Chronic disease surveillance and testing of clustering of disease and exposure: Application to leukemia incidence and TCE‐contaminated dumpsites in upstate New York

Abstract

AbstractSurveillance of large geographic areas for “clusters” of adverse health events, particularly cancers, has interested statisticians and epidemiologists. Recent statistical research involves so‐called “focused” clustering of disease, i.e. aggregation of incident cases of disease in the vicinity of prespecified, putative foci of increased risk. We introduce a statistical test of focused clustering with optimal, or near‐optimal power properties. This test, and two other recent tests of focused clustering are illustrated and compared using leukemia incidence data for 1978–1982 in a region of upstate New York, with inactive hazardous waste sites containing trichloroethylene acting as the suspected foci. Once potential problem areas are identified, more localized methods may be applied.