PENDEKATAN GLMM BINOMIAL NEGATIF DALAM MENGANALISIS KASUS KEMATIAN BAYI DI JAWA TIMUR TAHUN 2023

Abstrak

Infant Mortality Rate (IMR) is a critical indicator of the health and welfare of a community, particularly in East Java. This province has a concerning IMR, necessitating greater efforts to meet the Sustainable Development Goals (SDGs) target of 12 deaths per 1,000 lives births by 2023. Various factors contribute to infant mortality, including Low Birth Weight (LBW), limited exclusive breastfeeding, inadequate access to health services, low levels of maternal education, and economic disparities. These factors should be examined to understand their impact on the rising IMR. This study employs several statistical approaches, including the Generalized Linear Model (GLM), Generalized Linear Mixed Model (GLMM), and Integrated Nested Laplace Approximation (INLA), using three distributions; Negative Binomial, Poisson, and Gaussian. The GLMM using the Negative Binomial distribution proved to be the Best-Fit model for analyzing the relationship between IMR and its contributing factors, as indicated by the lowest values of the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and Deviance Information Criterion (DIC). The research findings reveal that the number of integrated village health posts has the most significant relationship with the IMR cases in East Java.