Abstract
BACKGROUND: Isoniazid is recommended for prevention of tuberculosis (TB) in HIV-infected adults, but its efficacy in children living with HIV (CLHIV) is not known. We performed a systematic review to assess the efficacy of isoniazid for the prevention of TB in CLHIV. METHODS: We searched PubMed, Cochrane Clinical Trial Registry and Google Scholar from inception to December 2016. Any randomized controlled trial assessing the role of isoniazid for the prevention of TB in CLHIV was eligible for inclusion. The primary endpoint was TB incidence; secondary end points were mortality, overall survival and severe adverse events. Dual independent extraction of all data was performed. Data were pooled under a random effects model and summarized either as risk ratio (RR) or hazard ratio along with 95% confidence intervals (CIs). RESULTS: Of 931 references, 3 randomized controlled trials enrolling 977 patients met the inclusion criteria. Pooled results showed a statistically nonsignificant reduction in TB incidence (RR: 0.70; 95% CI: 0.47-1.04; P = 0.07) and mortality (RR: 0.94; 95% CI: 0.39-2.23; P = 0.88) with the use of isoniazid compared with placebo. One study was stopped early because of excess deaths in the placebo arm. However, results from subgroup analysis restricted to only completed trials did not change the overall findings. CONCLUSIONS: Isoniazid did not reduce the incidence of TB in CLHIV. All included studies were performed in regions with high prevalence of TB making the overall generalizability limited.