Lisa Schnirring’s CIDRAP News article on this survey is also online.
What’s most remarkable to me about this survey is that the authors seem to see it as evidence of complacency among the doctors who responded. I don’t see it that way at all.
The question asked was the probability of a flu pandemic “in the next few years.” Since a flu pandemic occurs on average three times a century, the statistical probability of seeing one over, say, an eight-year period is pretty small. From the choices offered, a statistician would have picked “possible” or “unlikely” – certainly not “almost inevitable” or “very likely.” Yet 72 out of 161 respondents – 45% – did pick “almost inevitable” or “very likely.”
Despite the recent precipitous decline in media coverage of the bird flu strain of greatest current concern, H5N1, many survey respondents apparently continue to think a pandemic is more likely to occur in the current decade than in the average decade. The authors appear to agree. But instead of celebrating the respondents’ sustained concern, they criticize the bare majority who went with the statistical record and therefore didn’t think a pandemic was statistically likely in the next few years for not thinking so “despite evidence to the contrary.”
We are on solid ground when we say that another flu pandemic is nearly inevitable. We are on extremely weak ground when we say that it’s probably imminent. And we are doing both terrible science and terrible risk communication when we conflate the two claims, as the authors of this survey report do.
Some experts have a hunch that a pandemic really is imminent – that H5N1 looks like The Big One. Apparently Curtis and Pollard think so, and so do almost half of their respondents. They’re entitled to their hunch. But we make a big mistake when we ground the case for pandemic preparedness in that hunch. Pandemic preparedness is important because in our new world of global markets and just-in-time inventories, a severe pandemic could wreak unprecedented havoc. H5N1 was a much-needed wake-up call, but pandemic preparedness will remain important even if H5N1 disappears entirely.
We cannot predict when the next pandemic will come. We cannot judge how likely H5N1 is to be the flu strain that launches it. The more we connect our advocacy of pandemic preparedness with the hunch that H5N1 is about to go pandemic, the less preparedness we will end up with if a few years go by without an H5N1 pandemic.
It is good news that many of Curtis and Pollard’s respondents still think a pandemic may well be imminent. But it would be better news if they realized that nobody has a clue whether a pandemic is imminent or not, and if they understood that preparedness doesn’t depend on whether a pandemic is imminent or not.
Pandemics are a perpetual risk, and pandemic preparedness is a chronic issue. We were unwise to sell H5N1 as a crisis. That “pandemic crisis” pitch bought us a few months of media and public attention. Already the attention is waning, as we pay the price for having grounded our case for preparedness in a false sense of imminence. Now we urgently need to engineer a transition to chronic concern and evergreen preparedness – lest the pandemic risk become in the world’s judgment merely another supposed crisis that failed to materialize.
Two bottom lines in all this, Lisa:
- Because they conflate inevitability with imminence, Curtis and Pollard misinterpret their own data. The doctors they surveyed think a flu pandemic is likelier in the next few years than the historical record suggests they should. Far from being evidence that European physicians are complacent about pandemics, the data suggest if anything that European physicians share the hunch that H5N1 makes a pandemic likelier than usual.
- This is good news of a sort – in that it’s better to have doctors concerned about pandemics than unconcerned about pandemics. But to the extent that their concern is grounded in the imminence hunch, it is vulnerable; unless a pandemic materializes soon, the hunch will have been proved wrong. We need to work toward a transition to sustained pandemic preparedness that is grounded not in the hunch that a pandemic is imminent, but in the well-founded conviction that sooner or later one is nearly inevitable.
Copyright © 2007 by Peter M. Sandman
