Abstract

The aim of this study was to identify mortality rate, factors influencing mortality rate and high risk patients of respiratory intensive care unit (RICU). This prospective study was carried out at the respiratory ICU, department of chest diseases, the university hospital, Turkey, between 1999 and 2002. The following patients were excluded from the study: patients aged <18 years, those who died within the first 24 hours of ICU admission, and also patients presenting with trauma, malignancy, cardiac surgery and burn cases. 262 patients were included our study. Our patients were divided into seven cathegories: COPD patients with acute respiratory failure (202 patients), severe pneumonia (56 pts), congestive heart failure (115 pts), ARDS (38 pts), massive pulmonary thromboembolism (6 pts), postoperative respiratory failures (12 pts). Laboratory and physiologic parameters were recorded for all ICU patients based on the first 24 hours after ICU admission. Outcome of the patients were also recorded. Multivariate regression analysis was used for determination of factors affecting mortality. To determine the efficiency of APACHE II, ROC analysis was used. Mean age was 63.5±12.7 years. The mean length of ICU stay was 9.6 days. Rude mortality rate was calculated as 27.1%. The majority of our patients suffered from acute respiratory failure with chronic obstructive pulmonary diseases patients (70%). The highest mortality rates were patients with severe pneumonia and CHF (35.7%). Median APACHE II score for our patients was 21. Accordingly, predicted mortality rate was 29.7%. On the univariate analysis, 15 parameters were significantly effect on mortality rate. High APACHE II score (above 21), arrhythmias, need for mechanical ventilation, hypotension, high LDH level (>876 IU), length of ICU stay, need for cardiotonic medication, any complication, high BUN level and hypoproteinemia were identified as factors influencing mortality rate on multivariate regression analysis. In conclusion, patients with four or more of ten parameters having an independent effect on mortality rate could be determined to be those at risk upon RICU admission. Its sensitivity, specificity and area under the ROC curve were 83%, 89%, and 0.887, respectively. These were much higher than the APACHE II score itself.