So how has the U.S. fared in coping with West Nile virus? Human cases and fatalities have risen sharply since 1999. The CDC reported that in 2002, 4,156 Americans became infected. More than 280 of the 4,156 people died. The virus was found in 44 states.

This year, the CDC reported that the virus is spreading more quickly and widely than in the past. The latest figures for 2003 as of September 2 are 1,764 cases, including 31 deaths, in 34 states. And transmission is expected to increase over the next few weeks as the mosquito season peaks.

West Nile virus experts at HSPH caution against panic. Fewer than one percent of mosquitoes are thought to carry the virus, and fewer than one percent of people bitten by an infected mosquito become severely ill, explained Richard Pollack, instructor of tropical public health in the Department of Immunology and Infectious Diseases. According to the CDC web site, most people who are infected will not develop any symptoms.

In addition, said Pollack, "Once this infection has swept through a region, transmission becomes much less intense."

That point has been demonstrated this summer. New York, where the virus was first identified in the U.S. in 1999, has reported only two cases of human infection in 2003, as of September 2. Colorado, on the other hand, leads the nation in West Nile virus infection with 635 cases, after reporting only 14 cases in 2002. State health authorities have suggested that a rainy spring and hot summer created ideal conditions for mosquito breeding. Along with Colorado, the states of Nebraska (259 cases), South Dakota (250 cases) and Wyoming (140 cases) account for nearly 73 percent of cases nationwide so far.

One tantalizing explanation for the spikes in West Nile transmission in western states and for the decreases in eastern states where the virus first appeared in the country may be that birds build immunity to the virus over time, thereby reducing virus spread in areas that have already seen infections.

Certain kinds of birds and mosquitoes appear to interact in a way thought to propagate West Nile virus. Female mosquitoes feed on birds in order to produce the yolk that fills their eggs. In the process, such mosquitoes may ingest virus from an infected bird and later pass the virus through their saliva to other birds and animals, including humans.

Much remains to be understood about West Nile virus. Definitive epidemiologic studies have been difficult to implement. "There can be a two-week lag between individuals getting infected and reporting symptoms," explained Anthony Kiszewski, an instructor in the Department of Immunology and Infectious Diseases. "When you go back to investigate, the birds and mosquitoes that may have caused the problem are gone."

Kiszewski is studying sampling methods for analyzing mosquito populations, a complex process. Birds responsible for infection may die or migrate before they can be tested. More than a dozen kinds of mosquitoes may be present at any one trapping site, and the mix changes from week to week. Traps must be designed specially for many different kinds of mosquitoes.

Despite collecting a wealth of data, scientists are not completely confident that they know which kinds of birds and mosquitoes propagate the virus–and the culprits may differ according to geographic region. Andrew Spielman, professor of tropical public health in the Department of Immunology and Infectious Diseases, suspects that one communally roosting bird, the house sparrow, is most likely the major "reservoir" of West Nile virus.

"They form large roosts in the same places every night, which may aid transmission of the virus from bird to bird," he said. Mosquitoes can potentially bite an infected bird, then pass the infection along.

HSPH researchers are now attempting to determine whether these communal bird roosts contribute significantly to the spread of West Nile virus.

Spielman, Pollack, Kiszewski and their colleagues are also working to define the duration of the transmission season and to determine how the virus survives the winter in the bodies of hibernating mosquitoes.

In addition, they are collaborating with mosquito control program managers to assess the efficacy of various interventions.

"A great deal is being done to reduce risk due to West Nile virus, but we have no idea about the efficacy of such measures," asserted Spielman, who is seeking to devise improved intervention strategies.

He suggested that a persistent insecticide might be applied around the few communal bird roosts that form in any region, an approach that he regards as potentially both economical and environmentally benign. The insecticide, he said, would reduce the number of mosquitoes feeding on the birds, while not affecting the birds themselves.

If an effective oral vaccine can be developed, vaccinating birds against West Nile infection using specially treated food in backyard feeders is another possibility. Scientists are already tackling the challenge, with mixed results. A recently reported study found that an injectable formulation of a West Nile virus vaccine was more effective in preventing infection in birds than an oral formulation.

"Avenues like these are much more likely to do something meaningful against West Nile virus than simple reflexive measures, such as widespread insecticide spraying," said Spielman.

More information on West Nile virus is available at http://www.hsph.harvard.edu/mosquito/.

--Mark Dwortzan