An exploratory analysis of the spatial variation of malaria cases and associated household socio-economic factors in flood-prone areas of Mbire district, Zimbabwe

Although floods are well known for promoting the transmission of malaria, not much effort has been put to determine the spatial distribution of this disease at micro-geographical scale in flood prone areas. Therefore, this paper examines the spatial variation of malaria and associated socio-demographic factors in flood-prone areas of Mbire district, Zimbabwe. A cross-sectional household survey was conducted between 2018 and 2019 to collect data on malaria and socio-demographic factors. The study used the Bernoulli model to determine malaria hotspots, i.e. area with high number of cases compared to the surrounding areas. This was followed by a geographically weighted logistic regression model used to explore the spatial variation of malaria cases in relation to socio-demographic factors at household level. Descriptive statistics and chi-square showed that types of house material, water sources and age of the household head had a significant association with malaria cases (p < 0.05). The study has demonstrated that some local communities in flood prone areas experience increased numbers of malaria cases as indicated by one significant cluster (p < 0.05). There was high malaria risk (2.68) in the significant cluster compared to its outside. The GWLM model with water sources and the type of house material as exploratory variables showed the minimum corrected Akaike's Information Criterion compared to other models. Higher level of spatial variability was observed in type of house material (DIFF of criterion =  − 5.742) compared to water sources (DIFF of criterion =  − 1.064). The coefficients of these two exploratory variables were varying across the study area and significant (t-values ± 1.96) and high in other parts including the identified hotspots. The coefficients of the type of house material ranged from − 1.136 to 1.323. These were higher than those for water sources which were between − 0.781 and 0.605. These results may be used for development and implementation of place-specific malaria intervention strategies in flood prone areas.