作者: Crescenza Calculli , Alessio Pollice , Maria Serinelli
DOI: 10.1002/ENV.1138
关键词: Bayesian inference 、 Econometrics 、 Spatial variability 、 Confounding 、 Environmental science 、 Risk factor (computing) 、 Statistics 、 Logistic regression 、 Markov chain Monte Carlo 、 Parametric statistics 、 Environmental epidemiology
摘要: In Environmental Epidemiology studies, the effects of presence a source pollution on population health can be evaluated by models that consider distance from as possible risk factor. We introduce hierarchical Bayesian model in order to investigate association between multiple pathologies and single source. Our approach provides possibility incorporate spatial other confounding factors within logistic regression model. Spatial are decomposed into sum disease-specific parametric component accounting for point common semi-parametric interpreted residual variation. The is applied data case–control study evaluate incidence different cancers with residential location neighborhood petrochemical plant Brindisi area (Italy). Copyright © 2011 John Wiley & Sons, Ltd.