How accurate is the CDC's seasonal flu death estimate?

May 14, 2009  ·  Michael Fumento  ·  Weblog

All health agencies tend to exaggerate the threat of anything within their ambit, whether it's numbers or overall severity as we're currently seeing with the WHO and its threats to declare swine flu a pandemic. That's in addition to including with their ambit that which clearly doesn't belong, such as the CDC expanding into gun deaths and divorce.

For this reason, numbers coming from outside of the agencies and their officials tend to be more reliable. A 2008 study by Foppa and Hossain in Emerging Themes in Epidemiology analyzed data from 1995 to 2005 and comes up with an annual average of 23,710 U.S. seasonal flu deaths. By coincidence, until fairly recently, the CDC used a range of 24,000 to 36,000.

That said, Dushoff and others in the American Journal of Epidemiology in 2006 said the CDC figure was too low. Analyzing data from 1979 to 2001, they found an annual average of 41,400.

It would be interesting for somebody to contrast their methodologies. My gut feeling is that the CDC figure is too high, but gut feelings in science are for leads and not for conclusions.

All this said, when it comes to swine flu comparisons there isn't one.

The widely-used estimate for the U.S. seasonal flu death rate is one per one thousand infections (0.1%), though the CDC, using in part it's 36,000 death estimate, employs a range of 0.06% to 0.24%. Currently three Americans have died out of 4,298 "confirmed and probable cases," all of whom were chronically ill, for a rate of 0.07% without even counting hidden infections. (The "fourth" U.S. case was a Mexican national who sickened there and died under treatment here.)

But with any flu, each confirmed case represents many milder or even asymptomatic hidden infections. Indeed, despite its alarmist theme, the Ferguson et al. paper in Science magazine, prepared under the auspices of the WHO, says Mexico apparently has had hundreds of undetected infections for each confirmed swine flu case. Thus to peek below the tip of the swine flu iceberg would be to find the U.S. death rate dramatically lower than that of seasonal flu.