Monitoring Changes in Spatio-temporal Maps of Disease

The object of statistical surveillance is to detect a change in a process accurately and quickly as new observations keep adding to the observed part of the process. In this paper we discuss methodological issues in developing a rapid response in a spatial surveillance system. Simple exploratory statistical methods together with more sophisticated methods, based on hierarchical space-time models defined at small area level, are considered.