Title: A detection test for adjacent hotspot clusters
Authors: Kunihiko Takahashi - Tokyo Medical and Dental University (Japan) [presenting]
Hideyasu Shimadzu - Loughborough University (United Kingdom)
Abstract: Several statistical tests have been widely proposed and 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 potent elements of the cluster detection test based on the maximum likelihood ratio for detecting and evaluating spatial and/or temporal disease clusters; examples include Kulldorff's circular scan statistic along with the SaTScan software, and Tango and Takahashi's flexibly shaped scan statistic implemented in the FleXScan software. These scan tests usually deal with a hotspot cluster model assuming a constantly elevated disease rate within a single cluster. However, they can often detect adjacent hotspot clusters exhibiting different disease rates as though they are a single hotspot cluster. This talk proposes a new test procedure that can accurately evaluate these adjacent hotspot clusters as multiple clusters. We present practical examples applying the proposed procedure and compare the results with ones by conventional procedures.