during a quiescent period often leads to reduction in political, and hence financial, support for surveillance activities. Ironically, it is during these quiescent periods that surveillance would best be conducted, to detect the appearance of an infectious agent in environmental reservoirs and eradicate the agent before it infects humans. When surveillance is abandoned, an outbreak of human disease may be well under way before it is detected.
Of course, surveillance is of little use if not shared with other groups or individuals who can act on the information to prevent or diagnose disease. In Iowa, for example, the University Hygienic Laboratory has posted influenza surveillance data on its web site (www.uhl.uiowa.edu/HealthIssues/index.html) for the past several flu seasons. Data are updated automatically each night. Anyone wanting to know which viruses are circulating in their area can easily view a table that shows numbers of Influenza A and B detected during the current week, the past week, or all year. This information may influence a decision to administer prophylactic or therapeutic drugs or to control exposures. The web site also provides the latest data on a variety of other diseases, including some zoonoses. Sharing data with those who participate in its reporting and accumulation will encourage timely reporting and dialogue between the private and public health care communities.
In our efforts to improve surveillance for zoonotic agents, we can learn much by analyzing current programs, such as the Centers for Disease Control and Prevention (CDC)’s Emerging Infections Program and its Epidemiology and Laboratory Capacity Program. In addition, examining some of the diseases that public health laboratories now face may help to illustrate some possible new approaches to designing surveillance strategies, as well as possible means to sustain them.
Surveillance for arboviruses in the United States has focused primarily on three types of encephalitis viruses: St. Louis, Western equine, and Eastern equine viruses. One method of surveillance involves trapping mosquitoes, pooling and enumerating individual species, and checking extracts of the pools for the presence of these viruses. This activity is labor intensive and limited in sensitivity by the selection of the location and number of sites for mosquito collection. Another method of surveillance is to use birds, often chickens, as sentinels for the appearance of virus in a region, since the mosquitoes that transmit these viruses preferentially feed on birds. Surveillance can be accomplished by sampling wild birds or by housing chickens outdoors and bleeding them periodically during the summer to monitor for the appearance of antibodies to an encephalitis virus. Seroconversion requires time, however, and may not occur sufficiently early to signal the