Abstract

BACKGROUND: Previous evidence synthesis has suggested diabetes mellitus (DM) worsens tuberculosis (TB) treatment outcomes. However, these reviews are limited by the number, robustness and conflicting results among the studies included. We conducted a systematic review to update earlier analyses and explore heterogeneity among studies. METHODS: MEDLINE, EMBASE, AIM, LILACS, IMEMR, IMSEAR and WPRIM were searched between 1 January 1980 and 23 July 2018 unrestricted by language or region. All cohort and case-control studies investigating the difference in TB treatment outcomes amongst TB-DM patients compared to those with TB alone were included. Two reviewers independently assessed titles, abstracts, and extracted data. Culture conversion at two/three months, all-cause mortality, treatment failure, relapse and multidrug-resistant TB (MDR-TB) were evaluated using random effects meta-analysis with generic inverse variance. Heterogeneity was explored using subgroup analyses and meta-regression. RESULTS: One hundred and four publications were identified. Sixty-four studies including 56 122 individuals with TB-DM and 243 035 with TB, reported on death. Some outcomes showed substantial heterogeneity between studies, which we could not fully explain, though confounding adjustment and country income level accounted for some of the differences. TB-DM patients had higher odds of death (OR 1.88, 95%CI 1.59–2.21) and relapse (OR 1.64, 95%CI 1.29–2.08) compared to TB patients. More limited evidence suggested TB-DM patients had double the risk of developing MDR-TB (OR 1.98, 95%CI 1.51–2.60). CONCLUSION: DM is associated with increased risks of poor TB treatment outcomes, particularly mortality, and may increase risk of developing primary MDR-TB. Cost-effectiveness of interventions to enhance TB-DM treatment should be assessed.