Abstract:
Infant mortality is a major health problem remains a significant impact on individuals, their
families, and societies due to demographic and clinical factors observed in developing countries.
The main objective of this study was to determine socioeconomic, demographic and
environmental factors that may have impact on infant mortality in Ethiopia. The data source for
the analysis was the 2016 EDHS data from children in the age group of less than one years.
Survival analysis like KM, log rank, Cox PH and shared gamma frailty models were used in this
study. Cox PH and shared gamma frailty models were employed to identify factors affecting
infant mortality. AIC was used in order to compare models. A total of 2621 infants were taken, of
the total 496 were experienced the event. Regional states of children were considered as the
clustering effect in frailty model. The clustering effect was significant for infant with =.112(pvalue 0.000). Gamma shared frailty model was selected over Cox PH model based on their AIC.
The result showed that, the estimated risk factor for infant who didn’t used breast feeding and
type of birth multiple were increase infant mortality by the hazard of HR=12.65 and HR = 5.82
respectively. Based on the result Gamma shared frailty model, household size, breast feeding,
type birth, birth interval, place of delivery, occupation of the respondent and mother education
were statistically significant predictors for infant mortality. Thus, Police makers need to
consider factors that affect infant mortality during planning and developing policies in order to
reduced mortality
