Spatial autocorrelation analysis of tuberculosis incidence: Identifying geographical clusters and socio-environmental risk factors

Abstract

Background: Indonesia ranks second as the country with the highest number of tuberculosis cases in the world. The three most populous provinces on the island of Java (West Java, East Java, and Central Java) contribute the most TB cases in Indonesia. The Provincial Health Offices of West Java, East Java, and Central Java have data on tuberculosis incidence and influencing risk factors, but most of the data is processed manually and presented in tables and graphs. Spatial and mapping approaches can be used to visualize the distribution of tuberculosis incidence and its risk factors. Method: This study used an ecological study design with a spatial approach. The population in this study consisted of 100 districts/cities in the provinces of West Java, East Java, and Central Java. The data used were aggregated from annual publications issued by the health offices and the central statistics agencies of the three provinces for the period 2024. Findings: The tuberculosis case notification rate distribution in the three regions of Java Island exhibited positive spatial autocorrelation. Three independent variables had negative spatial autocorrelation with the TB CNR, namely the percentage of poor people, the percentage of households with access to proper sanitation, and the percentage of livable houses. Meanwhile, population density is the only variable that has positive spatial autocorrelation with TB CNR.  Conclusion: TB prevention, case finding, and intervention can adopt and modify the policy implications of these spatial analysis results by considering the conditions of each region. Originality of this article: This study applies spatial analysis using Moran's Index and LISA approaches in the regions of West Java, East Java, and Central Java, as well as its use of variables such as BPJS Health insurance ownership, access to proper sanitation, and livable housing, which have not been widely studied in previous research.