Lyme Disease Risk Map and Prioritizing Areas for Vaccine Deployment

Disease surveillance, including risk modeling, is beneficial for infectious diseases. Risk prediction maps can be used to prioritize areas to deliver vaccines reducing disease transmission. For zoonotic infectious diseases such as Lyme disease, potential vaccine deployment areas include cost-effective places key to vectors, wildlife hosts, and humans. This paper uses machine-learning techniques to develop models to predict Lyme disease risk, then uses the critical variables within the models, as well those considered principal from previous literature, to create a Lyme disease risk map and to prioritize areas for wildlife vaccine deployment. It was found that highest disease risk is in the eastern United States, especially the upper Midwest and Northeast regions, which coincides with previous literature. The study found national parks and counties within these areas to prioritize wildlife vaccine deployment. Future work includes ground-truthing of the risk map and dispersal of vaccine to priority areas.