Guidelines for Arbovirus Surveillance Programs in the United States
Click here for CONTENTS Page
For contact details regarding publication see Title Page

Types of Surveillance Systems

Vertebrate surveillance systems for arboviruses collect qualitative and quantitative information about the presence, distribution, intensity and temporal and spatial fluctuations in virus activity. Information can be obtained by testing specimens collected for some other purpose (passive system) or by collecting and testing specimens from vertebrates captured specifically for the surveillance program (active system). The data can be used as background information or to direct mosquito control operations to reduce the risk of human exposure. Examples of the use of vertebrate surveillance systems and useful sentinel hosts are listed below.

1. Presence and distribution of arboviruses in specific geographic area. This usually is a one time, simple, qualitative survey. It is useful to provide background information, usually detecting prevalence of antibody in free- ranging sentinels, at local, regional, or state level. The possibility of non-specific reactions should be kept in mind in this type of study.

a. Passively-collected specimens (i.e., collected for other purposes)

1)Hunter-killed wild ungulates - statewide (EEE, SLE, WEE, JC, LAC)

2) Trapped coyotes - predator control projects (WEE)

3) Trapped red fox - fur trappers (LAC, EEE, JC)

4) Rabbits or hares - trapped or hunter-killed (WEE, LAC)

5) Waterfowl - hunter-killed or trapped (WEE, EEE, SLE)

6) Cattle - after brucellosis testing or slaughter (WEE, JC)

b. Actively-collected specimens at selected locations 

1) Wild birds (including pigeons & house sparrows) (EEE, SLE, WEE)

2) Chicken flocks (EEE, SLE, WEE)

3) Raccoon (SLE, EEE, WEE)

4) Cotton rat (or other rodents) (SLE, EEE)

5) Eastern chipmunk and tree squirrels (LAC)

6) Domestic dog (SLE, LAC)

7) Equine (EEE, WEE, JC)

8) Farm flocks (WEE, EEE, SLE)

2. Annual changes in arbovirus activity. These systems detect changes in frequency or distribution. They may be qualitative or quantitative. These generally are passive systems, and use same animal species described above. Measures include the prevalence of antibody and sometimes virus isolation. The vertebrates are generally free-ranging sentinels, although captive sentinels like chickens are sometimes used at the local-state level
3. Seasonal changes in arbovirus activity. These systems detect changes in frequency of virus or antibody. They are generally active and quantitative. The prevalence of antibody or virus is monitored in both free-ranging and captive sentinels. Such programs are usually local or regional. They are important for establishing inter-epidemic prevalence rates.
4. Within season changes in arbovirus activity. These are active and quantitative systems that monitor the prevalence of antibody or virus in tagged, free-ranging, or captive sentinels. These programs are usually local in areas with history of disease. They are important for monitoring increasing and impending risk for the human population.
5. Investigation of an epidemic (unusual occurrence). Epidemic investigations are intensive, active and quantitative studies that measure the prevalence of antibody and virus in free-ranging sentinels. These investigations are usually local or occasionally regional in scope.

Two examples of well-established surveillance programs currently in operation at the local and state level are presented below. Both are effective surveillance systems. Surveillance programs must be structured to fit the specific expertise, resources, ecology, environmental conditions, and needs of the user.

A. LOCAL SYSTEMS - Memphis, Tennessee

1. This system relies on biweekly capture of free-ranging house sparrows with mist nets at 21 sites throughout the metropolitan area from April to November. Birds are aged, sexed and tagged and a blood specimen taken before they are released at the capture site.
2. From May to October, sentinel chickens are placed at selected sites with a history of human SLE. The chickens are bled biweekly, and positive birds are re-bled for confirmation and replaced.
3. Blood samples from house sparrows and chickens are tested for SLE viral antibody within 1 day of collection by the HI or ELISA test.
4. If immature house sparrows or sentinel chickens are antibody positive, additional house sparrows are sampled within the same week at positive and adjacent sites.
5. Rapidly increasing SLE viral antibody prevalences in either sentinel system will alert the mosquito control personnel to intensify insecticide application around the positive sites or throughout the city.
6. The advantage of this system is that the surveillance and testing of sentinel birds are under the same administration as the mosquito control operations. Therefore, there is little delay in sampling and testing. More important, there is no delay in communication of results. The efforts are coordinated. Re- sampling and testing of sentinels as well as initial mosquito control can be concentrated specifically in the problem areas. There is little delay in responding to an impending risk of human disease.
7. The disadvantages of this approach include the cost of equipment and supplies, problems in establishing and maintaining quality control, and the problem of test standardization among local agencies. The cost of upgrading or changing to new technologies can be prohibitive for a local agency. Data are generally available only for a small geographic area, and nearby focal activity may not be detected. Thus, a sense of security created by treatment of identified foci of transmission could be rudely interrupted by the spread of infection from un-monitored areas.

B. STATE SYSTEMS - California State Health Department

1. Sentinel chicken flocks are set out in early spring (April-May) in pre-selected areas throughout the state. Collaboration with local mosquito control districts is emphasized.
2. Flocks of 10 chickens are bled biweekly and tested for WEE and SLE antibody at the Viral and Rickettsial Disease Laboratory (VRDL) at Berkeley.
3. Mosquitoes, mostly Cx. tarsalis, are collected and pooled by the mosquito control districts and tested by the VRDL by means of an in situ ELISA test.
4. Seroconversions in chickens and virus-positive mosquito pools are reported to all agencies by telephone or facsimile, as well as in the weekly VRDL reports (which also are available through the "Mosquito Net" computer bulletin board service).
5. Mosquito control operations are intensified, emphasizing adulticiding in populated areas, depending upon the findings on vector abundance, virus isolations from mosquitoes and the human population at risk. Mosquito collections for virus isolations are intensified at the positive sites and in areas adjacent to population centers.
6. Passive reporting of suspected clinical WEE horse cases and submission of specimens for confirmation is encouraged. VRDL tests specimens for virus isolation and diagnostic rise in antibody, and reports results to the local health agency and to the veterinarian.
7. Virus surveillance activity and mosquito control operations are intensified at localities where early season (May-June) confirmed cases of WEE in horses are reported. If WEE virus is isolated from mosquito pools, local control agencies notify veterinarians and encourage them to vaccinate young and recently imported equines.
8. Advantages of this system include centralized access to advanced technology and highly trained personnel, greater ease of standardization and quality control, and state-wide comparability of results. Large geographic areas can be sampled on a routine basis. Use of the "Mosquito Net" BBS allows for rapid and widespread reporting of information to those agencies with access to the BBS.
9. Disadvantages of this system are mostly in turnaround time, particularly for seroconversion in chickens. There is a period of about 7 - 10 days after infection before antibodies are detected. Specimens are collected locally, packed, and sent to the state laboratory, which takes another 2 days. An additional 2 days are required for testing, for a turnaround time of 11 - 14 days. Since birds are bled biweekly, an additional 14 days are added for birds that have been infected but are not yet seropositive. Thus, delays of 25 - 28 days are possible between the infection of a sentinel chicken and detection of seroconversion.

Examples of Vertebrate Species Used in Surveillance Programs

Surveillance programs and epidemic investigations use many species to assess the potential for arboviral encephalitis in the United States. Table III-1 lists the most common species used.

Table III-1. Common birds and mammals for arbovirus surveillance in the United States.¹⁷⁵