[Prevalence of tuberculosis in incarcerated populations: systematic review and meta-analysisPrevalencia de tuberculosis en la población privada de libertad: revisión sistemática y metanálisis].

Objective: To estimate the prevalence of tuberculosis among incarcerated populations.

Method: A systematic review with meta-analysis was performed. The MEDLINE/PubMed, SciELO, and LILACS databases were searched for articles published from January 1997 to December 2016. The following search terms were used: ("tuberculosis"[MeSH] OR "tuberculosis"[TIAB]) AND ("prisons"[MeSH] OR "prisons"[TIAB]); or ("tuberculose" [DeCS] OU "tuberculose" [palavras] E "prisões" [DeCS] OU "prisões" [palavras]). The primary outcome was the prevalence of tuberculosis with 95% confidence interval (95%CI). In the metanalysis, variables associated with tuberculosis prevalence in incarcerated populations in the univariate analysis ( 0,20) were included in the final multivariate model.

Results: Based on the 29 studies included in the metanalysis, 2,163 prisoners with tuberculosis were identified. The combined prevalence of tuberculosis among prisoners was 2% (95%CI: 0.02-0.02). The prevalence among prisoners from countries with tuberculosis prevalence of 0-24 per 100,000 general population was below 1% (95%CI = 0.00-0.00). In countries with overall tuberculosis prevalence of 25-99/100,000, the estimate among prisoners was 3% (95%CI = 0.02-0.04); and in countries with overall prevalence ≥ 300/100 thousand, the estimated prevalence among prisoners was 8% (95%CI = 0.05-0.11).

Conclusion: The present results support the notion of a high prevalence of tuberculosis among incarcerated populations worldwide. The results also show a link between the prevalence of tuberculosis in the general and the prevalence of tuberculosis in prisons.