Predictors of Short- and Long-term Survival in HIV-Infected Patients Admitted to the ICU: Conclusion
A need for mechanical ventilation and the occurrence of renal failure requiring hemodialysis are of adverse significance for the short-term prognosis in ICU patients not infected with the HIV In our study, short-term survival was closely correlated with use of mechanical ventilation or hemodialysis. The duration of mechanical ventilation has been reported as an important prognostic marker in non-HIV-infected patients. Our multivariate analysis found that the duration of mechanical ventilation was significantly associated with a poor short-term outcome; in-ICU mortality rises from 23.7 to 67.6%, and in-hospital mortality goes from 54.2 to 76.2%.
We conclude that intensive care for HIV-infected patients should not be considered futile and that the appropriateness of intensive care for HIV-infected patients should be discussed on the same basis as for the general population. Our results strongly suggest that the prognosis in HIV-infected patients admitted to the ICU is dependent on the following: health status prior to ICU admission, as evaluated by a functional status score and by the percentage of weight lost; a number of HIV-related variables, including time since AIDS diagnosis, number of previous opportunistic infections, CD4 count, and the HIV disease stage; the severity of the acute illness, based on the cause of ICU admission and the SAPS I; and the need for and duration of mechanical ventilation while in the ICU. Whereas previous health status, acute illness severity, and duration of mechanical ventilation may be the most reliable prognostic markers for in-ICU mortality, in-hospital and long-term outcomes may be closely dependent on some of these parameters and, above all, on life expectancy as assessed based on HIV-related variables. Thus, the natural history of HIV disease may ultimately be the most powerful determinant of the long-term prognosis. Recent advances in antiretroviral treatments and the resulting improvements in life expectancy suggest that long-term outcomes in the near future may be better than those reported herein in HIV-infected patients discharged from an ICU.