Estimation of Environmental and Geographical Determinants of Acute Gastro Enteritis Using Geographically Weighted Regression Analysis

Disease mapping helps to investigate the geographical distribution of a disease burden on a certain population. Spatial regression differs from disease mapping in that the aim is to estimate the association between risk and co-variates, rather than to provide area specific relative risk estimates. Geographically Weighted Regression (GWR) is a statistical technique that allows variations in relationships between predictors and outcome variable over space to be measured within a single modeling framework. In the present paper we have combined rainfall and sanitation data to explain the exposure risk of Acute Gastroenteritis in Coimbatore district of Tamilnadu, India, using Geographically weighted regression model. All analyses were implemented using ESRI Arc GIS 10.0 and GWR 3.0 with 0.05 significance level. In the GWR model, the adaptive kernel with AICc estimated bandwidth was chosen. Ordinary Least Square regression (OLS) showed showed high incidence rates in Ikkaraiboluvampatti and moderates rates scattered in northern regions. GWR regression fitted best in villages Ikkaraiboluvampatti, Marudur, Chikkasampalayam, Odanthurai, Irrumbarai, and Muduthurai. This study provides further indications that the relationships of Incidence rates and rainfall were spatially non-stationary in Coimbatore region.