A1246
Title: Detecting disease clusters using functional additive models with spatial dependence
Authors: Michio Yamamoto - The University of Osaka / RIKEN AIP / Shiga University (Japan) [presenting]
Tatsuhiko Anzai - Institute of Science Tokyo (Japan)
Kunihiko Takahashi - Institute of Science Tokyo (Japan)
Abstract: Identifying spatial disease clusters is essential for characterizing disease patterns and informing prevention and treatment strategies. Spatial scan statistics are widely used for cluster detection with a variable scanning window size. When covariates influence the outcome and are not randomly distributed across space, cluster searches should adjust for them. Moreover, spatial correlation in the outcome, which is often overlooked in practice, can affect detection. The aim is to propose a new spatial scan statistic that accommodates multiple functional covariates summarizing longitudinal histories and explicitly accounts for spatial correlation in the outcome. These factors are modeled flexibly within a functional additive modeling framework. An optimization algorithm is developed for parameter estimation under Gaussian outcomes. Simulation studies and an application to real data show that, compared with existing methods, the proposed approach reliably detects disease clusters in the presence of longitudinal covariates and spatial correlation.
