Safe Birth Predictors: A Machine Learning Study in the Context of Bangladesh

Despite advancements, Bangladesh's maternal death rates are still too high. Interventions to enhance maternal health outcomes can be informed by identifying variables linked to safe births. The purpose of this research is to use machine learning to determine important factors that indicate safe deliveries in Bangladesh, considering the nation's distinct socioeconomic and medical circumstances. This study used data extracted from the Demographic and Health Survey (DHS) database, which included 4414 women of reproductive age. To achieve the analytical goals, a forward stepwise logistic regression approach was used. The odds ratios (OR) of safe deliveries were considerably higher for respondents who practiced contemporary contraception (OR = 1.9), lived in urban areas (OR = 1.4), had a secondary education (OR = 1.9), had a higher level of education (OR = 4.7), gave birth outside the home (OR = 1.1), and lived in urban zones (OR = 1.9). Our results offer important new perspectives on the intricate interactions between variables affecting safe deliveries in Bangladesh.