A diagnostic algorithm to serologically differentiate West Nile virus from Japanese encephalitis virus infections and its validation in field surveillance of poultry and horses.

The detection of West Nile virus (WNV) in areas endemic for Japanese encephalitis virus (JEV) is complicated by the extensive serological cross-reactivity between the two viruses. A testing algorithm was developed and employed for the detection of anti-WNV antibody in areas endemic for JEV. Using this differentiation algorithm, a serological survey of poultry (2004 through 2009) and horses (2007 through 2009) was performed. Among 2681 poultry sera, 125 samples were interpreted as being positive for antibodies against JEV, and 14 were suspected to be positive for antibodies against undetermined flaviviruses other than WNV and JEV. Of the 2601 horse sera tested, a total of 1914 (73.6%) were positive to the initial screening test. Of these positive sera, 132 sera (5.1%) had been collected from horses that had been imported from the United States, where WNV is endemic. These horses had WNV vaccination records, and no significant pattern of increasing titer was observed in paired sera tests. Of the remaining 1782 positive sera 1468 sera (56.4%) were also found to contain anti-JEV antibodies, and were interpreted to be JEV-specific antibodies by the differentiation algorithm developed in this study. The remaining 314 horses (12.1%) for which a fourfold difference in neutralizing antibody titer could not be demonstrated, were determined to contain an antibody against an unknown (unidentified or undetermined) flavivirus. No evidence of WNV infections were found during the period of this study.