Abstract
Background: Indonesia bears the second-highest tuberculosis burden globally, accounting for approximately 10% of all reported cases. However, national- and district-level evidence examining both univariate and bivariate spatial autocorrelation across all 514 districts/cities remains limited. Purpose: This study aimed to determine the spatial autocorrelation of TB and its determinants at the district level. Methods: This study analyzed district- and city-level data from 514 districts/cities in Indonesia using the Indonesian Basic Health Research (Riskesdas) and Statistics Indonesia reports. Univariate and bivariate Global and Local Indicators of Spatial Autocorrelation (GLISA) were applied to identify the spatial distribution patterns of TB and its determinants. Results: Significant positive spatial autocorrelation of TB prevalence was found in Sumatra (Moran’s Index=0.13, p value=0.01), Java (Moran’s Index=0.43, p value=0.00), and Bali–Nusa Tenggara (Moran’s Index=0.13, p value=0.03). LISA identified 41 high–high hotspot districts/cities and 78 low–low coldspot districts/cities, with hotspots mostly located in Java and Sumatra. Moran’s I bivariate analysis showed positive spatial associations between TB prevalence and poverty in Sumatra, Bali–Nusa Tenggara, and Moluccas; population density in Java; smoking in Java, Bali–Nusa Tenggara; adequate lighting in Celebes; and diabetes mellitus in Papua. Conclusion: TB prevalence in Indonesia exhibits heterogeneous district- and city-level clustering. These findings support geographically targeted active case finding, contact investigation, integrated TB–diabetes and tobacco-control interventions, and housing and social-protection strategies in high-priority clusters.

This publication has 24 references indexed in Scilit: