Spatial distribution and predictors of drug-resistant tuberculosis incidence in Mozambique: A nationwide Bayesian disease mapping study

Introduction: Mozambique is among the highest-burden countries for drug-resistant tuberculosis in the world. However, the spatial distribution of drug-resistant tuberculosis, in the country is unknown. Therefore, we aimed to analyse the spatial distribution, predictors, and hotspot districts for drug-resistant tuberculosis incidence in Mozambique.

Method: We carried out an ecological study using the district as the unit of analysis where we included all cases of drug-resistant tuberculosis diagnosed in Mozambique from 2016 to 2020. We obtained the data from the Minister of Health and other open sources. Parameters of interest were estimated through a spatial Bayesian Poisson regression model using Markov Chain Monte Carlo simulation.

Results: A total of 5092 people with drug-resistant tuberculosis in Mozambique were diagnosed during our study period. We found heterogeneity in the spatial distribution of drug-resistant tuberculosis incidence across the country. Higher incidence rates were mainly observed in the south and central regions, and 26 (16.9%) districts out of 154 were identified as hotspot areas. The incidence of drug-resistant tuberculosis increased with an increase in the HIV prevalence (Relative risk [RR]: 1.53; 95% Credibility Interval [CrI]: 1.32 to 1.76), electricity coverage rate (RR: 1.59; 95% CrI: 1.19 to 2.09), and population density (RR: 1.36; 95% CrI: 1.08 to 1.69) and decreased with an increase in the proportion of people with a bank account per district (RR: 0.71; 95% Crl: 0.51 to 0.96).

Conclusion: The incidence of drug-resistant tuberculosis was not homogeneous, and it was associated with social determinants of health. Targeting interventions in hotspot districts and addressing social determinants is crucial for tuberculosis elimination in Mozambique.