Imperial College, School of Medicine - Department of Epidemiology and Public Health
Disease mapping studies summarize spatial and spatio-temporal variation in disease risk. This information may be used for simple descriptive purposes, to assess whether health targets are being met or whether new policies are successful, to provide context for further studies (by providing information on the form and size of the spatial variability in risk) or, by comparing the estimated risk map with an exposure map, to obtain clues to etiology.
There are well-known problems with mapping raw risk rates and so a multilevel modelling approach is preferred. There are also a number of difficulties associated with the population and health data that are typically available for disease mapping studies. In terms of the health data we consider the problem of case under-ascertainment which for cancers in particular is a major problem and discuss a variety of approaches for rectifying this problem. The populations that are used for estimating disease risks have traditionally been treated as known quantities. In practice, however, these counts are often based on census data which are subject to inaccuracies (in particular under-enumeration) and are only available for census years. Inter-censual population counts must consider not only the usual demographic changes (e.g. births/deaths) but also migration. We propose a number of approaches for modelling population counts and investigate the sensitivity of relative risk estimates to the sizes of these errors. We illustrate the proposed methods using data for breast cancer in the Thames region of the United Kingdom, and compare our results with those obtained from more conventional approaches.