Sentinel Chicken Surveillance for West Nile Virus

Matt Yates raises important points that need to be carefully considered when embarking on a West Nile virus (WNV) surveillance program. The use of gravid traps to capture ovipositing females (= older females that have blood fed at least once) that are pooled to determine mosquito infection rates has clearly been essential for the identification of high risk WNV transmission foci in East Baton Rouge Parish Abatement and Rodent Control District. Well done EBRPMARC.

A few additional points need to be considered relative to WNV surveillance and disease transmission risk assessment. WNV-positive mosquito pools will always precede WNV antibody-positive sentinel chicken seroconversions when both collections are made in the same location. Seropositive sentinel chickens will always lag behind positive mosquito pools from the same location by at least two weeks -- the time needed for newly infected sentinel chickens to produce detectable levels of antibody. Hence, in the context of the EBRP time frames presented above, it is not surprising that WNV-positive mosquito pools preceded the WNV antibody-positive sentinel chickens assuming that all of the sentinel chickens maintained in EBRP were bled weekly. Therefore, whenever there are prior surveillance data that indicate repeated WNV transmission in a particular area, it is prudent to conduct mosquito surveillance and the pooling of gravid females for WNV isolation attempts to obtain a more timely assessment of active WNV transmission. This is precisely what is done by mosquito control workers in EBRP. Their observation that WNV-positive mosquito pools preceded the onset of West Nile human cases in urban settings, while WNV seropositive sentinel chickens did not, is likely a significant observation for WNV transmission dynamics within the EBRP.

Gravid mosquito pools certainly "detect" the presence of WNV prior to the sentinel chickens. This is important if the mosquito infection rates can be used as a gauge for actual later human risk and the magnitude of the risk as we have advocated for sentinel chicken seroconversion data (Tabachnick, WJ. 2004. Sentinel surveillance and human risk for West Nile. Buzzwords 4(2): 10-12). The first report of a WNV-positive mosquito pool or a WNV seropositive sentinel chicken (which may also precede the first human case purely by chance) is not always indicative of the risk of a significant epidemic of WNV within a particular surveillance zone. During low transmission years, these positive mosquito pools and/or sentinel chickens serve purely for "detection", and we should agree that WNV is here to stay in the western hemisphere and no longer needs to be detected. It is the magnitude and timing of mosquito infection rates, and/or the magnitude and timing of the sentinel chicken seroconversion rates that provide the essential information necessary to actually gauge the risk of a significant WNV epidemic in human populations on the local, county, statewide, or regional levels. The EBRPMARC bases its WNV transmission assessments and real-time vector control decisions on data obtained from pools of female mosquitoes collected in gravid traps. This mosquito surveillance system has worked well for the EBRP. Florida has yet to experience a major epidemic of WNV, so the surveillance data that will predict a "big event" in Florida remain unclear. However, many control programs in Florida will continue to rely on the sentinel chicken arboviral surveillance program that has been in place throughout the state since 1977. The resulting longterm surveillance data sets may be complemented by baseline mosquito pool data sets once they become established at already known arboviral transmission sites throughout the state.

The arboviral surveillance issues related to assessing WNV transmission by evaluating large numbers of pooled mosquitoes are discussed in an earlier BuzzWords column (Tabachnick, BuzzWords 6(4): 6-8). As outlined in this previous BuzzWords column, the bottom-line issue that needs to be considered by every arboviral surveillance program is the utility and efficiency of assessing large numbers of mosquito pools, which may provide an earlier indication of later WNV transmission to humans, weighed against the utility and efficiency of a sentinel chicken surveillance program capable of assessing actual transmission, knowing that there is a built-in two week lag between the infection date and the first detectable antibody in the sentinel chickens. The success of either arbovirus surveillance program is contingent upon the assumption that samples are collected from a location where there is actual ongoing WNV transmission. The intensity of that transmission, especially early in the transmission season, will be a major factor in determining how easily arboviral transmission is detected and how far in front of the eventual WNV epidemic curve vector control is initiated.

The value of any arboviral surveillance program is to provide enough lead time for local mosquito control and public health agencies to attempt effective, efficient, and environmentally sound vector control and epidemic mitigation. One thing we have learned during the tenure of WNV in North America is that there is a number of different ways to conduct sound arboviral surveillance. The method of choice used by the EBRP is mosquito pooling. Well done EBRP.

Walter J. Tabachnick, Professor and Jonathan F. Day, Professor
Florida Medical Entomology Laboratory, University of Florida/IFAS
Vero Beach, FL