RESUMEN
Objectives: To study the outcome of acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) in medical ICU, Ramathibodi Hospital and identify factors that contribute independently to mortality. Study design: Retrospective study. Methods: All patients admitted in medical ICU between January 1998 and December 2002, that met the criteria of the American-European Consensus Conference for ALI and ARDS were reviewed. The data collection included patient baseline characteristics, risk factors for ARDS, initial PaO₂/FiO₂, PaO₂/PAO₂, static lung compliance, best PEEP level, APACHE II score, associated early and late nonpulmonary organ dysfunction, mode of ventilator, ventilator days, outcomes and complications. The probability of death and median survival time were assessed by Kaplan-Meier method. Prognostic factors associated with mortality were determined by Cox Proportional Hazard method. Results: A total of 48 patients met the criteria of ALI and ARDS. The mean age of these patients was 46.8 \±18 yrs. Direct lung injuries were the most common causes of ARDS in this series (35/48), of which pneumonia attributed to the majority of cases (80%). The mean APACHE II score of the group was 20.9\±7.4, with 70.8% hospital mortality. Main cause of death was multiple organ dysfunction, while refractory hypoxemia was less common. Factors independently associated with mortality were initial APACHE II score of more than 20 (hazard ratio, 2.09; 95%CI 1.02 to 4.32) and the presence of circulatory dysfunction 24 h after the onset of ARDS (hazard ratio, 5.78; 95%CI 2.11 to 15.86). Conclusion: Mortality rate of ARDS in medical patients had been unchanged. The extreme high mortality (70.8%) in this group could be due to the high proportion of patients with pneumonia and sepsis. Only initial APACHE II score of more than 20 and the presence of circulatory dysfunction were found to be the independent predictors of mortality. These further confirmed and emphasized the concept of \“lung as a part of systemic inflammatory process\” in ARDS.
RESUMEN
Background: The sensitivity of sputum and bronchoalveolar lavage (BAL) for diagnosis of pulmonary tuberculosis (PTB) is low. Objective: To evaluate the value of sputum obtained post-bronchoscopically in the detection of AFB (smear and culture), with particular comparison to the standard pre-bronchoscopic culture, BAL and transbronchial biopsy (TBB). Method: Thirty-five active PTB patients (diagnosed clinically and radiologically) with 3 negative consecutive sputum AFB smears were recruited. Some of the sputums were also cultured. All patients underwent fiberoptic bronchoscopic examinations (FOB) with BAL and TBB performed from the indicated segments. The BAL specimens were processed for AFB stains and C/S. Post-bronchoscopic cultures patients were asked to expectorate 3 consecutive sputums for examinations (smear and C/S). Final diagnosis of PTB was defined as a recovery of AFB (smear ± C/S) at any step of the diagnostic procedures. In cases with negative AFB recovery, active PTB was diagnosed only if definite clinical plus radiologic improvement after chemotherapy was evidenced. Results: Of the thirty-five patients participated in this study, PTB was diagnosed in eighteen patients. The addition of post-FOB sputum smear alone did not increase the sensitivity over group 2 (p=0.11). The sensitivity of group 3 (72%) was higher than that of group 1 (30%, p=0.03) and that of group 2 (61%, p=0.016). No statistically difference was seen between group 1 and 2 (p=0.28). Conclusion: The addition of post-bronchoscopic sputum culture to BAL examination and TBB increased the sensitivity of PTB detection in sputum smear-negative PTB patients.