Spatial Confounding in Disease Mapping

A. Urdangarin Iztueta

Spatial disease mapping models are widely used to visualize spatial patterns of mortality and incidence risks. A further objective is to assess the association between covariates and the risks to identify possible risk factors. However, the estimation of fixed effects can be substantially biased and exhibit inflated variance if the covariates are spatially structured and spatial random effects are added to the model. This issue is known as spatial confounding, and the literature reflects differing perspectives on its definition and on the methods proposed to address it. The aim of this work is to assess the performance of several recently proposed methods for addressing spatial confounding in disease mapping, and to introduce a new approach for estimating unbiased fixed effects within this framework.

Palabras clave: Spatial confounding, spatial+, disease mapping

GT Bioestadística SEIO: Current Challenges and Advances in Spatial Biostatistics
2 de septiembre de 2026  11:20
Aula 28

C. Guarner Giner, D. V. Conesa Guillén

T. Kneib

A. Mateos Caballero, A. Domínguez Monterroza, A. Jiménez Martín, M. Martínez Pizarro