Residential inequalities in child mortality in Ethiopia: Multilevel and decomposition analyses

Abstract

Ethiopia is among the five countries which account for half of the global under-five deaths, with the under-five mortality rate of 67 deaths/1000 live births in 2016. Ethiopia had significant inequalities in child mortality between rural and urban areas where the risk of child mortality is largely higher in rural than urban areas. Inequalities in the distribution of factors influencing child mortality need to explain the gap between and within urban-rural areas. The study used the risk of child mortality as an outcome variable. Multilevel logistic regression was used as a standard model for assessing the effect of socioeconomic and contextual factors on child mortality. Furthermore, the Blinder-Oaxaca decomposition technique was used to explain the urban-rural, intra-rural, and intra-urban inequalities in child mortality. The birth order and sanitation type seem to be the most important explanatory factors, followed by wealth status in explaining the rural-urban inequality of 39 deaths/1000 children. Mean proportion indicates that there would be 47 deaths/1000 children for urban poor and 21 deaths/1000 children for urban non-poor, resulting in 26 deaths/1000 children change in urban poor when applying the urban non-poor coefficient and characteristics to urban poor behavior. The findings showed that some residential inequalities in child mortality occur at a level that could be addressed by targeting children, households, and some occurs at a community level that could be addressed by targeting regions. Therefore, any residential sensitive and specific interventions should consider child’s and household’s characteristics, and geographical location.