Spatio-temporal regression analysis of dengue cases with climate variables in Pulau Pinang

Abstract

In disease mapping, spatio-temporal regression analysis is essential in describing the relationship of certain covariates towards disease. Spatio-temporal regression analysis incorporates fixed covariates effects, spatial effect and temporal effect in the model. This paper applied spatio-temporal regression method to analyse the effect of climate variables which were the mean weekly temperature, the total weekly rainfall distribution and the mean weekly humidity percentage towrds dengue incidence in five districts in Pulau Pinang from 2015 to 2017. The model for this study was fitted within a hierarchical Bayesian framework and involved integrated nested Laplace approximation. For this analysis, the findings revealed that variable rainfall was not giving significant effect for dengue disease risk model in 2015 to 2017, while variable average weekly temperature was significant in 2015 and 2016 models and variable humidity only significant for dengue risk model in 2017.