Paradoxical Reaction to Tuberculosis Therapy among HIV-Negative Children: A Systematic Review and Meta-Analysis

Tuberculosis treatment is sometimes associated with clinical deterioration, referred to as paradoxical reaction (PR), especially in those with HIV coinfection. However, the burden and characteristics of PR in children without HIV coinfection are unclear. We undertook a systematic review to estimate the burden and clinical characteristics of PR in HIV-negative children. We searched PubMed, Embase, Web of Science, CINAHL, Scopus, Cochrane Library, ProQuest, and OpenGrey for studies reporting PR in HIV-negative children (<18 years old). We included observational studies including case series with at least five cases. Data on incidence/prevalence, clinical features, risk factors, management strategies, and outcome of PR were extracted. Risk of bias in the included studies was assessed using the NIH's quality assessment and Joanna Briggs Institute critical appraisal tools. We pooled the prevalence data using random effects meta-analysis. We identified 1,673 studies, of which 10 were eligible for inclusion. They described PR in 133 HIV-negative children. The pooled prevalence was 8.8% (95% CI: 2.9%, 14.6%). Owing to heterogeneity among studies, risk factors for the development of PR could not be identified. Limited data suggested that children developing PR were younger and had neurological or lymph node tuberculosis more often. Most children were treated with corticosteroids, but data were insufficient to identify the optimal management strategy. The review showed that PR affects nearly 1 in 12 HIV-negative children receiving tuberculosis treatment. The paucity of studies emphasizes the need for surveillance/studies to better characterize clinical features, risk factors, appropriate management strategies, and outcome.