Predictors of Short- and Long-term Survival in HIV-Infected Patients Admitted to the ICU: Outcome

Predictors of Short- and Long-term Survival in HIV-Infected Patients Admitted to the ICU: OutcomeWe found that HIV-related variables significantly influenced the in-hospital outcome and to a lesser extent, the in-ICU outcome. These variables were also closely associated with the long-term outcome. The number of previous opportunistic infections, as well as the stage and duration of AIDS, were significantly associated with the short- and long-term outcomes. The CD4 count is the prognostic marker most widely used in HIV-infected patients. However, as shown in Table 3 and Figure 2, differences in median and mean survival times across CD4+ lymphocyte count groups were modest. We believe that variations in CD4+ counts probably have little prognostic value in patients with CD4 counts <0.100X109/L. It has been reported that only logarithmic values are of interest in this population. HIV disease is a chronic disease that remains ultimately fatal, with a life expectancy that varies according to a number of clinical and laboratory marker 25,26 Clearly, the long-term outcome is closely dependent on this specific life expectancy. Bronchial Disease

The ICU admission cause group was a significant individual predictor of short- and long-term outcomes. Although the differences across admission cause groups may be ascribable to category assignment bias, another possibility is that they reflect true differences in the prognosis according to the type of organ failure that led to ICU admission. We believe that the cause for ICU admission can be used to predict the short-term outcome. In-ICU and inhospital mortality rates were clearly higher in the neurologic, cardiac, and severe sepsis groups, and as shown in Figure 1, survival curves in these groups fell more rapidly immediately after ICU discharge. After 3 years, however, cumulative survival rates were comparable across groups and survival curves intersected, making interpretation of the data extremely difficult. Nevertheless, admission to the ICU for a respiratory cause was associated with better short- and long-term survivals. These data corroborate recent studies and widely held beliefs.
The value of severity scores for predicting the short-term outcome has been firmly established. However, their predictive value in individual patients and their usefulness for decision making re main controversial. Furthermore, few data are available on their relevance for predicting the longterm outcome. Their accuracy in specific patient subgroups is also debated. It is therefore not surprising that the usefulness of severity scores in HIV-infected patients has been challenged. We used the SAPS I to evaluate the degree of physiologic disturbance caused by an acute intercurrent event, and we believe that this parameter is useful for evaluating HIV-infected patients admitted to the ICU. Our study provides evidence supporting the value of the SAPS I for predicting short- and long-term outcomes in HIV-infected patients.


So Many Advances in Medicine, So Many Yet to Come