A0324
Title: Detection and evaluation of multiple clusters in spatial epidemiology
Authors: Kunihiko Takahashi - Nagoya University Graduate School of Medicine (Japan) [presenting]
Hideyasu Shimadzu - Loughborough University (United Kingdom)
Abstract: A number of statistical tests have been proposed and are widely used in spatial epidemiology to investigate a regional or temporal tendency in the presence of certain diseases, whether the disease risk is relatively high to other surrounding regions or subsequent time periods. The scan statistic is one of the most powerful elements of the cluster detection test to detect and evaluate spatial and/or temporal disease clusters since it is based on the maximum likelihood ratio; examples include the Kulldorff's circular scan statistic along with the SaTScan software, and Tango and Takahashi's flexibly shaped scan statistic implemented in the FleXScan software. Although multiple clusters in the study space can be thus identified, current theoretical developments are mainly based on detecting a ``single'' cluster. The standard scan statistic procedure enables the detection of multiple clusters, recursively identifying additional ``secondary'' clusters. However, their p-values are calculated one at a time, as if each cluster is a primary one. Therefore, we proposed a new test procedure that can accurately evaluate multiple clusters as a whole, combining generalized linear models with an information criterion approach that selects an appropriate number of the clusters. This framework encompasses the conventional detection procedure as a special case. We present practical examples applying the proposed procedure and compare the results with ones by conventional procedures.
