Determinants of Infant Mortality in Tanzania

Abstract

This study set out to examine the determinants of infant mortality in Tanzania. Data from the Tanzania Demographic and Health Survey (TDHS), 2010 were used. A total number of 2,976 children who were born to women aged 15-49 years during the period of one year preceding the survey were involved during the study. The specific objectives of this study were; to describe the mothers‘ background characteristics, to establish the relationship between mothers‘ background variables and infant mortality and to make recommendations for infant centered health interventions to improve infant survivorship in Tanzania. Infant mortality was the dependent variables of the study. This variable was predicted by type of place of residence, zones of residence, educational attainment, household wealth index, place of delivery, and employment status as independent variables. The independent variables were assumed to work through intermediate variables which include; main floor materials, source of drinking water, type of toilet facilities, age of the mother, birth interval, birth order, sex of the child and breastfeeding. The Microsoft Excels, Statistical Package for Social Science (SPSS) and STATA software were used for the analysis of the data. Chapter one of this report is an overview of the problem at global to the area specific (Tanzania) level; the statement of the problem, objectives and the rationale of the study. The study hypothesized that; mothers with no education are more likely to experience infant deaths compared to mothers with secondary and higher education and mothers living in urban areas are less likely to experience infant mortality than women living in rural areas. The bivariate analysis was conducted at 95% confidence level to establish the relationship between infant mortality and background characteristics of the mother. During this stage of analysis, place of delivery, household wealth index, breastfeeding, birth interval and sex of child showed a significant association with infant mortality. Subsequently, the binary logistic regression was conducted at 95% confidence level to examine the independent effect of background variables on infant mortality. The results (Model III) obtained indicated that zones of residence, breastfeeding and birth interval were the significant determinants of infant mortality. Generally, after running the regression model (Model IV) which included only the variables which were significant at either bivariate and/or multivariate level; zones of residence, place of delivery, breastfeeding and birth interval emerged as the significant determinants of infant mortality. Therefore, this study is not an exceptional one, as other studies have come out with similar findings.