"AVIAN INFLUENZA: Pandemic Skeptics Warn Against Crying Wolf"

Read the full article by Dennis Normile in Science. (Subscription required.)

"Just as the threat of an influenza pandemic is finally being taken seriously by governments around the world, a small but increasingly visible number of scientists are questioning how great the danger really is. They acknowledge that another flu pandemic is inevitable--at least three major and several minor pandemics occurred in the last century--and they believe preparing for it is wise. But they are asking: Is the H5N1 virus now circulating in Asia really the one to watch? How soon will the next pandemic occur? And will it trigger a wave of mortality, as did the 1918 flu, or a just small ripple in the annual influenza death toll? If no serious pandemic emerges in the next few years, they warn, the current hype could backfire, undermining public support for efforts to prepare for an eventual pandemic, including developing and stockpiling better flu vaccines and drugs."

Some further excerpts:

"No H5 flu subtype has ever caused a human pandemic........Although human infections with H5, H7, and H9 subtypes have been documented.......these viruses have never been known to pass efficiently among humans. 'That doesn't mean they never will.....although H5N1 has been circulating widely among poultry for at least 8 years, it has not shown any signs of jumping more easily from chickens to humans or of spreading among humans."

"The six pandemics that have occurred since the late 1800s were caused by just three subtypes, which reappear in a repeating pattern: H2, H3, then H1. Roughly 68 years separated the reappearance of each subtype."

"Although the historical data are interesting......they simply aren't conclusive enough to rule an H5N1 pandemic in or out. 'We don't know what viruses circulated in the past [among humans], except for the most recent 150 years,' What's more, H5N1 is shattering historical precedents. Never before has a virus so highly lethal for poultry become so widespread and continued in circulation for such a long time. And with the virus continuing to spread......There are so many gaps in what is known about how virulence and pathogenicity evolve, 'there is no scientific basis to predict anything'........'We, as scientists, need to do a good job of something slightly tricky here, which is to convey that our predictions are probabilistic.'"

"Despite their differences over H5N1, flu experts on both sides of the debate agree that preparing for a pandemic is essential. Palese (Peter Palese, a virologist at Mount Sinai School of Medicine in New York City) says he strongly supports the pandemic preparedness plan recently announced by the U.S. government. Ewald (Paul Ewald, an evolutionary biologist at the University of Louisville, Kentucky) is in favor of tracking H5N1 and vaccinating exposed populations if the virus shows any tendency toward passing from human to human. 'This could provide an effective barrier to evolutionary increases in transmissibility,' he says. The plan is also similar to one of the strategies being pursued by WHO.

"Offit (Paul Offit, an immunologist and virologist at Children's Hospital of Philadelphia and the University of Pennsylvania School of Medicine) hopes the concerns about H5N1 will lead to efforts to strengthen the U.S. infrastructure for vaccine development and production, which he says has deteriorated over the last 50 years. He thinks the message scientists should be sending 'is not that we're going to protect you from the bird flu pandemic, but that we're going to be protecting you from a pandemic which may be 20 years from now.'"

Comment

The analysis here is not probablistic, but one of decision making under conditions of uncertainty. With so much unknown, and so little historical record, accurate estimates of probabilities are not possible. We know that as of now, we can not prevent a flu pandemic if a really dangerous virus evolves and if the conditions are propitious for its spread. There is sufficient historical experience to know that a flu pandemic can kill millions of people, and indeed tens of millions.

There are "no regrets" actions that can be taken, and indeed should be taken. These are actions that will be important if a pandemic virus arrives, but will also be useful even if a pandemic does not happen now. These include research on the flu, development of vaccines and antiflu drugs, improving vaccine production process technology, planning for control of viral respiratory epidemics, and building the capacity of epidemiological surveillance systems.

The hard part is deciding how much effort and money to put into efforts that will be wasted if not needed. Different approaches will have different costs and different payoffs (if and when a pandemic does occur, depending on how serious the threat turns out to be -- potentially killing hundreds of thousand, millions, or tens of millions). Experts, with both theoretical knowledge and practical experience in viral disease control, should help prioritize such efforts, and help develop an appropriate portfolio of actions for public health agencies to implement.

The U.S. government "cried wolf" about the swine flu in 1976. If it does so again in 2005 about the avian flu, over-reacting to a threat that never materializes, it will be that much harder in the future to mobilize against future threats. And, unless there is unexpectedly rapid and extensive progress on the research front, a flu pandemic will almost surely occur at some future time!