Incidence of HIV-Associated Tuberculosis among Individuals Taking Combination Antiretroviral Therapy: A Systematic Review and Meta-Analysis

Background: Knowledge of tuberculosis incidence and associated factors is required for the development and evaluation of strategies to reduce the burden of HIV-associated tuberculosis. Methods: Systematic literature review and meta-analysis of tuberculosis incidence rates among HIV-infected individuals taking combination antiretroviral therapy. Results: From PubMed, EMBASE and Global Index Medicus databases, 42 papers describing 43 cohorts (32 from high/intermediate and 11 from low tuberculosis burden settings) were included in the qualitative review and 33 in the quantitative review. Cohorts from high/intermediate burden settings were smaller in size, had lower median CD4 cell counts at study entry and fewer person-years of follow up. Tuberculosis incidence rates were higher in studies from Sub-Saharan Africa and from World Bank low/middle income countries. Tuberculosis incidence rates decreased with increasing CD4 count at study entry and duration on combination antiretroviral therapy. Summary estimates of tuberculosis incidence among individuals on combination antiretroviral therapy were higher for cohorts from high/intermediate burden settings compared to those from the low tuberculosis burden settings (4.17 per 100 person-years [95% Confidence Interval (CI) 3.39-5.14 per 100 person-years] vs. 0.4 per 100 person-years [95% CI 0.23-0.69 per 100 person-years]) with significant heterogeneity observed between the studies. Conclusions: Tuberculosis incidence rates were high among individuals on combination antiretroviral therapy in high/intermediate burden settings. Interventions to prevent tuberculosis in this population should address geographical, socioeconomic and individual factors such as low CD4 counts and prior history of tuberculosis.