Ecological comparison of tuberculosis and TB HIV coinfection in postconflict Liberia and Sierra Leone as markers of health security
Discuss this preprint
Start a discussionListed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background Tuberculosis (TB) remains a significant public health concern in fragile and post-conflict settings. Liberia and Sierra Leone, which are undergoing recovery from extended periods of civil unrest, continue to experience high TB rates, including TB-HIV co-infection. Examining temporal trends and testing coverage in these countries is critical for developing integrated surveillance and response strategies. Methods This retrospective comparative analysis used publicly available data for Liberia and Sierra Leone (2000-2022) from the World Health Organization, World Bank, and ACLED. The primary outcome was the TB incidence rate (per 100,000 population). Two-way fixed-effects panel regressions with country and year fixed effects were estimated, using Driscoll-Kraay standard errors (Bartlett kernel; plug-in bandwidth). Sensitivity analyses included a linear mixed-effects regression with a country random intercept and year fixed effects, and pooled ordinary least squares with country and year fixed effects and heteroskedasticity-robust (HC3) standard errors. Covariates were TB-HIV co-infection (percent), health expenditure per capita (US dollars), GDP per capita (2015 US dollars), and conflict events (count). A subset analysis (2016-2022) examined TB-HIV testing coverage, and a supplementary model evaluated TB mortality (excluding HIV-attributable deaths) from 2000 to 2022 using the same specification. Results Regression models consistently identified negative associations between GDP per capita and TB incidence across specifications. TB-HIV co-infection was inversely associated with TB incidence in most models, though not uniformly significant. In the subset analysis, TB-HIV testing coverage showed a positive coefficient but was not statistically significant ( b = 0.087, p = 0.160). Given two country clusters (G = 2), statistical inference was interpreted cautiously . Conclusions The findings suggest that economic conditions and TB-HIV dynamics may influence TB incidence in post-conflict settings. Although TB-HIV testing coverage was not significantly associated with TB incidence in this sample, its potential role warrants further investigation. Strengthening integrated TB-HIV surveillance and tailoring interventions to country-specific contexts may support more resilient health systems in fragile environments.