Effect of Household Dynamics on Infant Mortality in the Kassena_Nankana Districts of Rural Northern Ghana, 1999–2018

Background: Infant mortality rates still remains high, particularly in low- and middleincome countries. Several factors account for these high infant mortality rates. These factors require context-specific policy interventions to reduce the high level of infant mortality rates. This study seeks to examine the effect of household characteristics on infant mortality in rural northern Ghana. Methods: We used longitudinal data from the Navrongo Health and Demographic Surveillance System. In all 48,375 children born between January 1999 and December 2018 were included in the analysis. Person-months were generated, which produced 558,896 person-months. We used Multilevel Poisson regression to examine the association between the outcome variable and the independent variables. Results: Results showed that the presence of biological parents and a senior sibling, household size and polygyny are among several factors that are associated with infant survival in the area. Children who lost their mothers were more likely to die before age one compared to children whose mothers were alive. Similarly, children who lost their fathers were about 40% more likely to die before age one compared to children whose father were alive. Also, children from relatively larger households (4+ members) were more likely to die before age one compared to their counterparts from smaller households. Conclusions: The results show that the presence of biological parents and a senior sibling contribute to improved infant survival. Higher maternal education and smaller household size also contribute substantially to child survival. Policy interventions aimed at reducing infant mortality should target and intervene in these context-specific areas.