A shared component model for detecting joint and selective clustering of two diseases.

摘要: The study of spatial variations in disease rates is a common epidemiological approach used to describe the geographical clustering diseases and generate hypotheses about possible 'causes' which could explain apparent differences risk. Recent statistical computational developments have led use realistically complex models account for overdispersion correlation. However, these focused almost exclusively on modelling single disease. Many share risk factors (smoking being an obvious example) and, if similar patterns variation related can be identified, this may provide more convincing evidence real underlying surface. We propose shared component model joint analysis two diseases. key idea separate surface each into disease-specific component. various components formulation are modelled simultaneously by using cluster implemented via reversible jump Markov chain Monte Carlo methods. illustrate methodology through oral oesophageal cancer mortality 544 districts Germany, 1986–1990.