Generalized Poisson Regression Modeling on the Number of Infant Deaths in East Nusa Tenggara Province in 2022

The number of infant deaths in East Nusa Tenggara (NTT) Province is still above the national average. This research was conducted to investigate the Generalized Poisson Regression (GPR) model for addressing the overdispersion in the Poisson regression model of the number of cases of infant deaths and to explore the potential factors influencing the number of infant deaths in the province. The variable used is the number infant mortality as a response variable, and the number of predictor variables that are thought to influence the response variable. The data used is secondary data obtained from the publication by the Central Statistics Agency of ENT Province from each of the 22 cities/regencies. The study shows data on the number of cases of infant mortality experienced overdispersion with a ratio between deviance and degrees of freedom of 3.578. Modeling with GPR shows that the model with 5 independent variables produces an optimal model with an AIC value of 184.145. Those variables are the percentage of households with access to adequate sanitation, the percentage of births assisted by parties other than medical personnel, the number of teenagers who received reproductive health counseling, the percentage of the population aged 0-59 months who received incomplete immunization, and the number of health facilities.