
| HOME SCHEDULE AUTHOR INDEX SUBJECT INDEX |
|
Predicting the spatial dynamics of infectious diseases: raccoon rabies as a model system. Real, Leslie1, Smith, David2, Lucey, Brendan1, Waller, Lance1, Childs, James3, 1 2 3 ABSTRACT- Rabies in raccoons has been spreading north and south along the eastern US since its initial outbreak along the West Virginia/Virginia border in the mid-1970's. Cases of rabid raccoons have been reported by month to the CDC by all counties in the US. This database provides over 50,000 temporal and spatial locations of disease incidence for this epidemic. We used the database for the State of Connecticut to construct a predictive model of spatial dynamics across heterogeneous landscapes. The data were analyzed using both geostatistical and maximum likelihood techniques to uncover the relative effect of different habitat variables on the velocity of local propagation of the disease. Statistically significant variation in velocity of propagation can be associated with long-distance dispersal, river barriers, and human population corridors. We then constructed an Interactive Network Model (INM) to simulate rabies spread across heterogeneous landscapes. The INM associates transitions from a healthy township to a diseased township according to a probabilistic function of the disease state of neighboring townships and the geographical variables associated with the township. We used the statistical analysis to parameterize the INM. This sort of simulator represents an ideal approach to blending large databases and predictive theory of disease spread. KEY WORDS: infectious diseases, spatial dynamics, rabies |