Assessing Spatial Relationships Between Mosquito and Human West Nile Virus Cases Using Arcgis, Harris County, Texas

BACKGROUND:  West Nile virus (WNV) is endemic to Texas and an important mosquito vectored disease of humans in Harris County.  Efforts to curtail the spread of WNV to humans in Harris County began in 2002 when the virus was first identified in mosquitoes.  The Harris County Public Health & Environmental Services – Mosquito Control Division conducts surveillance of mosquitoes and dead birds to detect WNV and the Epidemiology Program gathers surveillance information on human cases.  Chemical control is used in parts of the county identified with WNV positive mosquitoes.  Using GIS to identify clusters of WNV positive mosquitoes associated with WNV human cases can help minimize human exposure by revealing focal areas for treatment of mosquito vectors and preventative measures in humans.

METHODS:  Geographical location data on WNV positive mosquitoes and humans from 2002 – 2013 were mapped and analyzed using ESRI ArcGIS 10.2.1 to determine spatial relationships between WNV positive mosquitoes and human cases.  General viral concentration was determined by Choropleth mapping.  More rigorous analyses were conducted using the Optimized Hotspot Analysis tool in ArcGIS.  This allowed identification of statistically significant hot and cold spots using the Getis-Ord Gi statistic.  Additionally, a Pearson correlation on frequencies of WNV positive mosquito and humans was calculated.

RESULTS: The overall trend for data from all 12 years revealed significant hotspots in the northwest and central parts of Harris County, with a large cold spot in the eastern part of the county.  The highest numbers of human cases were observed in the mosquito hot spots, with relatively fewer human cases in the cold spots.  There was a significant correlation between the number of positive mosquitoes and the number of human cases in the same areas (r = 0.235, p < 0.001, N = 268).  In 2012 there were a relatively large number of human WNV cases.  Hot spot analysis of this year alone reflected the overall results, with most human cases falling in the hot spot of positive mosquitoes. 

CONCLUSIONS:  These results show that mapping of vector mosquitoes and human WNV case locations can show spatial relationships that can be used to manage the vectors and help prepare humans in areas that have high WNV activity.  The use of the Optimized Hotspot Analysis provides statistical evidence of areas with high or low activity, which can be extremely useful in mosquito management and is critical for disseminating information to the scientific community.