Spatial Temporal Prediction of Malaria Risk in Western Kenya using Bayesian Geostatistical Approach

Malaria is a vector borne disease that occurs in areas where the climatic and environmental conditions are suitable for survival of Anopheles mosquitoes. The environmental and climatic factors that affect malaria transmission are, rainfall, temperature, humidity and vegetation. Deforestation, agricultural activities and population movements are anthropogenic factors that affects malaria transmission. Despite implementation of several strategies in controlling and management of malaria in western parts of Kenya, high number of malaria cases are still being recorded due to changing environmental and climatic factors. The aim of this study was to apply the geostatistical modelling to estimate and map the spatial and temporal changes in malaria risk by using the available time series climatic and environmental data and to estimate the population at risk at different time epochs. The data was prepared using python scripts and different ArcGIS tools. A spatial temporal model based on Bayesian approach was used to estimate malaria risk and was implemented in R using Intergrated Nested Laplace Approximation (INLA) package to estimate the malaria risk. INLA was preferred to Monte Carlo Markov Chain (MCMC) due to its efficient computation advantage.