reactive protein, CK-MB and troponin I, but not IL-6 ( Table 2). Conclusions Postoperative complications and biochemical evidence of myocardial cell damage after CABG were more frequent among on-pump patients, and this was correlated with higher serum levels of C-reactive protein.
Purpose To evaluate the heart rate (HR) variability at rest, during the deep breathing test (DBT) and during an exercise protocol, in patients with acute myocardial infarction (AMI) submitted to a hospital physiotherapy intervention 24 hours after the cardiac event. Additionally, to evaluate the safety of the protocol applied. Methods and results Eight male patients (mean age 50 ± 12 years), admitted to the Coronary Care Unit of the Irmandade Santa Casa de Misericórdia de São Carlos with noncomplicated AMI, were studied. All patients were hemodynamically stable and used conventional medications. The patients were submitted to the hospital physiotherapy intervention 24 hours after their admission. The instantaneous HR was acquired and monitored by an HR monitor (Polar ® S810i) during 10 min at rest pre-exercise, 4 min of DBT, 5 min of exercise protocol (active hand and foot exercises, and active-assisted lower extremities exercises) and 10 min at rest post-exercise in the supine position. The blood pressure (BP) was Critical Care June 2005 Vol 9 Suppl 2 Third International Symposium on Intensive Care and Emergency Medicine for Latin America * P < 0.05, DBT vs pre rest and post rest; † P < 0.05, DBT vs pre-rest; ‡ P < 0.05, exercise vs pre rest and post rest.
S9 measured before, during and after the exercise protocol. The R-R intervals (ms) were analyzed by time domain (RMSSD and RMSM) and frequency domain methods, and the power spectral components were expressed as normalized units (nu) at low (LF) and high (HF) frequencies, and as the LF/HF ratio. Data are presented as the mean ± standard deviation ( Table 1). The statistical analysis was performed by ANOVA and the Tukey posthoc test with the level of significance set at 5%. Conclusion The physiotherapy intervention protocol appeared to be effective as it induced hemodynamic repercussion and modification of the autonomic control of HR, without any clinical intercurrence. Acknowledgements Financial support from FAPESP-Proc. 04/05788-6 and CNPq-Proc. 478799/2003.

P20
The use of the SOFA score to analyze the profile and severity of organ dysfunction in patients with cardiovascular disorders

Measurements and main results
To assess the organ dysfunction, we collected data of assailed organ systems, individually, and the total maximum SOFA (TMS). We analyzed the association between basic cardiovascular pathology and admission diagnosis with higher dysfunction organ scores and their impact on mortality. The mean age was 61.1 years, and the length of CIU stay was 5.8 days, with CIU mortality of 13.3%. The median TMS was 3.86, significantly higher in non-survivors (12.0 vs 2.0, P < 0.001). The organ dysfunction, individually or in association, was correlated with higher mortality, and respiratory dysfunction was the highest prevalent (57.3%). With respect to cardiovascular pathology, although there was higher prevalence of ischemic, hypertensive and rheumatic pathology, just the dilated non-ischemic cardiomyopathy (24% non-survivors, P = 0.046) was correlated with higher mortality. Analyzing the admission diagnosis, the presence of cardiogenic shock (72.7% non-survivors, P < 0.0001), pulmonary infection (47.3%, P < 0.0001) and congestive heart failure (27.3%, P = 0.12) associated with higher scores of organ dysfunction and higher mortality. The acute coronary syndrome without ST-segment elevation and atrial arrhythmias, although of higher prevalence, were correlated with lower mortality. Conclusion The SOFA score allowed a simple and effective evaluation of organ dysfunction severity profile in patients with heart disease, identifying high morbidity and mortality diseases and high-risk groups, that will enable earlier therapeutic measures and increased monitorization.  Background Electrical cardioversion (CV) of patients with atrial fibrillation (AF) is a current practice. Recent data suggest that biphasic waveform shocks are more effective than monophasic ones for transthoracic CV, but the optimal protocol has not been defined.
Objective To determine the rate of CV success comparing biphasic versus monophasic waveform energies; which is more effective to achieve cardioversion at initial shock and to access the cumulative energy used by each waveform energy. Methods We conducted a prospective, randomized study of 43 patients with atrial fibrillation during an 18-month period. Success rates of CV with a defibrillator using monophasic waveform energy with a sequential energy of 200 J-300 J-360 J (Group 1) and using biphasic waveform energy with a sequential energy of 120 J-150 J-200 J (Group 2) were randomly compared. If the maximum energy used by one group did not achieve CV success, a crossover to the maximum energy of the other group was performed.

Results
The study population consisted of 22 patients in Group 1 and 21 patients in Group 2 with similar baseline characteristics. The rate of CV success was 95.5% in Group 1 and 85.5% in Group 2 (P = not significant). Group 1 achieved success at initial shock in 95.5% and Group 2 in 57.1% (P = 0.27). The mean cumulative energy was 200 J in Group 1 and 203 ± 135 J in the biphasic waveform group (P = 0.078).

Conclusion
In this study, AF cardioversion using biphasic waveform energy was less effective than a monophasic pulse. This result could be attributed to the initial energy of 200 J used by the monophasic group. . We used the ROC curve to establish the best cutoff for sensibility and specificity for inhospital death, followed by the chi-square test; and also the log rank test to analyze the Kaplan-Meier curve. We consider P ≤ 0.05 statistically significant.

Results
The best cutoff point of D-dimer in the ROC curve to predict inhospital death was 1433 mg/dl (P = 0.03), with sensibility = 80%, specificity = 69% and negative predictive value = 97%. After 1 year of follow-up we observed that patients with Ddimer ≥2000 mg/dl during initial hospitalization had the worst prognosis (event-free survival median = 295 days when D-dimer <2000 mg/dl vs 70 days when D-dimer ≥2000 mg/dl, P = 0.03).
Conclusions An elevated D-dimer at hospital admission in patients with decompensated HF seems to have clinical importance, indicating a higher probability of inhospital death and worse eventfree survival after 1 year.

Results
The sample was divided into three groups (group A -56.8%, group B -29.6% and group C -13.6%). There was no difference in age between the two groups (P = 0.14). Diabetes was more frequent in group B (overweight). Obese patients (group C) had lower BNP levels (P = 0.01) and D-dimer (P = 0.035). There were no differences between the three groups related to complications and inhospital mortality.
Introduction Inflammatory markers such as C-reactive protein (CRP) have shown a high prognostic value in the setting of coronary artery disease and heart failure (HF Objective To analyze the usage of hydroxymethylamide during volume reposition in the immediate postoperative period after heart surgery. Materials and methods A prospective study in which patients submitted to heart surgery had volume expansion with voluven 6%. Only patients with total preserved renal and hepatic function were included. None of them had any clinical or laboratorial signs of coagulation disturbances or any occurrence during surgical procedure. Voluven 6% was used during the initial phase of the postoperative period (first 2 hours), with a maximum volume limit of 1000 ml per patient. Other aspects were studied such as signs of bleeding, anaphylactic reactions and time for achieving clinical stability. Results IHP induced a reduction in SMVBF (579 ± 53 to 321 ± 10 ml/min) and SBF (44.7 ± 3.2 to 29.1 ± 5.3 ml/min) and an increase in D t-a pCO 2 (2 ± 2.8 to 20.5 ± 4.5 mmHg). No alterations on systemic metabolic or O 2 -derived variables were observed. The increase of the D t-a pCO 2 correlated with the grade of mucosal injury (Fig. 1).

Results
Conclusion IHP induces a proportional reduction on blood flow in all layers of the intestine, and none of the systemic markers of splanchnic ischemia predict the intestinal blood flow disturbances during the early phase of intestinal transplantation. In addition, intestinal pCO 2 measurement seems to be a useful way for monitoring graft perfusion and histological changes after hypothermic ischemia and reperfusion.  S15 mesenteric blood flow obstruction. In addition, we sought to obtain evidence that systemic markers of splanchnic hypoperfusion can detect the initial changes after intestinal ischemia induced by arterial or venous blood flow interruption. Methods Fourteen dogs were subjected to 45 min of superior mesenteric artery (SMA-O, n = 7) or vein occlusion (SMV-O, n = 7). Systemic hemodynamic was evaluated through a Swan-Ganz catheter and arterial catheters, while gastrointestinal tract perfusion was evaluated by superior mesenteric vein and serosal blood flows (SMVBF and SBF, ultrasonic flowprobe). Intestinal O 2derived variables, mesenteric-arterial and tonometric-arterial pCO 2 gradients (D mv-a pCO 2 and D t-a pCO 2 ) were calculated. Results A significant decrease in CO and MAP was detected in the SMV-O group; pCO 2 gradients presented a significant increase in both groups (Fig. 1). The histopathologic injury scores were 2.7 ± 0.5 and 4.8 ± 0.2 for the SMA-O and SMV-O groups, respectively.
Conclusion Temporary mesenteric congestion was associated with significant hemodynamic and metabolic disturbances. The D t-a pCO 2 changes can be detected by systemic markers of splanchnic hypoperfusion after temporary SMV occlusion.

P39
Inhospital mortality of patients with submassive pulmonary embolism submitted to thrombolytic therapy in a multicenter study  D mv-a pCO 2 (mmHg) S16

Conclusion
In the present study, clinical variables were similar in both young and elderly patients with severe sepsis/septic shock. However, patients over 65 years old had different hemodynamic characteristics. This finding should be considered for the diagnosis and management of elderly patients with severe sepsis and septic shock.

P41
In vivo leukocyte-endothelium interactions in rat mesenteric microvessels after ischemia/reperfusion and sepsis Objective A leukocyte-endothelium interaction is known to be a remarkable event at the beginning of systemic inflammatory response syndrome. The aim of this study was to evaluate leukocyteendothelium interactions in superfused mesenteric postcapillary venules after hemorrhagic shock/reperfusion and cecal ligation and puncture in rats.
Results Data of leukocyte-endothelium interactions in rat mesenteric microcirculation are presented as the mean ± standard deviation (Table 1).

Conclusions
The double-hit model (ischemia/reperfusion and sepsis) induced a severe inflammatory injury similar to sepsis alone. The inflammatory process was overcome by cecal resection and peritoneal lavage. Up to 72 hours of reperfusion with lactated Ringer's solution and 25% of the shed blood volume, inflammation is still evidenced by the increased number of migrated cells in the perivascular tissue. Acknowledgements Supported by PRONEX, FAPESP and UNICID. There is some evidence that central venous oxygen saturation (cSVO 2 ) can replace mSVO 2 . However, little is known about where it should be located: the superior vena cava (ccSVO 2 ) or the right atrium (acSVO 2 ). This study aimed at evaluating the differences between mSVO 2 and cSVO 2 (either from the cava or from the atrium) and the impact of them in patient management.
Methods We included patients admitted to a tertiary universitary ICU with septic shock that had a Swan-Ganz catheter and a central venous catheter in place. Each patient was submitted to three sets of hemodynamic and respiratory monitoring, with a minimal interval of 4 hours. Each set included a blood gas analysis of samples collected from the proximal (acSVO 2 ) and distal port of the Swan-Ganz catheter (mSVO 2 ) and also from the central line catheter (ccSVO 2 ). Each of these samples was analyzed by a blinded critical care physician who decided the hypothetical management for the patient. Statistical analysis was done using a paired Student t test. Results were considered significant if P ≤ 0.005.

Results
We studied 22 sets of measures in seven patients (five female and two male) with a mean age of 60.57 ± 23.25 years.
The mean values were 76.47 ± 8.02, 75.54 ± 11.96 and 70.90 ± 8.53 for ccSVO 2 , acSVO 2 and mSVO 2 , respectively. There was a significant difference between ccSVO 2 and mSVO 2 (P = 0.009) and acSVO 2 and mSVO 2 (P = 0.01), but not between ccSVO 2 and acSVO 2 (P = 0.60). The concordances in patient management were 63.2%, 68.2% and 78.9% between ccSVO 2 and mSVO 2 , acSVO 2 and mSVO 2 and ccSVO 2 and acSVO 2 . When only sets with a ccSVO 2 below 70 were considered, the concordance between ccSVO 2 and mSVO 2 was 75%. Conclusion Our results suggest that blood samples derived from a central catheter, even if it is located in the right atrium, may be not accurate enough to be used as a measure of tissue oxygenation and may lead to improper management of the patient, mainly when the values are above 70%. Background Pediatric septic shock is usually associated with multiple factors, including hypovolemia, myocardial depression, vascular failure, endocrine and metabolic disturbances. Eighty percent of the children with fluid refractory septic shock present with a low cardiac index. We emphasize the role of invasive and non-invasive monitoring for children with septic shock, leading to changes in treatment and prognosis. Case A 5-year-old boy presenting with pneumonia, respiratory failure and severe sepsis. At admission, he was tachypneic and tachycardic, with inaudible blood pressure, prolonged capillary S17 refill time and weak pulses. The patient received 60 ml/kg normal saline in 60 min, without recovery. After placement of a central venous catheter, he was started on continuous dopamine, reaching 15 µg/kg/min. Again without significant improvement in arterial blood pressure and perfusion, epinephrine infusion was associated, starting with 0.1 µg/kg/min and increasing until 1.5 µg/kg/min. At this moment, he had low blood pressure, tachycardia, superior vena cava saturation (SVO 2 ) of 65%, oliguria and prolonged capillary refill time and, 2 hours later, presented with desaturation, hepatomegaly and acute pulmonary edema. An echocardiogram revealed a cardiac index of 1 l/min/m 2 . Changing treatment strategy, he was started on milrinone infusion and monitoring with continuous SVO 2 and pulmonary artery catheter. Two hours later, he had normal urinary output, normal blood pressure, SVO 2 of 74% and cardiac index of 3 l/min/m 2 . Treated with milrinone and low-dose epinephrine, he progressively improved arterial blood pressure, perfusion, pulse and mental status. The patient was weaned off vasoactive drugs and mechanical ventilation after 6 days. Discussion Myocardial dysfunction is frequent in children with septic shock, and it persists even after correction of hypovolemia, acidosis and electrolyte disturbances. Most of the children with septic shock have a low cardiac index and increased systemic vascular resistance. Conclusion Monitoring cardiac function, combining an echocardiogram with invasive methods, such as SVO 2 or pulmonary artery catheter, can be lifesaving in pediatric refractory septic shock. Introduction Hemodynamic optimization based on tissue perfusion markers is a strategy considered adequate for the management of patients in shock in ICUs.

P43
Objectives To evaluate the variability and correlation between venous and arterial standard base excess (SBE) and lactate samples.

Materials and methods
The analysis of lactate levels was performed and the SBE obtained from the same blood of central venous and arterial samples of 115 patients. We compared these measurements (Wilcoxon signed rank test), and determined the correlation between these variables (Spearman rank order correlation). Results There was a statistically significant difference between the value of venous SBE: -4.3 mEq/l (-7.4 to -0.9) as compared with the arterial value: -3.2 (-6.9 to 0), P < 0.001; but there was no difference between the venous lactate: Introduction Qualitative alterations in the bone marrow morphology have been described after shock and fluid resuscitation. However, quantitative cellularity must also be addressed.
Objective To assess bone marrow cellularity after hemorrhagic shock and fluid resuscitation with hypertonic saline (HSS) and lactated Ringer's solution (LR).
Methods Wistar rats (250-300 g, n = 22), anesthetized with pentobarbital, were bled to a mean arterial pressure (MAP) of ±40 mmHg over 10 min and were maintained at this level for 50 min. The animals were randomized into four groups: Sham (cannulation, no shock, no treatment), NT (shock, no treatment), LR (shock, followed by the infusion of LR, 3 × shed blood volume), and HSS (shock, followed by the infusion of 7.5% NaCl, 4 ml/kg).
The shed blood was not reinfused. Animals were killed 72 hours after shock; the marrow cavity of the femur was washed with McCoy's Medium (2 ml), and total and differential leukocyte counts were performed in the perfusate. Blood samples for total and differential blood leukocyte counts were obtained from the cut tip of the tail of the animals. Results Resuscitation with LR and HSS restored the MAP to the basal levels. Table 1 presents the bone marrow counts. There was no difference among groups concerning leukocyte blood counts. Table 1 Sham NT LR HSS (× 10 7 /ml) (× 10 7 /ml) (× 10 7 /ml) (× 10 7 /ml) Conclusion After hemorrhagic shock, the bone marrow total cellularity decreased both in animals resuscitated with LR and in those not resuscitated with fluids. After resuscitation with LR, the number of myeloid and erythroblast cells was lower than that observed in the resuscitation with HSS. Background Cocaine abuse has been linked to penetrating and blunt trauma requiring surgical treatment. Although the cardiovascular effects of cocaine are well studied, particularly its association with arrhythmia, myocardial infarction and sudden death, little is known about the effects of cocaine abuse on splanchnic perfusion and the potential undesirable interaction with volatile anesthetics. We hypothesized that halothane would elicit more circulatory adverse effects than sevoflurane in an acute model of cocaine intoxication. Methods Mechanically ventilated Beagle dogs (n = 14, 12.8 ± 0.3 kg) underwent anesthesia induction with intravenous propofol. They were then randomly assigned to two experimental groups: 1.5% halothane (n = 7, Halo) or 2.25% sevoflurane (n = 7, Sevo). After 30 min (Baseline), intravenous cocaine was infused as a bolus (12 mg/kg over 5 min), followed by 0.22 mg/kg/min during 30 min (BL-T35), and followed for 60 min thereafter (T35-T95). Systemic hemodynamics were determined by arterial and pulmonary artery catheters. Portal vein blood flow was measured by a transit time ultrasonic flowprobe. The PCO 2 gap (gas tonometry), blood gases, arterial lactate and cocaine levels were measured at each timepoint.

P46
Critical Care June 2005 Vol 9 Suppl 2 Third International Symposium on Intensive Care and Emergency Medicine for Latin America  (T10-T30).
Results During shock we observed the predicted MPAP and PAWP drop and ∆Pp amplification, due to a decreased preload, and, consequently, cardiac output decreased. After total shed blood reinfusion and volume expansion (T30), ∆Pp increased when compared with either BL (P = 0.02) or T20 (P = 0.003). This increase was associated with increased PVR (P = 0.01) and MPAP (P = 0.05), when compared with BL (Table 1). Conclusion We conclude that ∆Pp can increase after rapid volume infusion as in hypovolemic states. An acute increase in pulmonary pressure and resistance may be responsible for the observed increment in ∆Pp.  [3,4], length of ICU stay, and need for mechanical ventilation. We used Student's t test and the Fischer exact test for a statistical analysis. We considered the significance level of 5%.

Results
The mean age of patients, of whom 60% were female, was 82 ± 9 years (minimum = 65 years, maximum = 99 years). The predominant diagnosis was septic shock in 67% of the cases, while 33% of the patients developed severe sepsis. On days 1, 3, 5, 7, 14, and 28, the SOFA score presented mean values of 7, 6, 4, 3, 2, and 2, respectively (minimum = 2 and maximum = 15), thus evidencing a significant relationship between the SOFA score on day 1 (P = 0.0001) and day 3 (P = 0.001), including (∆)SOFA score (P = 0.043), and mortality. The number of failures was also associated with mortality when two or more organ failures (P = 0.001) were present. The age, gender, APACHE II score, length of ICU stay, dependence level, presence of cognitive deficit and/or previous cardiovascular diseases, plasma glucose levels, troponin I, BNP and PCR were not associated with mortality. Conclusion A mean SOFA average above 5 as well as SOFA variation within the first 72 hours proved to be good predictive markers in elderly patients with septic shock and severe sepsis. The same occurred in the presence of two or more organ failures during the course of sepsis.
Available online http://ccforum.com/supplements/9/S2 Introduction Experimental and clinical studies have shown that bacterial translocation (BT) has been implicated in the pathogenesis of sepsis and multiple organ dysfunction syndrome (MODS). In this study we examined the roll of the intestinal lymph during the BT process on the clinical outcome in a pre-established sepsis state. Methods Adult female Wistar rats (200-250 g) were submitted to the combination of induction of BT plus sepsis (S), with and without mesenteric lymph flow into the systemic circulation, and were monitored in terms of bacterial quantification per compartments and mortality (n = 20/each group). Groups: sepsis group (inoculation of 10 7 , 10 9 or 10 10 CFU/ml/100 g body weight of Enterobacter cloacae 89 into the portal vein); BT group (5 ml Escherichia coli R-6 10 10 CFU/ml/100 g body weight confined to the small intestine for 2 hours); BT with lymphadenectomy group; and combination group (sepsis 10 7 or 10 9 plus BT-10 10 ) with and without lymphadenectomy. The interruption of the lymph flow was achieved by mesenteric lymph node resection 5 days prior to the experiments, which is not a sufficient time for the re-canalization of the lymph ducts. The observation period for mortality was of 30 days in all groups.

Results
The BT groups and S-10 7 group did not show any mortality; however, the combination of S-10 7 or S-10 9 with BT without lymphadenectomy significantly increased the mortality (50% within 32 hours and 100% within 13 hours, respectively) as compared with BT (0%), S-10 7 (0%) and S-10 9 (85% within 26 hours) alone. However, the combination group (S-10 7 + BT-10 10 ) with lymphadenectomy prevented death in all animals. In addition, the bacterial recovery in varying compartments of the combination groups was similar to the recovery of each group alone. Conclusion Overall data demonstrated significant deleterious synergistic effects of BT in combination with sepsis, suggesting that translocation of bacteria through the gut-associated lymphoid system might be the main factor for the aggravation of the host proinflammatory response. The BT process can thus be responsible for the installment of the MODS; moreover, this phenomena seems to not be related to the amount of translocated bacteria. Introduction A bacterial translocation (BT) event has been strongly related to the pathogenesis of sepsis as well as to the sepsis progression to the state of multiple organ failure, and increasing scientific findings have pointed out the beneficial role of BT by building a gut immune defense repertoire. In a previous study, we have demonstrated that a previous BT challenge reduces significantly the translocation index at the second BT challenge with the same bacterial strain. Thereby, in this study we sought to evaluate the effect of previous BT challenge on experimentally induced sepsis, examining the bacterial clearance index from the systemic blood circulation, in order to evaluate its influence on the host's immunological defense response. Methods Adult female Wistar rats (200-250 g) were submitted to BT (5 ml Escherichia coli R-6 10 10 CFU/ml/100 g body weight confined to the small intestine for 2 hours) and after 2 weeks were submitted to semi-lethal and lethal sepsis induction (inoculation of 10 9 or 10 10 CFU/ml/100 g body weight E. coli R-6 into the portal vein, respectively), and serial hemocultures were monitored at 0, 15, 30, 60 and 120 min. The BT control group received saline only (n = 8/group). In the other group (n = 6), we evaluated the BT capacity of inducing anti-E. coli R6 O-antigen antibody production after 14 days.
Results Animals submitted to previous BT in combination with semi-lethal sepsis demonstrated a significantly faster bacterial clearance index when compared with animals that were not submitted to previous BT. Also, specific antibody against E. coli R6 was detected in 4/6 animals submitted to BT only 14 days before, suggesting that BT challenge can induce a specific immune response and play a protective role against further second bacterial challenge. However, when animals were submitted to a high concentration of bacteria (lethal sepsis, DL 100 ), previous BT challenge could not play its beneficial role. Conclusion BT is able to build a host's specific immune response, although it is dependent on the sepsis severity. Introduction Bacterial translocation (BT) has been attributed as the causal hypothesis in the development of the multi-organ failure syndrome induced by sepsis, which has been known as 'the gut hypothesis of sepsis'. Such a process is believed to occur by lymphatic and/or hematological routes. In this study we evaluated the role of the lymphatic route in the genesis of microcirculation injury. Methods Adult female Wistar rats (200-250 g) underwent BT induction (BT), and BT induction 5 days (BT5) and 30 days (BT30) post-mesenteric lymphadenectomy, which provoked a complete obstruction of efferent mesenteric lymph flow and re-canalization of the lymph duct, respectively. In the other group, following 30 days post-lymphadenectomy, animals were submitted to the same BT process with exclusion of the efferent lymph duct performed by catheterization (BT30E) (n = 5/group). In these conditions, all animals were submitted to mesenteric microcirculation study by an intravital microscope method from 2 hours of BT up to 4 hours.
Results Animals submitted to BT showed significant injuries. Always, the first lesion was leukocyte adhesion followed by capillary and small venula obstructions and hemorrhage of low flow capillaries and venules 2 hours following the BT process. The BT5 group showed only leukocyte adhesion and the BT30E group showed similar lesions as BT-group alterations, although were much milder even after 4 hours of the BT process. The lesions in the BT30 group initiated around 3 hours of the BT process, basically with leukocyte adhesion followed by few capillary obstructions and rare hemorrhages. Conclusion This study demonstrates that the lymphatic route might carry factor(s) related to the microcirculation injuries induced by gut-associated lymphoid tissue during the BT process. Methods A contemporary cohort study involving adult patients with severe sepsis attending the emergency department in a tertiary hospital. We use the criteria of the SCCM/ACCP consensus conference to standardize the diagnosis of severe sepsis. Results A total 203 patients were enrolled, with mean age 74 ± 13.6 years; 51.7% of these patients were male. The hospital mortality rate was 65%. The mean AL was 3.0 ± 2.2 and the mean BE was -5.8 ± 6.8. Comparing survivors and non-survivors, the discriminative variables were age, APACHE II score and SOFA score, number of organ dysfunctions, AL ≥5 mmol/l, BE ≤ -4 mmol/l, lungs as site of infection, plus systolic and mean arterial pressure. The correlation between AL and BE using Pearson's coefficient of correlation showed an R 2 value of 0.40 (P < 0.0001). The Kaplan-Meier curve for AL ≥5 mmol/l was discriminative (log rank 0.002), but the same did not occur with BE ≤ -4 mmol/l (log rank 0.126). Logistic regression has shown that the variables considered as independent risk factors were age (P < 0. Conclusion During the first days of severe sepsis a moderatedegree metabolic acidosis is caused by unmeasured anions. The SID apparent value can be considered neutral on the first day and becomes more positive during the next few days. This could be the first compensatory mechanism for restoration of a normal metabolic acid-base status.

Bacterial translocation aggravates mesenteric microcirculation changes induced by sepsis
Introduction Considering that systemically exacerbated or impaired inflammatory response with alteration of the microcirculation flow is a universal feature related to septic shock, regarding the current theory of sepsis that considers bacterial translocation (BT) as the main etiological factor for the induction of systemic infection, we sought to examine the additional effect of BT in pre-established sepsis, evaluating intestinal microcirculation injury by intravital microscopy. Methods Adult female Wistar rats (200-250 g) were submitted to the combination induction of sepsis plus BT, and mesenteric microcirculation of the small bowel was monitored for up to 2 hours by intravital microscope under general anesthesia (n = 5/each group). Non-lethal, semi-lethal and lethal sepsis were induced by jugular vein inoculation of 10 7 , 10 9 or 10 10 CFU/ml/100 g body weight of Escherichia coli R-6, respectively. The BT process was induced soon after sepsis induction by oroduodenal inoculation of 5 ml E. coli R-6 10 10 CFU/ml/100 g body weight confined to the small intestine. In control groups, animals were submitted to BT or sepsis alone.
Results BT alone was able to provoke capillary hemorrhages and obstruction of capillaries and venulas of low flow, from 30 min after inoculation, which worsened up to 2 hours. In the sepsis group, although hemorrhagic lesions were not seen, obstruction of venules extended to even high-flow venules, and the severity of microcirculation obstructions were directly proportional to the intensity of sepsis. Only lethal sepsis showed arteriolar obstruction.
Most sepsis-related alterations initiated 25 min after inoculation and worsened up to 2 hours of the observation period. When the BT process was added to the sepsis, non-lethal sepsis microcirculation injuries were as intense as semi-lethal sepsis lesions, and the semi-lethal sepsis microcirculation injuries became as the lethal sepsis microcirculation injuries. Conclusion The BT process, when associated with the preexisting sepsis, augments significantly the mesenteric microcirculation injuries, showing that BT can be the additional triggering factor for the installment of multiple organ failure in sepsis shock. Methods Adult female Wistar rats (200-250 g) were submitted to PH induction by partial portal ligature (decrease of 50% of portal flow) or PH-sham surgery, and 2 days later were subject to a BT experiment (5 ml Escherichia coli R-6 10 10 CFU/ml/100 g body weight confined to the small intestine) or a BT-sham experiment (saline only) under general anesthesia. Following 2 hours of the BT process, samples of the mesenteric lymph node, liver, spleen and lung were collected for culture and the animals were sacrificed.
Groups: G1 (PH-Sham + BT-Sham), G2 (PH + BT-sham), G3 (PH-Sham + BT with inoculum) and G4 (PH + BT with inoculum) (n = 8/group). Results All cultures of G1 were negative; however, when PH factor was added (G2), the native Gram-negative intestinal bacteria translocated significantly to extraintestinal sites with exception of the lung: MLN (5/8, 62.5% positive); liver (2/8, 25% positive); spleen (1/8,12.5% positive) and lung (0/8, 0%). In the conventional BT experiment, all samples were 100% positive with exception of the lung (3/8, 37% positive). Again, when PH factor was added (G4), even the lung was 100% positive, showing that PH factor is a significant factor related to BT even for native microflora. Also, the increased BT index to the lung might be related to the portal blood shunt to the systemic circulation by bypassing the bacterial clearance role of the liver and reaching directly the lung microcirculation, even at the early phase of PH. This fact can be related to the high incidence of pulmonary insufficiency in cirrhotic patients.
Conclusion PH factor increases native flora BT and promotes a higher index of BT to the lung when intestinal bacterial overgrowth factor is present in the rat model of BT.
Introduction Bacterial migration to extraintestinal sites has been attributed as the central component of the gut hypothesis of sepsis. However, some studies have pointed out the beneficial effect of bacterial translocation (BT) on the host's acquired immune system. In this study we evaluated the role of previous BT on the subsequent BT challenge, examining the BT index to organs, changes in WBC count at the mesenteric lymph and blood, and clinical outcome. Methods Wistar rats (n = 60) were distributed into: BT group (n = 20), inoculation of 10 ml of 10 10 CFU/ml Escherichia coli R-6 confined to the small intestine; BT1-14 group (n = 20), submitted to the BT procedure on days 1 and 14; S1-BT14 group (n = 20), received 10 ml saline on day 1 and the BT procedure on day 14. One-half of animals were killed 2 hours following the BT procedure. Samples from different compartments were collected for culture, mesenteric lymph and peripheral blood for WBC count. The other half were subjected to the clinical outcome evaluation concerning weight gain and mortality.
Results Animals submitted to double BT presented a significantly lower index of bacterial recovery (liver, spleen and blood) as compared with animals submitted to a single BT (P < 0.05). The WBC count of mesenteric lymph cells post double BT was similar to naïve animals, and it was significantly lower when compared with single BT (P < 0.05). The clinical outcome was unchanged in double BT as compared with other groups. Conclusion A previous BT challenge was efficient in generating a host defense mechanism against a second episode of BT induced by intestinal overgrowth with the same bacterial strain.
Introduction Sepsis (S) produces regional perfusion abnormalities by causing vasodilatation and blood flow redistribution, and this process mostly affects the mesenteric circulation. In addition, gut mucosal hypoperfusion can perpetuate the inflammatory process and contributes to the multiple organ failure. Thus, following experimental sepsis induction we examined gut blood flow by intravital microscopy and the gut tissue perfusion by laser Doppler, in order to detect the onset of the intestinal microcirculation changes at the acute phase of sepsis. Methods Adult female Wistar rats (200-250 g) were submitted to sublethal sepsis (DL 85 27 hours) by inoculation of 10 9 CFU/ ml/100 g body weight of Escherichia coli R-6 into the jugular vein and were monitored at 3 hours (n = 4) and 6 hours (n = 5) periods, examining 50 villi mucosal blood flow at the distal ileum by intravital microscope and external and internal mucosal surface tissue perfusion by laser Doppler using a type-S probe (3 hours, n = 4 and 6 hours, n = 3). Saline was used in control groups and all procedures were realized under general anesthesia.
Results Intravital analysis showed more than 90% of normal villi microcirculation in all S groups at all time periods, similar to the control groups (100% normal). Concerning tissue perfusion analysis with laser Doppler at the gut external surface, no significant statistical difference could be seen among all groups. Nevertheless, there was a statistical difference at the mucosal surface between the sham and 6 hours S groups, showing that signs of gut hypoperfusion at the mucosal site were not detectable by the intravital microscopy method although they were detectable by laser Doppler at only the late period. These data highlight that intestinal microcirculation is quite stable even during severe sepsis conditions, in contrary to the concept of fragile gut microcirculation related to sepsis shock. Conclusion The laser Doppler mucosal tissue perfusion method can be a useful tool for the detection of microcirculation changes in sepsis. Ongoing studies are being performed to better evaluate microcirculation-measuring tools in sepsis. Methods Male Wistar rats, 21 days of age, were distributed into the control group (diet containing 50 mg/kg elemental iron, n = 12) and the anemic group (diet containing less than 5 mg/kg elemental iron, n = 12). The animals were housed in metabolic cages and received deionized water and diet ad libitum for 6 weeks, and were submitted to BT experiments. Rats were fasted for 24 hours prior to midline laparotomy under general anesthesia. Initially, the distal ileum was ligated and 10 ml saline containing Escherichia coli R-6 (10 10 CFU/ml) was inoculated by oro-duodenal catheterization, and confined in the entire small bowel by duodenal ligature. Afterwards, the abdomen wall was closed by suture. After 2 hours of the BT process, the mesenteric lymph nodes, liver, spleen, lung and blood were collected for culture under anesthesia and were sacrificed soon after. Results At the experimental day, the weight of the anemic group (187 ± 20 g) did not show a statistical difference (P = 0.863) in relation to the control group (193 ± 19 g). However, the hemoglobin (5.6 ± 1.1 g/dl) and hepatic iron (89 ± 15 µg/g) were statistically lower (P < 0.001) than the control group (14.8 ± 0.8 g/dl and 374 ± 60 µg/g, respectively). The median number of E. coli R-6 recovered in mesenteric lymph nodes in the anemic group (26.5 × 10 7 CFU/g) was higher than that of the control group (33.0 × 10 4 ; P = 0.049). The number of bacteria recovered in the liver, spleen, lung and blood were not statistically different between two groups, although there was a higher recovery in anemic group.

Conclusion Iron-deficiency anemia increases intestinal bacterial translocation in rats.
Introduction Sepsis has been considered a major healthcare problem, upheld by the resources consumed to care for patients with this disease and its high incidence and associated mortality rate. Although we are aware of the high total hospital costs associated with sepsis treatment, even post discharge, the difference in costs of sepsis treatment between survivors and nonsurvivors is an economic analysis that can provide more reliable and interchangeable data. Literature addressing the costs of sepsis management is scant. Objectives To assess direct costs of sepsis treatment in Brazilian ICUs, comparing survivors and nonsurvivors until ICU discharge. Design An observational cohort study. Setting Twenty-one ICUs of private and public hospitals. Patients and methods Patients admitted with sepsis, severe sepsis or septic shock were enrolled to the study. During 6 months, patients meeting these criteria underwent clinical and epidemiological evaluation. Hospital costs related to ICU stay were also estimated. Standard values were based on the Brazilian Medical Association (AMB) price index for medical procedures and the BRASÍNDICE price index for medications, solutions and hospital materials. The concept of direct costs was established considering clinical support services (pharmacy, physiotherapy, radiology and laboratory service), consumables (drugs, fluids, nutrition, blood and blood products), and staff (medical staff, technicians and nursing staff). The Kruskal-Wallis test was performed to test for differences in the medians of cost among groups defined according to quartiles of length of stay or tertiles of SOFA score. Simultaneous multiple pair-wise comparisons among groups were performed with the Conover-Inman test. All hypothesis testing was two-tailed; P < 0.05 was considered statistically significant.

Measurements and main results
A total of 524 patients were enrolled. The mean age was 60.5 years and 58% were male. The overall mortality was 43.8% and the median SOFA score was 7.6. Considering the length of stay, survivors and nonsurvivors had similar (median 13 and 10, respectively, P = 0.097). Costs did not differ significantly ($9352 for survivors vs $9116 for nonsurvivors; P = 0.763). However when comparing costs between survivors and nonsurvivors, dividing the length of stay into quartiles, we found a statistically significant difference between both groups (P < 0.0001), even considering the SOFA score when divided into tertiles, mainly comparing SOFA survivors <7 (P < 0.05).
Conclusions This century appears set to become the century of biotechnological advance in healthcare. Unfortunately, the general restrictions on resources make the introduction of new interventions difficult, even if it will not have any cost analysis. Our data reveal an expected reality, that we have differences in costs between survivors and nonsurvivors, mainly comparing length of stay. Moreover, we still have to access the specific areas that have more impact in these direct costs to apply strategies that should offer a better outcome in septic patients. Introduction and objective Elevation in the serum concentration of procalcitonin (PCT) has been proposed as a marker of disease severity and is associated with systemic infection [1]. This association has led to the proposed use of PCT as a novel biomarker of bacterial sepsis [2][3][4]. We sought to evaluate the PCT measurement with a semi-quantitative bedside method (PCTQ). Methods From April to July 2003 we evaluated 48 blood samples from 30 patients (14 males with median age 76.64 ± 13.66 years and 16 females with median age 82.06 ± 10.5 years) with sepsis or SIRS in the ICU. PCT levels were measured and grouped into four intervals (<0.5 ng/ml, 0.5-2 ng/ml, 2.0-10 ng/ml and >10 ng/ml) by a quick bedside semi-quantitative method (BRAHMS PCTQ), and the results compared with measurements performed by a quantitative luminometry method (PCTL) (BRAHMS LUMITEST PCT, Germany).

Results
The Kruskall-Wallis ANOVA analysis found a positive and reasonable correlation between the PCTQ and the PCTL for PCT levels >10 ng/ml. There was no significant difference between the other three intervals (<0.5 ng/ml, 0.5-2 ng/ml and 2.0-10 ng/ml) measured by the PCTQ. Conclusion This preliminary analysis suggested that the PCTQ can be used to accurately measure PCT levels above 10 ng/ml. Other studies with more samples are necessary to provide more information about levels below 10 ng/ml. Systemic penicilliosis is an extremely rare disease in the nonimmunocompromised host. In the HIV-infected patients it is predominantly a late occurrence, with CD4 count normally around 50 cells/mm. Fever and weight loss are the most common first clinical manifestations, present in up to 95% of cases. The respiratory system is the most common organ affected. Cutaneous manifestations are frequent (up to 75% of cases) and are very important clues to the clinical diagnosis of penicilliosis. Over the past 50 years, according to a recent review, a total of 34 cases of invasive infection have been described in the literature, most of them associated with non-immunocompromised hosts. Regarding CNS infection, only three cases had already been reported, two of them in non-immunocompromised patients [1].
Available online http://ccforum.com/supplements/9/S2 Ventilator and hemodynamic support, volume resuscitation and empirical antibiotic support with large spectrum were also employed. The choice of the antibiotics was based upon a probable community-acquired or hospital infection, and it was considered adequate when at least one effective drug had been included. Previous diseases, organic failures, and APACHE II scores were also evaluated. As for the statistical analysis, the t test, the chi-square test and Kaplan-Meier survival curve analyses were applied, considering 5% as the significance level.

Results
The average ranges were: for age (80 ± 7), for APACHE (19 ± 5), for ICU stay (18 ± 9 days), where 51% were women. Among the previous diseases one can point out systemic arterial hypertension in 40%, ischemic heart disease in 31%, stroke in 21% and the COPD in 30% of the cases. Pulmonary SS alone occurred in 70% of the cases, and in association with urinary SS in 27%. The blood cultures were positive in 10% of the samples. The Gram-negative pathogens were responsible for 79% of the infections, where 36% were due to Pseudomonas. The multidrugresistant microorganisms represented 8% of the cultures. Thirtynine deaths occurred during the stay in the ICU. The antibiotics used in the empirical form were correct in 87% of the patients and they were modified in around 72 hours when clinical worsening or inadequate antimicrobial susceptibility patterns result took place. There was no association between age (P = 0.22) or adequate empirical antibiotic therapy and mortality, but mortality was associated with APACHE score (P < 0.001) and organic failures (P = 0.006). The ICU length of stay was not correlated with the use of adequate empirical antibiotics (P = 0.66). Conclusion The adequate and early empirical antibiotic therapy was not associated with mortality or with the ICU stay of the elderly with SS. Possibly, the high level of correct choices of the antibiotic scheme and its modification due to clinical failure and inadequate antimicrobial susceptibility patterns have contributed to the results. .90 (P = 0.57) and the total mean for creatinine before and during treatment was 1.57 ± 1.6 and 2.24 ± 3.5 (P = 0.32), respectively.

Clinical impact of positive blood cultures in intensive
Conclusions In this small sample of critically ill patients, the use of Polimixin did not demonstrate a significant impact in renal function as observed by following the urea and creatinine levels.  Context Patients with renal dysfunction are at an increased risk for cardiovascular disease. Objective To evaluate the prognostic significance of serum creatinine for inhospital mortality in patients with acute coronary syndromes. Design An observational study. Patients and methods Included were consecutive patients with acute coronary syndromes admitted alive to the coronary care unit from February 2004 to January 2005. The patients were initially classified into three groups on the basis of serum creatinine concentration measured on admittance. Normal renal function and mild and severe renal dysfunction were defined as serum creatinine concentrations of <1.2 mg/dl, ≥1.2 but <2.0 mg/dl, and ≥2.0 mg/dl, respectively. Patients receiving regular hemodialysis were excluded from the study. Univariate and multivariate relative risks (RRs) were calculated for three renal risk quartiles using the serum creatinine concentration on presentation. Results This study included 227 patients. The mean age was greater in the severe renal dysfunction group (P = 0.031). Of the analysed variables -age, sex, diabetes, hypertension, dyslipidemia, previous myocardial infarction, previous coronary angioplasty and coronary artery bypass graft -only the creatinine level was an independent predictor of greater inhospital mortality. The inhospital mortalities of mild and severe renal dysfunction patients were greater (7.9% and 31.6%, respectively) than that of patients without renal dysfunction (2.8%), P = 0.000, with an increased risk of 10 times.

Conclusion
In this study, we showed that the creatinine level at admission is an independent predictor of inhospital mortality in patients with acute coronary syndromes. Introduction Acute renal failure (ARF) is frequent in severe patients, producing a poor outcome in the face of its multiple insults that are individually determined [1,2]. Technetium 99m ethylenedicysteine scintigraphy (Tc-EC) has a very low plasma protein binding and a large volume of distribution. Tc-EC is a good agent for renal function evaluation, providing an index of tubular function and yielding high-quality images [3][4][5]. Objective To evaluate the clinical impact of renal scintigraphy (RC) in the management of ARF.

Materials and methods
We evaluate all patients in ICUs of a general, private hospital, submitted to RC in the period of January 2003 to January 2004. Results Thirty-four patients (18 male), 79.56 ± 16.26 years old, with an APACHE II score of 16.94 ± 6.34 (expected mortality 25%) were studied. The observed mortality was 29.41%. Seventythree percent of the patients were anuric for 55.36 ± 13.43 days. Dialyses were used in 85.29% of the population and RC always carried through after 30 days. Seventy-four percent of the patients were mechanically ventilated during RC with no transport accident. RC changed the nephrology prescription, interrupted the method or indicated long-term access confection. The radiotracer most used was Tc-EC in 64.70% of the patients. Twenty percent of the patients continued in chronic dialysis and 17.64% had recovered renal function. Conclusion RC, mainly with Tc-EC, influenced and modified the clinical impression, nephrological decision, and medical care of severe ARF patients. RC seems to be a good cost-benefit with simple accomplishment method to evaluate glomerular filtration and tubular function with an important impact especially in anuric patients.
Objectives To evaluate the impact of RD in patients admitted due to DHF, and its relationship with clinical features and laboratory data, length of stay (LOS), and inhospital complications and mortality. Methods From January 2003 to December 2004, we studied a cohort of 137 patients admitted to the coronary care unit due to DHF (79.6% NYHA class IV). The mean age was 76.5 ± 11.08 years, 54% male, 29.9% diabetes mellitus, 74.5% systemic hypertension, 64% ischemic cardiomyopathy and the mean LOS was 14.2 ± 34.6 days. RD was defined as an estimated (Cockcroft) creatinine clearance less than 60 ml/min on admission. Baseline demographics, laboratory findings, LOS and complications (cardiac arrhythmias, hemorrhage, need of blood transfusions, hemodynamic instability and infections) and mortality rates were compared. The Mann-Whitney test (laboratory findings and LOS), the Student t test (age) and Pearson's chi-square test (other variables) were used. Results A total of 73.4% of the patients with DHF were considered to have RD. They were older (79.7 ± 9.5 vs 67.1 ± 10.4 years, P < 0.0001), with paradoxically less diabetes (18.1% vs 55.9%, P < 0.0001). On admission, B-type natriuretic peptide (P = 0.021), D-dimer (P = 0.024), hematocrit (36.2 ± 5.6% vs 38.2 ± 4.4%, P = 0.029) and hemoglobin (12.1 ± 1.89 vs 12.9 ± 2.08 g/dl, P = 0.057) were smaller. The need for blood transfusion (21.2% vs 5.9%, P = 0.041), and a significant increase of LOS was observed in the RD group (16.3 ± 41.2 vs 8.4 ± 5.7 days, P = 0.013), and higher inhospital mortality (9.57% vs 5.88%, P = not significant) was observed in the RD group. Conclusions RD is highly prevalent in patients admitted due to severe heart failure, with a clinical impact on blood transfusion, length of stay, and determining a trend to higher mortality. Introduction Acute renal (ARF) failure has a notable prevalence in the ICU. When patients are identified by small alterations in the creatinine levels, the percentage of ARF may reach 25% in the critically ill. Objective To evaluate ARF in two groups of patients: Group A (GA), patients with ARF at admission to the ICU; Group B (GB), patients developing ARF during the ICU stay. Materials and methods A prospective survey, carried out between July 2004 and February 2005 in a tertiary care private hospital. The sample was composed of two groups. GA: patients with ARF at admission, creatinine >2.0 mg/dl and normal renal function prior to the admission. GB: patients with normal renal function, with a 0.5 mg/dl increase in the serum creatinine during the ICU stay or if presented with chronic renal failure, defined as serum creatinine >2 and <4.9 mg/dl, an increase of 1.0 mg/dl in serum creatinine levels. Patients with serum creatinine level >5 mg/dl, history of ARF in another hospital or kidney transplant were excluded. Results Six hundred and eighty patients were admitted to the ICU during the study period. Sixty-five patients had ARF, 14 in GA and 51 in GB. The mean age was 62.79 years and 69.4 years, respectively. The causes of ARF in GA were hypoperfusion (28.6%) and sepsis (21.4%), whereas in GB they were hypoperfusion (35.3%), septic shock (29.4%) and sepsis (19.6%). The prevalence of contrast-induced ARF was 7.1% in GA, and 2.0% in GB (P < 0.05). The mean serum creatinine level was 4.2 mg/dl in GA and 2.7 mg/dl in GB (P < 0.05). The mean urine output measured 24 hours before the diagnosis of ARF was 551.08 ml in GA and 713.37 ml in GB. The APACHE II score was 33.94 in GA and 41.76 in GB. The mortality in GA was 28.6% (four patients) and was 58.8% in group B (30 patients) (P < 0.05). The mortality for the ICU in general was 13.2% (90 patients). Twenty-three patients went for dialysis: GA 35.7% (five patients) and GB 35.29% (18 patients). The mean number of sessions was 4.78 in GA and 3.2 in GB. Four patients in GA (80%) and 15 (83.3%) in GB died in the dialysis group. Conclusions The ARF mortality remains high despite the advances in technology, notably in patients on dialysis. In some cases the ARF diagnosis was delayed, as demonstrated by the serum creatinine levels in GA.
Introduction High pressures of the tracheal tube cuffs can impair damage for restriction of the air flow. Objectives To demonstrate the prevalence of high pressures of the tracheal cuff in orotracheal tubes (OTTs) and tracheotomies (TCT). To verify the impact of cuff pressure control on further cuff volume adjustment. Methods This is a transversal study, which evaluated every tracheally intubated (OTT or TCT) patient in a general or neurosurgery ICU. Daily we have measured cuff pressures with a mercury manometer. The pressure goal was 20 mmHg (27.2 cmH 2 O). Results One hundred and six patients were evaluated, 75 (71%) male. Their age ranged from 16 to 92 years (mode 85 years). Forty-two (39%) patients had undergone TCT and 64 (61%) had undergone OTT. The first-day APACHE II score ranged from 4 to 39 (mode 22) and the average mortality risk was 27.5%. The length of time of OTT ranged from 1 to 16 days (mode 8 days) and that for TCT from 5 to 70 days (mode 5 days). The initial pressures ranged from <20 mmHg to 140 mmHg (mode <20 mmHg). After initial reduction of the cuff volumes, the range of the pressures decreased (from <20 mmHg to 80 mmHg). Conclusion Increased prevalence of high pressures of the cuff in tracheal tubes justifies frequent monitoring of this parameter, aiming at brightening up the deleterious effect of the drawn-out use of these tubes. Background and objectives Indexes predicting the weaning outcome are frequently inaccurate. With the present study, we aim to evaluate the predictive performance of a new index for predicting the weaning outcome, which we called the integrative weaning index (IWI). Methods Two hundred and fifty patients of several etiologies in the weaning process that remained up to 24 hours in mechanical ventilation were evaluated (all with PaO 2 ≥60 mmHg with FiO 2 ≤0.4 and PEEP ≤8 cmH 2 O). All patients were submitted to a 2hour trial of spontaneous breathing. Those who sustained 2 hours of spontaneous breathing without return to mechanical ventilation in the following 24 hours were considered weaned, while those who could not sustain 2 hours of spontaneous breathing or returned to mechanical ventilation in the following 24 hours were considered not weaned. The frequency/tidal volume ratio (f/Vt ratio), the airway occlusion pressure at 0.1 s after the onset of inspiratory effort (P 0.1), the product of P 0.1 and f/Vt (P 0.1 × f/Vt), the respiratory rate (RR), the quasi-static compliance of the respiratory system (Cqst,rs), the PaO 2 /FiO 2 ratio and the new integrative weaning index (IWI = Cqst,rs × SaO 2 / f/Vt ratio) were evaluated in all patients. Arterial blood gas was collected with FiO 2 in 0.35. The sensitivity, specificity, positive predictive value, negative predictive value and the receiver operating characteristic (ROC) curves were calculated in order to evaluate the predictive performance of each index. The nonparametric method of Hanley and McNeil was used to compare the area under the ROC curves of each index. Results Two hundred and eighteen patients were weaned, while 32 patients were not weaned. The IWI presented the larger area under the ROC curves (0.97), followed by the f/Vt ratio (0.90), Cqst,rs (0.89), f/Vt × P 0.1 (0.85), RR (0.80), P 0.1 (0.70) and finally by the PaO 2 /FiO 2 ratio (0.60). The area under the ROC curves of the IWI was larger than those for f/Vt ratio (P < 0.0001) and also larger than those for the other indexes (P < 0.0001). Conclusion In our study, even compared with other essential indexes in the literature, the IWI showed the best criteria for predicting the weaning outcome. With the obtained results we believe that with the use of the IWI in other countries we may further prove its accuracy. Methods Seventy consecutive patients of several etiologies in the weaning process that remained up to 24 hours in mechanical ventilation were evaluated (all with PaO 2 ≥60 mmHg with FiO 2 ≤0.4 and PEEP ≤8 cmH 2 O). All patients were submitted to a 2hour trial of spontaneous breathing. Those who sustained 2 hours of spontaneous breathing without return to mechanical ventilation in the following 24 hours were considered weaned, while those who could not sustain 2 hours of spontaneous breathing or returned to mechanical ventilation in the following 24 hours were considered not weaned. MIP < -25 cmH 2 O, P 0.1 <4.2 cmH 2 O and P 0.1/MIP < 0.14 cmH 2 O were used to predict the success in weaning outcome. The predictive performance of each index was evaluated through the sensibility, specificity, positive predictive value, negative predictive value and diagnostic accuracy. The results were also evaluated by the area under the receiver operating characteristic (ROC) curves.
Results MIP presented an area under the ROC curves smaller than those for P 0.1 (0.52 ± 0.08 vs 0.76 ± 0.06, respectively; P = 0.004) and also smaller than those for P 0.1/MIP (0.52 ± 0.08 vs 0.78 ± 0.06, respectively; P = 0.0006). P 0.1/MIP presented excellent predictive performance in weaned patients, with sensibility of 98.08, but with the area under the ROC curves only slightly larger than those for P 0.1 (0.78 ± 0.06 vs 0.76 ± 0.06, respectively; P = 0.69). Conclusion In our study, Pi max was the criterion with the worst predictive performance. P 0.1 was shown to be a very important criterion to evaluate the respiratory center output, although with limitations in evaluating weaning failure. Patients with P 0.1/Pi max ratio >0.14 are not always associated with weaning failure, but values <0.14 were highly associated with success in weaning outcome.  Objective To analyze early extubation of patients submitted to heart surgery while in the ICU. Materials and methods Aleatory and prospective analysis of patients submitted to heart surgery and early extubation immediately post operation, with other aspects being studied such as: clinical evolution, cardiopulmonary complications, time of stay in the ICU and complications related to early extubation. As criteria for inclusion, the main determining factors were hemodynamic stability, consciousness level (Ramsay >3), motor function preservation, normal radiologic pattern and PO 2 /FiO 2 ratio >250. After 1 hour of ICU arrival, patients were submitted to superficialization of the consciousness level with an antagonist of benzodiazepinics, and then nebulization with a 'T' piece for 30 min, following extubation, and then intermittent noninvasive ventilation support (CPAP). All of the patients were monitored with pulse oximetry, series of gas analysis, clinical evaluation and radiologic control.
Results Out of the total number of patients analyzed (n = 100), 72% were male. Regarding the type of surgery, there was a higher prevalence of patients submitted to CABG (83%); other types such as valve surgeries and congenital corrections totaled 17%. The average time for extubation was 2 hours (ICU arrival until extubation), and the average period of nebulization was 30 min. In this group, 4% went through reintubation because of somnolence. There were no deaths among the patients studied.
Conclusion The determination of specific protocols for early extubation facilitates a satisfactory evolution with less morbidity among patients submitted to heart surgery. Method Twenty-four patients were randomly assigned into two groups: GEP (n = 8) performed respiratory exercises with positive expiratory pressure (EPAP) associated with PP1; and GP1 (n = 16) performed PP1 only. The PF was measured by spirometry preoperatively and at the fifth PO (5PO), and the IMS was measured by the maximal inspiratory pressure (PImax) preoperatively, at the first PO (1PO) and at 5PO. Results For spirometric variables, as a percentage of predicted, significant reductions were observed when the preoperative was compared with 1PO for GP1, but for GEP only a difference to the vital capacity was observed (P > 0.05). In relation to the differences between the treatments, smaller values were observed in GP1 for peak flow (PFl) at 5PO. We found significant reductions of PImax preoperatively to 1PO in both groups. However, PImax showed a reduction of the preoperative to 5PO situation only in GPI (P < 0.05).
Conclusion These data suggest that CS produces reductions of IMS and PF, and that EPAP associated with PP1 was more effective in the reduction of respiratory changes than PP1 isolated. However, the PF did not wholly reverse until 5PO, the continuance of treatments being necessary after this hospital period.
Background Previous studies have suggested that normal results of D-dimer and alveolar dead space are highly predictive tools to determine pulmonary embolism in patients with acute dyspnea.
Objective To assess the utility of a protocol to monitor the alveolar dead space fraction associated with D-dimer in the prediction of pulmonary embolism in patients with acute dyspnea. Methods A prospective study where a protocol was used to assess alveolar dead space and D-dimer in Q patients who presented to the emergency department with dyspnea. ; and nine men with COPD, aged 71 ± 7 years [G2]) were studied. The heart rate and R-R intervals (ms) were collected on a beat-to-beat basis from an electrocardiogram recording in the CM5 derivation with an electrocardiography monitor (Ecafix TC500). The signal was converted by an analogical/digital converser (Lab.PC+; National Instruments) and was processed by a specific routine. The record was collected in two conditions: 10 min at rest in the sitting position during spontaneous breathing, and 10 min during the same position with biBiPAP (BiPAP ® S; Respironics) application through a nasal mask. The BiPAP pressures slowly increased the inspiratory pressure up to 14-16 cmH 2 O and expiratory pressure up to 6 cmH 2 O, allowing voluntary adaptation. The BiPAP application was realized only in COPD. The heart rate and its variability were analyzed by the time domain (TD), through RMSSD (ms) and RMSM (ms) indexes, and the frequency domain (FD), which was expressed as normalized units of low frequency (LF), high frequency (HF) and the LF/HF ratio. The intragroup comparison was performed by the Wilcoxon test and the inter-group comparison by the Mann-Whitney test, with P < 0.05. Results See Table 1.
Conclusions Our results suggest that the patients with COPD present modifications of sympathetic-vagal balance, with parasympathetic predominance over sympathetic in the sinoatrial node when compared with healthy elderly men. Moreover, the BiPAP application in this study reduced the HRV by DT analysis in COPD patients.
Introduction Respiratory insufficiency is a common cause of ICU admission. In order to evaluate causes and mortality, a consecutive series of 44 patients were studied at the ICU of Santa Lucia Hospital, according to a protocol applied to collect data. Materials and methods From 6 November 2004 to 11 January 2005, 44 consecutive patients were admitted to the ICU of Santa Lucia Hospital with respiratory insufficiency. Data were collected prospectively through a specific protocol for all patients that needed invasive or non-invasive ventilatory support.
Available online http://ccforum.com/supplements/9/S2 Despite the initial treatment, he persisted with desaturation and marked hypercapnia. Chest X-rays revealed hyperinflation of the left lung, with significant reduction of the right lung area. The first CT scan showed extensive consolidation of the right lung and atelectasis. Bronchoscopy revealed reduction of the right bronchus caliber and was diagnosed as right lung hypoplasia. The child continued to have desaturation and it was not possible to wean him off mechanical ventilation. Fifteen days later he had another CT scan showing overdistension of the left lung apex, suggesting the diagnosis of congenital lobar emphysema. After the second CT scan, we performed a selective intubation to the right, trying to inflate the collapsed right lung, and there was a marked improvement on saturation; a chest X-ray showed aeration of the right lung. The child was submitted to superior left lobe lobectomy, and after the surgery he improved in pulmonary function and it was possible to wean him off mechanical ventilation. The patient was discharged from the ICU 20 days later. Discussion CLPE is characterized by postnatal overdistension of one or more pulmonary lobes, the left lobe being the most frequently affected. One-half of the cases have unknown causes. When a marked overdistension is present, lung herniation and collapse of the contralateral lung can occur. In the first instance, we considered the hypothesis of right lung hypoplasia, following bronchoscopy findings. CLPE was confirmed after a second CT scan.
Conclusion CLPE is a rare pulmonary condition, usually diagnosed before 6 months of age; it must be considered a differential diagnosis in cases of suspected contralateral lung hypoplasia. Background Vasospasm following subarachnoid hemorrhage (SAH) after rupture of a cerebral aneurysm is a major complication. If left untreated it leads to death or permanent deficits in over 20% of patients. Treatment includes triple H therapy and nimodipine given by the enteral route. In severe cases, angioplasty and superselective intra-arterial injection of vasodilators may be effective.

Pulmonary toxicity induced by bleomycin in a patient with
Objective To evaluate the effect on transcranial Doppler (TCD) velocities of the treatment of moderate to severe refractory vasospasm with intra-arterial nimodipine. Methods Three patients with Fischer IV SAH after aneurysm rupture were treated with coil embolization in the first 3 days after the hemorrhagic event, and enteral nimodipine (360 mg/day) developed moderate to severe vasospasm on the middle cerebral artery (detected by TCD) refractory to triple H therapy. They were treated with superselective injection of intra-arterial nimodipine, infused during a period of 30 min-1 hour, and were evaluated after 12-24 hours by a new TCD. Results All three patients showed a reduction of mean flow velocity greater than 30% and a Lindergaard ratio lower than 3. Conclusions This is a small series of patients but it suggests that treatment with intra-arterial nimodipine may be effective in reducing the severity of vasospasm refractory to clinical therapy. We intend to increase our cohort and correlate these effects with the development of ischemia and neurological outcome.
Introduction Systematic use of head CT scan after minor head trauma may lead to a high incidence of negative examinations and high costs. On the other hand, undetected intracranial lesions may be life threatening. Neuronal protein S-100b into the circulation has been suggested as a specific indication of neuronal damage. We tested the hypothesis that protein S-100b is a useful and costeffective screening tool for the management of minor head traumas.
Methods Fifty consecutive patients sustaining isolated minor head trauma were prospectively evaluated in the emergency room by routine head CT scan and blood sampling for protein S-100b measurement, using an immunoluminescence test kit. Seventeen normal healthy individuals served as negative controls. Data are presented as the median and 25th-75th percentiles.
Results Patients reached the emergency room 45 min (30-62 min) after minor head trauma. Six patients had relevant post-traumatic lesions at the initial head CT (12%) and were thereby counted as positive (CT+). The median systemic concentration of S-100b in those patients was 0.75 µg/l (0.61-6.5 µg/l), which was significantly different (U-test, P = 0.011) from the median concentration, 0.26 µg/l (0.12-0.65 µg/l), of those without post-traumatic lesions in the initial head CT scan (CT-). A sensitivity of 100%, a specificity of 20%, a positive predictive value of 15% and a negative predictive value of 100% were detected. Conclusions Protein S-100b has a very high sensitivity and negative predictive value, and could have an important role in ruling out the need for CT scan after minor head trauma. This may be of clinical relevance, particularly in countries in which trauma is epidemic and medical resources are limited, such as in Brazil.
Conclusion We should not estimate the protein necessity in a septic or trauma patient on CHVVD based exclusively on the amino acid loss. Other factors, such as the adsorption of amino acids, may contribute to the continuous catabolism seen in these critically ill patients. We found 32% of hyperglycemic levels and 0.28% of hypoglycemia in the whole population. In the diabetic group, the incidence of hyperglycemia and hypoglycemia were, respectively, 37% and 0.44%, while in the nondiabetic group they were 29% and 0.21%.

Discussion
The literature provides similar trends in the safety profile of IIIP in the critically ill. It seems necessary to observe the differences between diabetic and nondiabetic patients. Objectives To analyze when to start EN and the BEE achieved in the elderly hospitalized with a septic shock diagnosis, and to verify its association with mortality in these patients. Methods A prospective cohort conducted within 32 months and with 67 patients over 65 years old in the ICU with septic shock, where 59 of these patients had EN. The APACHE II score was determined in every patient. The following variables were analyzed in this group: the time taken to start the NE, the BEE (achieved or not), and the time to reach the BEE in those who had it. These variables were correlated with death and it was still observed whether there was correlation between the starting time point of Critical Care June 2005 Vol 9 Suppl 2 Third International Symposium on Intensive Care and Emergency Medicine for Latin America S39 EN and the achieved BEE. The statistical tests used were the t test and the chi-square test, considering 5% as the significance level. Results The average age was 80 ± 7 (minimum = 66, maximum = 96) years, the APACHE II average score was 18 ± 5 (minimum = 8, maximum = 28), and the time average to start EN was 80 ± 53 hours (minimum of 12 hours and maximum of 240 hours). The achieved BEE occurred in 69.5% (n = 41) of the patients and the time average to reach it was 115 ± 56 hours (minimum of 72 hours and maximum of 360 hours). Death was associated with time to start EN (P = 0.001) and with the non-achieved BEE (P < 0.001). However, there was no correlation with time to reach BEE (P = 0.22). The time to start EN did not show association with the achieved BEE (P = 0.08).
Conclusions Initiating EN as soon as possible and the BEE when achieved in this group of patients showed some benefit. The time to achieve the BEE does not seem to have correlation with mortality in these patients. The time to start NE did not have association with the BEE achieved in this sample.

Results
The main indication for the use of PICC was the administration of antibiotics, followed by difficulty in venous access and the administration of medications that act in the vascular system. Eighty-five percent of the catheters were used in the Semi-ICU. The great majority of patients were taken off the catheter just after the completion of the treatment (85%). There were two cases of phlebitis, three cases in which the catheter was accidentally removed and one case of obstruction. Conclusions The PICC has importance and application in intensive therapy, being one more therapeutic option, with a low range of mechanical and infectious complications. It is necessary for institutional training to have adequate maintenance and manipulation.
Avf, II and III. In the following, they were analysed by our software tool to estimate the variation of the area of the ST segment. First, the .pdf files were converted to eight-bit .bmp files, as images with 2000 × 1600 pixel resolution, in a grey scale. Axes are defined in the image to extract a data file, which may be converted into any desired format. This data file was processed by two different methods, in order to estimate the variation of the ST-segment area. The first method employs vertical axis scanning whereas the second method carries out standard integration, based on a finiteinterval technique. Results The two methods presented the following average errors for the estimation: 1.80% for derivation Avf, 1.27% for derivation II and 1.38% for derivation III. Average variances were 7000 for method 1 (axis scanning) and 7200 for method 2 (standard integration). Conclusion Simple and cost-effective methods for the analysis of ECG recordings arising from a telemedicine system, which may be used in current computer configurations in the ICU, were successfully employed to estimate the variation of the ST-segment area for myocardial infarction patients. Method  Among the relatives, the parents had larger values of TSS. The respiratory therapists were the professionals that obtained a larger TSS and the smallest average was found among the doctors. The average of the patients' TSS was lower than that of the relatives and that of the health care professionals. There was no statistically difference between the latter two groups (family and professionals). Conclusion The perception of the main stressful factors was different among the three groups, especially when patients were compared with the relatives and health care professionals. These groups (family and professionals) believe that some factors are more stressful for the patient than the patient themself would have stated.

P109
Intensive care medicine teaching in a Brazilian medical school: the student's perspective S41 in their curriculum. As a result, students often seek exposure to ICM in extracurricular activities.
Objective To analyze the interest and contact with ICM among students of a Brazilian public medical school. Methods This is a descriptive study. We applied a questionnaire to enroll students between the sixth and the final semesters.

Results
We studied 216 students. The mean age was 22.8 ± 1.7 years, and 61.7% (n = 129) were men. Most of them (56.5%, n = 122) had never frequented an ICU despite classifying the usefulness of an apprenticeship in this area as high (average of 4.3 ± 0.9, in a scale from 1 to 5). The main reason for not frequenting an ICU was lack of opportunity to do so (80.9%, n = 93). Among students that had already frequented an ICU, 81.9% did so exclusively as part of extracurricular activities; the main reason for seeking this exposure was interest in ICM as a future specialty (37.7%, n = 26). Almost all students (98.6%, n = 212) thought that ICM topics should be more explored at their university. The main causes for students' dissatisfaction with ICM teaching at their university were: disinterest of the manager (65.5%, n = 135), disinterest of the teachers (24.3%, n = 50) and lack of qualified teachers (14.4%, n = 31). Although most students (55.3%, n = 119) had already participated in a discussion to send a patient to the ICU, 32.4% (n = 36) thought they were not capable of identifying a patient with the need for intensive care. On a scale from 1 to 5, the mean interest in ICM was 3.6 ± 1.0. The most popular topics were: shock (4.74 ± 0.60), cardiopulmonary resuscitation (4.73 ± 0.64), and SIRS/sepsis (4.63 ± 0.66). The procedures that more students had contact with were: peripheral venous access (40.9%, n = 88); cardiopulmonary resuscitation (28.4%, n = 61); and passage of a vesical probe (25.6%, n = 55).
Conclusions This study revealed a high interest in ICM topics among medical students. However, the majority of the population studied had not frequented an ICU before. Moreover, of those who had already frequented an ICU, most had done so in units outside the academic atmosphere. Objectives Portable, supine chest roentgenograms (CXRs) are the most commonly used noninvasive method to identify pulmonary edema in the ICU. We conduct a study to evaluate the influence of cardiomegaly, pulmonary edema and vascular pedicle width (VPW) on diagnosing hypervolemia using digital CXRs. Methods A radiologist selected eight CXRs to assure equal proportions of: cardiothoracic ratio (CTR) >0.55; presence of Kerley's lines and VPW >70 mm. Assuming that VPW >70 mm is indicative of hypervolemia, two groups of 10 intensivists (≤10 years and >10 years of activity) were enrolled without any additional clinical information.
Results See Table 1.
Conclusions The overall agreement of VPW and the diagnostic of hypervolemia is poor among young and older intensivists. Older intensivists are prone to admit hypervolemia even in the absence of cardiomegaly or pulmonary congestion. Objective To retrospectively review dexmedetomidine infusion for more than 24 hours in acutely ill patients and to evaluate its safety profile. Design and setting A retrospective observational study in a tertiary general hospital, mixed ICU. Patients Fifty-seven patients who were assigned to dexmedetomidine use for more than 24 hours over a 12-month period.
Results From January to December 2004, 57 patients were included. Most patients were male (71.9%) and the mean age was 55.6 years (19-93 years); the mean APACHE II score was 17.5 (6-41). The mean ICU and hospital lengths of stay were 15.4 and 43.5 days, respectively. There were 32 medical patients and 25 emergency and elective postoperative care admissions. The mean duration of drug infusion was 82.2 hours (25-408 hours), and 100% of the patients were on coadjuvant sedative agents when dexmedetomidine was started. Hypotension episodes occurred in five patients (8.7%) receiving the drug for more than 24 hours, and no bradycardia episodes were noted. Six patients (10.9%) died during the hospital stay, but they were not timely related to the dexmedetomidine infusion.
Conclusions The safety profile of dexmedetomidine use, as a coadjuvant sedative agent, for more than 24 hours in the ICU setting is comparable with its use for less than 24 hours. No clinically relevant side effects directly related to the study drug were observed after a 24-hour infusion period.
Introduction Severe sepsis and septic shock are characterized by a high incidence, mortality and cost. Actually, sepsis is a major healthcare problem, upheld by the resources consumed to care for patients with this disease. Although we are aware of the high total hospital costs associated with sepsis treatment, even post discharge, the heterogeneity of the health care system (private or public hospitals) makes any estimate of costs directly attributable to sepsis a real challenge. Besides direct and indirect costs, 'hidden costs' like education, staff training and comorbidity-related issues can be significant in a major disease. Direct costs are defined mainly by the physicians and nursing fees, medicines, blood products and equipments used for monitoring and organ dysfunction support in sepsis. This simplified type of economic analysis can provide more reliable and interchangeable data.
Objective To assess the direct costs of sepsis treatment in Brazilian ICUs, comparing private and public hospitals. Design An observational cohort study. Setting Twenty-one ICUs of private and public hospitals.
Patients Patients admitted to one of the ICUs with sepsis, severe sepsis or septic shock, according to SCCM/ACCP Consensus Conference criteria, were enrolled to the study. During 6 months (1 October 2003 to 30 March 2004) the collected data were analyzed. Patients meeting these criteria underwent clinical and epidemiological evaluation. Hospital costs related to ICU stay were also estimated. Indirect cost estimates like administrative issues, electrical energy, depressed state of the equipment and facility maintenance were not included in the economic analysis. The TISS (Simplified Therapeutic Intervention Scoring System) score was also used for cost estimation. To compare the groups (public and private) we used the Mann-Whitney test and the Student t test. Standard values were based on the Brazilian Medical Association (AMB) price index for medical procedures and the BRASÍNDICE price index for medications, solutions and hospital materials.

Measurements and main results
For the 619 patients included, only 85% were enrolled, considering 37.6% from private institutions and 62.4% from public institutions. From these data, 58% were male, the mean age was 60.5 years and the overall mortality was 43.8%. For public and private hospitals we found a median SOFA score of 7.5 and 7.1, and the mortality rate was 49.1% and 36.7%, respectively. Public and private hospitals had similar lengths of stay (median 10 and 9, respectively, P = 0.091). The total direct costs did not differ significantly ($9260 for public hospital vs $8776 for private hospital, P = 0.328). Conclusions Sepsis remains a major world health problem. Our data have not shown a significant difference in direct costs between public and private hospitals, perhaps our length of stay was also comparable. An accurate estimate of the cost of hospital care for septic patients would be essential to encourage physicians and healthcare managers to develop and implement evidence-based strategies to improve quality of care and to reduce costs in the ICU. Thrombotic thrombocytopenic purpura and hemolytic uremic syndrome are uncommon diseases that frequently overlap their features, together having an incidence of three to seven cases per million. Pancreatitis may be, in some cases, part of both clinical presentations. The onset of pancreatitis followed by manifestations of these diseases has few reports in the literature. We report the case of a 27-year-old man who presented to the emergency department of our hospital with 2 days of epigastric pain associated with nausea and vomiting, mild jaundice and moderate dehydration. He had no fever and his physical examination was unremarkable except for a light tenderness of the epigastric area and abdominal distention. An initial amylase level of 612 U/l, lipase of 2678 U/l together with a leukocyte count of 12,000 were the only altered laboratory data. The abdominal sonography showed a heterogeneous pancreas, with a slight increase in its size, and a lightly distended gallbladder without any signs of gallstones. We also performed an abdominal CT, where there was an undefined and irregular contour of the pancreas, increased density of adjacent plans, a few liquid collections within and bilateral pleural effusion. A diagnosis of pancreatitis was made and 2 days after his admission the patient had complete clinical relapse and was about to start an oral diet again. Nevertheless his laboratory data showed progressive thrombocytopenia with a fall from 333,000 to 9000 platelets and a rise on LDH from 363 to 5711, together with a progressive loss of renal function (peak creatinine, 3.8 mg/dl). He became oliguric but had no need for renal replacement therapy; he also presented light drowsiness and confusion. After the confirmation of erythrocyte fragmentation in a peripheral blood smear, we established a diagnosis of thrombotic thrombocytopenic purpura and started plasmapheresis daily for 6 days until his platelets reached 150,000. After that he had four more sessions on alternate days and was discharged from hospital with a platelet count of 269,000, without any neurologic or renal impairment. He has been followed in the hematologic department of our hospital and has so far maintained remission (3 months).

Surgery/trauma
Critical Care June 2005 Vol 9 Suppl 2 Third International Symposium on Intensive Care and Emergency Medicine for Latin America Introduction and objectives The objective of this trial was to compare abdominal hypertension and abdominal compartmental syndrome (ACS) in patients under risk for these pathologies in the ICU.
Methods Eighty-nine patients with monitoring indications for intraabdominal pressure (IAP) based on clinical or surgical admission diagnosis in the ICU were included. Measurements via an indirect method through the vesical catheter were made at three distinct timepoints: at admission, and at 6 hours and 12 hours. Results The incidence of ACS was 14.6% (13 patients); in the total sample, the male sex prevailed at 63% (56 patients). The patients had been distributed into two groups: with ACS and without ACS. Gastric surgery diagnosis predominated in both groups (84.6%/57.9%; P < 0.067). Global mortality was 30.3% (27 patients): 53.8% (seven patients) in the group with ACS and 26.3% (20 patients) in the group without ACS; P < 0.046. Table 1 describes the sample. Conclusion DEX is an effective and safe sedative agent that has analgesic properties and predictable cardiovascular effects, and may benefit a heterogeneous population of critically ill patients for short-term (early postoperative period) and long-term sedation. We believe that a continuous infusion of a α 2 -agonist agent for sedation-induced withdrawal syndrome management is highly applicable. Reference Introduction Trauma patients have a high mortality and more often develop organ failure with poor outcome.
Objective To study the incidence of organ dysfunction and failure, and the impact on mortality of patients who were admitted to an ICU with abdominal trauma without other lesions.

Materials and methods
We studied patients admitted to a trauma ICU from 1 January 2001 to 31 December 2003. The Sequential Organ Failure Assessment score at admission was measured; we considered points 1 and 2 of the score as dysfunction and points 3 and 4 as failure.
Results A total of 77 patients were studied and the mortality was 31%; the mortality of the general population with multiple traumas (725 patients) was 39%. The incidence of organ dysfunction and failure, with the correspondent mortality, is illustrated in Table 1.

Conclusion
The frequency of coagulation and renal dysfunction and failure is higher in patients with abdominal trauma than in patients with multiple trauma. The presence of renal, cardiovascular and hematological failures was associated with higher mortality in patients with abdominal trauma than in patients with multiple trauma.
Critical Care June 2005 Vol 9 Suppl 2 Third International Symposium on Intensive Care and Emergency Medicine for Latin America