To judge from reports that appeared in The New York Times and Washington Post, as well as Science magazine, it appears that the World Health Organization will recommend, over U.S. objections, full publication of the bird flu studies showing that an air-transmissible, mammal-to-mammal form of the H5N1 virus can be engineered. A Times story that appeared on Feb. 17 seemed to indicate the the decision to publish was final, though actual publication would take place only in a few months. A Reuters story published by the Post on Feb. 20 was somewhat more circumspect, indicating that publication would occur only after further review of security concerns.
Whatever the nuances, it seems clear that the potential benefits of publication–helping scientists develop means of guarding against a human-to-human air transmissible virus–are being carefully weighed against concerns about putting dangerous information into the hands of maniacs. Those concerns are not trivial, considering that once an engineered virus has been created, it obviously would be possible–albeit very difficult–for somebody to replicate the work. But nor should the concerns be over-stated (or the benefits of publication minimized.)
Today, Science magazine’s Science Express carries a scientific report suggesting that the fatality rate from the bird flu virus may have been over-estimated. Taia T. Wang and colleagues at the Mount Sinai School of Medicine in New York City point out that current fatality estimates are based on 573 WHO-confirmed cases involving hospitalization. They surmise that only the most severely infected are hospitalized, and that many more individuals may be infected but undetected.
The Mount Sinai scientists did a meta-analysis of studies evaluating serological evidence of H5N1 infection, including studies specifically of poultry workers. They found a general incidence of mild infection of 1.2 percent, with a 95 percent confidence interval of 0.5 to 3.4 percent, and estimated incidence among poultry workers ranging from 0.5 to 3.2 percent in several different groups.
Wang and colleagues conclude that incidence of the bird flu virus may be not merely in the hundreds or thousands, as assumed so far in fatality estimates, but in the millions.
Another reason not to over-react: Linked-In’s Security and Terrorism group has been hosting a discussion prompted by the observation that “since 1978 there have been approximately 20 separate biological attacks in the U.S.” With such a low rate of occurrence,” asks national security analyst McArthur Billing Jr., “are federal, state, and local biodefensive measures and centers justified?”