Multilevel Modeling of Utilization of Maternal Health Care Services in Ethiopia

Abstract

This paper has aimed to find out the determinants of utilization of maternal health care services in Ethiopia. The study used data from 2011 Ethiopian Demographic and Health Survey (EDHS) which has a two-level hierarchical structure, with over 16,515 women nested within eleven geographical regions. The bivariate analysis result showed that maternal education, husband education, maternal age, visited by community health workers, marital status, women’s  exposure to media, place of residence, birth order, maternal work status, wealth index, and religion have shown significant variations in the usage of antenatal and delivery care services.. The study shows that the coverage of antenatal care service uptake is only 35.1 percent while that of delivery care is 9.6 percent. Multilevel logistic regression models have been used to analysis the data. The significant deviance based chi- square test  shows  that random  coefficients  model  better fit  to the data. Generally,  the usage  of maternal care services in Ethiopia is far from the expectation. So, attention must be given to educating women, providing relevant information, shaping the traditional believes of mothers and improving service quality. It is also recommended that multilevel models better than traditional single level models when the data structure is hierarchical, like EDHS data