AIDS in a medical intensive care unit: immediate prognosis and long-term survival.

OBJECTIVE: To help physicians decide whether to admit patients with acquired immunodeficiency syndrome (AIDS) to the medical intensive care unit (MICU).
DESIGN: Case series study of AIDS patients admitted to the MICU between October 1990 and October 1992 and followed up until April 1993 (median follow-up, 1 year).
SETTING: The MICU in a 970-bed teaching hospital in Paris, France. PATIENTS: A total of 120 consecutive AIDS patients with acute respiratory failure (50%), central nervous system dysfunction (22.5%), pneumothorax (12.5%), shock (10.8%), or miscellaneous conditions (4.2%). A total of 86 patients were discharged alive from the MICU. MAIN OUTCOME MEASURES: Predictive factors for mortality during and after MICU stay.
RESULTS: Multivariate analysis identified 3 factors predicting poor MICU outcome: Simplified Acute Physiology Score I (SAPS I) above 10 (relative risk [RR], 6.1; 95% confidence interval [CI], 1.5-26.6), time between AIDS diagnosis and MICU admission more than 1 year(RR, 6.0; 95% CI, 2.1-17.5), serum albumin level less than 30 g/L (RR, 4.9; 95% CI, 1.3-18.2). The CD4 cell count, beta2-microglobulinemia, and previous opportunistic infections had no influence on MICU mortality. After MICU discharge, survival rates were 86% at 1 week, 82% at 1 month, 53% at 6 months, and 39% at 1 year. The Karnofsky scale score and the number of previous opportunistic infections were simultaneously associated with post-MICU outcome. Predictive factors for MICU survival did not influence post-MICU survival.
CONCLUSION: The MICU mortality was related to immediate severity (assessed within 48 hours of admission) and the time between AIDS diagnosis and MICU admission. Long-term survival after MICU discharge depended only on the severity of AIDS. We conclude that AIDS patients should be admitted to the MICU on the same basis as other patients.