Modeling of infant mortality in west sulawesi using zero inflated poisson regression method

Abstract

The number of infant deaths, one of which can be reflected in the Infant Mortality Rate (IMR) is one indicator to determine the development of a country’s development. The AKB target in the 2015-2019 National Medium-Term Development Plan is 24 deaths per 1,000 live births. Based on the SDKI 2017, the IMR value in West Sulawesi is 42 deaths per 1,000 live births. Factors suspected of affecting infant mortality in West Sulawesi are not only in the health but also in socio-economic fields. Data on the number of infant mortality in West Sulawesi is a data count that contains a value of 0. Therefore, the analysis used is Zero Inflated Poisson (ZIP) Regression to determine what factors influence infant mortality in West Sulawesi. The results of the ZIP regression analysis showed that the best model with the smallest AIC and AICc is 324,3407 and 326,7407 were models with a combination of the variables of the percentage of active maternal and child health services, percentage of society with good drinking water adequate access and percentage of society with proper sanitation access. The variable that has a significant effect on the log model is the percentage of active maternal and child health services, percentage of society with good drinking water adequate access and percentage of society with proper sanitation access. In the variable logit model that is influential is the percentage of active maternal and child health services.