Spatial Patterns of Under-Five Pneumonia and Socio-Environmental Factors in West Java: A LISA and Geographically Weighted Regression Approach, 2019–2023

Pola Spasial Pneumonia Balita dan Faktor Sosio-Lingkungan di Jawa Barat: Pendekatan LISA dan Regresi Tertimbang Geografis, 2019–2023

Abstract

Introduction: Pneumonia remains a leading cause of mortality in children under five in Indonesia, with significant geographical disparities. Objective: This study addresses the urgent need to understand the complex spatial distribution of pediatric pneumonia and its relationship with local socio-environmental factors in West Java, Indonesia’s most populous province. Methods: This study uses an ecological study design, utilizing aggregated district/city level data from 2019 to 2023, and employs a quantitative spatial analysis approach. As the analysis is at an aggregated level, the findings reflect area-level associations, and conclusions cannot be inferred regarding individual-level risk, thus avoiding potential ecological fallacy. Result: The study reveals that while the overall distribution of pneumonia was not globally clustered, significant local patterns exist. Geographically Weighted Regression (GWR) analysis demonstrated superior performance over traditional Ordinary Least Squares (OLS) models, with an R² value reaching 0.800 in 2019, indicating that relationships between pneumonia and its determinants are highly variable across space. Getis-Ord Gi* analysis identified persistent coldspots (low-incidence clusters) in the southeastern region and dynamic hotspots (high-incidence clusters) along the northern coast. Furthermore, Bivariate LISA analysis confirmed that population density and poverty are significant clustered determinants spatially correlated with pneumonia incidence. Conclusion: These findings provide critical evidence for developing targeted, location-specific public health interventions to effectively reduce the burden of childhood pneumonia.