Abstract Body

Cotrimoxazole (CTX) is recommended for all children recently diagnosed with HIV in order to prevent pneumocystis pneumonia. Although it improves survival in HIV-infected children, the impact of CTX on mortality in those without severe immune deficiency is unclear. We evaluated temporal trends in CTX use at antiretroviral therapy (ART) initiation and mortality rates while on CTX in the absence of severe immune deficiency.

Pediatric data collected between 1 Jan 2006 and 31 Mar 2016 in the International Epidemiology Databases to Evaluate AIDS global cohort consortium were analyzed from regions that routinely report information on CTX use (East, West, Central, and Southern Africa, Asia-Pacific). All 24 countries included are considered low- or middle-income by World Bank criteria. Severe immune deficiency was defined by the age-specific CD4 thresholds in the 2006 WHO ART guidelines. Logistic regression was used to evaluate factors associated with using CTX at ART initiation. Competing risk regression was used to assess factors associated with mortality. Sensitivity analyses used multiple imputation to account for missing data.

A total of 54,113 children aged 1 month to 15 years were included. Median age at ART initiation was 5.7 years and 51% were female. CTX use increased from 67% in 2006 to a peak of 86% in 2010 and then declined to 49% in 2015/16. The proportion of children with severe immune deficiency dropped from 70% in 2006 to 43% in 2015/16 (figure). In our adjusted analysis, age (odds ratio [OR] 1.2 for <1 year vs. 1 to <5 years, 95% confidence interval [CI] 1.1-1.3), anemia (OR 1.1, 95%CI 1.0-1.2), severe immune deficiency (OR 1.3, 95%CI 1.2-1.3), height-for-age z-score (OR 1.2 for <-3 vs. >-2, 95%CI 1.1-1.2), year of ART initiation (OR 2.8 for 2010 vs. 2006, 95%CI 2.5-3.1 and OR 0.7 for 2015/16 vs. 2006, 95%CI 0.6-0.8) and region (OR 9.0 for East vs. Southern Africa, 95%CI 8.2-9.9) were associated with CTX use. The rate of death in children without severe immune deficiency was 0.5 per 100 person-years and did not differ with CTX use (adjusted hazard ratio 1.0 for non-users versus users, 95%CI 0.8-1.3). Sensitivity analyses yielded similar results.

Recent declines in pediatric CTX use coincide with declines in the proportion of children starting ART with severe immune deficiency. CTX did not improve survival among children without severe immune deficiency although use was not randomized and unmeasured confounders cannot be excluded.