Friday, June 30, 2006

Lessons from the 1976 Non-pandemic

In 1976 a soldier at Fort Dix, N.J. died of what was diagnosed as a new and virulent strain of swine flu. This event led the CDC to recommend a $135 million campaign to immunize Americans. President Ford served as the poster child and got his shot on TV.

Almost forty million men, women, and children were immunized before public health athorities acknowledged that there was no pandemic. Very sadly, the swine flu vaccine caused at least 25 deaths. The vaccine was also linked to substantial increase in the incidence of Guillain-Barré syndrome, a serious (and otherwise rare) nerve disease.

The immunization campaign is widely regarded as one of the worse public health humiliations in US history, and its consequences can not be far from the mind of any public health official who now warns of the potential of a H5N1 pandemic.

Is it different this time?

Mark Twain once said, "History may not repeat itself, but it rhymes." There is something to be learned from the 1976 experience; somethings have already been learned. One lesson is that producers of vaccine have a dicy business. This is why the hundred-plus US flu vaccine producers in the US in 1976 have been driven down to just two --- an economic story that deserves its own page.

One of take-aways from the 1976 senario is simply a reminder that human beings are remarkably poor at dealing with uncertainty. As a species, we have been well served by assuming that "things happen for a reason," and evolution has left us without much natural talent for dealing with things that genuinely depend on chance.

Then and Now

Let's look at some of the changes that have taken place since 1976 to see how they might draw a distiction between pandemic predictions then and pandemic preditions now.
  • World-wide monitoring is vastly improved
  • Understanding of the mutation pathways of viruses is greatly extended
  • Information sharing between nations has improved; in particular, the mental models that now frame the pandemic risks are not primarily US models. They are genuinely international models
One can cerainly add to this list of factors; in fact, commentors are encouraged to do so! Even just these three might be enough to suggest that we are now much better prepared to make a reliable prediction. On the flipside, we are not more prepared to provide timely distribution of a vaccine, nor are we substantially more able to minimize the side effects of a vaccine.

Bottom Line?

The events of 1976 underscore the real consequences of reputaion risk, and public health officials every where are keenly aware of the risks of crying wolf. If the cry goes out now, we know that they know the risk that is being taken. To ignore the cry today would be to fail to learn what 1976 has to teach us.

Reference: Fluwiki has a useful page on the 1976 pandemic, including a insightful quote from one of my favorite public policy books, Neustadt and May, Thinking in Time: The Uses of History for Decision Makers (1988).


Blogger Dan McCarthy said...

Stratfor did what I thought to be a not bad analysis of the bird flu situation back in Oct 2005, with the basic conclusion that it is overblown in light of the following considerations:

1) Trailing 3 year # of total cases was 115 and first ever human case of H5N1 was back in 1997-- it has been around for a while.
2) The "W" nature of the mortality curve for the 1918 pandemic (mortality rate was for some reason highest for people in the prime of their lives) made the virus more effective at killing people.
3) WW1 was fought heavily in the trenches-- extremely unsanitary environment increased the probability of catching it and increased the mortality rate.
4) Circulation of military personnel was one of the key drivers of its spread to the US-- 2% involvement in WW1, .005% involvement in Iraq.
5) Overall health is better now than it was then.
6) For our sake, the mortality rate was more than twice as high for the poor than the well to do because of living conditions (people live closer together) and working conditions (farmers, etc).
7) Antiobiotics-- 1918 pandemic didn't directly kill most of the people who died directly-- it was typically through secondary infections that trigger pneumonia. Penicillin was discovered in 1929.

Doesn't eliminate the probability, but perhaps the differences between then and now can at least mitigate the expected outcome and the probability both.

4:17 AM  
Blogger gs said...

dan mccarthy, your points 1)-7) don't convince me.
1) despite being around since 9 years, H5N1 does still spread and develope and increase a lot
2) 1918 H1N1 and 2006 H5N1 are similar here
3),4) it was even worse in non-war countries
5) H5N1 does hit the healthy people
6) population (i.e. urban) is larger conditions in 3rd world are worse
7.) most people died from ARDS (see 2))

6:09 AM  

Post a Comment

Links to this post:

Create a Link

<< Home