Haemoglobin behaviour in critical patients

Hypocalcemia has a prevalence of 88% in general intensive care units (ICUs). Calcium (Ca) administration may be associated with hemodynamic improvement, but with increased mortality in animal studies.

Background Intra-myocardial injections of BMMCs have shown promising initial results regarding improvement in myocardial ischemia. Experimental models have depicted the potential of some cell phenotypes in differentiating into blood vessels. BMMCs are a heterogeneous cell subpopulation group and the individual contribution of each cell subpopulation to favorable clinical outcomes remains unclear.
At 6 months, the global mortality was greater in group I (12%) versus 5% in group II (P < 0.02).

Conclusion
Patients with AU/NSTIMI with troponin levels more than 0.2 µg/ml had more risk of death in 6 months. Background This study evaluated the hypothesis that transendocardial injections of autologous mononuclear bone-marrow cells in patients with end-stage ischemic heart disease could promote neovascularization and improve perfusion and myocardial contractility.

Methods and results
Twenty-one patients were enrolled into this prospective, non-randomized, open-label, controlled study (first 14, treatment; last seven, control). Baseline evaluations included complete clinical and laboratory evaluations, exercise stress (ramp treadmill), two-dimensional Doppler echocardiogram, SPECT perfusion scan, and 24-hour Holter monitoring. Bone-marrow mononuclear cells were harvested, isolated, washed, and resuspended in saline for injection by NOGA catheter (15 injections of 0.2 cm 3 ). Electromechanical mapping (EMM) was used to identify viable myocardium (unipolar voltage ≥ 6.9 mV) for treatment. Patients underwent 2-month noninvasive and 4-month invasive (treatment group only) followup using standard protocols and the same procedures as baseline. Patient population demographics and exercise test variables did not differ significantly between the treatment and control groups; only creatinine and BNP levels varied in laboratory evaluations. At 2 months, there was a significant reduction in total reversible defect within the treatment group and between the treatment and control groups (P = 0.02) on quantitative SPECT analysis. At 4 months, there was improvement in ejection fraction from a baseline of 20% to 29% (P = 0.003) and a reduction in ESV (P = 0.03) in the treated Introduction Atrial fibrillation (AF) has a high prevalence in the elderly population. Nevertheless, it has been found in young patients.
Objectives To show the clinical and epidemiological aspects of a population of patients with AF and age < 60 years old in the emergency room (ER), evaluating symptoms, triggering factors, related diseases and recurrence of AF.
Methods From March 2000 to October 2002, 236 patients with AF were seen in the ER. Fifty-seven patients (24.1%) were aged < 60 years old. Forty-six patients (80%) were male, mean age 49.4 ± 8.3 years old. The patients were set on an algorithm for AF.

Results
All the patients were hemodynamically stable. Forty-five patients (78.9%) presented palpitation and 10 patients (17.5%) precordial pain to admission. Twelve patients (21%) had the first reported incident of AF; 39 patients (68.3%) had recurrent AF, six patients (10.5%) had > 10 admissions per AF in the past year. Twenty-five patients (43.8%) indicated stress as the main triggering factor of the event and 23 patients (40.3%) indicated alcohol intake. Thirty-nine patients (68.4%) started AF at a rest period, 13 (22.8%) at activity and five patients (8.7%) after food intake. Among the risk factors for embolic events, 20 patients (35.1%) were hypertensive; two patients (3.5%) had previous stroke; three patients (5.2%) had mitral disease; four patients (7%) had hypertrophic cardiomyopathy; four patients (7%) had coronary artery disease; one patient (1.7%) had diabetes mellitus; and seven patients (12.3%) had thyroidal disease. Twenty-two patients (38.5%) had been using anti-arrhythmic medications regularly. Forty-one patients (71.9%) showed < 48 hours of symptoms, and the others an unknown time or > 48 hours. Thirty patients (52.6%) had arrhythmia reversed with oral medication, with mean reversion ∆t of 5.7 hours. Thirteen patients (22.8%) had successful ECV with an average charge of 200 J. Ten patients (17.5%) had spontaneous reversion; three (5.26%) had unsuccessful. In a follow-up of 5 months to 2 years, 32 patients were observed. Fifteen patients (46.8%) had recurrence of AF despite use of anti-arrhythmic medication. Eighteen patients (31.5%) did not use anticoagulant or anti-agglutinant. There was an embolic event in one patient (3.1%).
Conclusions Our patients develop with hemodynamic stability to admission and present an elevated reversion rate in the ER (75.4%) with mean ∆t < 6 hours. Hypertension was the main risk factor without correlation to recurrence (P = not significant). Stress was the factor correlated to recurrence (P = 0.038). Patients with ∆t < 48 hours showed a higher reversion rate of AF in the ER (P = 0.009). The recurrent rate of AF in this population was high even with anti-arrhythmic medication, but the number of thromboembolic events was low. Background Although data have shown that B-type natriuretic peptide (BNP) levels correlate with the severity and prognosis of heart failure, there are few studies regarding its levels in cardiac surgery patients.
Objectives We sought to correlate the clinical and hemodynamic features in postoperative (PO) stay and the levels of BNP.
Methods A prospective and observational study. We assessed the level of BNP (imunofluorescence -Triage ® ) at 1 and 24 hours PO. A BNP level above a cutoff point of 100 pg/ml was found to be highly sensitive and specific for the diagnosis of cardiac heart failure. The population consisted of two groups: group A had levels below 100 pg/ml and group B was above this level. Evaluated at 1 hour of PO stay were: time of cardiopulmonary bypass (CBP), hidric balance (HB), mean arterial pressure (MAP), heart rate (HR), central venous pressure (CVP), pO 2 and FiO 2 ratio (P/F), mechanical ventilation time (MV) and O 2 central venous saturation (VO 2 SAT). The left ventricular function was assessed by two-dimensional echocardiography in the preoperative period (Simpson method) and values under 40% were considered ventricular dysfunction. The Student t-test was used for comparison between the found means.

Results
We investigated 17 patients (three women, median age 58.4 years old; standard deviation = 9.7). Group A was composed of 11 patients and group B of six patients. No statistical difference was found regarding CPB, HB, MAP, HR, CVP, P/F and VO 2 SAT, whereas the MV time in group A was 211.3 ± 229 min, with regard to group B being 520.8 ± 332.9 min (P = 0.038). At 24 hours PO, the BNP mean level (327.8 ± 206.9 pg/ml) was found in 13 patients (76.4%).
Left ventricular dysfunction was observed in two patients of group B.
Conclusions Although there was a reduced number of patients, these findings suggest that the BNP levels were related to the mechanical ventilatory time. Methods From June 2001 to December 2002, we selected 273 patients who had undergone coronary angioplasty, and divided them into two groups. Group I consisted of 173 patients who used subcutaneous or intravenous enoxaparin, following a protocol. Group II consisted of 130 patients who used intravenous heparin during angioplasty. The protocol of enoxaparin consisted of administering intravenous enoxiparin (after insertion of a catheter) to patients who were not using subcutaneous enoxaparin and those patients who used enoxaparin more than 6 hours before the angioplasty. To the patients whose last dose of subcutaneous enoxaparin had been administered in the 6 hour interval, no anticoagulant needed to be added.

Results
There was no significant difference between the groups in relation to age, sex, risk factors, drugs in use and obstructive coronary arteries.
In 30 days the great bleeding was less in group I (1.7%) compared with group II (3.1%). The incidence of death and myocardial infarction was not different between the two groups within 30 days.
Conclusion The use of enoxaparin as an anticoagulant during PTCA did not increase ischemic or hemorrhagic complications after coronary angioplasty. Background The chronobiological rhythms have been shown to cause an impact in the occurrence of a variety of cardiovascular disorders like acute myocardial infarction, sudden death and stroke. However, the effects of the chronobiological rhythms in patients with acute aortic dissection (AAD) have not been well studied. The International Registry of Acute Aortic Dissection (IRAD) observed that the frequency of AAD was significantly higher between 6:00 am and 12:00 noon, during the winter with a peak in January, and no variation was found for the day of the week. The objective was to know the chronobiological rhythms of our population with AAD.

Patients and methods
We evaluated in a cross-sectional and prospective study patients admitted to a chest pain unit, between March 1997 and May 2001, with a diagnosis of AAD. The authors carried out a descriptive analysis in the sample and they compared the proportions of the categorical variables between types A and B (Fisher test). Values of P < 0.05 were considered significant. Background Differentiating congestive heart failure (CHF) from lung disease is extremely important in patients evaluated in the emergency department (ED). Therefore we sought to assess the utility of B-type natriuretic peptide (BNP), which is secreted by the left ventricle in response to volume or pressure overload, in differentiating CHF from lung diseases in elderly patients presenting to the ED with acute dyspnea.

Results
Methods From April to July 2001, 70 patients presenting to the ED of a tertiary cardiology hospital with acute dyspnea were included. Mean age was 72 ± 16 years and 33 (47%) were male. BNP was measured in all patients at the moment of admission in the ED using a rapid bedside test. Emergency-care physicians were required to assign a probable diagnosis, blinded to BNP values. A cardiologist retrospectively reviewed patients' data (blinded to BNP measurements) and assigned a diagnosis that was considered the gold standard to assess the diagnostic performance of the BNP test.

Results
The mean BNP concentration was higher in patients with CHF (n = 36) than it was in patients with lung diseases (n = 29). Such values were 990 ± 550 vs 112 ± 59 pg/ml, respectively (P < 0.001). The pulmonary diseases and their respective BNP levels were: chronic obstructive pulmonary disease, 98 ± 69 pg/ml (n = 5); asthma, 38 ± 30 pg/ml (n = 3); acute pulmonary embolism, 158 ± 35 pg/ml (n = 2); and pneumonia, 80 ± 52 pg/ml (n = 19). In patients with a history of lung disease but whose current complaint of dyspnea was found to be CHF, BNP levels were 898 ± 456 pg/ml. Those patients with a history of CHF but a current diagnosis of pulmonary disease had a BNP of 98 ± 47 pg/ml. The area under the receiver operating curve for BNP levels in differentiating CHF from lung diseases was 0.98.
Conclusion A rapid bedside test for BNP is useful in differentiating lung diseases from decompensated CHF in elderly patients presenting to the ED with dyspnea. Background Levosimendan (Ls) is a novel inotropic agent, calcium sensitizer and vasodilator indicated for the treatment of patients (patients) with acutely decompensated heart failure (ADHF). Randomized trials show Ls to be an effective and safe option for the management of ADHF.
Objective To analyze the hemodynamic effects of intravenous Ls in patients with ADHF.
Method Data from 10 Argentinean hospitals in a multicenter registry were collected. Eligibility criteria were clinical ADHF, ejection  ) were blacks, presenting an average age of 63.5 ± 13.5 years. It was observed that the schedule of the day for an incidence of AAD was between 6:00 pm and 12:00 midnight (41.2%), and only 11.7% occurred in the period between 6:00 am and 12:00 noon. The day of biggest occurrence was Monday, with 26.4% of the cases. The months of May and July (14.7% each) were the most frequent, and the season of the year was winter (32.3%).
Conclusions Like other cardiovascular conditions, AAD could exhibit chronobiological rhythms. In our population we observed the incidence of AAD in the nocturnal period of 6:00 pm and 12:00 midnight, on Mondays, and in the period of winter.
fraction ≤ 40%, cardiac index ≤ 2.5 l/min/m 2 , and pulmonary capillary pressure ≥ 15 mmHg if a Swan-Ganz (SG) catheter was used. We analyzed the data of the 41 patients monitored with a SG catheter. Complete clinical, radiographic, EKG, and laboratory examinations were performed before and after Ls. Ls was administered as a loading dose of 6-24 µg/kg over 10 min, followed by a continuous infusion of 0.1-0.2 µg/kg/min for 24 hours. Hemodynamic measures were recorded at baseline, 30 min, 2, 6, 24, and 48 hours. Data were compared using the t test or Wilcoxon rank-sum test.  Table 1; similar results were obtained at 24 hours.

Results
Conclusion Ls significantly improved pulmonary pressures, cardiac index and output, with no significant effects on heart rate in patients with ADHF. Ls is an effective and safe option that should be considered for the management of ADHF. The sample was composed of seven patients whose age varied from 32 to 78 years old (mean 51.85 years), and there were five male and two female patients. Coronary artery bypass grafting was performed in six patients and mitral valve replacement in one. Four patients presented failure of one ventricle and three had failure of both ventricles in the immediate postoperative period.
The system most commonly used was the Bio-pump ® centrifugal pump, used in all cases; in one case biventricular support with the Bio-pump was changed after 48 hours to a biventricular DAV-InCor ® system (temporary pulsatile ventricular assist device). Intraaortic balloon pumping was used as secondary support in four cases with the aim of delivering a pulsatile flow.
The mean support time was 51 hours and 38 min/patient, the shortest time was 4 hours and 6 min, and the longest was 151 hours and 20 min.
There were five deaths during CS and the cause of them all was multiple organ failure. Two patients were discharged from CS (28% removal), one was bridged to emergency heart transplant and the other recovered ventricle function (bridge to recovery). The patient bridged to recovery is now in 12 months of followup, and in New York Heart Association class II.
Although the mortality index is still high (86%), we were able to bridge 28% of the patients to a more definitive treatment or status (transplant or recovery). The overall survival was 14%, but longterm survival of bridged patients is 50% and we expect to increase it in the near future.

Table 1
Type of circulatory support Number of patients Left atrium-aorta 3 Right atrium-pulmonary artery 2

Biventricular 2
Introduction Although the use of vasoactive drugs is widespread in the critical care setting, its use is associated sometimes with an undesirable hemodynamic outcome. The dynamic subaortic stenosis is a phenomenon described in the echocardiogram stress testing in which patients are submitted to the use of dobutamine.
Similarly it could happen in the critical setting where high doses of vasoactives drugs are frequently prescribed, but to our knowledge this has never been described previously.
Case report A male patient, 66 years old, with no mentioned cardiovascular disease was submitted to an elective surgical correction of an infrarenal abdominal aortic aneurysm. The surgery was complicated with hemorrhagic shock, with the necessity for large amounts of volume (crystalloids, colloids and blood products) and high doses of vasoactive drugs. He was admitted to the intensive care unit (ICU) where a pulmonary artery catheter was placed. On the fourth day he was using 16.67 µg/kg/min dobutamine and was submitted to the first cardiac echocardiogram, which revealed a normal aortic valve/ventricular gradient (Ao/LV) (< 25 mmHg) and an ejection fraction (EF) of 0.61. Despite the aggressive treatment, he had progressive hemodynamic worsening that prompted progressive elevation of the vasoactive drugs dosage. On the seventh day, with a dobutamine dose of 20.80 µg/kg/min, a second echocardiogram was performed that revealed an Ao/LV of 100 mmHg, an EF of 0.59 and an image suggestive of subaortic stenosis. During the recovery period of his clinical status, on the eighth day, a new echocardiogram was performed and showed an Ao/LV lower than 25 mmHg and an EF of 0.61. The patient was discharged from the ICU 64 days later. Results A total of 30 × 10 6 BM-MNC were injected at 15 sites in each patient. Left ventricular total reversibility (TR), percentual of myocardial rest defect at 50% and ejection fraction at baseline and 8 weeks. The results of TT are shown in Figure 1.

Conclusion
In this pilot study of BM-MNC endocardial injections, our data suggest that this procedure seems to be safe to be performed in this very sick population, and suggest an improvement in objective assays of myocardial perfusion and exercise capacity in some patients listed for HTx. Future studies must be conducted to evaluate the role of BM-MNC transplantation as an alternative therapy for these patients. Severe sepsis is a complex process that involves a number of host immune responses with an orchestration of various specific and nonspecific soluble factors and cellular elements that may result in a completely different outcome. Among organ dysfunction induced by sepsis, heart failure can occur in up to 40% of cases [1]. Myocardial depression in shock probably was first described in 1947 [2]. Its onset may be extremely early, but is most evident in the first 3 days of the disease. Normalization usually happens over the following 7-10 days in the survivor patients [3]. We have described an atypical sort time reversible myocardial dysfunction, in a patient with rapid evolution to a multiple organ dysfunction syndrome (heart, renal, pulmonary, and hematological). An 87-yearold man with hypertension and diabetes mellitus type 2 was admitted to our intensive care unit (ICU) with severe sepsis caused by community-acquired pneumonia. On the following day of his ICU entrance he developed septic shock which was associated with an increase in the cardiac enzymes, particularly troponin I and CK mass (fluorogen immunoassay). We started the infusion of Xigris ® , a recombinant version of human activated protein C, according to the PROWESS [4] protocol. His baseline examination characteristics before and 3 days after Xigris ® infusion are summarized in Table 1.

P23
The patient was discharged 10 days after the Xigris ® end-of-infusion to a step-down unit. The physiology of the myocardial dysfunction that occurs in systemic inflammatory response syndrome is not well understood, although there are several theories to explain it [5][6][7][8].
We are reporting an unusual behavior of reversible nonischemic myocardial dysfunction possibly related to Xigris ® treatment.
Perhaps this communication could be tested in a well-designed study to address a new hypothesis for new applications of activated protein C outside the setting of severe sepsis.

Results
One hundred and twenty-five patients (78.6%) have achieved the primary endpoint. The same proportion was seen in the CD patient subset, when compared with the non-CD group (75% vs 79.3%, P = 0.813). CD patients had the same mean age, sex distribution and clinical features as the non-CD patients. The CD group had more frequent hypotension when compared with the non-CD group (28.5% vs 15.4%, P = 0.0195). The dilated myocardiopathy group also had the same incidence of hypotension as the CD group (28.5% vs 34.2%, P = 0.826).
Conclusion CD patients with acutely decompensated HF had the same hemodynamic benefits as HF patients with other etiologies treated with Levo. Three-quarters of CD patients have been discharged from the hospital without the need for inotropes after Levo treatment. Although hypotension was slightly more frequent in the CD group, it did not reduce the Levo response in this group of patients.
Introduction Bearing in mind the importance of the early detection and correction of tissue hypoxia to avoid the progressive organic dysfunction and death, SVO 2 has been used as a prognosis index and as a therapeutic answer. Therefore, this study aims at the assessment of the correlation between the SVO 2 drawn in the superior cava vein, which shows itself as an easier access route and also of lower cost, and the SVO 2 drawn in the pulmonary artery, aiming to facilitate and optimize the cost of one more index of assessment of the use of oxygen.

Materials and methods
We are dealing with a transverse study with the record of pieces of information on the mixed venous blood samples drawn in the superior cava vein and pulmonary artery for comparative analysis of the mixed SVO 2 in the period of January through May scheduled until November 2003. The samples were chosen randomly and through simple analysis according to a collecting protocol which consists of using 5 ml mixed venous blood, collected with a previously heparinized (0.1 ml heparin) syringe and with a total suction time of 20 s. The analysis was made in a gasmeter brand Radiometer (model ABL5) immediately after collecting. The data are evaluated through the hypothesis test in which the void hypothesis equals the average and the alternative is the difference between them. The statistic is confirmed through the Student t method and the correlation through the Pearson index. The correlation between the SVO 2 value collected in the superior cava vein and the pulmonary artery as a comparative method is evaluated.

Conclusion
The preliminary results indicate a Pearson coefficient of 0.6. The Student t test shows that the probability of results coming from a different distribution is 0.4, indicating that the void hypothesis is true.
Introduction Esophageal Doppler is a noninvasive method used to guide fluid loading, resulting in clinical outcome benefits, especially during anesthesia. Its role in critically ill patients is still controversial.
Objective To compare cardiac output (CO) obtained from esophageal Doppler with thermodilution, using a Swan-Ganz catheter.
Methods Data was obtained from two medical intensive care units between February and March 2003. An esophageal probe for cardiac output monitoring was introduced in severe sepsis and septic shock patients when a pulmonary artery catheter was indicated by the attendant physician; four CO measurements were done for each patient with 6-hour intervals. CO determination by esophageal Doppler was performed simultaneously with thermodilution.

Results
Eight consecutive patients with a mean age of 61.1 years were included (five females and three males). Twenty-five measurements were done; as shown in Figure 1 there was no correlation (R 2 = 0.02) between thermodilution and esophageal Doppler CO measurements. Figure 2 shows that no correlation was seen when we analysed only the variations of CO measurements (R 2 = 0.4; P = 0.11).
Conclusion In this study, in concordance with published data, there is no evidence to support esophageal Doppler as a technique for CO continuous monitoring in the intensive care unit.

Figure 1
Thermodilution (TH)    Live E. coli injection in dogs promotes low cardiac output, systemic and regional lactic acidosis and severe splanchnic hypoperfusion. The RL solution promoted only modest and transient improvement in blood flows but not in systemic and regional acidosis. There were no differences between the resuscitated groups.
P27 Systemic and regional hemodynamic effects of fluid resuscitation in experimental septic shock Methods BT groups: midline laparotomy was performed on Wistar rats under ketamine + hydroxychloral anesthesia (4:1). Rats were inoculated with 10 ml 10 7 /10 10 cfu/ml Escherichia coli R-6 (n = 20/group) by oroduodenal catheterization, which was confined to the small intestine by ligation of both the duodenum and ileum. S groups: inoculation of 10 7 /10 9 /10 10 cfu/ml/100 g body weight Enterobacter cloacae 89 into the portal vein (n = 20/group). BT + S groups: 10 10 BT + 10 7 /10 9 S (n = 20/group). From 10 animals of each group, samples of mesenteric lymph node, liver, spleen and blood were collected 2 hours post inoculation and cultured in MacConkey agar medium. The remaining animals in each group were observed for mortality for 30 days (n = 10/group).
Results BT-10 7 /10 10 did not cause death and only 10 10 inoculum promoted BT (mean = 1.8 × 10 5 cfu/g tissue). S-10 7 was not lethal, but promoted a transient bacteremia state, S-10 9 was LD 85 within 25 hours, and S-10 10 showed LD 100 within 5 hours. Bacterial recovery from these groups/g tissue were, at the most, 10 4 cfu at S-10 7 , 10 7 cfu at S-10 9 and 10 8 cfu at S-10 10 . BT-10 10 in combination with S-10 7 showed significantly increased mortality (LD 50 within 32 hours) as compared with BT-10 10 (LD 0 ) and S-10 7 (LD 0 ) alone (P < 0.05), and the mortality rate was statistically similar to the severe sepsis group (S-10 9 ). In addition, the association of BT-10 10 + S-10 9 also provoked a significant increase in mortality (LD 100 within 13 hours) as compared with BT-10 10 (LD 0 ) and S-10 9 (LD 85 within 25 hours) in terms of length of time to cause mortality (P < 0.05). Besides, sepsis in combination with BT showed a decreased rate of translocation in all groups as compared with the BT group alone. Overall data demonstrated significant deleterious synergistic effects of BT in combination with all states of sepsis, suggesting that translocation of bacteria through the gut-associated lymphoid system (GALT) favors the activation of the host systemic inflammatory response, even though the total quantification of internalized bacteria in the host compartments did not change at all by the addition of the BT process. Therefore, BT appears to provoke an exacerbated inflammatory state due to the bacterial challenge to the GALT during their traffic through mesenteric lymphatic tissue rather than the quantitative physical presence of the bacteria in the systemic compartment, therefore suggesting a distinctive GALTrelated host inflammatory response associated with the BT phenomena. Background Increasing evidence has implicated bacterial translocation (BT) as the main source of the so-called gut hypothesis of the pathogenesis of sepsis progressing to multiple organ failure. Others have shown that mesenteric lymph content in the course of BT promotes increased pulmonary permeability. In previous work we have shown significantly increased tumor necrosis factor alpha and lymphocytes in the mesenteric lymph during BT. In this study we examined the correlation between microcirculation injury and mesenteric lymph exclusion during the BT process.
Methods Female Wistar rats were distributed in the following groups: BT, inoculation of 10 ml of 10 10 cfu/ml Escherichia coli R-6 confined to the small intestine; BT-E, submitted to BT without the influence of the mesenteric efferent lymph (diverted away from systemic circulation by catheterization of the mesenteric lymph duct); BT-R, submitted to the same procedure as BT-E group followed by re-inoculation of the collected lymph into the systemic circulation; BT-N, inoculation of lymph collected during BT into naïve animals. All animal mesenteric microcirculations were examined for 2 hours (n = 6/group) using an intravital microscope. The same number of BT-R and BT-N animals was injected with either lymph cells or lymph supernatants into the systemic circulation (immediately after light centrifugation of lymph collected during BT). A midline laparotomy was performed on Wistar rats under ketamine + hydroxychloral anesthesia (4:1). Rats were inoculated with 10 ml E. coli R-6 10 10 cfu/ml by oroduodenal catheterization, which was confined to the small intestine by ligation of both the duodenum and ileum. All lymph samples were submitted to culture in MacConkey agar medium.

Results
All lymph cultures were negative. During BT, the onset of injuries in the mesenteric microcirculation was mainly focal hemorrhages in capillaries and small venules beginning around 30 min, which progressed quantitatively up to 2 hours. After 1 hour of translocation, focal thrombosis of capillaries and small and medium venules were observed. In contrast, in animals where the lymph was diverted away from systemic circulation by catheterization of the mesenteric lymph duct, no microcirculation injury occurred (BT-E group). The re-inoculation of the collected whole lymph (BT-R) promoted similar injuries to the microcirculation as seen in the BT group in the same time period. Interestingly, the injection of whole lymph collected (negative culture for bacteria) from animals submitted to BT in the naïve animals (BT-N group) provoked similar mesenteric microcirculation damage within the same period. In addition, only lymph supernatant was able to promote microcirculation injuries in both the BT-R and BT-N groups. These findings allow us to speculate that BT-induced alterations in the mesenteric microcirculation are possibly due to the gut-associated lymphoid system activation by the BT process with the release of proinflammatory factor(s) and not due to the existence of bacteria in the lymph. Therefore, this might be the BT mechanism for the aggravation of a pre-existing state of sepsis or for the installment of infectious disease. Ongoing experiments are in progress to better elucidate this hypothesis.

Background
To improve resectability of several hepatobiliopancreatic tumors, the vascular structures with cancer invasion could be resected and reconstructed. The liver is submitted to a global hypoxia during the hepatic artery reconstruction, since almost 50% of oxygen delivery to this organ is maintained through this vessel. This study addresses the initial impact of prolonged hepatic artery occlusion on liver hemodynamics and oxygen metabolism.
Methods Seven pentobarbital anesthetized mongrel dogs (19.7 ± 1.2 kg) underwent laparotomy. The gastroduodenal artery was ligated and the common hepatic artery was occluded during 60 min, followed by 120 min of reperfusion. Systemic hemodynamics were evaluated through a Swan-Ganz catheter and arterial catheters. Splanchnic perfusion was assessed by portal vein blood flow (ultrasonic flowprobe), hepatic artery blood flow and liver enzymes (ALT, AST, DHL). Systemic and hepatic oxygen delivery (DO 2 s and DO 2 h, respectively) were calculated using standard formulae.

Results
The results are presented in Table 1.

Conclusion
We conclude that temporary hepatic artery occlusion induces a progressive decrease in portal vein blood flow during ischemia, which is maintained during reperfusion. The hepatic artery blood flow was promptly restored after arterial unclamping. This effect was associated with a significant and progressive reduction in hepatic oxygen delivery that could contribute to the development of postoperative hepatic failure in critically ill patients with a borderline of established preoperative hepatic dysfunction. The impact of acute normovolemic hemodilution (HD) on splanchnic perfusion was evaluated in 21 anesthetized (fentanyl and vancuronium) mongrel dogs (16 ± 1 kg). They were randomized to controls (n = 7, no HD), moderate HD (hematocrit 25 ± 3%, n = 7) or severe HD (hematocrit 15 ± 3% ml/kg), through an isovolemic exchange of whole blood and 6% hydroxyethylstarch at a 20 ml/min rate, to the target hematocrit. The animals were followed 120 min after HD. Cardiac output (ml/min), portal vein blood flow (ml/min), portal vein-arterial CO 2 gradient (mmHg) and PCO 2 gap (gas tonometry, mmHg), and splanchnic perfusion were evaluated through portal vein blood flow and gas tonometry.
Results Exchange blood volumes were 33.9 ± 3.3 and 61.5 ± 5.8 ml/kg for moderate HD and severe HD, respectively. Controls maintained a hematocrit of around 41% throughout the study. Arterial pressure remained stable for all animals.
Conclusion Global and regional hemodynamic stability were maintained after moderate and severe HD. However, a moderate gastric mucosal acidosis was induced with a hematocrit of 15%, which may become relevant after major surgery or trauma.

Introduction
Homocysteine is a sulfur-containing amino acid formed during methionine metabolism that has been appointed as a marker of cardiovascular disease. The mechanisms involved are unclear, but include an increase in oxidative stress, excessive thrombogenesis, mitotic alterations in smooth muscular cells and endothelial dysfunction. Some of these mechanisms are present in septic patients, suggesting that total homocysteine (tHcy) levels might be implicated in the pathogenesis of organ dysfunction. The objective of this study is to correlate tHcy levels and the severity of septic process, evaluated by SOFA score.

Methods
In this prospective clinical trial, patients admitted in a tertiary universitary intensive care unit with severe sepsis were included, before 48 hours of organ dysfunction diagnosis. Patients with acute renal failure were excluded. Blood samples were collected after 8 hours of starvation, and SOFA parameters, on days 1, 3, 7 and 14 after inclusion. Statistical analysis was done using the Kappa test.

Conclusion
Our results suggest that homocysteine levels could not be related to organ dysfunction in this septic patient population. These results must be confirmed in a larger population of septic patients.
Introduction Multiple organ failure is the main cause of mortality in intensive care units (ICUs). Initial treatment is probably the most important step for critically ill patient outcome. Many scores have been proposed to assess organ dysfunction evolution and outcome. However, the ideal time to apply organ dysfunction scores has not been clearly established. We hypothesized that the SOFA score, when applied after initial treatment (24 hours later on), is more valuable than at the ICU admission to predict outcome in critically ill patients.
Objective To evaluate whether SOFA score calculation after treatment is more adequate to predict outcome than SOFA score calculation at ICU admission in critically ill patients.
Method This is a prospective analysis from the BASES study, which is an epidemiological, observational study performed in 1379 patients from private and public Brazilian ICUs. From this databank, we selected only patients with a length of stay longer than 24 hours (n = 884). From those patients, we calculated the daily SOFA score at admission and 24 hours later, and we also collected patientrelated outcome. The Youden test was calculated to choose the best cut-off value. Receiver-operating characteristic (ROC) curves were built for SOFA scores in those days. Finally, we compared the areas under ROC curves throughout a Hanley and McNeil test to estimate the most appropriate day to apply this organ dysfunction descriptor. P = 0.0002 was considered significant.

Results
The mean age was 62 ± 19 years, 59% were male, and the overall mortality rate was 22%. The best cut-off value for SOFA score at day of admission was 8 and for the next day was 5. Areas under the ROC curves were 0.716 and 0.775 for day of admission and the next day, respectively (P < 0.05).

Conclusion
The SOFA score applied after initial resuscitation is more accurate to predict outcome than the SOFA score applied at ICU admission. Introduction Hypocalcemia has a prevalence of 88% in general intensive care units (ICUs). Calcium (Ca) administration may be associated with hemodynamic improvement, but with increased mortality in animal studies.
Objective To describe the incidence of hypocalcemia in septic patients, and its associated morbidity and mortality in the first 14 days of ICU stay.
Methods A prospectively collected database was retrieved for Ca and SOFA score at days 1-14, APACHE II, lactate, creatinine, albumin, and mortality. All severe sepsis and septic shock patients were included from June 2000 to June 2001. Patients were classified as presenting severe hypocalcemia (SH) (< 1.0 mmol/l) or moderate hypocalcemia/normocalcemia (> 1.0 mmol/l).
Results Fifty-five patients had a mean (± standard deviation) age of 50.5 (± 18.0) years, and an APACHE II score of 21.4 (± 9.1). At entrance, 27.3% of patients were in septic shock, and subsequently 70.9% developed criteria for septic shock. The incidence of hypocalcemia was 80%, and that of SH was 41.9% (23 patients). APACHE II scores, lactate and creatinine at D1 where higher, and albumin was lower in SH (P < 0.05). Vasoactive drug use was higher in SH (87% vs 59.4%; P < 0.05). Mortality and morbidity (SOFA maximum) and are shown in Figures 1 and 2.
Conclusions Hypocalcemia is common in septic patients. SH is associated with increased organ dysfunction, and a trend towards increased mortality. It probably represents a sign of more severe disease. More studies are needed to establish the role of calcium supplementation in septic patients.

Figure 1
Mortality in severe hypocalcemic patients.

P35 Evaluation of blood transfusion effects on mixed venous oxygen saturation and lactate levels in patients with systemic inflammatory response syndrome (SIRS)/sepsis
Introduction Blood transfusions continue to be a controversial therapy in intensive care units, mostly in patients with SIRS/sepsis, with conflicting thresholds for transfusion and different results in the literature. The present study is aimed at evaluating the effects of blood transfusion in two parameters of organ perfusion, mixed venous oxygen saturation (SvO 2 ) and serum lactate levels in patients with SIRS/sepsis who presented with hemoglobin levels < 9.0 g/dl.
Methods All patients admitted to the intensive care unit with SIRS/sepsis, as defined by Consensus Conference 1992, and hemoglobin levels < 9.0 g/dl were included. Hemoglobin levels, mixed venous oxygen saturation and lactate levels were collected before (BT) and up to 1 hour after blood transfusion (AT). These variables were analyzed through the paired Student t test and results were considered significant if P ≤ 0.005.

Results
Twenty-nine patients (17 male, 12 female) with mean age of 61.9 ± 15.1 years (21-85 years) and a mean APACHE II score of 12.5 ± 3.75 (7-21) were transfused with a mean of 1.41 packed red cell units. Although a significant increase in hemoglobin levels was achieved by blood transfusion (BT, 8.14 ± 0.64 and AT, 9.4 ± 0.33; P > 0.0001), this was not accompanied by a significant change in lactate levels (BT, 1.87 ± 1.22 and AT, 1.56 ± 0.28; P = 0.28) or in SvO 2 (BT, 64.3 ± 8.52 and AT, 67.4 ± 6.74; P = 0.13). The results were similar when the analysis was performed only with those patients with hemoglobin levels < 8.0 g/dl (n = 9).
Conclusions These results suggests that blood transfusions, despite a significant increase in hemoglobin levels, are not associated with an improvement in tissue oxygenation in patients with SIRS/sepsis with hemoglobin levels < 9 g/dl.  Introduction Cardiovascular dysfunction (CD) occurs frequently among patients with septic shock and it is more severe in the first 3 days, relapsing at the end of the first week among those who survive. Fifteen percent of the patients with septic shock, who evolve to death, die as a result of CD, whose physiopathologic mechanisms are not fully understood. Levosimedan (LEVO), a new inotropic drug which acts as calcium sensitizing agent and also has effects on potassium channels in vascular smooth muscle cells leading to vasodilatation action, has shown benefits in patients with cardiac failure. An experimental study on the endotoxic shock model has shown improvement on cardiac output, and systemic and regional oxygenation. The aim of this study is to prove the acute hemodynamic effects of LEVO administration on septic patients with severe CD.

P36 Severe hemodynamic compromise, respiratory failure and disseminated skin lesions due to Strongyloides stercoralis
Materials and methods SLS, 70 years old, male, with a history of ischemic cardiomyopathy was admitted with pulmonary infection sepsis and multiple organ dysfunction. The patient received LEVO without bolus dose at an initial rate of 0.1 µg/kg/min in the first hour, followed by 0.2 µg/kg/min during the next 24 hours, monitored with a pulmonary arterial catheter during the observation.

Results
The main hemodynamic and metabolic data are shown in Table 1.

Conclusion
The administration of LEVO in patients with sepsis and septic shock and severe CD has proved to be safe, with immediate improvement of hemodynamic and metabolic parameters that was maintained after discontinuation of the drug. It is necessary that a randomized, controlled trial be done to compare the use of LEVO and other inotropic drugs in the treatment of CD associated with sepsis in order to validate this new indication for the drug. The usual diagnostic approach to patients with sepsis in an intensive care unit (ICU) or coronary care unit is based on clinical, biochemical, microbiologic and pathologic data and on radiologic imaging (radiograph, ultrasound and computed tomography), which are used to determine the most common sites of infection. In many cases, however, the site of infection is difficult to determine. Nuclear medicine provides various noninvasive scintigraphic methods for the imaging of focal sepsis, based on the intravenous administration of a radiotracer that accumulates at the site of infection or inflammation. The combined use of total body scintigraphy with ultrasound and computed tomography is considered a useful tool for the diagnosis of occult sepsis in ICU patients, and allows the targeting of aggressive measures against infections. The aim of this study is to evaluate the diagnostic value of technetium 99m-white blood cell ( 99m Tc-WBC)-labeled scintigraphy and gallium-67 citrate ( 67 Ga) scintigraphy in the detection of focal sepsis in the ICU. We reviewed seven patients affected by sepsis of unknown origin. After the usual diagnostic approach, five patients were submitted to a total body scan using the 99m Tc-WBC and two patients using 67 Ga. The patients had complete clinical and instrumental data, but none of the radiological image detected the site of infection. The 99m Tc-WBC scan showed typical patterns of increased tracer accumulation in six different sites. Four of these sites were studied histopathologically, confirming the infection: one case of left kidney abscess was associated with concomitant infection of the psoas muscle in the same patient, one frontoparietal osteomyelitis, and one acute cholecystitis. The other two sites corresponded to pulmonary accumulation of the tracer, which was interpreted as pneumonia. Of the two patients who underwent 67 Ga scintigraphy, one had decubitus ulcerated infection associated with sacrum and left femoral osteomyelitis, and the other had clinical suspect of pulmonary vasculitis and diffuse pulmonary 67 Ga accumulation.
Conclusions Analysis of our results suggests that scintigraphy with 99m Tc-WBC and 67 Ga can be considered a powerful tool in the detection of the source of infection in patients with sepsis in the ICU and the coronary care unit. Introduction Hypoperfusion in sepsis may be identified by lactate levels, but there are many other unmeasured acids that may be better represented by negative base excess (BE). Successful resuscitation should be followed by increased BE.

P39 Acidosis and mortality in severe sepsis and septic shock evaluated by base excess variation
Objective To evaluate the utility of BE variation in mortality in severe sepsis (S) and septic shock (SS).
Methods A prospectively collected database was retrieved for BE at days 1 and 3 (D1 and D3), APACHE II, lactate, creatinine, albumin, and mortality at 28 days. Patients with S or SS were included, except if renal failure was diagnosed at D1 (creatinine > 3.5 mg/dl; diuresis < 500 ml). Patients were classified as increased (less acidosis) BE vs decreased BE, based on the difference between D1 and D3.
Results Forty patients had a mean (± standard deviation) age of 48.4 (± 19.8) years, and an APACHE II score of 19.6 (± 9.1). At D1 and day 14, 20% and 65% of patients were in SS, respectively. Table 1 summarizes the main findings. Binary logistic regression analysis showed that only the APACHE II score (odds All values are shown as mean (standard deviation) unless indicated otherwise.

Figure 1
Twenty-eight day survival. Conclusions In patients with S and SS, increased BE from D1 to D3 seems to be a good predictor of morbidity and mortality, and may be considered a possible goal. Objective To evaluate the correlation between CVP and Pw in hepatic transplantation.

Materials and methods
Twelve postoperative hepatic transplantations were studied from January to November 2002. All analyses were performed 24 hours after surgery.

Results
We analyzed 69 simultaneous measurements and found a good correlation between CVP and Pw ( Fig. 1).

Conclusion
The CVP can be used to estimate Pw in the postoperative period of hepatic transplantation. Probably, CVP can be used as a preload variable to guide volume infusion in this group of patients.

Figure 1
Correlation between central venous pressure (CVP) and wedge pressure (Pw) in hepatic transplantation. R, Pearson correlation coefficient; R 2 , Square Pearson correlation coefficient. Introduction Sepsis is a very frequent, severe multisystemic disease with a myriad of signs and symptoms that can mimic several critical illnesses. Only the optimal treatment of sepsis-associated conditions is able to reduce the high morbidity and mortality associated with such a complex disease. Evidence-based medicine concepts have been developed to manage those conditions, improving patient care. Particularly regarding sepsis, there is a gap between evidence-based literature and clinical application. Moreover, different intensive care units (ICUs) may have heterogeneous facilities with unequal health care providers. Therefore, we have been developing a tool able to diagnose current sepsis ICU management and to translate evidencebased concepts into bedside practice.
Objectives To use sepsis as a model to highlight the optimal clinical practice that may influence outcome based on known pathophysiological mechanisms and life-saving interventions in critical care medicine. Also, to implement a tool capable to find major drawbacks in critical care management.
Methods Initially, evidence-based interventions were reviewed in order to choose six inexpensive interventions, which could be taught to most physicians and easily applied in ICUs, even those with minimal resources. First, the MEDLINE database was systematically reviewed, addressing established evidence-based interventions and preparing guidelines and protocols to conditions that reduce sepsis-induced organ dysfunctions, and thereby morbidity and mortality. Second, evaluative questionnaires before and after the proposed interventions are characterizing individualized ICU current standards of care and how they will be impacted by the SETUP Project. Third, adherence to the proposed protocols will be evaluated. Finally, patient outcome evaluation will demonstrate the overall impact of the SETUP Project. Introduction Oxygen delivery depends on cardiac output and oxygen arterial content. Adequate preload is important to optimize cardiac output. Optimal preload can be obtained through Frank-Starling curves, which are difficult to determine at the bedside. Thus, optimal preload is difficult to determine in the intensive care unit.

P42 Is it possible to obtain Frank-Starling curves in the intensive care unit? FS Machado, ACKB Amaral, MC Souza, GL Büchele, ACB Sogayar, GL Oliveira, CJ Fernandes Jr, V Michels Jr, E Silva, E Knobel
Objective To evaluate the correlation between the wedge pressure and systolic volume in different subsets of shock. Results See Figure 1.

Materials and methods
Conclusion A higher wedge pressure is not associated with a greater systolic volume. Thus, the absolute value of the wedge pressure cannot be associated with optimal preload.

Figure 1
R, Pearson correlation coefficient; R 2 , square Pearson correlation coefficient.  Background Patients with acute renal failure (ARF) often represent a diagnostic challenge to physicians in the intensive care unit. A noninvasive test would be valuable to assist the clinicians in making differential diagnoses for these patients. Technetium-99m-L,L-ethylenedicysteine ( 99m Tc-EC), a new renal radiopharmaceutical, provides high-quality images in patients with renal failure and its clearance resembles that of hippuran, which makes it a desirable agent to evaluate tubular function, such as in patients with acute tubular necrosis (ATN). Adequate renal uptake seemed to predict recovery of renal function.
Aim To evaluate the diagnostic value of 99m Tc-EC renography in the diagnosis of acute renal failure in critically ill patients in an intensive care unit.
Methods After the usual diagnostic approach, two patients were submitted to renal scintigraphy with 99m Tc-DTPA and 99m Tc-EC, and one patient to 99m Tc-EC renal scintigraphy.

Results
The first patient was hospitalized with hypertension and left renal abscess. She was submitted to left nephrectomy and evoluted to ARF, requiring hemodialysis. The 99m Tc-EC renography showed normal flow to the right kidney, and delayed and decreased activity compatible with ATN. The clearance values of 99m Tc-EC and 99m Tc-DTPA were 28.4 ml/min and 6.1 ml/min, respectively. The hemodialysis was suspended and the 99m Tc-EC clearance increased to 35 ml/min. The second patient had diabetes and hypertension, and was hospitalized with acute pulmonary edema. The 99m Tc-EC renography showed decreased flow and uptake to both kidneys without any evidence of significant excretion into collecting systems. The findings were consistent with the clinical suspect of renal cortical necrosis. The clearance values of 99m Tc-EC and 99m Tc-DTPA were 16.9 ml/min and 9.4 ml/min, respectively. The patient required hemodialysis treatment. The third patient had myocardial infarction and evoluted to shock with ARF. The 99m Tc-EC renography showed normal flow to both kidneys, and delayed and decreased activity compatible with ATN. The patient required hemodialysis. The clearance value of 99m Tc-EC was 39 ml/min.

Conclusions
For patients with acute renal failure, the 99m Tc-EC renal scan may facilitate decisions regarding the management of therapy. The use of the 99m Tc-EC renal scan in the differential diagnosis of acute renal failure appears promising in the setting of the intensive care unit. Methods All patients admitted to the ICU with a length of stay greater than 48 hours were prospectively followed. ARF was defined by a serum creatinine higher than 1.5 mg/dl. The Acute Physiology and Chronic Health Evaluation (APACHE) II score, LODS score, pre-admission and admission characteristics, and clinical evolution were registered. The results were analyzed by t test, Wilcoxon test and, in multivariate analysis, logistic regression and discriminant analysis. They were considered significant if P < 0.05 or if the confidence interval differs from one. Conclusions The incidence, morbidity and mortality of ARF in the ICU are high. The main factors associated with its development and prognosis are those associated with hemodynamic instability, suggesting that ARF is a part of multiple organ dysfunction syndrome. The PE was confirmed in 37 of those patients (of which 32 were older than 65 years old) with one of these confirmatory methods: ventilation-perfusion lung scan with a high probability result (56%), the identification of a clot by spiral computer tomography (41%), or a positive gadolinium-enhanced magnetic resonance angiography (3%).

Results
The mean age of our patients was 78 ± 5 years old (maximum 98 years) and the mean Acute Physiology and Chronic Health Evaluation II score was 11 ± 6 (maximum 20). Eighty-four percent of our patients were female, and the most prevalent risk factors were age, immobilization (31%), ejection fraction < 40% Introduction It is known that mechanical ventilation with the adoption of high tidal volumes (V T = 10-15 ml/kg) has a relationship with the increase of the death rate in patients with acute respiratory distress syndrome (ARDS) [1]. Since then, different ventilatory strategies have been investigated adopting small tidal volumes; however, there is no agreement about which of them would cause a minimal pulmonary aggression to the preexisting injury.
Objective To compare two methods of mechanical ventilation employed in septic patients with ARDS (controlled volume and controlled pressure), both adopting permissive hypercapnia, evaluating the hemodynamic and respiratory effects of those patients.

Methods
Research previously approved by the Hospital's Medical Ethics Committee, prospective and randomized, performed in the intensive care unit of the Hospital of UNICAMP. Seven patients

Conclusion
The authors warn of the necessity of a clinical suspicion of pulmonary thromboembolism as the first step of a differential diagnosis with other prevalent pathologies among the elderly, and highlight the good results with therapy, including the thrombolitics.
were ventilated with controlled volume and nine patients with controlled pressure (Bird 8400 ® Sti ventilator for both groups), the tidal volume variation being between 6 and 8 ml/kg, accepting PaCO 2 until 80 mmHg with pH > 7.2. The ideal positive end expiratory pressure was calculated based on the higher compliance level through the method of progressive positive end expiratory pressure. The selected patients were submitted to the Murray scale [2] with LIS ≥ 2.5, and the Sepsis-related Organ Failure Assessment [3] and Acute Physiology and Chronic Health Evaluation II scale were applied to evaluate their gravity on the moment of the protocol inclusion. Hemodynamic and respiratory monitoring was carried out by Swan-Ganz catheter, gasometry (arterial and venous) and capnometry sampling, twice a day during three subsequent days.

Results
There were no parameter differences between both mechanical ventilation groups. There was a significant difference for both groups, from the first to the second day of collection, only on the following parameters: pH rise and reductions of PaCO 2 and RVP.
Conclusion All septic patients with ARDS studied, when ventilated with low V T (6-8 ml/kg), both in controlled volume and controlled pressure, did not present differences in the lung function nor in the hemodynamic state.
Introduction Bronchopulmonary hygiene physical therapy refers to techniques that promote clearance of airway secretion. One of these techniques is manually assisted coughing (MAC), which could be applied unilaterally, bilaterally or thoracic-abdominally.
The first research about MAC was published in 1966; nevertheless, until now, no study has tried to compare the three different forms of applying this technique. Studies about MAC have just analysed the effects of secretion clearance on oxygenation. The aim of this study was therefore to compare the peak expiratory flow (PEF) produced by MAC when accomplished in different application forms, in mechanically ventilated patients.
Design A prospective, experimental and blinded study.
Methods Twenty-six patients (mean age 60 ± 13.5 years) admitted to the Hospital Português Clinical-Surgery Intensive Care Unit requiring controlled mechanical ventilation and with positive end expiratory pressure of 5.5 ± 1.44 cmH 2 O were assessed. Patients with a history of pulmonary disease, hemodynamic instability, rib cage and/or abdominal abnormalities, scoliosis, pregnancy, obesity, a cardiac pacemaker, pneumothorax, unstable thorax and positive end expiratory pressure higher than 10 cmH 2 O were excluded. The PEF was measured by the Navigator Graphics Monitor (Newport Medical Instruments Inc., Newport Beach, CA, USA). MAC was performed five times in each hand position (unilateral, bilateral and abdominal-thoracic) in an alleatory way, with intervals of three breathings between each application of the technique. The Mann-Whitney rank-sum test was used to compare the groups. Differences with P < 0.05 were considered significant.

Results
Means of PEF variations are presented in Table 1. When comparing the different modalities of MAC, no statistical significant differences were noted.
Conclusion MAC was efficient to increase the PEF in all tested modalities. Nevertheless, these results confirm that the efficacy was no different between the modalities, suggesting that it is a personal choice involving the patient and the therapist on the moment of technique performance. Introduction Bronchopulmonary hygiene physical therapy refers to techniques that promote clearance of airway secretion. One of these techniques is manually assisted coughing (MAC), which could have early smaller airways narrowing or collapsing as an undesirable effect. Positive end expiratory pressure (PEEP) promotes small airway stability, and increases functional residual capacity and pulmonary volume at the end of expiration. The purpose of this study was to analyze whether the PEEP effects described are able to optimize peak expiratory flow (PEF) during MAC.
Design A prospective, experimental and blinded study.
Methods Twenty-six patients (mean age 60 ± 13.5 years) admitted to the Hospital Português Clinical-Surgery Intensive Care Unit and requiring controlled mechanical ventilation were assessed. Patients with previous pulmonary disease, hemodynamic instability, rib cage and/or abdominal abnormalities, scoliosis, pregnancy, obesity, a cardiac pacemaker, pneumothorax and unstable thorax were excluded. The PEF variation during MAC with a baseline PEEP (mean 5.5 ± 1.44 cmH 2 O; n = 26) was compared with the same modality with a PEEP of 12 cmH 2 O (n = 16). PEF was measured by the Navigator Graphics Monitor (Newport Medical Instruments Inc., Newport Beach, CA, USA). MAC was performed five times in each hand position (unilateral, bilateral and thoracic-abdominal) in an alleatory way. The Mann-Whitney rank-sum test was used to assess differences between the groups. Differences with P < 0.05 were considered significant.

Results
Means of PEF variations are presented in Table 1. When all variations of PEF in each modality of MAC were compared with the variations after addition of PEEP, statistical significant differences were noted. When comparing the different modalities of MAC with adding PEEP, no statistical significant differences were noted.
Conclusion PEEP was efficient in MAC optimization and in increasing the PEF. The association of PEEP during MAC in mechanically ventilated patients is recommended.

Introduction
The intra-abdominal pressure (IAP) usually elevates in critically ill patients and must be monitored to avoid compartmental syndrome. The mechanical ventilation may increase the IAP even more by the transmission of the thoraxic pressure from the diaphragm.
Objective To assess the effect of the optimization of the positive end expiratory pressure (PEEP) on the increase of the IAP in patients with a diagnosis of intra-abdominal hypertension.
Patients and methods Fifteen patients needing PEEP optimization and with intra-abdominal hypertension. The measurement of the IAP was obtained by intravesical pressure at five different moments: before and after neuromuscular blockade, right after PEEP optimization, and 6 and 12 hours after this procedure.

Conclusion
The increment of the PEEP does not alter the levels of intra-abdominal pressure in the first 12 hours after PEEP optimization. Objectives To analyze gas exchange, respiratory mechanics and cardiovascular monitoring parameters during mechanical ventilation weaning, using pressure support ventilation (PSV) and T-piece techniques, and to compare these variables in subgroups of patients with heart disease (HD) or nonheart disease (NHD).

Materials and methods
A randomized crossover clinical trial comparing PSV and T-piece techniques was performed. Twenty patients, aged 57 ± 15 years, 13 (65%) male and seven (35%) female, who were on mechanical ventilation for a period ranging from 2 to 54 days were studied. The following were analyzed: peripheral oxygen saturation (SaO 2 ), partial carbon dioxide pressure in the exhaled air (PetCO 2 ), respiratory rate, tidal volume (V T ), minute ventilation (V E ), heart rate (HR), systolic blood pressure (SBP), diastolic blood pressure (DBP), mean blood pressure (MBP), changes in ST segment and presence of arrhythmia at the electrocardiographic evaluation. Data were recorded at times 0, 15 and 30 min after the start of the randomized weaning technique, with a 30 min resting interval before starting the second technique. Patients were also grouped as having HD (n = 11) and NHD (n = 9), and compared relative to cardiovascular parameters.

Results
The comparison between PSV and T-piece techniques demonstrated that total SaO 2 and PetCO 2 were significantly higher during PSV, at all times (P < 0.001 and P < 0.05). As for respiratory rate, it was reduced when subjected to PSV at times 0 and 15 min (P < 0.05). V E and V T were significantly increased while at PSV, at all three times (P < 0.001). There were no differences between the PSV and T-piece techniques for the values of MBP, SBP, DBP and HR. The comparison between PSV and T-piece techniques in patients with HD and NHD has shown that total HR values in patients with HD were significantly lower at time 30 min in PSV (P < 0.05), with no significant difference in the other measurements. The HR was greater in NHD patients, both during PSV and T-piece (P < 0.05). When comparing patients with HD (n = 11) versus NHD patients (n = 9), ST segment changes were observed more often in those with HD (64%) than in NHD patients (11%) (P < 0.05). Arrhythmia occurred in 27% of the patients with HD and in 11% of those with NHD; sinus tachycardia was observed only in NHD patients, in five (56%) of them (P < 0.01).
Conclusions When comparing PSV and T-piece techniques, the measurements of respiratory parameters and oxygenation displayed better results with the use of PSV. There were no significant differences in the measurements of cardiovascular and EKG parameters. When patients with HD and NHD were compared, a reduction in HR at 30 min on PSV was observed only in those with HD. Also, a greater number of ST segment changes, a smaller occurrence of sinus tachycardia and a trend toward a greater occurrence of arrhythmia in patients with heart disease were observed in both weaning modes. Objectives Ventilation by bilevel positive airway pressure (BIPAP) was been applied in chronic obstructive pulmonary disease (COPD) with the aim of increasing oxygenation, and reducing dyspnea and respiratory work. However, the positive pressure may produce harmful effects on the cardiovascular system. The aim of this study was to evaluate the autonomic modulation of the heart by the heart rate response and heart rate variability analysis (HRV) during BIPAP ventilation in COPD patients.

P56
Methods Seven patients aged 65.2 ± 6 years with FEV 1 < 50% predicted were selected for the study. The COPD diagnostic was based on the clinical history and spirometry test. The study protocol was approved by the Universidade Federal de São Carlos Human Ethics Committee. All patients were informed of the nature of the study and signed a consent document for participation (CNS no. 03/2000). The patients were submitted to clinical and resting electrocardiograph evaluation before the study. The experimental protocol consists of the collection of the heart rate beat to beat and RR intervals (RRi) (in milliseconds) using a cardiac rate meter (Polar-Vantage). Heart rate data and RRi were collected over a period of 10 min during rest in the supine and sitting positions. After this, the BIPAP (Respironics) was applied with inspiratory pressure (IPAP) of 14 cmH 2 O and expiratory pressure (EPAP) of 6 cmH 2 O, during 20 min with a nasal mask, and then HR and RRi values were collected. The data analysis evaluated the RRi values and HRV using the calculation of the RMSSD index of the RRi (in milliseconds), the square root of the mean sum of squares of the differences between the adjacent normal RRi in the record divided by the number of RRi within a given time minus one RRi. Data were analyzed statistically by the Friedman and Dunn test, with the level of significance set at P < 0.05.

Results
There were no significant differences in the absolute results of RRi and HRV by the RMSSD index in different situations (P > 0.05).

Conclusion
The results suggest that the BIPAP with pressures levels applied did not promote changes in the vagal-sympathetic control of the heart and in HVR. These procedures thus do not involve additional cardiovascular risks for the patients with COPD. Objective To determine the hemodynamic profile and the response to vasopressor therapy in ARDS patients, and its relationship with mortality.

Materials and methods All patients who met American-European
Consensus criteria for ARDS between January 1999 and December 2002 were included in the analysis, dividing into survivors (SV) and nonsurvivors (NSV). The following data were collected prospectively: age, Acute Physiology and Chronic Health Evaluation (APACHE) II score at admission and at diagnosis, heart rate, mean arterial pressure, central venous pressure, mean pulmonary artery pressure, pulmonary artery occlusion pressure, systemic vascular resistance index and pulmonary vascular resistance index, left and right ventricular stroke work index, cardiac index, and doses of dopamine, dobutamine and norepinephrine, between the first and seventh days. The differences between groups were analyzed with a t test and considered significant with P < 0.05.

Results
The results showed a mean age of 79 ± 9.7 years and a preponderance of the male gender (58%). The mean APACHE II score was 16.7 ± 7.1 and the average time of requiring the IMV was 12.6 ± 8 days. Although the cohort mortality was 13%, this rate in patients submitted to IMV was 52%. Among all the variables studied, the following had a correlation with death: pulmonary septic shock (P = 0.01), cardiogenic shock (P = 0.01), inotropic drugs use (P = 0.01), acute coronary syndrome (P = 0.01) and cardiac failure during ICU hospitalization (P = 0.01).

Conclusions
In patients older than 65 years IMV was associated with significant mortality, and the inotropic drugs used, pulmonary septic shock, cardiogenic shock, acute coronary syndrome and the cardiac failure during ICU hospitalization were the only variables studied that were related with death. Automatic pressure support reduction based on a targeted respiratory frequency or MRV is disposable in the TAENA ventilator for an automatic reduction of pressure support during weaning of patients in the intensive care unit (ICU). We studied 23 patients (63.52 years) in the postoperative period (14 cardiac, two thoracic and seven abdominal surgeries) in a prospective, randomized protocol comparing automatic pressure support weaning with the traditional manual reduction of pressure support to 5-7 cmH 2 O in our ICU. After arriving in the ICU after cardiac, thoracic or abdominal surgery, the patients were randomly assigned to traditional weaning consisting of manual reduction of pressure support (the pressure support was decreased every 30 min, keeping the RR/TV(L) < 80 until 5-7 cmH 2 O pressure support ventilation) or to the automatic pressure support reduction (MRV) with a respiratory frequency target of 20/min (the TAENA ventilator automatically decreased the pressure support ventilation level by 1 cmH 2 O every four respiratory cycles if the patient's RR was less than 16/min). Twelve patients were assigned to manual weaning whereas 11 patients were assigned to the automatic pressure support reduction weaning. The weaning mean time for the manual group was 3.18 hours while the weaning mean time for the automatic pressure support reduction group was 2.24 hours. There was no reintubation in both groups.

Conclusion
The automatic reduction of pressure support is effective and without complications, and it can be useful for weaning patients in the postoperative period in the ICU.

Results
One hundred and ten patients met the ARDS criteria, of which 96 were managed with a pulmonary artery catheter and included in the analysis. In SV and NSV, the mean age was 36 ± 16 and 51 ± 19 years (P = 0.0001), and the APACHE II score at admission was 16.6 ± 6.3 and 20.1 ± 8.3 (P = 0.031) and that at ARDS diagnosis was 17.1 ± 5.5 and 22.3 ± 6.7 (P = 0.001), respectively. The hemodynamic profile and doses of the vasoactive drugs are presented in Table 1.
Conclusions NSV were older and sicker according to the APACHE II score than SV, and they presented significant vasopressor dependency in the first 4 days. The absence of statistical difference in hemodynamics is probably due to purposeful interventions aimed to optimize the cardiovascular status of patients based on invasive hemodynamic data, and not related to the pulmonary artery catheter's inability to change patient outcome. The need for higher doses in the vasopressor reflects a vasodilatory state that could contribute to mortality. Introduction Central venous catheters (CVC) are frequently used in the intensive care (ICU) setting. A wide number of strategies have been studied to prevent colonization and infection related to CVC. One of them is the use of antimicrobial-impregnated catheters, but no positive effect has been shown primarily for triplelumen catheters.

INFECTION P60 Prevalence of cytomegalovirus infection among patients in an intensive
Objective To compare the incidence of CVC colonization in two groups of patients using a double-lumen CVC impregnated with chlorhexidine and silver sulfadiazine, or the standard one.

Patients and methods
Patients undergoing insertion of a doublelumen CVC in the ICU were randomized to receive either a VCV impregnated with chlorhexidine and silver sulfadiazine, or the standard one. The catheter tips were cultured by the roll-plate method after removal.
Results Sixty patients enrolled with successful insertion of 60 catheters, 24 of them impregnated (group 1) and 36 standards (group 2). There were no statistically significant differences between the groups in age, seven infection-related risk factors, ICU diagnosis, mean Sepsis-related Organ Failure Assessment score, insertion sites, duration of catheterization, wrong location at X-ray, signs of allergy, and catheter colonization rates. The mean times of duration of catheterization in group 1 and group 2 were 14.5 ± 8.3 days and 13.8 ± 6.2 days respectively (P = 0.8). The mean Sepsis-related Organ Failure Assessment scores in groups 1 and 2 were 5.4 ± 3.4 and 5.2 ± 3.3, respectively (P = 0.8). The colonization rates were 23.1% (six catheters) in group 1 and 29.4% (10 catheters) in group 2 (P = 0.5). The impregnated catheters presented a trend of being removed more frequently due to systemic infection suspicion (P = 0.05). Fifteen catheters presented Gram-positive cocci, four of them associated with Gram-negative bacilli and two with fungi. One catheter presented two Gram-negative bacilli in the roll plate.
Conclusion This preliminary analysis of the comparative study between a double-lumen CVC impregnated with chlorhexidine and silver sulfadiazine and the standard one did not show any statistically significant difference in colonization rates between the two groups. Improvement of the endocarditis was achieved but the patient deteriorated due to osteomyelitis in the left femur and sacral region. All of these were detected by gallium-67 citrate scintigraphy in spite of standard vancomycin therapy. Therefore, we changed the therapeutic strategy to oral linezolid (600 mg twice daily) because of some difficulties to maintain safe vascular access [2]. An achievement of good outcome was shown by the second gallium-67 scintigraphy 5 weeks later. Furthermore, long treatment with linezolid was very well tolerated.

P64 Resolution of
In conclusion, control of the bone infection with staphylococcus MRSA after a 5-week course of oral therapy with linezolid was attained. Treatment of osteomyelitis associated with a susceptible bacterial strain (MRSA) with this class of antibiotics taken orally appears to be safe, effective, and yielding a good outcome. Our case report supports the arguments of those who advocate the utilization of this kind of therapy, although there has not yet been consensus in the literature [3]. Introduction It is of essential importance to be acquainted with the microbiological predominance profile (bacteria and fungi) in an intensive care unit (ICU), and with this to better lead the therapy offered to the patient, to guide the treatment, to establish the procedures, preventive actions and prolonged education, to influence clinical decision making, to understand problems connected to hospital contamination and to take administrativedirected decisions.

P65 Microbiological predominance in an intensive
Objective Being acquainted with the predominance tendency and the fungal and bacterial profiles in the Albert Einstein Jewish Hospital Intensive Care Unit. Results See Tables 1 and 2.

Materials and methods
Conclusion Pseudomonas aeruginosa predominance is very high. Acnetobacter baumanni and Candida albicans are similar to the worldwide literature. Staphylococcos aureus has a low value, in our point of view, due to the impact of medical actions, educative procedures and administrative rules inserted in the service. Objective Catheter-related bloodstream infection is a major cause of morbidity and mortality in intensive care units (ICUs). The purpose of this study is to assess the risk factors associated with these infections.
Methods Thirty-one patients admitted to an ICU of our hospital were enrolled for a retrospective study. In these patients, a total of 64 central venous catheters were inserted. Data was collected and submitted to a univariate analysis.

Results
The mean Acute Physiology and Chronic Health Evaluation score and age were 19 and 49 years, respectively. Sixteen patients were imunosuppressed, 24 patients were under mechanical ventilation and 10 patients died during the ICU stay.
The mean duration of catheter permanence was 9 ± 4 days. Among the catheters, eight were dual-lumen hemodialysis catheters, 12 were Swan-Ganz catheters and 44 were duallumen catheters. The sites of insertion were 37 in the internal jugular vein, 22 in the subclavian vein and five in the femoral vein. Fourteen dual-lumen catheters were the source of bloodstream infection. The majority of these infections involved Gram-positive aerobic organisms.

Conclusion
In this study, some classical risk factors for catheterrelated bloodstream infection were not found, but some new risk factors were identified. The technique of insertion and the care at the site of catheter insertion must be further assessed, seeking for the main risk factors for catheter-related bloodstream infection. Discussion In this present study, incidence and mortality rates did not differ from literature data, despite appropriate and early treatment of these infections. Hantavirus Cardiopulmonary Syndrome was first described in southwestern USA as a condition that involves mainly the lungs, leading to a high mortality rate (over 50%), caused by a new Hantavirus species. The pathology of this disease involves the direct viral damage of lung endothelial vessel cells, leading to leaking and alveolar inundation. Huge fluid expansion may lead to deterioration of pulmonary function and increased mortality. Therefore, one has to be cautious about fluid reposition, and treatment must take into account the use of vasoactive drugs.

P67 Mobile septic intra-atrial masses in premature infants
We report four cases of such a syndrome admitted to our intensive care unit from December 2000 to March 2002. All patients came from the same county. They were rural inhabitants, living near rice and soybean storehouses that allow rodents entrance. The patients' age ranged from 31 to 57 years (median 36.5 years), and two of them were female. The initial symptoms (coughing, muscular pain, fever and dyspnea) set in from 48 to 96 hours before admission. Two patients required mechanical ventilation (and one of them also required the use of a pulmonary artery catheter).
Three patients survived (intensive care unit discharge before 1 week after admission). One patient died due to pulmonary atelectasis on the seventh day after admission.
Chest X-rays demonstrated in all cases mainly alveolar compromise, both bilateral and symmetrical, reminiscent of acute respiratory distress syndrome patterns in the latter stages. Blood sample serology confirmed the suspected diagnosis. Treatment consisted of general support, cautious fluid replacement and vasoactive drugs for the treatment of shock. We did not use ribavirin (although its use is well described in the literature).
In conclusion, diagnosis is based on high clinical suspicion, taking into account the epidemiological aspects that are highly important in such cases. Clinical manifestation is mainly noncardiogenic pulmonary edema, due to the direct endothelial vessel cell damage, leading to leaking and alveolar inundation. Huge fluid expansion may increase mortality. Use of a pulmonary artery catheter plays an important role in selected cases, in which fluid replacement can be targeted according to the pulmonary capillary wedge pressure. Through a retrospective study of 100 patients with IE based on Duke criteria, our group evaluated the clinical and echocardiographic factors that most strongly correlate with intrahospital mortality.

P71 Mortality predictor parameters in infective endocarditis C Magalhães, R Gomes, W Almeida, J Amino, B Tura, A Siqueira
The study included patients with native valves and prosthetic valves as well as patients with congenital cardiac disease.
The results were obtained using the chi-square and Fisher exact tests.
In conclusion, age, anemia, neurological events, number of echocardiographic abnormalities and Janeway lesions are strongly correlated with an increased mortality risk. The presence of cardiac heart failure, renal insufficiency, arthritis, mitral valve involvement, abscess and surgical indication show a tendency for the same correlation. Background The modern management of traumatic brain injury (TBI) resulted in a significant reduction in mortality and functional deficit. These advances are imputed to the introduction of neurosurgical intensive care units (ICUs) and the implementation of guidelines to prevent secondary insult. However, in Brazil, most patients with TBI are managed in general ICUs. The results of the treatment of patients admitted to nonspecialized ICUs must be compared with those obtained in neurosurgical ICUs.
Objective We evaluated retrospectively a group of patients with TBI managed at a general ICU. Epidemiological aspects, severity evaluation, monitoring and the impact of therapeutic interventions were analyzed. The extended Glasgow Outcome Scale [1] was used to evaluate the long-term prognosis of patients discharged from the unit.
Patients All adult patients with TBI admitted to a general ICU from February 2000 to December 2002 were included. Patients who were discharged and those that died in the first 24 hours after admission to the ICU were excluded.

Measurements and results
Thirty patients (46.9%) had a Glasgow Coma Scale of 3-8 on admission. The 10 patients that died in the ICU belonged to this group. Thirty-three patients were victims of motor vehicle accidents and 26 (40.6%) had major extracranial injuries. In 50 patients, tomographic results could be classified according to the Traumatic Coma Data Bank. Diffuse injury I (14 cases), diffuse injury II (14 cases) and nonevacuated mass lesion (10 cases) were the most frequent findings. It was not possible to identify any impact of admission glycemia > 110 mg/dl, PaO 2 /FiO 2 and mean PaCO 2 in the first 72 hours on mortality. The intracranial pressure was monitored in six patients and the jugular bulb oxygen saturation in seven patients. The Glasgow Outcome Scale was evaluated in 40 (62%) patients. Assessment was carried out ≥ 6 months after the date of injury (median 29 months). Thirteen patients died, 10 while at the ICU. Sixteen patients made a good recovery; eight (20%) had moderate disability and three (7.5%) had severe disability. No patient remained in a vegetative state.

Conclusions
Our results, compared with studies that analyzed patients managed at neurosurgical ICUs [2], demonstrated that patients with TBI admitted to a general ICU with resources to prevent and treat secondary injury have mortality and functional results after discharge comparable with patients managed at specialized ICUs. Objectives To assess the predisposing factors, the incidence and the influence of ACS in patients undergoing CS, and to assess the impact of ACS on the length of stay, morbidity, and mortality.

Results
The univariate analysis showed significance of the following variables: chronic obstructive pulmonary disease (P = 0.03), advanced age, and multiple organ dysfunction syndrome (MODS) score. After logistic regression, only the MODS score (P = 0.01) and age (P = 0.005) showed significance. In regard to length of stay in the ICU, the results were as follows: up to 3 days, 41.17% of the patients with ACS and 65.91% of those without ACS; from 4 to 7 days, 38.23% of the patients with ACS and 21.42% of those without ACS; and longer than 8 days, 20.58% of the patients with ACS and 12.65% of those without ACS, with significance (P = 0.00002).
Conclusions ACS relates to age, MODS score, and longer length of stay in the ICU. Materials and methods A retrospective series of 33 cases from January 1998 to December 1999, in which dissociative anesthesia with ketamine plus midazolan was used to provide anesthesia to prehospital trauma patients, most of them trapped in vehicles. A protocol was developed in 1997, and its use was the decision of the attending physician in patients with a revised trauma score 11 or 12. The data were collected from the patient form.

Results
Ketamine was administrated by intravenous route in 96.9% of cases. In 66.6% of the cases, patients received ketamine parallel to the extrication procedure. All patients became unconscious. The most frequent complications were agitation (9.09%), clonic eye movements (3.03%) and transient ventilatory depression (3.03%). None of the patients necessitated a definitive airway.
Conclusion Dissociative anesthesia with ketamine plus benzodiazepine is efficacious and safe in a prehospital setting, in patients with a revised trauma score of 11 or 12. Design, setting, and patients A prospective observational study of 200 neurological and neurosurgical patients admitted between April and October 2002 at a NICU located in São Paulo city's central area hospital. Subsets of this population were also studied: tumoral excision, aneurysm surgery, laminectomy, stroke, neurological intravascular procedures, GCS ≤ 8, GCS > 8 and GCS > 14, and APACHE II mortality ≤ 10 and APACHE II mortality > 10. A bicaudal analysis was made, and P <0.05 was considered significant.

Results
The LOS in NICU and in hospital was significantly higher for SUS patients. We found a significantly larger proportion of patients with a GCS ≤ 8 in the SUS group. This trend was maintained in the majority of subgroups of patients studied. The predicted mortality (APACHE II), mortality rate, readmission rate, age, APACHE II score, proportion of clinical patients and emergency surgery rate was similar among the two main groups and in the majority of subgroups studied.

Conclusion
Our data indicate a relationship between the insurance coverage and LOS (NICU and hospital). We also found a larger proportion of patients with reduced consciousness level, higher in the SUS group. These trends are maintained in the majority of subgroups studied, strongly suggesting a true variability in the process of care among the main groups studied. We believe that, at least in countries with similar health system structures, insurance coverage could play a substantial role in explaining care variability, and should be better studied. Dexemedetomidine (DEX) is a potent α 2 -adrenoceptor agonist with an α 2 : α 1 ratio of 1300:1 that produces stable tranquility with rousability. DEX permits haemodynamic stability by effectively blunting both cathecolamine and haemodynamic responses to endotracheal intubations, surgical stress, and arousal from anaesthesia. We evaluated DEX indication, time of usage and dosage, necessity of other sedating drugs, and reasons for DEX interruption.

Results
The patient age average was 62.6 years and the Acute Physiology and Chronic Health Evaluation II score was 12.73. Indications for DEX were sedation for collaboration for weaning from mechanical ventilation, sedation of agitation in the ICU, adjuvant treatment of delirium and adjuvant to analgesia. The average dose used was 0.31 µg/kg per hour (0.17-1.0), and the loading dose was used in only four patients (3.73%). The reasons for interruption of DEX were: arterial hypotension, eight cases (7.47%); sinus bradycardia, three cases (2.80%); bradycardia + hypotension, two cases (1.86%); and weaning failure, 22 cases (20.5%). All cardiovascular events disappeared immediately after DEX interruption. The mean time of usage was 3.25 days (range 1-13 days). Concomitantly sedating drugs had to be used in 25 patients (23.3%): fentanyl in seven (6.54%), haloperidol in seven (6.54%), haloperidol + prometazine in four (3.73%), midazolam in five (4.67%), and midazolam + fentanyl in one (0.93%).

Conclusions
The use of DEX beyond 24 hours appears to be safe and effective for the sedation of ICU patients. The need for other sedating/analgesic drugs occurred in less than one-quarter of the patients and was well tolerated, with no extrapyramidal signs seen with antipsychotic drugs or no respiratory depression with opiates. Objectives The M-mode transcranial color Doppler technique became an important tool for neurointensivists in the followup of patients with subarachnoid hemorrhage and after its corrective surgery. Through the two-dimensional color Doppler (TDCD) technique we could obtain the same sort of information obtained with M-mode, but additionally it would be possible to visualize anatomically intracranial vessels. Therefore our purpose was to acquire color images from the intracranial vessels, the identification of arterial segments under vasospasm, and the recognition of aneurysms using the color Doppler technique.

Methods
We used a two-dimensional color Doppler ultrasound with a 2 MHz transducer. We obtained two-dimensional color images from the Circle of Willis, recording vessel velocities and analyzing the flow, resistance index and pulsatility index of patients admitted to the intensive care unit with subarachnoid hemorrhage. We used transtemporal, occipital and ocular windows in order to register the arterial flow. We defined vasospasm when mean velocities were higher than 120 cm/s. Images of vasospasm and probable aneurysms were recorded, and afterwards compared with cerebral angiography.

Results
From May 2000 to August 2001, 16 patients were admitted to our surgical intensive care unit, where 13 were considered Fisher stage IV on head computed tomography scan. We diagnosed three aneurysms that were later confirmed by cerebral angiography. One was on the top of the basilar artery and the two others were in the middle cerebral artery. Arteries under vasospasm were also identified. Sixty-one percent of patients in Fisher stage IV had vasospasm initially diagnosed by TDCD and confirmed in the angiogram afterwards. Flow and velocities were recorded; these data helped us to understand and to use appropriate therapeutic intervention.

Conclusions
There was technical feasibility on obtaining twodimensional color Doppler images of intracranial vessels. There was no need to use contrast in the identification of the vessels. Immediate better management of cerebral vasospasm, including percutaneous angioplasty in selected cases, were possible thanks to color Doppler. Finally, good anatomical correlation between images from TDCD and angiography were also noted. Methods Fourteen dogs were randomized to two groups: group 1, VF without AV (n = 7); group 2, VF with bag ventilation (15:2) (n = 7). A 10-min CPR followed 1-min unassisted cardiac arrest. After CPR, animals underwent defibrillation and advanced cardiac life support. Blood samples from the cerebral transverse venous sinus and the pulmonary artery were collected for gas analysis and lactate. Ventilatory parameters were measured by a flow transducer.

Results
There was no significant difference between groups in baseline measurements and successful CPR (three in each group). Systemic and cerebral oxygen extractions were significantly higher in group 2, although there was no difference in lactate between the groups. During CPR, PaO 2 and PaCO 2 were, respectively, higher and lower in group 2. There was no difference in the minute respiratory volume during the first 5 min of CPR. Thereafter, the minute respiratory volume decreased significantly in the group without AV.
Conclusions In this experimental model of CPR for VF, assisted ventilation (15:2) maintained higher arterial oxygen saturation and higher systemic and cerebral oxygen extraction but did not result in higher return of spontaneous circulation. After the first 5 min of CPR, AV maintained significantly higher ventilation and oxygenation parameters. during insertion, arterial puncture, bleeding, and misplacement; during catheter permanence, obstruction and infection (daily examination); and after decannulation, thrombosis detected through Doppler examination, which was performed after 24 hours. All the catheter tips were sent for bacteriological examination.

Results
The mean time of cannulation was 5 days. The thrombosis rate detected by Doppler examination was 31.6% (without clinical compromise). The catheter obstruction rate was 15.8% and the infection rate was 10.5%.

Conclusion
Strict control with Doppler examination is very important to warrant optimal flow. The catheter must be changed every 5 days in order to avoid infection. Diffuse axonal lesion 8 9

Background
The occurrence of decubitus ulcers in the intensive care unit (ICU) is a significant comorbidity and reflects quality of care. In addition, it has great impact on the patient's quality of life. The skin is the first line of defense in protecting the body from constant changes in the environment. Far too often, the attention needed for keeping skin integrity is only realized after it has been disrupted.
Objective To evaluate the new cutaneous integrity protocol (CIP) implemented in our critical patients.
Methodology A total of 542 patients were studied (253 before and 289 after the initiation of the protocol) between May and October 2002 in a 24-bed medical/surgical ICU. Patients who received the new CIP care were submitted to pressure-reducing devices such as a pyramidal mattress, topic fatty acids after corporal hygiene, the Braden scale and changes in bed position as indicated by the Lowthien revolving clock.

Results
Both groups were comparable regarding sex, age and origin of admission (P > 0.05). In the group before implementation of the CIP 45.7% of patients were male and 48.6% were between 60 and 80 years old, whereas in the CIP group 63.1% were male and 73.7% were between 60 and 80 years old. The main sites of ulceration were, in order of importance: the sacrum region, the head, the auricular pavilion and the intergluteous fissure. The number of ulcers developed in the ICU decreased from 13.8% to 6.57% (P < 0.05). The mean number of ulcers developed by patients was 2.49 versus 1.63 before and after the implementation of the protocol, respectively (P < 0.05).
Conclusion These data suggest that simple and low-cost measures can significantly improve the quality of patient care in the ICU by reducing the occurrence of decubitus ulcers. All patients should benefit from a well-designed program to keep skin integrity and prevent the development of pressure sores.
Introduction After the publication of the TRICC trial [1] almost 3 years ago, we expected transfusional practices to change, becoming more restrictive. However, since then no survey has been made in order to observe these practices in our intensive care units (ICUs).

Background
We have reported a pilot study of ABM-MCT to areas of ischemic myocardium [1], and in humans at 8 weeks follow-up after ABM-MCT ACC'03. The purpose of this study was to assess whether the changes in the QOL in patients with end-stage ischemic heart failure (ESIHF) submitted to catheter-based ABM-MCT persists in the 6 month followup.
Methods A prospective assessment of the QOL before and 8 and 24 weeks after ABM-MCT guided by the NOGA system, targeting hibernating myocardium in ESIHF, using the Minnesota Living with Heart Failure Questionnaire and the Medical Outcomes Study Short Form-36. The cardiac evaluation included Canadian Cardiovascular Society class, VO 2 max on the treadmill test, ejection fraction and end systolic volume on two-dimensional echocardiogram and total reversibility defect on MIBI-SPECT.

Results
There were 14 patients (two females). The changes in parameters 8 and 24 weeks after the procedure are presented in Table 1. According to these objective data we observed a QOL improvement, based on the scores of Minnessota varying from 46 ± 19 to 30 ± 17 (P = 0.002) at 8 weeks and to 18 ± 14 (P = 0.003) at 24 weeks, and an increase of all eight dimensions of the Medical Outcomes Study Short Form-36 as presented in Table 2.

Conclusion
The data suggest that ABM-MCT can improve the QOL of patients with ESIHF. Prospective studies with a greater number of patients will be necessary to confirm these initial data.  Methods Medical students interviewed 45 family members of patients in the ICU of a university hospital, using a 12-query questionnaire on the quality of care (physicians and nursing).
Results Forty-five families answered the questionnaire. The mean length of stay by the time of the interview was varying from 3 to 120 days; the most frequent complaint was the noise and reduced visit period (only 30 min/day), in 26% (12) of the answers. Twelve (26%) of the families reported that the patients had not complained of pain and nine (20%) did report pain, most of minimum intensity (66.6%) with quick relief after medication. There were two complaints (4.4%) of pain during blood examination sampling. Variations of temperature troubled seven (15.5%) of the patients. The visit period (30 min) was considered to be unsatisfactory by 18 (40%) of the families; 20 (44.44%) families suggested amplifying the visit period and the number of visitors.
The medical attendance was qualified as 'very good' by 24 (53%) and 'good' by 21 (47%), and the nursing attendance was qualified as 'very good' by 12 (26.6%) and 'good' by 27 (60%). Unsatisfactory information provided by the nursing staff was the complaint of 14 (31.11%) families. Thirty-eight (84.4%) families reported great hope in the treatment instituted at the ICU.

Conclusion
To know the anxieties of the families in the regard of treatment instituted in the ICU allows correction of mistakes and improvement of the quality and humanization. Purpose A haemoglobin decrease is common in critical patients. We investigated the haemoglobin behaviour in nonbleeding patients during the first week of stay in the intensive care unit (ICU).

Methods
The study was retrospective and included 23 patients (16 men and seven women) in a seven-bed ICU from January to March 2001. Data were collected on days 1, 2, 4 and 7. We analysed the haemoglobin range and the possible causes related to it. We used the median and interquartile range to show the data.

Results
Patients ages were 63 years (41.76 years), the Acute Physiology and Chronic Health Evaluation (APACHE) II score was 22 (16.32), and the body mass index was 24 kg/m 2 (23.28 kg/m 2 ). Haemoglobin medians were: day 1, 12 g/dl (10, 13 g/dl); day 2, 11 g/dl (9, 13 g/dl); day 4, 11 g/dl (9, 13 g/dl); day 7, 10 g/dl (8, 13 g/dl) (Friedman P < 0.001, post-hoc P < 0.05, day 1 > day 2, day 4, day 7). Fluid balance between day 1 and day 2 was 586 ml (-263, 2129 ml). The most important haemoglobin reduction occurred between day 1 and day 2. Conclusions In our sample, we verified a major fall in haemoglobin between the first and second day of stay in the ICU. Female gender and, with least importance, the APACHE score were linked to a major fall in haemoglobin during the time of study. Background The CR has been used as routine in our hospital, after thoracic tube withdrawal in HS-PO. This study evaluates the utility of the CR after thoracic tube withdrawal.

P86 Chest roentgenogram (CR) performed as routine after thoracic tube withdrawal in heart
Methods A retrospective study, in which 407 patients admitted to HS-PO with one or more thoracic tubes were evaluated. A total of 310 patients were submitted to coronary artery bypass graft and the other 97 patients were submitted to other cardiac surgeries. The incidences of pneumothorax and pneumomediastinum were evaluated, including clinical repercussion, the treatment used, and the relationship with thoracic tube withdrawal.

Results
Thirteen patients (3.17%) had a diagnosis of pneumothorax in HS-PO. Five patients had a pneumothorax diagnosis at a later time (6 days), related to central venous catheterization (subclavian vein). Eight patients (1.9%) had the diagnosis related to surgery and thoracic tube withdrawal. In two of these patients, an air escape through one or more tubes were previously detected. In two other patients, subcutaneous emphysema was detected before the tube withdrawal. Three other patients had dyspnea, chest pain and low oxygen saturation after tube withdrawal and before the roentgenogram. One of these last three patients also presented subcutaneous emphysema. These seven patients were submitted to pneumothorax treatment using a 'pig-tail' catheter. The eighth patient, despite having no signs and symptoms that could suggest any complication, had a pneumothorax diagnosed by roentgenogram, performed after thoracic tube withdrawal. This patient was kept in observation and was dismissed from hospital 6 days after surgery. Among the eight patients, two did not have the complication diagnosed by roentgenogram, but through thorax computed tomography.
Conclusion Among the 407 patients included in this study, CR was not essential for an early detection of complications after thoracic tube withdrawal in CS-PO. From the eight patients who had a pneumothorax diagnosis (1.9%) related to surgery and tube withdrawal, seven presented some kind of symptom suggesting the necessity of doing the examination. In only one patient did the examination detect the complication without any previous sign or symptom. We suggest a similar evaluation in other surgical intensive care units. Introduction Intrahospital transportation of mechanically ventilated patients is a recognized high-risk situation. Our objective was to determine whether transportation could be safely performed using a defined transport routine.
Methods Between February and March 2003, all mechanically ventilated patients who needed transportation out of the intensive care unit (ICU) were included. All transports were done using a microprocessed ventilator (Microtak 920 plus-Takaoka ® , São Paulo, Brazil) and an oxymeter, noninvasive arterial blood pressure and an eletrocardiography monitor (M.3000-Morrya ® ), together with a transport team composed of a physician, a nurse and a physiotherapist. Hemodynamics and respiratory parameters were measured immediately before disconnection from the patient's basal ventilator (BT) and, after returning to the ICU, immediately before disconnection from the transport ventilator (AT). All the complications during transport were registered. Statistical analysis was carried out using variance analysis and the paired Student t test. Results were considered significant if P ≤ 0.005.

Results
We studied 33 transports of 22 patients (eight female and 14 male) with a mean age of 46.6 ± 15.7 years. The main causes of ICU admission were trauma (42.4%) and elective neurosurgery (24.2%). Patients with pulmonary disease comprised 42.4% of all the transports. Patients were ventilated with positive end expiratory pressure higher than 5, with FiO 2 > 0.5 or were using vasoactive drugs before transportation in 24.2%, 24.2% and 33.0% of the cases. The mean duration of the transport was 43.4 ± 18.9 min and performing a tomography was the reason in 96.9% of the cases. Complications occurred in only 27.3%, mainly (72.7%) agitation easily treated with an increase in sedation. A significant decrease in CO 2 was found (BT, 46.6 ± 15.7 and AT, 38.75 ± 16.14; P = 0.005) together with a trend towards a better PO 2 /FiO 2 ratio (BT, 303.6 ± 137.4 and AT, 346.4 ± 126.7; P = 0.06). There was a trend towards an increase in cardiac rate (BT, 80.96 ± 18.7 and AT, 85.45 ± 17.6; P = 0.08) with no significant changes in mean arterial blood pressure (P = 0.93).

Conclusion
These results suggest that intrahospital transport can be safely performed. Our low incidence of complications is possible related to the presence of a multidisciplinary transport team together with proper equipment to monitor vital functions and close control of the patient's ventilation. Intensive care unit (ICU) admission of such patients carries a significant mortality. The mortality rate of critically ill patients with acute leukemia may be higher than 80%.

P88 Evaluation of patients with acute leukemia admitted to an intensive care unit
Objective To evaluate prospectively the characteristics of acute leukemia patients admitted to an ICU.

Materials and methods
During the period from January 1998 to December 2002 we evaluated patients with the diagnosis of acute leukemia admitted to a medical-surgical ICU of a 560-bed tertiary hospital in southern Brazil. The variables evaluated were: age, Acute Physiology and Chronic Health Evaluation II score classification, time of mechanical ventilation, time from ICU admission to shock, admission time before admission to the ICU, length of stay in the ICU and total hospital length of stay, incidence of septic shock, use of invasive hemodynamic monitoring through a Swan-Ganz catheter, and ICU and hospital survival. Statistical analysis was performed with the SPSS 11.0 software package using a t test and Kruskall-Wallis statistics where appropriate, with a significance level set at 5%.

Results
In the period from January 1998 to December 2002 there were 44 patients admitted with a diagnosis of acute leukemia. The findings obtained from these patients were compared with those of 1753 patients without the diagnosis of acute leukemia admitted to the ICU in the same period. In the acute leukemia group, the mean age was 48 ± 18 years, the mean Acute Physiology and Chronic Health Evaluation II score at admission was 22.7 ± 6.4, and the incidence of septic shock was 34.1%, with an overall ICU survival of 18%. Summarized data are presented in Table 1.

Conclusions
Patients with acute leukemia admitted to an ICU present an elevated risk of death, despite the progress in the care of critically ill patients achieved in recent years. The ICU admission of acute leukemia patients remains a controversial issue, based on conflicting data in the literature [1,2]. Because there is no survival prediction tool accurate enough to evaluate these patients, their admission to an ICU must be dependent on an individualized assessment of the current critical illness.   Intensive care historical development is always related to increasing high technological patterns and, as consequence, increasing financial costs. Economic and technological assessment of intensive care units has therefore become a very important analytical tools for efficient management of this kind of therapy.
The mainstream of the literature about economic and technological evaluation analysis focuses on technological innovation and, on the other hand, on cost containment. Our aim in the present article is to discuss the way in which ethical dimension should be incorporated to economic and technological evaluation analysis. In such a perspective the question is: How can ethical dimension be an endogeneous part of those evaluations, and a ground for decision making in intensive care procedure?
Our critical view over the mainstream of this specialized literature refuses both the analytical reduction from a wide view of technology to a simple absorption of technical innovations, and also the reduction from a extensive view of economy to a simple matter of cost containment. We think that the ethical dimension can integrate those evaluations analysis and render possible a solid management of the process of the intensive care unit. We are sure this process could not be reached by a nonintegrated evaluation analysis. So, in that perspective, ethics can be the link between investments, costs and technological innovation in intensive care therapies.

Introduction
In the past 5 years we have observed a progressive increase in nonmet demands of intensive care unit (ICU) beds. Most referrals to the ICU are emergencies or prebooked surgical cases. Some patients are refused admission because the units are full. The objective of this work is to identify the frequency of refused admission due to the lack of beds, the waiting time for admission, and the evolution of those patients.

Materials and methods
Data of all referrals consecutively made to the HURNP's ICU collected daily for a year (February 2002-February 2003) were collated. The referrals were categorized into clinical and surgical. The referrals criteria adopted was the hierarchy through a request order. The statistics was carried out through the EpiInfo program.
Results Throughout the observation period 1210 referrals to the ICU were made, 43.7% of patients being immediately admitted. Out of 681 referrals initially refused, 49.8% of the surgical cases were admitted and 312 clinical referrals were refused because of the lack of beds. The surgical referrals made were major elective surgeries that were prebooked ranging from 1 to 7 days. Of all clinical patients initially refused, 11.9% were cancelled due to patients' clinical recovery; of the remaining 275 patients, 62.9% were admitted, 37.1% were never admitted into the ICU, and 8.4% died before an available bed. Of clinical patients later admitted into the ICU, 25% had a waiting time longer than 1 day. Their age median was 63.5 years (45-72 years), 60.1% being from the Emergency Room and the rest from the wards. Most clinical patients (24%) were diagnosed as having severe sepsis, and 31.8% were in mechanical ventilation when the referral was made. The mortality of patients admitted into the Emergency Room waiting for a bed in the ICU was no higher than those of the ward (P = 0.37).

Conclusion
In this population, 56.3% of the demand for ICU beds was not admitted immediately. The waiting time for admission ranged from 1 to 10 days. Fifty-five patients stayed longer than 1 day on mechanical ventilation outside the ICU. The mortality among the referrals expected to be about 60% was smaller then the ICU overall mortality (31.85%). This suggests that probably some patients reported as a ward evolution actually died. We suggest a study to analyse the need of adult ICU beds for this population, and the development of triage criteria. Method A retrospective study was carried out, based on data collected from patient's records (those treated in the ICU of Israeli Hospital Albert Einstein). The study comprised 81 patients with dialytic ARF, treated from 1 January 1996 to 31 December 1998. Age, sex, origin, length and type of internment, associated chronic pathologies, intervals for dialysis indication, APACHE II and ATN-ISS scores and the hospital mortality were all analyzed.

PROGNOSIS
Outcome The prognostic scores APACHE II and ATN-ISS were carried out on the day of the realization of the first dialysis. The average age was 69, 14.72% were men and the mortality rate was 69%. In the survivors group there were 25 patients and 56 patients evolved for death internment. Both groups presented very similar characteristics, regarding the average prognostic scores and demographic data, differing only in time of hospital stay which was significantly longer in the survivors. Out of the evaluated prognostic scores, only the risk of death (APACHE II) presented a statistically significant difference among the survivors and nonsurvivors. The risk of death (APACHE II) and the ATN-ISS presented good discrimination. For the calibration, the risk of death underestimated the mortality in the lower tierces of seriousness significantly, while the ATN-ISS underestimated the mortality in all tierces of seriousness, although without significant difference.
Conclusion It is possible to use both scores in our ICU, however by its easy obtainment we have opted to use the ATN-ISS score. Background For identification of cardiac prognostic risk markers in the emergency room, in patients with ischemic heart syndrome without ST elevation, it is important to choose the best and most cost-effective therapeutic strategy.

P92 Prognostic risk markers at 180 days in patients with ischemic heart syndrome without ST elevation
Aim Evaluation of clinic, laboratorial and eletrocardiographic prognostic markers in nonselected patients with acute ischemic syndrome without ST elevation admitted to the emergency room.
Methods A prospective study took place from June 1998 to March 2000, with 124 patients with acute ischemic syndrome without ST elevation admitted to the emergency room of a tertiary hospital. Most patients were male (58%), with an age average of 68.9 ± 12.3 years; 62.9% have had previous coronary heart disease.

Results
Left ventricular heart failure was the most important prognostic risk factor for events, with a relative risk of 3.16 (95% confidence interval, 2.28-4.04). Troponin did not indicate risk, with a relative risk of 2.14 (95% confidence interval, 0.95-3.32).
Conclusion Left ventricular heart failure was the best risk marker of events in this population, which was older and had a higher incidence of previous coronary disease than the average. , and previous cardiovascular disease (n = 34). The significance level was 5% using analysis of variance, and a descriptive analysis was performed.

Results
The Acute Physiology and Chronic Health Evaluation mean score was 20 ± 5.28 (range 8-29). The mean age was 80 ± 7.43 (range, 65-96) and its influence on these patients' mortality was not significant (P = 0.31). The presence of the previous cardiovascular disease and/or cardiac failure (P = 0.001) and of the previous cardiovascular disease (P = 0.001) showed a relevant correlation with mortality, while the lack of the previous cardiovascular disease and/or acquired cardiac failure presented a correlation with ICU discharge (P = 0.013).

Conclusion
The age itself did not affect the mortality of the critically ill elderly population. The presence of cardiovascular disease seems to play an important role in the elderly patients' mortality. Objectives To identify first postoperative day (FPOD) markers for length of stay in the SICU using preoperative, perioperative, and FPOD variables. Results Based on the CART, 16 relevant variables were selected. The model had an accuracy of 69% for group A and 80% for groups B and C. Analyzing patients with length of stay in the SICU up to 7 days, the accuracy increased to 87%.

Conclusions
The CART may provide interesting solutions regarding patient allocation and also quality assessment. Variables Multivariate analyses demonstrated that BE at entrance was the best predictor of multiple organ dysfunction (coefficient β = -0.416, P = 0.017).
Conclusions Lactic acidosis is not the major factor responsible for MA in critically ill patients. BE is the best predictor of multiple organ dysfunction when compared with the other methods used, and is less demanding.

Conclusion
The APACHE II method showed an excellent discrimination and an inadequate calibration in this nontraumatic NICU population. A total 42.6% of patients were classified as 'other neurological/neurosurgical disease' as the main cause of admission, which could give room for improvement of the method. Objectives To assess the mortality and LOS in the POU of patients undergoing MR, who had MI on admission or more than 28 days before admission. Conclusion In this small case series no correlation between MI on admission and mortality or length of stay in the POU was observed, although a greater incidence of emergency MR was found in group A. Objective This study analyzes the influence of nutritional support on morbidity and mortality of critically ill patients.

Case series and methods
Methods Included in the study were all patients admitted to a 13-bed general intensive care unit (ICU) in the period from 1 June 2000 to 31 July 2001 who remained in the ICU for at least 7 days and who received at least 4 days of nutritional support (parenteral or enteral). Patients were classified into two groups according to calories received in the study period (7-10 days after admission): group A, patients who received at least 70% of their resting energy expenditure; and group B, patients who received less than 70%. Patients were also classified according to calories received on the third ICU day: group C, those patients who on the third day received 70% or more of the resting energy expenditure; and group D, those patients receiving less than 70%. We analyzed the length of stay in the ICU, mortality, incidence of nosocomial pneumonia and other infectious complications.

Results
The score created, shown in Table 1, provides the following risk prediction: 0-4, low risk; 5-8, medium risk; and 9-13, high risk. The results had significance (P < 0.0001) and a linear trend (P < 0.0001). The area under the receiver operating characteristic curve was 0.78.

Conclusions
This prognostic score shows the strength of first postoperative day variables, as does the Sepsis-related Organ Failure Assessment score [2], and the need for high doses of amines. The combined valvular surgery was the only perioperative marker. The 50-year cutoff point for age shows the precocity of valve disease in our country.  Background Readmission rates have been used as a measure of the results of the quality of care. It is proposed that a significant number of readmissions are potentially avoidable. The only legitimate basis for using precocious readmission as a quality indicator is that it demonstrates one relationship between readmission and the care process during the previous hospital stay. The interest in readmission is driven by the hypothesis that an improvement of care can result in a reduction of the readmissions to and in the costs of the intensive care unit (ICU).

Reference
Objective To identify risk factors for patient readmission for urinary tract infection.

Materials and methods
A cohort study, based on data of patients from the ICU. The internments of 1 January 1999-31 December 2000 for UTI were analyzed. We established two groups: the first group was composed of patients that just presented one admission, and the second group was composed of patients that presented two admissions to the ICU (they were analyzed for the first internment).

Results
In the study period 3034 patients were interned for UTI. In this population, the readmission rate for UTI was 10.7% and the rate of medium occupation was 86.76 ± 4.16%. The readmitted patients presented a medium age of 67.5 years (15.5 years), and for those patients with just a first admission the medium age was 63.2 years (17.2 years) (P < 0.0001); although age was divided into classes it also presented a significant difference. With relationship to sex there were 172 (60.6%) in the group with more than one admission and 1646 (61.9%) in the group with just one admission, without a significant difference. The indexes Acute Physiology and Chronic Health Evaluation II prognostics and Severe Acute Physiology Score II were, on average, for patients with one admission 12.0 (6.3) and 27.7 (11.9), respectively, and for readmitted patients 18.7 (6.4) and 33.1 (12.9), respectively; both with a significant difference. The hospital lengths of stay were 12.8 days (14.9 days) and 39.2 days (41.1 days), in the group with one admission and in the group that presented readmission, respectively; a fact also repeated in the ICU lengths of stay, both with a significant difference. In the multivariate analysis, the presence of creatinine A > 2.0 and systolic pressure < 90 mmHg, in the moment of internment, presented a significant difference.

Conclusion
The patients' precocious identification with risk factors and the care at the moment of discharge for UTI can be decisive to reduce the readmission rates. Background Several studies on prognostic scores (PS) in cardiac surgery (CS) and their comparisons have been published in the literature. However, these studies have been carried out with populations whose demographic characteristics and prevalence of pathologies differ from those found among ours. We know no study aiming at predicting length of stay (LOS) in surgical intensive care units (SICU).

P102 Comparison between prognostic scores of patients undergoing cardiac surgery
Objectives To compare the three following PS of inhospital mortality in patients undergoing CS and admitted to a public (A) and a private (B) SICU, analyzing preoperative, perioperative, and first postoperative day variables: EuroSCORE [1], pre-Cleveland [2], and post-Cleveland [3].

Results
The three PS had significantly different prediction of mortality and of LOS in the SICU (P < 0.0001). In predicting mortality and LOS in the SICU longer than 7 days, the three PS analyzed did not provide a good correlation (Nagelkerke's R 2 of 0.134 = 13.4% and of 0.226 = 22.6%, respectively). The less significant PS for prediction of mortality is the pre-Cleveland (P = 0.054) as compared with the EuroSCORE and the post-Cleveland (P < 0.0001). Comparing the receiver operating characteristic curves of the three PS for LOS in the SICU longer than 7 days and prediction of mortality, the following was observed, respectively: EuroSCORE, 0.575-0.745; pre-Cleveland, 0.550-0.769; and post-Cleveland, 0.769-0.769.

Conclusions
Prognostic scores are not intended to predict LOS in the SICU. With regard to prediction of mortality, the three receiver operating characteristic curves are similar, and the logistic regression is worse for pre-Cleveland. Although none of the PS analyzed seemed adequate to be used in this group of patients undergoing CS, post-Cleveland was the best among the three.

Introduction
The multiple organ dysfunction syndrome (MODS) is the main cause of mortality in intensive care units. The number of patients older than 65 years in such units has progressively been increasing. The aim of this study is to evaluate the frequency and evolution of the MODS in that population and its impact on mortality. Methods A prospective analysis of the data of all patients admitted to the ICU during the study period in order to classify all the first admission origins into four subgroups (ward, operating room, Emergency Department and specialized coronary and gastrohepatology ICU), and to evaluate the different patient needs of ICU resources in these different patient populations (see Table 1). Conclusions Compared with previously reported data, these data suggest that a large part of the available resources for intensive care in our hospital are devoted to the inhospital patient care. One suggested hypothesis is that this could result mainly from the lack of a subcritical care area. Objective To assess the variation of the Sequential Organ Failure Assessment (SOFA) index and the Multiple Organ Dysfunction Score (MODS) index as morbidity predictors after cardiac arrest.

Patients and methods
The indexes SOFA and MODS were calculated for 40 patients that suffered cardiac arrest in the intensive care unit; the indexes were calculated based on laboratory values and clinical data obtained 24 hours before and after cardiac arrest.
Results Forty patients, 17 (42.5%) female and 23 (57.5%) male, whose age varied from 17 to 84 years (mode 76 years). The causes of cardiac arrest were shock and metabolic disorders in 18 patients (45%), hypoxemia in 16 (40%), and myocardial ischaemia and poisoning by drugs in six patients (15%). The modality of arrest was asystolia in 17 patients (42.5%), pulseless electrical activity in 14 (35%) and ventricular fibrillation in nine patients (22.5%). The Acute Pysiology and Chronic Health Evaluation II score varied from 2 to 47 (mean 21, mode 11), and the mean risk of mortality was 32.54%. The previous SOFA score varied from 6 to 16 (mode 8) and that after arrest from 8 to 18 (mode 14). The MODS score varied from 3 to 16 (mode 7) and that after arrest from 5 to 21 (mode 12). In statistical analysis using the Wilcoxon test, the increase of the SOFA and MODS indexes after cardiac arrest was significant (z calc = 5.33 or P < 0.001), but in a nonparametric comparison between the two indexes we noticed that the proportional increase of each one in the same patient occurred only in 21% of the patients.

Conclusion
The SOFA and MODS indexes were separately demonstrated to be good predictors of major morbidity of patients after cardiac arrest, but they did not as correlate variables for the same situation.
Introduction 'Damage control' therapy has been efficient in the control of haemorrhagic situations, particularly in serious trauma; assisting in the temporary control of severe bleeding in situations in which hypothermia, acidosis and coagulation disturbances aggravate the immediate prognosis.
Case report A male, 55 years old, admitted to the emergency room wounded by a gunshot. Subdued exploratory laparotomy found hepatic injury, gastric injury and splenic blast. The surgical procedure was interrupted because of hypothermia (33.3°C) and acidosis (pH 7.20), temporary haemostasis carried out with compresses and haemostatic surgical instruments. The patient was admitted to the intensive care unit hypothermal, in metabolic acidemia, haemodynamically unstable, and needing vasoactive drugs and mechanical ventilation. After 8 hours, when the acid-basic balance, temperature and coagulation were normal, the patient underwent surgery for correction of the hepatic wound with epiplonplasty and peritoneostomy. When back at the intensive care unit, the patient was monitored with a pulmonary artery catheter, received large spectrum antibiotic therapy and reversal of the multiple organ dysfunction. After 14 days of internment, the patient was discharged in adequate clinical condition.
Conclusion 'Damage control' therapy has been demonstrated to be a promising therapy for temporary bleeding control under disturbance of coagulation in the presence of hypothermia, acidosis and prolonged hypotension in major surgical procedures. The intensive care unit must be ready for quick treatment of these disturbances, allowing the patient to undergo definitive surgery as soon as possible. Results With data obtained in the sample, a histogram of postoperative thoracic blood drainage was made. Adjusting through the likelihood ratio and distributing in the exponential form, its 95th percentile was determined, and the value of 0.97 ml/kg per hour was obtained in the postoperative period. When the BDI was correlated with the variables studied, greater drainage values were found in the following conditions: valve replacement associated with myocardial revascularization and surgery of the aorta (P = 0.0000 and P = 0.00002, respectively); patients with left atrium > 4.5 (P = 0.003); longer extracorporeal circulation (P = 0.0001); platelet count lower than 100,000 (P = 0.001); multiple organ dysfunction syndrome score greater than 4 (P = 0.00000); Sepsis-related Organ Failure Assessment score > 4 (P = 0.00000); longer length of stay in the ICU (P = 0.001); and a greater death index (P = 0.00000). No statistical difference between the results of the two hospitals was found.
Conclusions A normal BDI of 0.97 ml/kg per hour was established in our population in the postoperative period of CS, and a poorer outcome and longer length of stay in the ICU were observed in patients with drainage greater than the BDI found. The following clinical variables can predict greater BDI in the postoperative period of CS: the type of CS; left atrium > 4.5; prolonged extracorporeal circulation; platelet count < 100,000; and greater multiple organ dysfunction syndrome and Sepsis-related Organ Failure Assessment scores. Objective To assess the impact on hospital costs of the use of NA initiated in the first 12 hours in patients with a short length of stay in the ICU in the PO period of MR, and to compare other variables between the groups using or not using that drug.

Patients and methods
The use of NA initiated in the first 12 hours and the costs of hospitalization were studied in 268 adult patients undergoing MR and discharged from the ICU within the first 48 PO hours (94 patients received NA and 174 patients did not receive the drug). Other variables, such as fluid balance in the operation room and fluid balance in the first 24 hours (FBD1) of the PO period, extracorporeal circulation time (ECCt), mortality prediction score of the American Heart Association (mAHA), serum level of lactate in the postoperative period (first day), and postoperative multiple organ dysfunction syndrome and Sepsis-related Organ Failure Assessment (SOFA) scores, were also analyzed. The following statistical tests were used: Student t test, Wilcoxon text, rank sum text, and linear regression text.

Results
The use of NA was not an independent predictor of hospital costs in this group of patients. The ECCt and the preoperative mAHA score were cost predictors, but influenced only 9.2% of the variation. The comparisons between the groups of the mAHA score, the ECCt, the fluid balance in the operation room, and the serum level of lactate in the early PO period did not show any statistically significant difference. The group receiving NA had a significantly greater FBD1 (0.9 × 0.2 ml/kg per hour) with P < 0.00001. The MODS and SOFA scores were also significantly greater with P = 0.01 and P < 0.00001, respectively.

Conclusion
The early use of NA in the PO period of MR was not an independent predictor of cost in this group of patients who stayed in the ICU less than 48 hours. The preoperative variables were similar in the groups. The postoperative MODS and SOFA scores were greater in the group receiving the drug, but their values were impaired because the use of the drug was one of their components. The FBD1 was significantly greater in the group receiving the drug, which may be a marker of a different outcome, justifying further studies.
Introduction The percutaneous tracheotomy first described by Ciaglia in 1985 has been widely used since, showing it to be fast, safe and easy to perform.
In 1999 a new method arose, the Ciaglia Blue Rhino, which was done only by one person (hydrophilic dilatator). We compare the two procedures in our services.
Objective Registration of our experience with percutaneous tracheotomy, and a comparison between two procedures. Results A total of 565 arterial cannulations were analyzed (see Table 1).

Conclusion
Despite the lower utilization of the axillary artery, the number of complications favored this site for monitoring over the mostly used radial artery.

Materials and methods
The Vivid 3 vascular echocardiographic and ultrasound device (General Electric) equipped with a 10 MHz linear probe and pulsed color Doppler was used. Bovine thrombin (component of BERIPLAST ® P; Aventis) in a 100 U/ml solution was used. A color duplex scan of the lower limbs was performed for diagnostic confirmation of the PA, identification of the vessel related to the PA, measurements required by the procedure, and assessment of the arterial anatomy of both lower limbs. Ultrasound-guided puncture of the PA was performed, and contrast medium (agitated 0.9% saline solution) was injected to show the exact position of the needle tip inside the PA, which should be as far as possible from the PA neck, avoiding thrombin embolism. Then, the thrombin solution was slowly injected until closure of the PA (i.e. cessation of the systodiastolic flow through the PA neck), assessed on ultrasound. The amount of solution required for this was 0.5 ml, corresponding to 50 U thrombin. The duplex scan was repeated with the patient resting for 1 hour, and again on hospital discharge, 24 hours after the procedure, which confirmed the good result of the technique used.

Conclusion
The injection of a low dose of thrombin guided by ultrasound and aided by contrast medium use was effective in closing the PA. The hospitalization time was reduced even in patients undergoing anticoagulation. Objectives To report five cases of emergency thoracotomy in a prehospital setting and its indication with an algorithm.

Materials and methods
A case report of five emergency thoracotomies for penetrating thoracic trauma, in a prehospital environment, in an advanced life support unit with a physician.

Results
All patients were male with a median age of 24 years and presented by a left penetrating thoracic trauma: two were gunshots, two were vehicle debris and one was stabbed. All of them had a hemothorax, two of them with cardiac injury (only one tamponated). Internal cardiac massage was performed in all patients, with return of a spontaneous circulation (ROSC) in two of them. One patient died at the scene, one in the emergency room, two in the operation room and the other 6 hours after surgery. Three patients had signs of life before the procedure and two of them had ROSC.
Conclusion Emergency thoracotomy has a dismal result in a prehospital environment. Patients with signs of life before the procedure have the greatest chance of ROSC. An algorithm for rational use of emergency thoracotomy is proposed. Introduction Pulmonary Embolism (PE) continues to have a high mortality despite advances in diagnosis and therapy. We hereby present two patients with massive PE that underwent successful pulmonary thromboendarterectomy with embolectomy (PTE).
Case 1 A 56-year-old white male presented with a 10-day history of progressive dyspnea. Massive PE was diagnosed and the patient was started on anticoagulation. An inferior vena cava filter was placed because of extensive internal iliac thrombosis. Finally, PTE was undertaken because of recurrent hemodynamic instability in spite of thrombolytic therapy. In the postoperative period, the patient developed hemoptysis followed by status epilepticus and the appearance of petechiae on the legs. Serologies confirmed antiphospholipid antibody syndrome, which was managed with corticosteroids and immunoglobulin because of the aggressive presentation. After a long inhospital stay with several infectious, renal and hematological (bleeding) complications, the patient was discharged with no ventilatory assistance.
Case 2 A 46-year-old white, heavy smoker and obese male was admitted after a 3-day history of dyspnea. Initial examinations showed hypoxemia, a S 1 Q 3 pattern on ECG, chronic pulmonary hypertension and right ventricular dysfunction on echocardiogram. A chest computed tomography confirmed massive PE in both pulmonary arteries with calcification over the thrombi, leading to the diagnosis of an acute episode complicating chronic PE. An inferior vena cava filter was placed because the patient was considered to have a high risk of death after recurrent PE. Because of a worsening clinical condition despite adequate anticoagulation, the patient was submitted to PTE. He was also discharged after complete resolution of a nosocomial pneumonia.
Conclusion Nowadays surgical embolectomy is rarely performed. These two cases underwent this unusual form of therapy with a good outcome; however, it is certainly an alternative form of treatment for PE. Although whether considered a last resource to be reserved for desperate situations, some authors suggest it as one of several available treatments that could be also used for anatomically extensive PE without hemodynamic compromise. We evaluate the clinical value of in vitro labeling of red blood cells with technetium 99m (99mTc-RBC) scintigraphy for the detection of gastrointestinal (GI) bleeding sites. Ten patients referred with clinical evidence of GI bleeding and negative colonoscopy underwent 99mTc-RBC scintigraphy after endovenous administration of 925 MBq (25 mCi) 99mTc-RBC labeled by a simple technique, previously described by ourselves. Dynamic images of the abdomen were taken at 10 s intervals for 40 min. Then, 5 min images were obtained 1, 3 and 5 hours after cell administration. Delayed images up to 24 hours were obtained when early results were negative. All the patients with suspected GI bleeding were confirmed to have active hemorrhage up to 24 hours. The identification of bleeding sites was 40% (40 min), 20% (60 min) and 30% up to 24 hours. Of the nine patients with definite active hemorrhage, the bleeding sites were identified by surgery in all of them; and in the remaining patient, without active hemorrhage, the bleeding site was not identified by surgery. In conclusion, the simplicity, reproducibility and reliability of this technique of in vitro labeling of red blood cells, particularly when bleeding rates are low and intermittent, make it, in our point of view, the first line of investigation in any patient with suspected bleeding from the colon or upper GI tract if endoscopic evaluation is not possible in the latter. Cells release microparticles following apoptosis (apoptotic bodies) or for signaling purposes (exosomes). In contrast to the bigger apoptotic bodies (> 400 nm), exosomes (diameter 100 nm) do not present phosphatidylserine on their surface and expose major histocompatibility complex components, CD9 and CD63. Prognosis of some thrombotic-inflammatory diseases has been related to microparticle release. Sepsis is an abnormal immuno-inflammatory response to an infection, including dysregulation of apoptotic mechanisms in vascular cells. The major redox signaling pathway in vessels involves the enzymatic complex superoxide generating NADPH oxidase. In previous work we showed that, in sepsis, there is augmented platelet release of microparticles when compared with healthy controls. Those septic microparticles also possess greater NADPH oxidase activity, which can be responsible for vascular cell apoptosis. Our objective was to better characterize those microparticles obtained from septic patients and determine possible pathways related to their release. Through sequential filtration and centrifugation we separated microparticles from septic plasma (n = 16, 24 hours of diagnosis accordingly to ACCP/SCCM 1992 criteria) or from healthy controls (n = 6). Apoptotic bodies were obtained from the medium of cultured endothelial cells exposed to ultraviolet light for 30 min. Laser light scattering revealed particles with diameter between 82 and 112 nm. In contrast to the microparticles, apoptotic bodies do not have NADPH oxidase activity as assayed by lucigenin 5 µM luminescence. Western blot analysis revealed greater NADPH oxidase subunit expression in septic particles when compared with healthy controls, and none on apoptotic bodies. Flow cytometry disclosed positive phosphatidylserine exposure on apoptotic bodies, while microparticles were positive to CD9 and CD63. Washed, fresh platelets from single donors were stimulated with thrombin (0.1 U/ml), tumor necrosis factor alpha (10 µg/ml), lipopolysaccharide (LPS) (0.1 µg/ml) and with the nitric oxide (NO) donor sodium nitroprusside (2 mM) for analogous enzymatic complex is found within the membrane or cytosol but it has primarily signaling purposes. Both oxidases are inhibitable by thiol oxidants, not affected by the global redox cell state. Thiol oxidoreductases are effective modulators of the thiol redox state on cell membranes, in the ER and cytosol. Considering that spontaneous thiol-disulfide exchange reactions occur at rates too slow to be involved with regulatory pathways, we postulate that thiol oxidoreductases may control the redox state of important thiol sites on the NADPH oxidase, affecting its superoxide generating activity. Protein disulfide isomerase (PDI) is a ubiquitous multifunctional enzyme of the thiol oxidoreductase family, involved in regulation of diverse cellular mechanisms. Our objective is to investigate a possible interaction between PDI and the oxidases in vascular smooth muscle cells and neutrophils. Western blot analysis disclosed PDI in neutrophils within granules and the membrane, the same place where cytochrome b558 is found. In smooth muscle cells PDI was also found on the membrane fraction, and immunofluorescence disclosed a spatial colocalization between PDI and the oxidase. Superoxide generation was evaluated by superoxide dismutase-inhibitable cytochrome c reduction spectrophotometric assay using the cell free system (isolated membrane and cytosol from human neutrophils). Inhibition of PDI activity caused a 60% reduction of NADPH oxidase activity. With vascular smooth muscle cell homogenates, PDI inhibition caused similar oxidase activity. Angiotensin II, a known vascular NADPH oxidase agonist, induced in vascular cells a parallel increase in PDI activity.

BASIC
PDI modulation of thiol redox state on the vascular and phagocytic oxidases may thus represent a new regulatory mechanism of reactive oxygen species generation. Aim This study begins to characterize the expression patterns of the aforementioned genes in cultured MC.
Methods Primary MC were obtained from adult male Wistar rats. Cells were grown in 20% fetal bovine serum until confluence and then kept in serum-free medium for 24 hours. Total RNA was extracted from quiescent cells and cDNA synthesized using oligo-(d)T primers. Analysis of mRNA gene expression was performed through quantitative real-time polymerase chain reaction.

Results
Initially we confirmed the constitutive expression of Gremlin in rat MC. In addition, two other BMP antagonists, the head-inducing factor Cerberus and the tumor suppressor Dan, were also found at very similar expression levels. Quantitative realtime polymerase chain reaction analysis also demonstrated for the first time the presence of ID in quiescent rat MC. In fact, all four members of this family of genes (ID1-ID4) were expressed at relatively high levels in quiescent MC.

Conclusions
Our study demonstrates the expression of proproliferative and antiproliferative genes that play a significant role in TGF-β1/BMP-activated pathways in MC. The simultaneous, constitutive expression of these genes in adult, nontransformed rat MC suggests the presence of a novel autocrine loop, which may modulate MC proliferation. Additional studies are underway to characterize the functional role of BMP antagonists and ID-cultured rat MC. The elucidation of functional interaction between BMP antagonists and ID may enhance our knowledge about the molecular circuitry required for MC proliferation. Several studies have shown that endothelium-dependent vascular relaxation is altered in experimental diabetes and in diabetic patients, with peripheral vascular disease being almost twice as frequent in diabetic women as compared with diabetic men. Moreover, the mesenteric microvessels exhibit impaired responses to acetylcholine, bradykinin, histamine, and platelet activator factor in diabetic female rats. The aim of the present study was to investigate the effects of inhibition of cyclooxygenase upon the reduced response to endotheliumdependent vasodilator agents in diabetic female rats, to verify the possible involvement of prostaglandins in that alteration. The changes of arteriolar and venular diameter after topical application of acetylcholine (17 nmol), bradykinin (30 pmol), and histamine (2.7 nmol) were measured in vivo by means of a closed video circuit coupled to a microscope before and after acute treatment with diclofenac, a cyclooxygenase inhibitor (2.5 mg/kg, intramuscularly). The cyclooxygenase inhibition corrected the decreased response of arterioles and venules to bradykinin, but only corrected the decreased venular response to acetylcholine. Diabetes impaired the response to histamine in arterioles but not in venules, and that response reduction was also corrected by diclofenac (see Table 1). These data suggest that increased Table 1 Arterioles Venules Aim The purpose of the present study is to characterize the expression patterns of BMP-antag in VSMC, and their regulation by well-known growth factors.

P123 Expression and regulation of bone morphogenetic protein antagonists in vascular smooth muscle cells TT Maciel, N Schor, AH Campos
wall shear rate using an optical Doppler velocimeter, the number of circulating leukocytes, L-selectin or CD11/CD18 integrin expression on granulocytes by flow cytometry, and the intracellular adhesion molecule-1 (ICAM-1) or P-selectin expression in endothelial cells by imunohistochemistry. The left carotid artery of each anesthetized SHR was catheterized and the mean arterial blood pressure (MAP) was measured. The tail-cuff blood pressure (TBP) was measured in unanesthetized SHR.
Results E and L reduced TBP and MAP, whereas D increased the TBP and MAP levels. The association of D did not interfere with the TBP and MAP lowering effect of E and L. The number of rollers was reduced by D, by L and by D + L. E and D + E did not modify the number of rollers. The number of adherent and migrated leukocytes was reduced by all treatments. Neither treatment increased the venular wall shear rate or modified the venular diameter, circulating leukocytes and L-selectin or CD11/CD18 integrin expression. All treatments decreased ICAM-1 expression. P-selectin expression was reduced by D, by L and by D + L, but E and D + E did not modify the P-selectin expression (Table 1).

Conclusion
Our data allow us to suggest that the association of D did not interfere with the antihypertensive effect of E and L, but E interfered with the anti-inflammatory effect of D on leukocyte rolling and P-selectin expression. The reduction in ICAM-1 expression is involved and might explain the reduction of the number of adherent and migrated leukocytes observed.
Aim The present study analyzes the effect of the triterpenoid tormentic acid (TA, from the plant Rubus sieboldii), a DNA polymerase inhibitor with anti-inflammatory properties, on cultured VSMC.
Methods Subconfluent cultures of embryonic rat aortic VSMC (A7r5) were exposed to increasing concentrations of TA or vehicle (DMSO) for 24 h, in the absence or presence of 10% fetal bovine serum (FBS), and apoptosis rates were evaluated through chromatin morphology analysis following DNA staining with the fluorescent dye HOE33342. In a separate series of experiment, A7r5 cells were cultured in 10% FBS continuously incubated with TA or vehicle. Cells were electronically counted (Coulter counter) at 2-day intervals.

Conclusions
Our data indicate that TA is a VSMC apoptosis inducer and proliferation inhibitor. The absence of a pro-apoptotic effect associated with an anti-mitotic action in the presence of serum suggests that TA may be useful in preventing proliferative vascular diseases without negatively affecting normal vasculature.
In vivo models of vascular lesion are currently being employed in order to investigate this hypothesis.