A birth cohort study in South-West Ethiopia to identify factors associated with infant mortality that are amenable for intervention

Abstract

A one year live birth cohort of all identified children born in Jimma, Keffa and Illubabor Zones of Southwest Ethiopia, was followed from birth to the end of the first year of life, or to an earlier death, to investigate factors associated with mortality. All identified live-born infants were visited as soon as they were born and then bimonthly to the age of one or early death or lost for follow-up. Information was collected on socio-economic, behavioral, biological, and environmental factors for infants, mothers, and family immediately after birth and in consecutive visits. Overall, infant mortality was 106.2/1000, with estimates of 97.0/1000 and 113.5/1000 for urban and rural areas, respectively. Based on the results of the final Cox Proportional Hazards Model, mortality was associated with mothers' education and with antenatal care follow-up: there was better survival with at least one antenatal care follow-up. It was not associated with the mother's age per se but with marital status. Birth weight was strongly associated with infant mortality, and boys were more likely to die than girls. Twins were much more likely to die than singletons, even after taking their birth weight into account. Once adjustment had been made for these variables, there was no independent effect of the mothers' ethnic group or religion, or of family income. Family size was found to be significantly associated with infant mortality and higher mortality rate to families with the 2-4 persons. The type of water supply to the home was not associated with mortality, but habit of use of soap by the mother in washing her hands was. These results suggest that improvements in survival could be brought about by attempts to improve maternal education, antenatal care attendance, and habit of soap use in washing her hands which are amenable for change with local resources.


(Ethiopian Journal of Health Development, 2000, 14(2): 161-168)