Estimating Prevalence of Coronary Heart Disease for Small Areas Using Collateral Indicators of Morbidity

摘要: Different indicators of morbidity for chronic disease may not necessarily be available at a disaggregated spatial scale (e.g., small areas with populations under 10 thousand). Instead certain only more highly aggregated scale; example, deaths recorded areas, but prevalence considerably higher scale. Nevertheless estimates area level are important assessing health need. An instance is provided by England where and hospital admissions coronary heart known as wards, relatively large authority areas. To estimate CHD in such situation, shared random effect method proposed that pools information regarding contrasts over different (deaths, hospitalizations, prevalence). The approach also incorporates differences between risk factors income, ethnic structure). A Poisson-multinomial equivalence used to ensure sum the total. illustration data London using ward level, together totals larger local involved spatially correlated common factor, accounts clustering latent factors, provides summary measure morbidity.