Detecting clusters of disease with R

One of the main concerns of Public Health surveillance is the detection of clusters of disease, i. e., the presence of high incidence rates around a particular location, which usually means a higher risk of suffering from the disease under study (Aylin et al. 1999). Many methods have been proposed for cluster detection, ranging from visual inspection of disease maps to full Bayesian models analysed using MCMC. In this paper we describe the use and implementation, as a package for the R programming language, of several methods which have been widely used in the literature, such as Openshaws GAM, Stones test and others. Although some of the statistics involved in these methods have an asymptotical distribution, bootstrap will be used to estimate their actual sampling distributions.We would like to thank co-editor Dr. Manfred M. Fischer and four anonymous referees for their suggestions and comments to improve this paper. The help of Dr. Roger Bivand has also been of great value. Furthermore, this work has been partly funded by Consellería de Sanitat and EUROHEIS Project (code SI2.329122, 2001CVG2-604). The authors wish to express their regard and gratitude to Prof. Juan Ferrándiz-Ferragud who died during the revision of this paper. Juan was the main researcher of the Spanish EUROHEIS group, and was really a master for all the people involved in the project.