West Nile in North America: A Florida Perspective

We in Florida have watched the 2000 West Nile virus (WNV) outbreak in the northeast with intense interest. WNV is apparently established in North America, and likely will spread to other regions. Given Florida's experiences with St. Louis encephalitis (SLE), most predict that WNV will arrive in Florida. It is only a matter of time.

Facts About WNV in the Northeastern U.S.

  • What have we learned from the northeast outbreak? Here are the facts:
  • Mosquitoes transmit WNV to a wide variety of wild and domestic avian species.
  • Several mosquito species have been found with WNV or WNV nucleic acid.
  • There have been numerous reports of WNV positive birds from several states. Hundreds have been identified, most with polymerase chain reaction (PCR) that detects WNV nucleic acid.
  • There have been numerous reports of WNV positive mosquito pools from several states, many based on PCR.
  • Sentinel chicken flocks have largely failed to detect WNV in the northeast.
  • Human cases have been reported. The majority this year from Staten Island which is apparently an epicenter. Scattered horse cases have also been reported.

These facts are based on information derived primarily from the media, various web sites, ProMED-mail, and CDC. How do we interpret these facts and are they useful to us in Florida?

Needed Information:

It has been difficult to obtain meaningful data during the northeast WN outbreak. Detailed data are necessary to evaluate mosquito transmission to birds, assess the role of specific mosquito species in transmission, and to evaluate real-time transmission in specific areas. We need this information to assess potential transmission patterns if and when WNV arrives in Florida.

Donald Shroyer, of the Indian River Mosquito Control District, recently asked some probing questions on ProMED-mail about sentinel chicken surveillance in the northeast. At this time he has not received a single reply. What diagnostic tests were used to detect WNV? How many were PCR positive? Have there been WNV isolations? How many? What serological tests have been used?

Here are three additional questions:

  1. What is the correlation between PCR-positive birds and live WNV isolations? A PCR-positive dead bird says nothing about where or when the bird was infected. Even PCR-positive mosquito pools may be misleading.
  2. Could a blood meal containing WNV nucleic acid, but not infectious WNV, be reported as a "positive"? It is essential to know the number of actual virus isolations, and particularly how many attempts to isolate virus from PCR-positive samples have failed.
  3. What are the denominators in the reporting system? It is essential to know the frequency of positive birds and positive mosquitoes. Without information about the percent of positive birds and mosquitoes in a region, how are we to gauge the meaning of 200 infected birds (or mosquitoes) vs. 500? Differences could simply be a matter of relative sampling effort or some other artifact.

Sentinel Chickens and WNV

Florida's extensive experience with SLE virus shows that surveillance using sentinel chicken flocks can forecast SLE transmission to humans. The Florida surveillance program provided real-time data that predicted the 1990 SLE epidemic that caused 226 clinical cases and 11 deaths. What do the reports of the failure of sentinel chickens to detect WNV in the northeast mean to Florida? Again we need information and answers to many questions. What diagnostic tests are being done on chickens? Where are flocks placed? Are the chickens of the correct variety to attract vectors? Given the apparent ease in detecting WNV nucleic acid in the northeast, why the failure of sentinel chickens? Could WNV transmission be focal, as is eastern equine encephalitis (EEE) in north Florida? If so, then flock placement for WNV detection, as for EEE virus, becomes critical. Indeed humans in the northeast seem to be serving as sentinels for the chicken flocks! If WNV transmission is focal, then the extensive spread of WNV in the northeast, as indicated by PCR-positive wild birds and mosquitoes, might be misinterpreted as evidence of real-time transmission at the locations where the positive birds and mosquitoes were collected. The birds/mosquitoes may be infected focally and then move. If WNV transmission in the northeast is similar to SLE transmission in rural Florida, and there is a massive ongoing avian epizootic, why has the virus not burned out in the avian amplification cycle where susceptibles should be rapidly declining in numbers? If WNV is not focal, then why the failure of the sentinel chickens? Could the current outbreak be a collection of dispersed but focal transmission areas? Would sentinel chickens be a useful indicator of more dispersed transmission? Would involvement of the sentinels serve as the forecaster of an even larger epidemic involving hundreds or thousands of cases? We need information!

Recommendations for Florida:

In Florida we must keep our eyes on the ball. The best indicator of real-time transmission is a sentinel surveillance program, whether the sentinels are chickens, or horses. It is unacceptable to use humans as sentinels. Florida relies on an established and interpretable sentinel surveillance program that includes chickens, vectors, known wild avian amplification hosts and meteorological patterns that allow us to predict human cases of SLE. Chickens can serve to measure real-time transmission. However, 1) they must be placed so they are exposed to vectors, 2) they must be seronegative when set out, 3) they must be monitored frequently using an appropriate laboratory test to determine real-time transmission and 4) seropositives must be replaced with seronegatives as soon as seroconversion is detected. The number of sentinel chickens that are infected relative to the number exposed provides a basis to make comparisons, assess transmission, and make predictions.

The information delivery system during the 2000 northeast WN outbreak is focused primarily on detection, not on surveillance and to date there are few details that help evaluate the true extent and magnitude of the current outbreak. Detection systems only provide limited information. The information we now need must come from an extensive surveillance program. Hopefully, we will eventually be provided with useful surveillance data from the northeast outbreak in the form of scientifically peer reviewed data. This surveillance data will allow a better capability to predict the risk of WN transmission to humans and horses. For more information on these issues see "West Nile Virus in North America: What's happening in 2000" on the West Nile virus page of the FMEL home page at

Walter J. Tabachnick, Director
Jonathan F. Day, Professor
Florida Medical Entomology Laboratory