"Bird Flu" Could Be Slowed at the Source, Study Says

Brian Handwerk
for National Geographic News

August 3, 2005

Amid fears that a possible "bird flu" pandemic could kill millions, a new study examined potential ways to slow such an outbreak at its source—and buy precious time for vaccine production.

Some results are encouraging, but success remains dependent on improved surveillance, widespread drug availability, effective quarantine, and good old-fashioned luck.

Experts from U.S. universities and Thailand's Ministry of Health used a computer model to simulate an outbreak in rural Southeast Asia—the most likely place for a new, more deadly avian influenza strain to emerge.

"If we end up with a pandemic like [previous catastrophic pandemics], we'll have a lot of people dead," said study team member Elizabeth Halloran, professor of biostatistics at Emory University in Atlanta, Georgia.

Halloran added that the simulations show that it should be possible to contain an outbreak at its source. But the results are unpredictable. "We have shown in these simulations that—even given the same [hypothetical] situation—sometimes when we intervene it's successful and sometimes it's not," Halloran added.

A global influenza pandemic like the infamous "Spanish flu" of 1918 and 1919 could leave tens of millions dead. Many experts warn that the world is unprepared, and overdue, for just such a catastrophe. The leading candidate to spark such an event is H5N1, a strain of avian influenza, or bird flu.

Pandemics occur when viruses mutate into forms that are easily passed from human to human and that cannot generally be stopped by the human immune system.

H5N1 is found primarily in domestic and wild birds but also in mammals, such as cats and pigs. The virus has yet to become easily transferable between humans.

Model Mimics Spread of Asian Outbreak

The study team used data from the Thai 2000 census and from previous studies of social networks in Thailand's rural Nang Rong district. The researchers attempted to replicate the ways in which an outbreak would spread through a rural population of some 500,000 people.

The model simulated a flu's movement and person-to-person interactions at all levels of society, from "close contact" environments—like homes, schools, and workplaces—to more casual contacts in places such as markets and temples.

Beginning with a single infected person, the model projected numerous scenarios of viral spread and development. The spread scenarios were largely based on the average number of people that each infected person is likely to infect.

In the computer model, simulated treatments included the targeted use of antiviral medications (which typically slow the disease's spread), quarantine, and pre-vaccination in selected areas—despite the use of relatively inefficient vaccines.

While results varied, they often showed a reasonably high probability that an outbreak could be contained in a variety of scenarios.

Early intervention was key to each successful outcome.

"What we show is that the ability to implement something fairly quickly in a couple of weeks is important, otherwise [the virus] will likely get out," Halloran said.

The study will be reported in Friday's issue of the journal Science.

Buying Time at the Outbreak's Source

Anthony Fauci directs the National Institute of Allergy and Infectious Diseases at the National Institutes of Health in Bethesda, Maryland. He called the model valuable but noted that "any model is only as good as the assumptions put into that model, and these really are only assumptions."

Still, such models can give experts "some specifics about options available that might be able to contain an epidemic early on, " he said.

Fauci said that the new study confirms that even a limited supply of antiviral drugs can help slow the spread of a virus in a particular region. "The same is true with a partially effective vaccine. That can also slow the spread of a disease."

Harder to diagnose is how outbreaks behave in the face of unknown variables such as the makeup and behavior of the virus, as well as human factors, like the ability to recognize and monitor a developing outbreak.

"If you find [these things] out a couple of months after it's started, you're in big trouble, and the types of measures recommended in the model won't help," Fauci said.

"This underscores the need for surveillance and the need for transparency—namely countries letting people know immediately when they have a situation, because the longer you wait, the less effective the intervention becomes," he said.

To make matters worse, in cases where the disease is very contagious from the onset, treatment has very little effect.

Real-World Benefits?

The large number of unknowns leaves some experts skeptical that models can help contain virus outbreaks.

"In this case there are any number of issues," said Michael Osterholm, director of the Center for Infectious Disease Research and Policy at the University of Minnesota in Minneapolis.

"It's not just having in place the ability to move antiviral drugs to an area—though that is a concern—but the entire system. What kind of surveillance do you have to spot the appearance of an illness? How quickly can you do lab tests to find out what's actually happening?" Osterholm said.

"What kind of system do you have in place to really control population movement [for quarantine]?" he continued. "Look at the model from the early days of SARS. One individual left a given area and suddenly overnight you had [the illness] in multiple countries."

Osterholm has been a vocal proponent of coordinated international efforts to prepare for a potentially devastating pandemic.

"I worry that these kinds of studies leave policymakers with the sense that there is something meaningful we can do at the local level to stop this, which may prevent them from addressing the global issues pertaining to a pandemic," he said. "And we can't for one minute ease the pressure on them to do that"

Free E-Mail News Updates
Sign up for our Inside National Geographic newsletter. Every two weeks we'll send you our top stories and pictures (see sample).