Spatial–temporal heterogeneity in tuberculosis incidence and its association with socioeconomic factors in South Sulawesi

TB remains a major public health challenge in Indonesia, yet the district-level spatial–temporal dynamics of reported tuberculosis (TB) incidence and its socioeconomic determinants are not well understood. South Sulawesi exhibits persistent epidemiological disparities despite gradual improvements in development indicators. This study aimed to characterize spatial–temporal patterns of reported TB incidence across 24 districts in South Sulawesi from 2018 to 2023 and to identify the key socioeconomic factors associated with these patterns. Reported TB case incidence rate (CIR) and district-level socioeconomic indicators were analyzed using descriptive statistics, Jenks natural-breaks classification, Pearson correlation, Global Moran’s I, and the Breusch–Pagan test. A Nonparametric Geographically and Temporally Weighted Regression with Truncated Spline (NGTWR-TS) model was applied to examine spatial nonstationarity, nonlinear relationships, and temporal variation in district-level associations between CIR and socioeconomic conditions. Reported TB incidence varied markedly between districts and showed an overall increasing trend over time, with widening disparities and shifting high-burden areas. Spatial classification indicated that district-level incidence patterns were unstable over time, with frequent reclassification across incidence categories rather than persistence of fixed high-burden areas. Correlation analysis identified mean years of schooling (positive), open unemployment rate (positive), and the human development index (positive) as the most consistent socioeconomic correlates of CIR. Moran’s I indicated no global spatial clustering, while the Breusch–Pagan test confirmed strong spatial heterogeneity. NGTWR-TS demonstrated nonlinear threshold effects and substantial spatial–temporal variation in the significance of these socioeconomic associations. Districts with no significant predictors declined steadily, while those with simultaneous significance of all three factors increased from 6 to 18 over the study period. Reported TB incidence in South Sulawesi appears to reflect dynamic, localized socioeconomic conditions rather than fixed geographic patterns. These findings suggest that district-specific strategies integrating educational, economic, and development-oriented interventions may help address the evolving structural vulnerabilities associated with reported TB incidence.