Mortalidade e fatores prognósticos em pacientes HIV positivo internados em Unidade de Terapia Intensiva de hospital especializado em doenças infecto-parasitárias, Belo Horizonte, Minas Gerais, Brasil

Abstract

Introduction: Highly active antiretroviral therapy for HIV infection has produced significant decline in morbidity and mortality from AIDS. The outcome of HIV-infected patients requiring critical care has also improved during the course of the AIDS epidemic. However, in developing countries, insufficient data are available about HIV-infectedpatients admitted in ICUs. Objectives: To evaluate the prognosis of HIV-infected patients admitted to ICU and to indentify factors predictive of short- and long-term survival. Two disease severity scores (APACHE II and SAPS II) applied to those patients were also compared. Methods: A retrospective study from January 1, 2006, to December 31, 2006,including all consecutive HIV-infected patients admitted to the ICU of a HIV/AIDS referral center for the first time was conducted. ICU survivors were followed up to 24 months after ICU discharge. Demographic, clinical and laboratory data; disease severity scores and mortality were evaluated. Data were analyzed using survival and regressionmodels. Results: One hundred twenty-five HIV-infected patients were studied. Respiratory failure was the main cause of admission (43,2%), followed by neurological disorders (24,8%), severe sepsis (20%), heart failure (1,6%) and miscellaneous disorders (10,4%). In-ICU and in-hospital mortality rates were 46,4 and 68%, respectively. Multivariateanalysis showed that the in-ICU mortality was significantly associated with APACHE II and SAPS II scores, need for mechanical ventilation, PCP or tuberculosis treatment, use of antiretroviral therapy in ICU, need for inotropic support, need for inotropic support during the first 24 hours and septic shock in ICU. In-hospital mortality was significantlyassociated, in the multivariate analysis, with APACHE II and SAPS II scores, need for inotropic support, septic shock in ICU, serum albumin level, AIDS-associated admission diagnosis and time on hospital wards before ICU. The mortality after 24 months from ICU discharge was independently associated with time on hospital wards before ICU, severesepsis during admission to ICU, septic shock in ICU, inotropic support, length of signs and symptoms before admission to hospital and time of HIV diagnosis. The two severity indexes, APACHE II and SAPS II, were moderately accurate to assess the short-term prognosis. Observed ICU mortality was over-predicted by SAPS II and under-predicted byAPACHE II. Conclusions: The ICU and in-hospital mortality rates of HIV-infected patients who require intensive care in our study were higher when compared with developed countries. Most factors associated with poor outcome in HIV-infected patients admitted to the ICU we have found are in agreement with the literature. The two severity indexes were independently associated with prognosis; however the accuracy to predictoutcome was just moderate. The high mortality reported in this study can be in part explained by clinical characteristics of our population. Most of our patients were admitted by AIDS-related diagnosis, had higher severity indexes, lower albumin levels and were more frequently admitted by respiratory failure and sepsis.