Free bird: Plans to contain a flu pandemic will most likely fail, experts say. Credit: Sabah Arar/afp/Getty Images

As the H5N1 avian flu virus continues its relentless march westward, scientists are scrambling to prepare for the possibility that it will mutate into a form that can jump between—and kill—humans. Top among the strategies is a plan to snuff out an outbreak right where it begins, but few experts believe that will work.

Quarantine and other measures limited outbreaks of the severe acute respiratory syndrome in China, but won't contain influenza, says Yoshihiro Kawaoka, professor of virology at the University of Wisconsin in Madison (see page 489). “There's no way, it won't work. Influenza is different,” he says.

If there is a chance to stop a pandemic, we have a responsibility to try it. Maria Cheng, World Health Organization

In August 2005, epidemiologists modeled outbreaks in Southeast Asia and concluded that, given early warning and enough Tamiflu, a human outbreak can be contained (Science 309, 1083–1087; Nature 437, 209–214). But in February, Harvard University researcher Marc Lipsitch argued that the virus is likely to jump to humans multiple times, making containment nearly impossible (PLoS Med. 3, e135).

“If [the jump] can happen in one place, it can happen in more than one place,” Lipsitch says. “If it is introduced into more than one site, containment won't work.”

Still, the World Health Organization (WHO) is embracing containment in its strategy, and plans to fund it out of the avian flu monies pledged at a conference in Beijing in January. A draft containment plan released in March lays out guidelines for national authorities to investigate a cluster of ill people, as well as for launching a full-blown containment effort. The two-phase plan will focus first on tracing the chain of infection, giving antivirals to those closest to the ill, monitoring health and preparing hospitals for highly infectious patients. The second phase may include voluntary and involuntary quarantine, closing of schools, churches, public transport and borders, and the large-scale distribution of antivirals.

The pharmaceutical company Roche has donated 5 million courses of Tamiflu to be rushed to the site of an outbreak. The goal is to get the doses distributed within 21 days, a possible tipping point into chaos identified by the models.

Even the skeptics, including Lipsitch, say the plan is worth a shot. The strategy might buy time to make more vaccines, for instance, might slow the pandemic down, and might improve the infrastructure of the countries, notes Imperial College London epidemiologist Neil Ferguson, who published one of the August models.

“If there is a chance to stop a pandemic, we have responsibility to try it,” adds Maria Cheng, a spokeswoman for the WHO. “It is a worthwhile effort, even if it fails.”