Integrating Environmental Monitoring and Mosquito Surveillance to Predict Vector-borne Disease: Prospective Forecasts of a West Nile Virus Outbreak

The objective of this study was to evaluate a predictive model of West Nile virus in South Dakota (SD), which was used to make whole-year forecasts at the beginning of the 2016 WNV season, as well as short-term weekly forecasts throughout the season. This modeling approach was unique in that it incorporated not only weather data but also WNV infection data to predict spatial and temporal patterns of WNV cases. Because the model was specified and parameterized prior to the start of the 2016 WNV season, this exercise provided an opportunity to make a priori predictions and then validate them using independent data. The model’s estimates were generally accurate for 2004-2015, as were its predictions for the 2016 WNV season. In particular, warmer weather in early 2016 indicated that the WNV season would begin earlier than usual, and high mosquito infection rates detected early in the WNV season provided further evidence to support predictions of a high WNV human case burden. Materials and Methods Data sources Human case data: Records for 1,249 human WNV cases from 2004 to 2015 were provided by the South Dakota Department of Health (SDDOH). While there were 1,076 human cases in 2002-2003 alone, the large number of cases in these initial epidemic years reflected the unique conditions when WNV was first introduced into the region 27. As a result, these atypical years were excluded and we focused on endemic WNV from 2004 to the present. All disease cases were laboratory-co...
Source: PLOS Currents Outbreaks - Category: Epidemiology Authors: Source Type: research