chapter  6
Case Event and Count Data Modeling
ByAndrew B. Lawson
Pages 12

In this chapter, the basic forms of data arising in disease mapping studies are considered and common models for the variation in disease risk are described. The chapter is broken into consideration of case event data that arises when residential addresses of cases are available, and count data that is usually found as an aggregation of case events into small areas and fixed time periods. With case event data, the data takes the form of geographic coordinates, such as latitude and longitude, of a residence. Often the (x, y) residence coordinate may not be available due to medical confidentiality restrictions, however. In that case, aggregated count data is often the only available form. For some studies, in particular those dealing with putative health hazards (see Chapter 14), a fine spatial resolution is the focus, and so address locations may be at the resolution level essential for the study to proceed. In what follows, I examine common likelihood models for both case events and count data and also consider how these can be simply extended to include extra heterogeneity in disease risk.