The Bush administration, along with the EU, the UN and the general public, is finally taking seriously the possibility of an outbreak of an avian influenza pandemic. Health experts are working on identifying the strains of avian flu most likely to mutate into human transmissible form, and drug companies are busily producing vaccines to ward off their effects. Together, these actions are encouraging; at the very least, should an outbreak of human transmissable bird flu occur, we'll be better off than if it had appeared out of the blue, as SARS did in early 2003.
But while it's reassuring to think that we won't be caught off guard this time, the government's overwhelming focus on vaccine production is worrisome. Although understandable, it downplays other, potentially more effective prevention strategies.
At present, human-to-human transmissible avian flu virus does not exist outside the lab. However, such a virus could emerge any day as a result of the mixing of a bird flu -- such as the virulent strain H5N1 that is currently decimating bird flocks throughout southeast Asia -- with a human influenza virus within a single human. Alternatively, human transmissibility could also evolve directly out of the bird-flu virus itself. Recent studies indicate that the virus responsible for the 1918 influenza pandemic, for example, which killed 20–50 million and infected up to a billion people worldwide, arose in precisely this way.
Unfortunately, there is no good way of calculating the likelihood of either of these scenarios occurring -- hence what particular strain of the virus to anticipate. Even plentiful vaccine stockpiles (assuming they can be mustered in time) could be of little benefit.
In the absence of an ironclad vaccination defense, it is difficult to predict the numbers affected by such an epidemic. The upper limits for avian flu have been placed at 100 million deaths, a plague of biblical proportions; but these numbers are effectively educated guesses, arrived at by analogy with the 1918 pandemic, not from detailed models. In fact, as simple as it sounds, the question of how large any influenza epidemic is likely to become is one that mathematical epidemiology is ill equipped to answer.
Against this background, what can be done to mitigate the spread of an incipient influenza epidemic, if and when it occurs? Returning to the SARS example, our experience with SARS suggests that epidemics can be fought quite effectively even without vaccines. The key is to employ several strategies simultaneously -- and to enlist the public's support.
According to recent studies, SARS had roughly the same potential to infect people as the 1918 influenza virus -- which was estimated at up to a billion people. How was it, then, that SARS affected only about 8,000? Although part of the answer is sheer luck -- for instance, it was fortunate the virus spread first to Toronto and Frankfurt, instead of to cities with weak public healthcare systems like Bombay and Cairo -- a big part of the explanation involves the steps taken by the World Health Organization and other governments to detect new outbreaks immediately, to coordinate information between public health agencies, to isolate known and suspected infectives and trace their recent contacts, and to instruct airlines to identify and isolate potentially infected passengers and crew -- thereby targeting perhaps the most dangerous mechanism for rapid global dissemination.
But there is another feature of the SARS epidemic that has received relatively little attention and in my view was of enormous importance -- that is, how the population itself responded. The distribution of people in space, and even more importantly, the way they move around -- whether it is between buildings, neighborhoods, cities or nations -- can have a dramatic impact on the ability of an epidemic to grow from a small, localized outbreak to a global pandemic. How people respond to the knowledge that an epidemic is spreading in their midst, or far away, therefore ought to be a critical parameter both in mathematical models of disease spread, as well as in public health strategies.
For example, a recent model developed by my research group at Columbia suggests that while most infections occur between people in highly localized neighborhoods, the size of an epidemic ultimately depends almost entirely on the movement of just a few individuals who, in traveling from infected to uninfected regions of the global network, allow the epidemic to spread simultaneously in many places at once.
If these people refrain from traveling, or are actively discouraged from doing so, even a potentially dangerous epidemic may be forced to spread locally, and therefore less rapidly, allowing public health authorities more time to apply other, more direct containment measures.
Applying that lesson forward to an avian flu outbreak, people in affected cities can mitigate the spread of the disease by avoiding work and school and not flooding to hospitals unless they are sick. Those in nearby regions can go about their lives but should avoid traveling into the affected cities or even out of the region. And people far from affected areas can maintain their distance (that is, more or less how they reacted to SARS). Airline travel should be discouraged, or even actively shut down until the extent of spread is ascertained; but other forms of transport, like rail and road, should also be treated warily.
Measures such as these will no doubt inflict considerable economic cost on affected regions -- as indeed happened with SARS -- but it is a small price to pay for averting major catastrophe.
Social response to disease epidemics remains very poorly understood and ought to be a far more urgent research priority than it is. Nevertheless, the 2003 SARS experience provides at least some hope that if governments provide early and accurate information, self-interested individuals can reduce much of the risk to themselves and to others, simply by proceeding with caution.
Duncan Watts is an associate professor in the department of sociology. He is the author of Six Degrees: The Science of a Connected Age. The initial results of his research on the effects of population structure and movement on the spread of large epidemics were published recently in the Proceedings of the Academy of Sciences, available at http://www.pnas.org/cgi/content/full/102/32/11157.
Additional Faculty Perspectives on Avian Flu
This is the story about avian flu: First, it is a legitimate cause for real concern, even though we can't say for certain right now when it will break loose or how lethal it will actually be.
Second, it is essentially an incontrovertible reality that we were just asleep at the switch in terms of enhancing preparedness for a major pandemic. If the H5N1 "goes pandemic" this year, we won't have the vaccine, we'll be ridiculously short of anti-viral medications and we do not have a healthcare system in the United States even remotely close to being able to accommodate the volume of people who'll need care.
So Duncan Watts' advice to help control the spread of what could be a living nightmare is sound.
Will those measures work? Well, folks, that's about all we'll have.
-- Irwin Redlener, professor of Clinical Public Health & Pediatrics; associate dean and director, National Center for Disaster Preparedness
Most virologists consider another influenza pandemic to be inevitable, although no one can predict exactly when or how it will occur. The techniques Watts and his group have pioneered show great promise for improving predictive capabilities.
Meanwhile, in the face of uncertainty, early warning and public education are essential. Early warning can help public health organizations contain the infection at its origins if virus transmission is spotted in time.
Public education is important, because too often, a period of intense concern can give way to complacency if the pandemic doesn't materialize exactly as expected. And inappropriate use of antiviral drugs by an anxious public can render these defenses useless by selecting for resistant virus strains.
The media, still learning to strike the right balance, has an essential role in the educational process. And, as Watts notes, appropriate individual actions can reduce risk while also protecting the world. Taking the right actions at the right time will be crucial.
-- Stephen S. Morse, associate professor of Clinical Epidemiology; founding director & senior research scientist, Center for Public Health Preparedness, National Center for Disaster Preparedness