Evaluation of the West Nile Surveillance System for the State of Kansas

Abstract

Abstract: Background: West Nile virus (WNV) is an arboviral disease that has caused an estimated 29,624 clinical illnesses and 1,161 deaths in the United States since its emergence in 1999. A national WNV surveillance program was established by the Center for Disease Control and Prevention by providing states with grant funds to construct surveillance systems in 2000. Kansas launched statewide surveillance efforts in 2001. This project describes the evaluation of the WNV surveillance system in Kansas to determine its level of effectiveness as a public health tool including recommendations to improve the system.

Methods The surveillance system was evaluated utilizing the CDC’s 2001 MMWR Updated Guidelines for Evaluating Public Health Surveillance Systems. The surveillance system was also compared to the CDC’s Epidemic/Epizootic West Nile Virus in the United States: Guidelines for Surveillance, Prevention, and Control published in 2003. Key personnel in Kansas and neighboring states were interviewed during this evaluation. Mosquito pool collection data was evaluated for 2008 and 2009 for time lapse between collection and reporting of results. Records from Kansas’s Electronic Disease Surveillance System 2003-2009, were analyzed using SAS 9.1.3 to determine number of days between non-human cases and human illness onset dates. A WNV surveillance system survey was created and utilized to interview public health officials in 4 surrounding states. Results Mosquito pool collection is conducted in 13 of 105 Kansas counties by the Kansas State University Entomology Department between May and October of each year. For 2008 and 2009, the combined range of time between collection and reporting of mosquito pool results was 6-87 days with a median of 23 days. When positive mosquito data was compared to human onset dates for 2003-2009 the time between positive mosquito pools and positive human cases, ranged from 36 days prior to human illness onset to 82 days after with a median of 24 days after human onset dates.

Conclusions WNV is now considered endemic in the state of Kansas and is an established seasonal health threat for its residents. Mosquito pool collection data was shown to be a poor predictor of human disease. The timeliness of testing, reporting of results, and the evidence of human cases prior to detection in mosquito populations indicates that this method of surveillance is not providing adequate information to implement public health interventions. Resources would be better utilized if they were focused on educational efforts in disease prevention and mosquito control measures.