Multilevel analysis of infant mortality and its risk factors in South Africa

The study analyzed infant mortality and its risk factors in South Africa. It aimed to examine infant mortality in the country by taking into account the hierarchical nature of the problem and investigate the with-in country variation in modeling. In addition to the usual individual level risk factors of infant mortality, living standard, mother’s education, and income inequality were defined at municipal level, while HIV prevalence was fixed at province level. A multilevel logistic regression model was then fitted with Bayesian MCMC parameter estimation procedure using the 2011 South African census data. Most of the demographic and socioeconomic variables identified at individual level were found significant. More remarkably, the result indicated that communities with better living standard and women's education were associated with lower infant mortality rates, while higher income inequality and HIV prevalence in the communities were associated higher levels of infant mortality. The changes in infants’ odds of death were estimated to be 26%, -21%, 13% and 8% respectively for HIV, women’s education, income inequality and level of the living standard. In addition, unobservable municipal and province level random effects significantly affected the level of infant mortality rates.