Inayati Ngoro, Henny Pramoedyo, Ani Budi Astuti
Maternal Mortality Rate (MMR) is a key public health indicator that reflects spatial variation across districts in East Java. This study aims to model the spatial distribution of MMR using Geographically Weighted Poisson Regression (GWPR) with an Adaptive Gaussian Kernel weighting function. Secondary data were obtained from the 2022 East Java Provincial Health Profile, covering 38 districts and municipalities. The results indicate that GWPR outperforms the classical Poisson regression. The intercept β=2.889 (exp=17.95) suggests an average of 18 maternal deaths in the absence of predictor effects. The coverage of the fourth antenatal care visit (K4) has a significant negative effect (β=-0.027; exp = 0.973), indicating that a 1% increase in K4 coverage reduces MMR by approximately 2.7%. Conversely, obstetric complications managed by midwives show a significant positive effect (β= 0.0173; exp =1.017), meaning that a 1% increase in complications raises MMR by 1.7%. Other predictors—first antenatal care visit (K1), iron–folic acid (IFA) supplementation, and number of health workers—are not statistically significant. This study underscores the importance of expanding K4 coverage and strengthening complication management as priority strategies to reduce maternal mortality. Furthermore, GWPR-based mapping enables more targeted maternal health interventions tailored to local characteristics. © 2025 by the Author(s).
Department of Statistics, Faculty of Mathematics and Natural Sciences, Brawijaya University, Indonesia
