Predicting Antenatal Care Utilization in the Philippines: A Chi-Squared Automatic Interaction Detector Analysis

Abstract

Complications associated with pregnancy and childbirth are still considered as leading causes of deaths in the reproductive age group of women in many developing countries. This problem may be addressed when pregnant women utilize antenatal care services. However, despite the benefits of this service to the mother and child, it is not fully utilized in the Philippines. Using the 2013 National Demographic and Health Survey, an exhaustive chi-squared automatic interaction detector model- a data mining approach was employed to analyze the predictors of antenatal care (ANC) utilization of reproductive women age 15 to 49 in the Philippines. Findings reveal that the most significant predictor of ANC utilization was the region where a woman resides. Women who were living in NCR, CAR, Ilocos Region, Cagayan Valley, Central Luzon, CALABARZON, Bicol, Western and Central Visayas, Davao and Caraga had the highest likelihood of utilizing ANC than those who lived in MIMAROPA, Eastern Visayas, Zamboanga, Northern Mindanao and SOCCSKSARGEN with the lowest in ARRM. The second most influential factor is the woman’s religious affiliation. If a woman is a Muslim, her utilization is lower than a Christian, regardless of what her socio-economic status is or in what wealth quintile she belongs. The result further shows that wealth quintile and educational attainment are also significant predictors. Policy makers and health providers should intensify their efforts to address differentials between these groups of women and to widen the campaign of at least four antenatal visits during pregnancy as a requirement in all parts of the country.