Modeling spatial risk of the Foot-Mouth-Disease epidemic in South Korea

The 2010/2011 food and mouth (FMD) disease epidemic in South Korea spread nationwide and resulted in substantial economical damage. Its rapid transmission throughout the country despite the government's control policy argues for the need to improve our understanding of its spatial dimensions. The spatial point patterns of FMD incidences revealed that FMD incidences formed statistically significant spatial clusters during the early phase of epidemic but its significance and the spatial dimensions dynamically evolved over the course of epidemic. We developed a log-Gaussian Cox process (LGCP) model to quantify and to make inferences on the spatial distribution of FMD risk, while accounting for both known risk factors and unexplained variation. The FMD risk surface estimated from the LGCP model captured the spatial heterogeneity of FMD risk driven by environmental and landscape factors and unknown risk variations at both local and regional scales. This study demonstrated that point pattern analyses and statistical models enabled investigators to improve our understanding on the spatial dynamics of FMD, which may provide useful information for decision makers to make efficient and effective control strategies for future outbreaks.