Spatial and spatial-temporal clustering analysis of hemorrhagic disease in white-tailed deer in the southeastern USA: 1980-2003.

Abstract

We used the space-time K function and Kulldorff's scan statistic to analyze the spatial and spatial-temporal clustering of hemorrhagic disease (HD) in white-tailed deer in Alabama, Georgia, South Carolina, North Carolina, and Tennessee. The HD occurrence data were binary presence/absence data acquired annually on a county basis from 1980 to 2003. Space-time K function was employed to globally examine the existence of spatial-temporal clustering in the HD data. Three approaches of Kulldorff's scan statistic, i.e., spatial clustering analysis for the entire period, spatial-temporal clustering analysis, and spatial clustering analysis by individual years, were applied to detect potential HD clusters. Statistically significant spatial clusters and spatial-temporal clusters were detected in the five southeastern states during the 24-year study period. Some clusters were observed in multiple years. Clusters were most evident in west Alabama, south Alabama, central South Carolina, and along the border between South Carolina and North Carolina. The identification of HD clusters may provide a means to better understand the causal factors related to the HD outbreaks. Results also have potential application in improving or designing effective surveillance programs for this disease.