Creating a Predictive Model for Monitoring West Nile Virus in Denton County, Texas

Keywords: West Nile Virus, Texas, Predictive Analysis, Public Health, GIS
Abstract Type: Paper Abstract

Abstract

Since 2002, West Nile Virus (WNV) has become the leading vector-borne disease in Texas. While annual case counts remain relatively low, spikes do occur, including the 2012 outbreak with a total of 1,868 cases. To monitor the threat, public health departments survey mosquito populations caught in strategically placed traps. However, outbreaks are unpredictable, and this makes accurate forecasting difficult. Predictive risk modeling provides a useful tool for monitoring WNV risk, especially as changes in climate patterns, such as increasing temperatures, can accelerate mosquito populations and transmission rates. This paper presents a predictive risk model of Denton County using mosquito surveillance data over a ten-year period sourced from the county’s public health department. Variables collected include mosquito trap location, date of collection, WNV test results, total mosquito count, and species. Using GIS spatial analysis and predictive analytics this research creates a predictive risk model for visualizing clusters of high-risk zones throughout the county. The result provides a useful tool for estimating future WNV prevalence and forecasting spatial patterns of cases.