Early Warning System
In order to improve our ability to identify areas of human risk for acquiring West Nile virus, we have developed a tool for prospective, daily analysis of dead bird reports to detect small area clustering in time and space. We have used this tool in New York City prospectively since 2001 to target public education and surveillance activities as well as timelier larval and adult mosquito control. The system requires minimal resource inputs and utilizes routinely collected data to provide real-time support for decision making to public health practitioners.
The methods described on this web page and in the recent Emerging Infectious Diseases article (Vol. 9, No. 6, June 2003 pages 641-646) are based on census tract aggregation of dead bird reports. We have made substantial improvements to the methods described here, in particular the use of real space point data for each dead bird rather than census tract aggregation, examination of elliptical clusters as well as circular windows, and flexible windows in time as well as space. We are in the process of evaluating these different methods and preparing a manuscript for publication. As such, this web page will be updated frequently and we ask you to check back occasionally for additional results and SAS code.
