This blog post claims that the CDC over-reports annual flu deaths in the U.S. by 1000x or more, in an effort to sell flu vaccinations.

Yet, their own numbers show 500 deaths in 2010, not 36,000.

It goes on further to explain how even the low number isn't "confirmed flu deaths."

They claim that many pneumonia deaths stem from the flu. They use a computer model to make the calculations.

But in conventional medical literature, there are at least 10 different types of pneumonia. Trying to model which ones stem from flu cases is a fool’s errand.

It also says the CDC claims between 3,000 and 49,000 deaths annually, depending on severity of the flu season, a claim I found on the CDC web site, too:

CDC estimates that ... flu-associated deaths ranged from a low of about 3,000 to a high of about 49,000 people

Is the CDC over-stating the danger of the flu?

  • 1
    It is worth noting that the CDC page explicitly steps away from the quoted 36,000 estimate as being based on old data, during an particular severe period. The blog post's argument, in essence, boils down to an argument from personal incredulity (how could statisticians possibly separate out different sorts of pneumonia?) – Oddthinking Jun 21 '13 at 10:29

The CDC reports 500 deaths from influenza. (Table 10 of http://www.cdc.gov/nchs/data/dvs/deaths_2010_release.pdf)

The CDC reports flu-associated deaths ranging from from a low of about 3,000 to a high of about 49,000 people.

The CDC says:

It has been recognized for many years that influenza is infrequently listed on death certificates and testing for seasonal influenza infections is usually not done, particularly among the elderly who are at greatest risk of seasonal influenza complications and death. Some deaths — particularly in the elderly — are associated with secondary complications of seasonal influenza (including bacterial pneumonias).

Thompson et al. review the CDC's methodology.

Because a broad range of respiratory and cardiac diagnoses have been associated with influenza virus infections and these infections are often not confirmed by virologic testing, statistical models based on vital statistics data have been used for decades to estimate the overall burden of influenza in the United States.

In general, the estimates from each of these models suggest that seasonal influenza epidemics from 1976 through 2000 were associated with substantial morbidity, including >200,000 annual hospitalizations and an annual average of >30,000 influenza-associated all-cause US deaths.

They review the range of statistical models that have been used to estimate the mortality caused by influenza.

The report mentions:

  • a linear-regression approach,
  • Serfling-type models, using Poisson regression techniques, and
  • simpler models, based on rate differences, [that] have been used for many years to estimate influenza-associated morbidity and mortality.

In my opinion, the CDC is making a sincere attempt to accurately estimate the number of deaths attributable to influenza.

I could find no evidence that the CDC is intentionally over-reporting annual flu deaths.


Thompson, William W. and Comanor, Lorraine and Shay, David K., Epidemiology of Seasonal Influenza: Use of Surveillance Data and Statistical Models to Estimate the Burden of Disease J Infect Dis. (2006) 194 (Supplement 2): S82-S91 doi:10.1086/507558

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .