Cancer Cluster Morphology

DESCRIPTION (provided by applicant): This project will develop a new, meta-analytic approach for evaluating cancer clusters of flexible shape called Cluster Morphology Analysis (CMA). To date, two of the major deficiencies of geographic studies of cancer a re that they often assume clusters have a specific shape (e.g. circle or ellipse) and do not evaluate statistical power using the geography, at-risk population, demographics, covariates and numbers of observed cases of the cancer under investigation. These limitations are overcome by this project. Power analyses will be conducted for 11 clustering techniques using a suite of plausible clusters of different sizes, relative risks and shapes. The results are then ranked by statistical power and by the proporti on of false positives, under the rationale that the objective of cluster-based cancer surveillance should be to (1) find true clusters while (2) avoiding false clusters. CMA then synthesizes the results of those clustering methods found to have the best st atistical performance. This approach is applied to pancreatic cancer incidence and mortality in Michigan, focusing on three counties that comprise a significant cluster that persists and grows from 1950 to the present day. CMA is a significant advance over clustering approaches that assume just one shape and rely on only one clustering method. The major innovation is the creation of methods and software for analyzing cancer incidence and mortality data to accurately identify flexibly shaped clusters defined by geographic sub-population of excess cancer risk. PUBLIC HEALTH RELEVANCE: The techniques and software from this project will provide a more concise and accurate description of cancer clusters via (1) the accurate detection of clusters founded on fle xible shapes, rather than on arbitrary shape templates such as circles and ellipses; (2) the automated evaluation of the statistical power of clustering techniques for the specific geography, cancer and sub-population being scrutinized by the software us er; and (3) Cluster Morphology Analysis that synthesizes results across clustering approaches to more accurately identify true clusters. To our knowledge the techniques and software from this project will be the first to address all of these factors within a single, comprehensive framework.