Geographic modeling of weighted regression of self-reported malaria cases associated with environmental risk factors in Benin during the rainy season

Abstract

Background: Geographically Weighted Regression (GWR) is a technique applied to capture variation by calibrating a multiple regression model, which allows different relationships to exist at different points in space. With malaria elimination at the top of the health agenda, integrated action on all elements of the malaria system that contributes to improved knowledge and local capacity building for positive effects on the health of the local population is needed.Methods: Several variables were collected for 192 sampling points in 12 communes in Benin, one per department. A questionnaire was sent to the head of the household to analyze the impact of environmental factors on reported malaria cases. Numerous GIS classification software for spatial analysis, remote sensing, data analysis/modeling and GPS management, R and MGWR software were used for geographic modeling.Results: An abundance of malaria cases reported in crop areas than in non-crop areas and in rural areas than in urban areas. The Hot Spot Analysis shows the localities of South Benin and Malanville as priority issue areas with a remarkable increase in crop diversity favorable to malaria vector proliferation. The spatial autocorrelation z-score of 4.83653470763 shows that there is less than a 1% probability that this clustered pattern is the result of chance. Conclusion: The observed non-stationarity means that the relationship between the variables studied varies from location to location depending on the physical factors of the environment that are spatially autocorrelated. Environmental factors therefore influence the intensity of transmission, seasonality, and geographic distribution of malaria. With minimal funding, we plan to correlate these data with parasitological and entomological da.