Biostatisticians at Fred Hutchinson Cancer Research Center are racing to plot swine-origin influenza A H1N1's path to predict whether the new pathogen appears headed for possible pandemic — or may peter out.

Share story

On Wednesday afternoon, Ira Longini, a biostatistician at the Fred Hutchinson Cancer Research Center, made a hypothetical wager: $10,000 that swine flu would appear in Washington by the end of June.

Longini should have bet the house.

Less than six hours later, local health officials issued a hasty alert that they’d detected the state’s first probable swine-flu cases in six people, including three in King County. Seven more potential cases popped up Thursday, temporarily shuttering affected schools and triggering massive civic hand-washing.

The statewide tally as of Saturday evening stood at 25 cases, all of which were awaiting confirmation from the federal Centers for Disease Control and Prevention.

Predicting potential pandemics is a complicated science. And from his narrow office at “The Hutch” overlooking boats on Lake Union, Longini is at the forefront of it.

The Hutch is part of a national network that includes six universities which, since 2004, have been building computer models to predict how infectious diseases, from SARS to smallpox, might spread. Their biggest target, of course, was avian influenza H5N1, which everyone had expected would become the next pandemic.

Instead, Longini and his counterparts now are scrambling to revise their calculations for an entirely different virus — swine-origin influenza, A H1N1. Their analysis will give public-health authorities critical guidance in how best to combat the novel pathogen.

But more than a week after H1N1 burst into a global health scare, essential data about characteristics of the virus remain elusive. A key missing piece is information about H1N1’s contagiousness, which is measured in what’s called a basic reproduction number — the average number of new infections produced by a single infected person, assuming no one is immune to the virus.

If the number is less than 1, an outbreak will eventually extinguish itself.

The 1918 influenza pandemic, which killed millions, had a reproduction number of 2, Longini said. For the 1957 and 1968 pandemics, it was about 1.7, just above the threshold needed to turn an epidemic into a pandemic.

The World Health Organization has raised H1N1 to a pandemic alert level of 5, on a scale of 1 to 6. A 5 means that widespread human-to-human transmission is occurring in at least two countries in the same region — a strong signal that a pandemic is imminent.

Computer simulations

Longini’s expertise is in simulating how an infectious agent would spread in real life. As facts emerge about who is being infected with H1N1, where and how, he is feeding the data into his computer to discern how its trajectory is being shaped by school closures, quarantines, treatments with antiviral drugs and other actions by public-health authorities.

He is working with computer scientists and other investigators from Harvard University, Virginia Tech and four other institutions in a project called MIDAS — for Models of Infectious Disease Agent Study — that’s funded by the federal National Institutes of Health.

At the Hutch, Longini is joined by his longtime collaborator, Dr. Elizabeth Halloran, professor of biostatistics at the University of Washington.

Longini said investigators are close to gaining some firm insights about H1N1, at least in the U.S.

“We have quite a bit of detail on a number of school and family clusters from around the country,” Longini said. “More than six or seven linked cases in a school usually indicates that there could be some substantial transmission.”

So far in Washington, only two clusters of probable H1N1 infections have turned up in King County. Both clusters are within families; one is a pediatrician and her two children, and the other is a Seattle elementary student and his two siblings.

Despite the slow trickle of reliable numbers emerging from Mexico about the extent of H1N1 infections and deaths, the outbreak has given investigators a chance to run their statistical models through a real-time exercise.

“We’ve never done this at the beginning” of a possible pandemic, Longini said. “In 1968, there were large outbreaks even before we became aware that transmission was occurring.”

Halloran said H1N1 could go in several directions. The current outbreak could peter out, only to emerge in a second wave months from now. Or it could flare up in an epicenter other than Mexico.

But Halloran emphasized that any pandemic doesn’t necessarily have to be lethal. A pandemic simply is an epidemic that is occurring in two or more regions of the world. And seasonal flu, after all, is a regular epidemic.

“People shouldn’t be panicking. This doesn’t look like a very deadly disease,” Halloran said.

Containing an outbreak

If a pandemic were to unfold, Longini and Halloran’s analysis of avian flu suggests quick action could contain it. Particularly among children, the most formidable tool is mass vaccination, which likely won’t be possible until fall. But even poorly matched, relatively ineffective vaccines could be combined with other steps to corral a runaway pathogen.

“Without a vaccine, there is nothing to do except react,” Longini said.

Children excel as disease transmitters, so closing schools also can be effective, he said.

Travel restrictions, on the other hand, are all but impotent against a pandemic, Longini said. Banning flights or travel can be an appealing political option and may slow an outbreak at the onset. But they ultimately make no difference in the number of infections.

Even if H1N1 appears to be subsiding, the possibility remains that it could morph into something far more contagious or lethal, said Dr. Christopher Murray, director of the Institute for Health Metrics and Evaluation at the UW and a former World Health Organization officer.

H1N1 is a new virus never seen in humans before, so people lack natural immunity to it.

Longini said viruses can be wily foes. Despite all the predictive tools available, he said, “The truth is, we don’t know what’s going to happen anywhere at this point.”

Kyung Song: 206-464-2423 or