Identifying priority areas for tuberculosis mortality in Brazil, 2010–2024: a Bayesian spatiotemporal modelling study

Tuberculosis (TB) mortality remains a major public health concern in Brazil, marked by spatial inequalities and recent temporal changes. We aimed to characterise municipal TB mortality patterns from 2010 to 2024, examine contextual factors associated with mortality, and develop a Bayesian prioritisation framework integrating risk magnitude, temporal dynamics, spatial patterns, uncertainty, and robustness. We conducted a nationwide ecological study including all 5570 Brazilian municipalities. Age-standardised mortality rates, Joinpoint regression, spatial autocorrelation, Bayesian spatiotemporal smoothing, hierarchical models, and posterior probabilities were used. Among 73,432 TB-related deaths, mortality declined until 2020 (APC − 2.27; 95% CI − 3.13 to − 1.56) and subsequently increased through 2024 (APC 7.94; 95% CI 4.95 to 12.62). Positive spatial autocorrelation was observed (Moran’s I = 0.305; p < 0.001), with higher-risk concentrations in the North, Central-West, and parts of the Northeast. Mortality was associated with urbanisation, household crowding, income inequality, and operational factors, while spatiotemporal patterns remained largely stable after adjustment. The index classified 70.7% (3938) of municipalities as low priority, 12.5% (695) as low priority with potential risk, 8.2% (459) as moderate priority, 7.2% (403) as high priority, and 1.3% (75) as very high priority; among the latter, 0.3% (18) met the additional robustness criteria. These findings support more precise resource allocation and targeted TB control strategies.