Geographically weighted negative binomial regression model to analysis of factors that influence on maternal mortality in Central Java Province

Abstract

The maternal mortality rate in Indonesia is still above the target set by the Millennium Development Goals (MDGs). Central Java Province included in five provinces in Indonesia with the highest cases of maternal deaths. The maternal mortality is the death during pregnancy or within 42 days after the end of pregnancy. This study uses the number of maternal deaths in Central Java Province in 2017 in the form of count data, so the analysis used is a Poisson regression model. In the Poisson regression analysis, we find cases of overdispersion and negative binomial regression can be used to overcome the case. However, because there is heterogeneity in residue, the appropriate model applied is the Geographically Weighted Negative Binomial Regression (GWNBR) model in each district and city in Central Java Province. The modeling analyzes the factors that significantly influence maternal mortality. The results of this study obtained factors that influence the number of maternal deaths in all district and city is the percentage of clean and healthy living behavior household and the number of community health centers (puskesmas). Besides, three classification groups of districts and cities were obtained with each variable influencing them.The maternal mortality rate in Indonesia is still above the target set by the Millennium Development Goals (MDGs). Central Java Province included in five provinces in Indonesia with the highest cases of maternal deaths. The maternal mortality is the death during pregnancy or within 42 days after the end of pregnancy. This study uses the number of maternal deaths in Central Java Province in 2017 in the form of count data, so the analysis used is a Poisson regression model. In the Poisson regression analysis, we find cases of overdispersion and negative binomial regression can be used to overcome the case. However, because there is heterogeneity in residue, the appropriate model applied is the Geographically Weighted Negative Binomial Regression (GWNBR) model in each district and city in Central Java Province. The modeling analyzes the factors that significantly influence maternal mortality. The results of this study obtained factors that influence the number of maternal deaths in all district and city is t...