Bivariate Poisson Regression in Modelling Infant and Maternal Mortality Rate in West Java Province 2014

Abstract

Reducing infant and maternal mortality are two of the eight of the Millenium Development Goals (MDG’s).
Nevertheless, the two goals are hardly to achieved particularly in West Java Province. Instead of decreasing, infant
mortality rate in West Java Province increased from 26 infant deaths per 1000 live birth in 2010 to 34 infant deaths per
1000 live birth in 2014. The maternal mortality rate in West Java also increased from 226 maternal deaths per 100.000
live birth in 2010 to 369 maternal deaths per 100.000 live birth in 2014. Accordingly, identifying infant and maternal
mortality important factors are required for designing effective development programs to meet the two Millenium
Development Goals (MDG’s). In this paper, we identified the important factors of infant and maternal mortality by means
at Poisson regression. Particularly, we used bivariate Poisson regression model to accommodate interrelation between
infant and maternal mortality. Based on 2014 West Java Health Profile data, we found that percentage of health workers
is the most important factor. The testing parameters for bivariate Poisson regression model of infant mortality and
maternal mortality show that, the pregnant women get a tablet Fe3, treatment of obstetric complications , households
clean and healthy behavior, health workers, and women with the age marriage less than 18 years are significant to infant
and maternal mortality model.
Keywords: Infant Mortality, Maternal Mortality, Poisson Regression, Bivariate Poisson Regression.