Inference from ecological models: Estimating the relative risk of stroke from air pollution exposure using small area data

摘要: Abstract Maheswaran et al. (2006) analysed the effect of outdoor modelled NO x levels, classified into quintiles, on stroke mortality using a Poisson Bayesian hierarchical model with spatial random effects. An association was observed between higher levels and at small area (enumeration district) level. As this is framed in an ecological perspective, relative risk estimates suffer from bias. In paper we use different specification based Jackson (2008) , modelling number cases due to as binomial variable where p ( i ) probability dying . The within-area variation used determine proportion population falling each five exposure categories order estimate individual given k th level assuming homogeneous across study region. inclusion variability regression has been demonstrated help reduce bias al., 2006, 2008 ). Revised are obtained compared previous estimates.