| Literature DB >> 8272658 |
Abstract
Classical tests for clustering rarely have a major role in the investigation of disease clusters at the neighbourhood level performed by a health department. When examining a single cluster, the boundaries are often ill-defined, and the dimension (geographic, occupational, iatrogenic) is rarely obvious. Furthermore, using cluster tests to routinely scan disease registries looking for clusters is undesirable for several reasons. A better approach is to be vigilant for unusual environmental exposures, and to evaluate the impact they may have. In this context, cluster techniques may serve as part of a larger investigation that includes other epidemiologic approaches.Mesh:
Substances:
Year: 1993 PMID: 8272658 DOI: 10.1002/sim.4780121904
Source DB: PubMed Journal: Stat Med ISSN: 0277-6715 Impact factor: 2.373