Objectives: Identifying factors associated with mortality among acutely ill HIV-infected children presenting with advanced HIV disease may help clinicians optimize care for those at highest risk of death.
Design: Using data from a randomized controlled trial (NCT02063880), we determined baseline sociodemographic, clinical, and laboratory cofactors of mortality among HIV-infected children in Kenya.
Methods: We enrolled hospitalized, HIV-infected, antiretroviral therapy-naive children (0-12 years), initiated antiretroviral therapy, and followed up them for 6 months. We used Cox proportional hazards regression to estimate hazard ratios (HRs) for death and 95% confidence intervals (CIs).
Results: Of 181 enrolled children, 39 (22%) died. Common diagnoses at death were pneumonia or suspected pulmonary tuberculosis [23 (59%)] and gastroenteritis [7 (18%)]. Factors associated with mortality in univariate analysis included age <2 years [HR 3.08 (95% CI: 1.50 to 6.33)], orphaned or vulnerable child (OVC) [HR 2.05 (95% CI: 1.09 to 3.84)], weight-for-age Z score <-2 [HR 2.29 (95% CI: 1.05 to 5.00)], diagnosis of pneumonia with hypoxia [HR 5.25 (95% CI: 2.00 to 13.84)], oral thrush [HR 2.17 (95% CI: 1.15 to 4.09)], persistent diarrhea [HR 3.81 (95% CI: 1.89 to 7.69)], and higher log10 HIV-1 viral load [HR 2.16 (95% CI: 1.35 to 3.46)] (all P < 0.05). In multivariable analysis, age <2 years and OVC status remained significantly associated with mortality.
Conclusions: Young age and OVC status independently predicted mortality. Hypoxic pneumonia, oral thrush, and persistent diarrhea are important clinical features that predict mortality. Strategies to enhance early diagnosis in children and improve hospital management of critically ill HIV-infected children are needed.