The last pandemic in the U.S. occurred in 1918.  Hospitals like this one -- set up at a military camp in Kansas -- were set up to deal with the emergency.  (National Museum of Health & Medicine photo)

Public health officials across the globe want answers to the same urgent question: What will they do if the deadly H5N1 avian flu virus mutates and sparks a human pandemic? Since 2003, the virus has spread to Asia, the Middle East, Europe and Africa killing millions of birds and infecting more than 200 people -- half of whom have died because of contact with diseased poultry.

Researchers at Johns Hopkins University developed computer models to analyze a variety of public health emergency scenarios, including restrictions on travel, isolation of infected patients, stockpiling of antivirals and the use of less effective but more available generic vaccines. The model simulated the spread of a pandemic across the United States and England using data based on population density, demographics and human travel patterns.

Derek Cummings

Researcher Derek Cummings says the model shows that no single response can control a pandemic, should one develop. "We found that you really need suites of interventions in order to significantly alter the course of the pandemic."

The spread of the flu by a few weeks, Cummings says. He adds, however, that an approach that includes antiviral use for infected and non-infected individuals living in the same household, combined with school closures, could lower infection rates by more than 50 percent.

"But we also found that the resources… [required] to do that are more than most nations have accumulated thus far. In order to employ the antiviral strategies that we found most effective, you would need antivirals for 50 percent of the population," he says. "Thus far, the United States and the UK have targeted 25 percent of the population as their antiviral stockpile."

No specifically targeted vaccines are yet available for H5N1. But there are plenty of vaccines developed for other similar strains of avian flu. While they're less effective against H5N1, Cummings says they could help reduce the impact of a pandemic -- though, again, only in combination with other responses.

He says the most effective strategy is to stop the pandemic in its tracks, at its source. "Even though it is a challenging task, given that this could emerge in any of a number of locations that might have very poor surveillance systems and very poor public health systems, I think that both the U.S. and the WHO (World Health Organization) have recognized that we should support nations across the world in shoring up the surveillance and getting ready to mount a response if there is an emergence [of the virus]."

Should a pandemic emerge, the models suggest the public health scenario could be grim. The sick would swamp the hospitals, millions would die and basic social services would be interrupted.

Anthony Fauci heads the National Institute of Allergy and Infectious Diseases

The U.S. avian flu plan unveiled by the Bush Administration last week outlines specific recommendations for businesses, schools and the health care industry in a worst-case scenario. Anthony Fauci with the National Institutes of Health says preparing for the worst makes sense. "Given the track record of what we know pandemic influenza can do historically from our experience in the 20th century (the 1918 flu pandemic), it would be irresponsible not to prepare for the worst. And that doesn't mean to panic," he says. "That doesn't mean to think that the world is coming to an end. That means to pay attention to what is going on and to prepare ourselves."

Derek Cummings from Johns Hopkins University says the simulated epidemic models he and his colleagues have developed can help nations better assess their risks and direct scarce public health resources where they are needed the most.

The project is a joint effort by Johns Hopkins University's Bloomberg School of Public Health, the Imperial College, London, and RTI International in North Carolina. The research was published in the journal Nature.