Abstract
One of the most powerful uses of GIS in the field of public health is as an exploratory data analysis tool. By combining the three post-input defining components of a GIS (data manipulation, data investigation, data analysis), the spatial understanding of a disease can be furthered by identifying patterns of cases, or associations between disease and other spatial phenomena (such as elevation). This chapter sets the groundwork for one such exploratory tool that could be used to identify the spatial and temporal patterns of an infectious disease. The disease in question is raccoon rabies in West Virginia during 1999-2000. The exploratory tool, animation, has the potential to give insights into an evolving disease pattern that current spatial cluster techniques could miss. The current raccoon rabies epizootic presents a complex spatial surface as multiple disease foci may be present. Added to this could be a residual “background” or enzootic level of rabies. In order to reduce the impact of multiple foci, an appropriate “scale” of animation is needed. This scale has to be of a small enough geographic area that only one disease focus is considered, and is of practical use so that other meaningful spatial information (such as land cover or elevation) can be interpreted. The purpose of this chapter is to decide on an appropriate method of identifying this scale of animation for an infectious disease of this type. This chapter will select one commonly used technique, Nearest Neighbor Hierarchical (NNH) spatial clustering, to identify the correct scale and location on which to perform an animation. NNH spatial clustering will be applied to three combinations of Raccoon Rabies data for West Virginia, for 1999, 2000 and both years combined. NNH cluster analysis will also be performed on a four-county area identified as having the highest intensity of rabies cases in the state. These results will then be compared to a preliminary animation of rabies cases in West Virginia from which subjects were asked to identify dynamically evolving disease clusters. An animation was also run for the same area of high disease intensity. Cluster and animation results were compared for similarities. It was found that a spatial cluster technique, such as NNH spatial clustering, provides an adequate means of identifying the scale and location on which a more sophisticated animation can be based. The chapter concludes with a discussion of how, once a scale has been decided, a more sophisticated animation can be constructed and ultimately used to guide the placement of interventions such as oral vaccine barriers.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have