Geospatial analysis of childhood morbidity in Ghana

INTRODUCTION: Childhood morbidities are common in Ghana. The present study sought to geospatially analyze morbidities among children (0–23 months of age) using five different survey datasets (1993–2014) from the Ghana Demographic and Health Survey. METHODS: Logistic regression was used to examine childhood morbidity within a place of residence. Then three spatial statistical tools were applied to analyze morbidities among children (0–23 months of age). These tools were: spatial autocorrelation (Global Moran’s I)—used to examine clustering or dispersion patterns; cluster and outlier analysis (Anselin’s local Moran’s I)—to ascertain geographic composition of childhood morbidity clusters and outliers; and hot spot analysis (Getis-Ord G)—to identify clusters of high values (hot spots) and low values (cold spots). RESULTS: Children in rural areas were much burdened with the occurrence of childhood morbidity. The study revealed positive spatial autocorrelation for childhood morbidity in Ghana. Childhood morbidity (diarrhoea, ARI, anaemia, and fever) clusters were identified within districts in the country. Children in rural areas were more likely to be morbid with diarrhoea, anaemia, and fever compared to those in urban areas. Hot spot districts for diarrhoea, anaemia and fever were mainly situated in semi-arid areas and those with ARI were located both at the semi-arid areas and coastal portions of Ghana. CONCLUSION: Rural children are much exposed to have higher burden of a childhood morbidity compared to their urban counterparts. Most semi-arid districts in Ghana are burdened with diarrhoea, ARI, anaemia, and fever. To minimize the occurrence of childhood morbidity in Ghana, designing of more context-based interventions to target hot spots districts of these morbidities are required in order to use scarce resources judiciously.