Covariate Effect in Spatio-temporal Bayesian Model with Two-level Spatial Structure

Abstract

Spatio-temporal models suffer from comparability problems of relative risks (RRs) based on the removal of the covariate effect as a confounding factor on the risk esti- mate of the study population through distribution of standardized mortality ratios (SMRs). Two spatio-temporal models with two-level spatial structure with different approaches are considered in this study for comparison. The first model followed SMRs procedure by re- moval of the effect of the confounding factors on the risk estimate in the study population through distribution standardization., while the second model included covariate effect as confounding factors on the risk estimate in the study population. The two models were fitted within a hierarchical Bayesian framework with integrated nested Laplace approxi- mation (INLA) estimation procedures. The objectives of this study are to compare both models in terms of their performance and identify the age-group(s) of women with sig- nificant higher risk due to breast cancer disease. The models are applied to female breast cancer mortality data in Nigeria.