Are there clinical and laboratory predictors of 5-year mortality in HIV-infected children and adolescents with hemophilia?

To determine factors associated with survival in a cohort of HIV-infected children and adolescents with hemophilia, an analysis of the 5-year mortality data for 207 HIV-infected young men was performed to examine the effect of selected clinical covariates on survival. The subjects were enrolled into the Hemophilia Growth and Development Study cohort from 1989 to 1990. Estimated mean time since infection at baseline was 6.7 years and mean estimated age at infection was 6.5 years. The baseline characteristics examined for their association with the hazard of death over the 5-year follow-up period were the following: absolute CD4+ cell count, hemoglobin status, skin test anergy, results of brain magnetic resonance imaging, non-hemophilia-related muscle atrophy (NHRMA), height for age, and impaired neuropsychological functioning as measured by the Vineland Adaptive Behavior and the Pediatric Behavior Scales. In all, 66 deaths occurred over the 5-year follow-up, 62 of whom met the 1987 (n = 56) or 1993 (n = 6) U.S. Centers for Disease Control and Prevention (CDC) definition of AIDS. Although each of the characteristics listed previously significantly increased the hazard of death by Cox proportional hazard regression models, only NHRMA remained a significant predictor of AIDS-related death when added to models that included each of the other cited baseline covariates.