Bayesian Model Averaging (BMA) Modeling In The Analysis Of Factor Affecting Childhood Pneumonia Cases In Medan City

Abstract

Pneumonia is one of the leading causes of death among toddlers worldwide, including in Indonesia. Medan, as a densely populated city, faces significant challenges in managing pneumonia in toddlers, with a still low case detection rate. The high incidence of pneumonia is influenced by various risk factors such as environmental density, air pollution, malnutrition, smoking habits, and lack of public awareness. Therefore, this study aims to identify the most influential factors affecting the incidence of pneumonia in toddlers in Medan City. This study uses the Bayesian Model Averaging (BMA) approach to address model uncertainty and produce more accurate estimates by considering various possible combinations of risk factors. The BMA method is considered superior in the context of complex health data because it can probabilistically combine information from multiple models. The results of this study are expected to provide data-based recommendations for more effective strategies for the prevention and control of pneumonia in toddlers in densely populated urban areas such as Medan City.