[A critical review of epidemiological methods for spatial analysis].

Since the 1950's a number of methods have been developed whose aims is to provide enhanced mapping tools, which allow the depiction of the distribution of disease, and thereafter, to assess risks due to environmental factors (or other spatial-defined factors). Our review of statistical-epidemiological methodology for spatial analysis of diseases shall outline the distinction between global tests for clustering (detection of non casual distribution of events in an area) and cluster detection's techniques for the identification and location of abnormal aggregation of disease's cases. In this paper we evaluate each-one of the proposed methods, on general grounds, with regard to whether they meet the aims of the analysis and how successful they are in overcoming problems.