Determinants of Under-Five Mortality in Ethiopia: an Application of Cox Proportional Hazard and Frailty Models

Abstract

Objective: Under-five mortality is an essential indicator of the development of a country. In Ethiopia under-five mortality rate is among the highest in the world. Nearly one out of 10 babies born in Ethiopia does not survive to celebrate its first birthday. Material and Methods: The data for the study was obtained from Ethiopian Demography and Health Survey data conducted in 2016. The Kaplan-Meier, Cox'sproportional hazards and gamma shared frailty models were employed for the analysis of under-five children data. Results: Results obtained by fitting both Cox-proportional hazard model and gamma shared frailty model: place of residence, Type of Birth, Birth order, sex of a child and preceding birth intervalwere found to be significant factors. Further more a high risk death of under-five children was found to be associated with place of residence at rural, multiple births, birth order at fifth and above, male children and preceding birth interval less than 24 months. From gamma shared frailty model (θ=0.145) we had enough evidence that the existence of unobserved heterogeneity at the regional level. Conclusion: The findings of this paper highlighted the potential associated with under-five child mortality in Ethiopia. The shared frailty model provided better estimates and also justified the presence of unobserved heterogeneity at regional level. Therefore, special attention should be given to these significant predictors, which ultimately reduce the under-five mortality. (English) [ABSTRACT FROM AUTHOR]