Mortality descriptors in HIV inpatients

Objective To assess the clinical-epidemiological descriptors of inpatient mortality in persons infected with the human immunodeficiency virus (HIV). Method All adult HIV/AIDS patients hospitalized at a university hospital in 1990, 1992 and 1994 were studied retrospectively Descriptive statistics for all variables of inter-est were generated. Chi-square test and Fisher's exact test were performed to compare categorical variables. Means were compared using the Student t test. Stepwise logistic regression was used to identify the odds of dying associated with each risk factor. Results Two hundred and forty patients were included in the study. Between 1990 and 1994 the mean age rose from 35 to 36.9 years, the male/female ratio decreased from 9.8 to 2.0, the non-white proportion increased from 18.5 to 41.3% and the mean time between detection of HIV infection and hospitalization rose from 0.7 to 2.5 years. In addition, length of stay dropped from 31.3 to 25.3 days and the proportion of patients being followed up in the outpatient clinic of the Universitary Hospital increased from 47.8 to 83.3%. Respiratory infections were the main cause of hospitalization (58.0%). Oral candidiasis (27.1%), tuberculosis (18.3%), Pneumocystis carinii pneumonia (15.4%) and toxoplasmic encephalitis (10.4%) were the most frequent opportunistic infections. Multivariate analysis showed that the factors associated with a worse outcome included the length of stay less than or equal to 7 days (Odds Ratio [OR]=3.88; p=0.02) and no outpatient follow-up at the Hospital (OR=3.29; p=0.01). Conclusion Identification of independent risk factors for death may help in the implementation of more efficient interventions directed towards inpatients with HIV/AIDS.