Comment on Calculating U.S. Influenza Deaths, by F. Edward Yazbak, MD, FAAP written by TMN on May 8th, 2009

This comment is on a paper written by F. Edward Yazbak, MD, FAAP (the link is given below this comment) who says that the figures often cited by the CDC that 36,000 deaths per year due to flu is preposterous. I found the paper by Dr. Yazbak very interesting but a bit difficult to understand for the layman. This comment is sort of a reader's digest version. This explanation should help make Dr. Yazbak's paper easier to understand.
Readers Digest Version (Abstract)
Until recently, deaths due to influenza and pneumonia have been lumped together but there was enough information to figure out how many deaths were due to each cause. In the past the ratio of influenza to pneumonia deaths has been around 1-10%. In other words, many more people die of pneumonia than influenza. Currently, however, the CDC lumps these deaths together and says that an average of 36,000 people die from influenza-related complications. It is less than a half-truth to call pneumonia an influenza-related complication when 90% of the time it has nothing whatsoever to do with influenza (pneumonia is often caused by a bacteria while flu is due to virus).
A more detailed explanation of the paper (Comment)
Yazbak focuses on two years in which the number of cases of influenza and flu can be unraveled from the government data (although if you read Yazbak's paper you will see that it is not that easy).
In 2001, the government said that there were only about 280 deaths from flu. These deaths were spread more or less evenly over different age groups. During this same year, however, there were 61,777 deaths related to pneumonia of which 48,686 were in people over the age of 75 years. The official estimates of death rates for 2001 was 21.7 deaths for every 100,000 due to pneumonia and 0.1 deaths out of every 100,000 for flu. Since the population was 284 million people in 2001, these numbers fit well with the stated number of deaths for each disease. During this year flu accounted for about 0.5% of the deaths and pneumonia accounted for about 99.5% of these deaths.
The article next discusses data in which 52,720 deaths in 1980 were attributed to the combined causes of pneumonia and influenza. The death rate caused by both influenza and pneumonia increased from 11.1 people in every 100,000 in 1979 to 12.6 in every 100,000 in 1980. Yazbak takes a worse case assumption that all of this increase or 1.5 people in every 100,000 was due to the influenza epidemics in 1980. Since the U.S. population was around 226.5 million in 1980, 1.5 deaths per 100.000 would translate to around 4,000 additional deaths that year due to an influenza epidemic. Even in this worse case assumption, the flu deaths were only about 10% of the combined flu and pneumonia deaths
Both of Dr. Yazbak's examples show that a number like 36,000 deaths/year is a reasonable number for the combined flu and pneumonia deaths but a best guess is that only about 10% could be attributed to flu. Therefore a more reasonable number for the deaths caused by flu each year is probably closer to 4,000.
Yazbak goes on to describe when CDC Director Julie L. Gerberding, MD, carefully testified that "CDC scientists estimate that an average of 36,000 people die from influenza-related complications each year in the United States.", before the committee on government reform of the U.S. House of Representatives on Feb. 12, 2004. Yazbak infers that she is lumping flu and pneumonia together as "influenza-complications". This explains why the number, 36,000 is a similar order of magnitude to the numbers for 1980 and 2001 when the data were a bit easier to understand and it was clear that these two causes of death were being lumped together.
The author ends the paper with, "... and the propaganda machine is still talking about "an average of 36,000 deaths" a year. How preposterous!"

Personal Comment
I think that Dr. Yazbak may be onto something. It seems curious if it is true that the government has been slowly but steadily lumping these two distinct causes of death together since 2001. Why would they do this? Is it too much to ask for two different numbers, one for pneumonia and another for flu? Heck as long as we are asking for two numbers why not splurge and ask for enough information to understand what is going on. How about a list like this:

The number of deaths per 100,000 and the total number of deaths for

Flu

Pneumonia

The % overlap between these two numbers

The rate of infection per 100,000

Flu

Pneumonia

The % overlap between these two numbers

I personally have no idea if 36,000 deaths/year can reasonably be attributed to flu. However, it is a bit suprising that I have never known anyone who died from flu. A retired pediatrician that I know personally recently commented to me,
"As a neonatologist I encountered bacterial, fungal and viral pneumonia deaths but I do not recall an influenza death."
Many people focus on the politics or the commercial ramifications of this question (are the numbers being skewed to sell vaccine, etc.), however, it would be very useful to know what death rate can be expected from the swine flu and the observed case grwoth rate if it has just "normal" flu viruility. Apparently even this simple number is difficult to fish out of the murky system.
Here is a link to the original paper written by F. Edward Yazbak, MD, FAAP