Spatial Regression Analysis of Leprosy in East Java 2025 Using Spatial Autoregressive and Spatial Error Models with Queen Contiguity

Abstract

Leprosy remains a public health issue in Indonesia, including in East Java, which has a relatively high number of cases. Differences in social, economic, and health characteristics between regions can cause variations in leprosy cases and spatial clustering. Unlike previous studies, which generally used absolute case numbers and applied one spatial regression model, this study compares spatial autoregressive (SAR) and spatial error (SEM) models using the new case detection rate (NCDR) as the response variable. The study uses Queen Contiguity weighting to analyze the factors that influence the NCDR of leprosy in East Java in 2025. The predictor variables include population density, the percentage of households without toilet  facilities, the percentage of impoverished residents, and the number of health centers. Secondary data from 38 districts/cities in East Java were used for the analysis. The analysis revealed significant positive spatial autocorrelation, with a Moran's I value of 0.6059. Of the predictor variables, the percentage of impoverished residents was the only one that significantly affected the NCDR of leprosy. Based on Akaike's information criterion (AIC) and log-likelihood values, the SAR model was selected as the best model. These findings suggest that controlling leprosy requires social and economic interventions, particularly poverty reduction efforts.