Poisson Spatial Regression Model to Study Gastric Cancer in Boyacá Department, Colombia in 2005

Gastric cancer is the most prevalent and the leading cause of cancer death in Colombia. It has been identified as possible risk factors associated the altitude in the Andean region and the acquisition of the bacterium Helicobacter (H.) pylori in childhood. The aim was to explain the behavior of this pathology in the department of Boyacá, Colombia for 2005, to assess the effect of incorporating a structure of spatial dependence in the model, and identify the most significant etiological factors that may explain the occurrence of cases. In this study, two models were estimated, the first one, a generalized linear model with Poisson response without spatial structure (GLMp), and the second one, a generalized linear mixed model Poisson with spatial structure using Moran eigenvectors (GLMMp_ME) to remove the spatial autocorrelation of the residuals. Additionally, spatial aggregation tests for model residuals were made. The work was developed in R 3.1.1. The results showed aggregations of local type in some municipalities of the department and very few of the explanatory variables were significant in the model. The importance of considering spatial effects in these models was evident and some etiological hypotheses referenced in related studies were refuted.