Some Hard Truths About Bird Flu

The issues surrounding the possibility of a pandemic of the H5N1 strain of avian flu are extraordinarily complex, encompassing medicine, epidemiology, virology and even politics and ethics. Moreover, there is tremendous uncertainty about exactly when H5N1, which now primarily affects birds, might mutate into a form that is transmissible between humans, and how infectious and lethal it might be.<?xml:namespace prefix = o ns = “urn:schemas-microsoft-com:office:office” />

It is thus hardly surprising that commentaries about avian flu often miss the mark. A recent New York Times editorial, for example, decried the “me first” attitude of wealthy countries toward a possible H5N1 pandemic, because “[t]he best hope of stopping a pandemic, or at least buying time to respond, is to improve surveillance and health practices in East Africa and Asia, where one would probably begin.”

To be sure, good surveillance is needed in order to obtain early warning that a strain of H5N1 flu transmissible between humans has been detected, so that nations around the world can rapidly initiate a variety of public health measures, including a program to produce large amounts of vaccine against that strain. But the massive undertaking required to “improve health practices in the poorest countries of the world” plays better on the editorial page than on the ground.

Intensive animal husbandry procedures that place billions of poultry and swine in close proximity to humans, combined with unsanitary conditions, poverty and grossly inadequate public health infrastructure of all kinds, make it unlikely that a pandemic can be prevented or contained at the source. It is noteworthy that <?xml:namespace prefix = st1 ns = “urn:schemas-microsoft-com:office:smarttags” />China's chaotic effort to vaccinate 14 billion chickens has been compromised by counterfeit vaccines and the absence of protective gear for vaccination teams, which might actually spread disease by carrying fecal material on their shoes from one farm to another.

In theory, it is possible to contain a flu pandemic in its early stages by performing “ring prophylaxis”—using anti-flu drugs and quarantine aggressively to isolate relatively small outbreaks of a human-to-human transmissible strain of H5N1.

According to Johns Hopkins University virologist Donald Burke, “it may be possible to identify a human outbreak at the earliest stage, while there are fewer than 100 cases, and deploy international resources—such as a WHO stockpile of antiviral drugs—to rapidly quench it. This tipping point strategy is highly cost-effective.”

However, a strategy can be “cost-effective” only if it is feasible. Although ring prophylaxis might work in Minneapolis, Toronto, or Zurich, in the parts of the world where flu pandemics begin, the probability of success approaches zero. In places like Vietnam, Indonesia and China—where the pandemic strain will likely originate—expertise, coordination, discipline, and infrastructure are lacking.

The response in Turkey—where as many as 50 possible cases have appeared in the eastern part of the country—is instructive. Officials in that region warned the government on Dec. 16 of a surge in bird deaths, but it took 12 days for an investigation to begin. When a 14-year-old boy became Turkey's first avian flu mortality last week (soon followed by two siblings), a government spokesman criticized doctors for mentioning the disease because they were “damaging Turkey's reputation.” This is ominously reminiscent of China's initial response to SARS in 2003.

For now, it seems that all of the human H5N1 infections have been contracted from contact with infected poultry. But the situation in Turkey is what the outbreak of a human to human pandemic could look like at its earliest stages: the rapid spread of confirmed cases (and deaths) from an initial site to nearby villages and cities. We would expect to see a large number of illnesses among both employees and patients in hospitals where the victims are treated, and soon someone [perhaps even a carrier who is not ill] would spread it to Ankara, Istanbul, Tbilisi, Damascus, Baghdad and beyond.

The anti-flu drugs Tamiflu and Relenza are extremely expensive and in short supply. History suggests that if we were to make these drugs available to poor countries for ring prophylaxis, they would often be administered improperly—such as in sub-optimal doses—in a way that would promote viral resistance and only intensify a pandemic. Or perhaps they would be sold on the black market to enrich corrupt government officials.

A politically incorrect but rational strategy would be for rich countries to devote resources to developing countries primarily for surveillance. They would obtain timely warning of the existence of an H5N1 strain that is transmissible from human to human, but would focus the vast majority of their funding on parallel, low and high-tech approaches—vaccines, drugs and other public health measures—that would primarily benefit themselves.

If the pandemic were to begin relatively soon—say, within a year or two—there would be little that could be done to attenuate significantly the first wave of infections. But, if we're ready to rush the pandemic strain into an emergency program to manufacture vaccine, we could possibly blunt the second wave.

A flu pandemic will require triage on many levels, including not only decisions about which patients are likely to benefit from scarce commodities such as drugs, vaccines and ventilators, but also broader public policy choices about how best—among, literally, a world of possibilities—to expend resources.