Low Carb Charlie

From Morbidly Obese to Wow!

Low Carb Charlie - From Morbidly Obese to Wow!

Final Flu Redux: Defining the Problem


This is my last post for some time about flu vaccines, though in retrospect, perhaps this should’ve been my first.

If mass flu vaccines every year are the answer, then just how severe is the flu problem exactly, specifically as it relates to its worst possible outcome, death?  We know people die from the flu each year, but how is this data known, confirmed, and compiled to measure the mortality burden the flu imposes on society?

My initial thoughts before researching answers to these questions were as follows: since many illnesses can masquerade or mimic the flu virus, a lab-confirmed test should be performed to confirm a flu infection, and of those lab-confirmed flu infections, death certificates should state flu as the primary cause when tests corroborate a medical autopsy as such.  These statistics would then just be confirmed and consolidated from local and state levels into a national figure for a given year.  At least, that’s how I think an ideal way to aggregate this data would be more or less performed in order to accurately determine deaths attributed to the flu.

Post 2003, I’ve read anything from 23, 607 (Centers for Disease Control–CDC) to 36,000 (most quoted figure) and 41,400 (National Institute of Allergy and Infectious Diseases–NIAID) for the number of flu-caused deaths per year in the US.  But prior to 2003, the CDC used 20,000 as the annual figure of flu deaths.  However, the reason for this wide variance is simple: since most US death certificates don’t list the flu as a primary or secondary cause of death and of those that do, even fewer are lab-confirmed for the flu virus, the amount of deaths attributed to the flu each year must be guessed.  Different assumptions and even different computer models can yield vastly different estimates.

Also, a peculiar thing to note with the CDC’s annual mortality data from its National Vital Statistics Reports, is that influenza is grouped along with pneumonia, but the overwhelming number of deaths in this “Influenza and Pneumonia” category was from pneumonia instead of the flu (together, the 7th or 8th leading cause of death in the US, depending on year):

CDC Excerpt

The data above is from the 2006 final version of the NVSR.  The final versions of these mortality data reports seem to be produced on a 2- or 3- year lag and unlike preliminary versions, details the breakouts of the “Influenza and pneumonia” category.  Currently, 2009 is the latest version for finalized data, and going back 12 years to 1998, deaths attributed to influenza by age grouping and year are as follows:

CDC Table

So, grouping the flu with pneumonia (even though the two don’t always correlate) and using different modeling techniques and assumptions over time (instead of measuring more directly), it’s difficult to obtain an accurate, unbiased picture of how severe flu deaths are each year.  The CDC doesn’t agree with itself all the time either.  And I’m far from the only one to notice these discrepancies.

In 2006, the Journal of American Physicians and Surgeons voiced criticism of the CDC’s methodologies for estimating and classifying flu deaths:

The CDC and news media frequently proclaim that there are about 36,000 influenza-associated deaths annually.  Review of the mortality data from the CDC’s National Vital Statistics System(NVSS) reveals these estimates are grossly exaggerated.  The NVSS reports preliminary mortality statistics and distinguishes between influenza-related deaths and pneumonia-related mortality.  When the final report is issued, influenza mortalities are combined with the far more frequent pneumonia deaths, yielding an exaggerated representation of influenza deaths.  Pneumonia related mortality due to immunosuppression, AIDS, malnutrition, and a variety of other predisposing medical conditions is therefore combined with seasonal influenza deaths.  The actual influenza related deaths for the years 1997 to 2002 ranged from 257 to 1,765 annually.  These values are further overestimated by combining deaths from laboratory-confirmed influenza infections with cases lacking laboratory confirmation.

In the prior year, a review article titled “Are US flu death figures more PR than science?” was published in the British Medical Journal noting similar criticisms:

Meanwhile, according to the CDC’s National Center for Health Statistics (NCHS), “influenza and pneumonia” took 62 034 lives in 2001—61 777 of which were attributed to pneumonia and 257 to flu, and in only 18 cases was flu virus positively identified. Between 1979 and 2002, NCHS data show an average 1348 flu deaths per year (range 257 to 3006)…

William Thompson of the CDC’s National Immunization Program (NIP), and lead author of the CDC’s 2003 JAMA article, explained that “influenza-associated mortality” is “a statistical association between deaths and viral data available.” He said that an association does not imply an underlying cause of death: “Based on modelling, we think it’s associated. I don’t know that we would say that it’s the underlying cause of death.”…

Before 2003 CDC said that 20 000 influenza-associated deaths occurred each year. The new figure of 36 000 reported in the January 2003 JAMA paper is an estimate of influenza-associated mortality over the 1990s. Keiji Fukuda, a flu researcher and a co-author of the paper, has been quoted as offering two possible causes for this 80% increase: “One is that the number of people older than 65 is growing larger…The second possible reason is the type of virus that predominated in the 1990s [was more virulent].”…

If flu is in fact not a major cause of death, this public relations approach is surely exaggerated. Moreover, by arbitrarily linking flu with pneumonia, current data are statistically biased. Until corrected and until unbiased statistics are developed, the chances for sound discussion and public health policy are limited.

None of this is to be misconstrued as denying that people die from the flu or that the flu isn’t a serious issue, only that there are serious methodological issues in how this is defined, classified, and estimated on a large scale since it’s not measured in a direct, straightforward manner.

Also, taken in context with my previous post about how the CDC/FDA/HHS is more of a cheerleader and business partner (with Supreme Court blessing) with vaccine manufacturers than an independent regulator that places public health over private profits, like the media advertising of flu vaccine benefits (my first post about the flu), the magnitude of flu deaths has so far been greatly exaggerated to boost more vaccine sales through fear-mongering rather than cold hard facts and evidence that should guide public health policies.

More efforts should be invested in having better surveillance of flu-caused deaths instead of relying so much on assumptions and indirect guesses that are heavily sensitive to bias in either direction.  Lab tests should confirm flu virus infections and there should be more effort to have flu as a primary cause of death in instances where viral lab tests are positive and corroborate a medical professional’s judgment cause of death.  This would make deaths attributed to flu more reliable and easy to tally.  Also, the flu should be its own mortality classification separate from pneumonia.  All of these measures would provide us with a more accurate, consistent, and reliable picture of how big of a burden flu deaths are on society each year and hopefully guide public health policy decisions in a more rational manner.

Final Conclusions at this Time

Before concluding this off-topic series of posts about my own thoughts related to flu vaccines, I thought it’d be interesting to look at the CDC’s Vaccine Adverse Event Reporting Sytem (VAERS) database, more specifically using the query tool from the National Vaccine Information Center’s (NIVC) Medalert (it’s an easier, more intuitive interface).

I was just curious how many different adverse reactions are reported each year just for the flu vaccines for 1998-2009, and here is what I found:


Considering that a typical flu season utilizes about 100 million vaccines, I thought it was pretty impressive that only 46,000 to 76,000 different adverse reactions are reported each year.  The 1986 National Childhood Vaccine Injury Act that was discussed in my second post about the flu, requires doctors and vaccine makers to report adverse reactions, but the NVIC estimates compliance with this requirement to be between 1% and 10% in any year.  Outside of it being part of the law, I’m unclear what the penalties, if any, there are for non-compliance.  But if we take the NVIC’s estimate seriously, that means the average number of adverse reaction reports where at least the flu vaccine is implicated over this 12 year period (64,437) is probably an under-estimation of the adverse reactions.

I also noticed that the 2001 CDC figure for flu deaths (257) didn’t seem to cohere well with the corresponding VAERS death records (432).  Given that the 257 is an indirect guess and VAERS records can be submitted by anyone (though mostly it seems to be some type of medical professional), it’s hard to make perfect sense of these data sets.

In conclusion,

1. The magnitude of flu-caused deaths in the US seems to be relatively small according to the CDC’s own data and greatly exaggerated (and inconsistently over-estimated and confusedly classified) as well;

2. The benefits of flu vaccines are also greatly exaggerated in the media compared to the results of independent, controlled studies; and

3. The number of adverse reactions in context of the number of flu vaccines given in an average flu season is also small by comparison.

How did such molehills become towering mountains?

Category: Uncategorized
  • Andrea Long says:

    excellent post!

    02/26/2013 at 5:31 AM
  • Todd Barton says:

    You’re in good company with the JAPS authors and Peter Doshi (the Harvard grad student who wrote the BMJ opinion piece) in misconstruing the CDC stats. As explained in the Science Blogs post you linked to and the online responses to the BMJ piece – including the one by 7 CDC scientists – those are causes off the death certificates, which are a physician’s best guess at what caused the death and not a definitive, scientific assessment. That’s why influenza and pneumonia are lumped together since they present similarly. Since flu compromises the immune system and puts a large strain on the body in general, secondary conditions often result that then are the cause of death. That includes stroke and cardiac arrest (http://www.ncbi.nlm.nih.gov/pubmed?term=22798684), which wouldn’t show up in the “Influenza and pneumonia” category at all. The bottom line is that those statistics are not grounds for doubting the CDC’s flu-caused death estimates of 25k/yr, since they are apples to oranges.

    By the way, I’d steer clear of referencing JAPS articles in the future, given the lack of scientific integrity that organization and its journal has: http://www.sciencebasedmedicine.org/index.php/the-journal-of-american-physicians-and-surgeons-ideology-trumps-science-based-medicine/

    Interestingly, The Atlantic published an article in 2009 that makes pretty much the same arguments that you have been making (albeit specifically in regards to the swine flu). Hopefully you will find the rebuttal here http://www.sciencebasedmedicine.org/index.php/yes-but-the-annotated-atlantic/ more impressive and enlightening than what I’ve been able to provide. The author definitely hits the biggest nail in this whole debate on the head here:

    “What if the 14,400 plus influenza vaccine articles on Pubmed are wrong, all the biology and virology and pharmacology and clinical trials about influenza are wrong? What if every brick in the wall was an illusion and the edifice of flu treatment is wrong. What if a few brave souls can see the real truth. I look forward to a review of the 50 years of influenza research in all its complexity. It’s a huge literature, with multiple lines of evidence all converging on the conclusion that vaccines and antivirals are effective against influenza. Because I would hate to make decisions based on the opinions of a few people reading a narrow sampling of the literature.”

    Apparently you’re still unable to see the bias leading you to continue “reading a narrow sampling of the literature” that results in you seizing on JAPS articles and misinterpreting statistics. Rather than re-invent the wheel – or trying to say it’s not round! – why don’t you start reading through the mounds and mounds of evidence in support of the flu vaccine and the many lives it saves?

    02/27/2013 at 4:07 PM
    • admin says:

      The lumping of influenza and pneumonia is a little awkward, but it’s still easy to parse the deaths CDC attributes to each within that category by age and year as I have been able to do; however, that’s not the primary reason to doubt the CDC’s figures as Mr. Doshi argues in his BMJ piece: it’s the fact that the published CDC data estimate anywhere between 257 to 3,006 flu deaths depending on the year while their PR statements on flu death estimates have ranged from 20,000 to 36,000 per year. Wouldn’t you question this as well? Which published CDC estimates of flu deaths are correct, if any? Would you use different assumptions than theirs in your statistical modeling?

      In Logic 101, a common form of fallacious reasoning is called the “genetic fallacy” where a conclusion is suggested based solely on someone’s origin or reputation rather than its actual substance or context. As a classic example of this, you’re suggesting that because JAPS is an organization of poor reputation that their paragraph I cited somehow makes my whole argument invalidated. However, anyone with an Internet connection and the CDC website address can independently verify what they wrote about the CDC’s statistical methodologies in regard to annual flu deaths. In fact, this is part of what I did with my research as evidenced in some of the screen prints used in this post. Did you independently verify the substantive accuracy of what was quoted from them?

      Another form of fallacious reasoning is called “argument from popuarlity” which is based on the premise that since a large consensus of people can’t be wrong, that if this consensus believe X is right and Y is wrong, then X must be right. We both know this is unsound reasoning, but this has essentially been your argument as demonstrated again in the rhetorical paragraph you quoted. If 14,400+ Pubmed vaccine articles are wrong (hypothetically), it’d still be just as wrong as if there were only 1 or even 1,000,000 wrong articles.

      In response to your question “Rather than re-invent the wheel – or trying to say it’s not round! – why don’t you start reading through the mounds and mounds of evidence in support of the flu vaccine and the many lives it saves?” I have done a fair amount of my own research, expecting at first to find evidence of the great problem the flu posed to society and the efficacy of such vaccines, but found it lacking and changed my mind, reserving the right to do so again in the future should better evidence become available. I’ve learned that one cannot always reliably outsource one’s own thinking to consensus opinions in many contexts, so that’s why I decided to decipher the data for myself.

      02/27/2013 at 10:22 PM
  • Todd Barton says:

    Actually, my remark about the poor reputation JAPS has was in response to what appeared on your part to be a “genetic fallacy” by dropping a fancy-sounding name. ;-)

    However, I committed no such fallacy in the main thread of my argument since I responded directly to the claims given in the JAPS article where they – like you continue to do – conflated the two sets of statistics. One set – the ones shown in your chart above – is what physicians put on death certificates as their best guess between “flu” and “pneumonia.” The other set takes several of the causes reported from those certificates – not just “influenza and pneumonia”, but also respiratory and circulatory, since flu exacerbates those as I explained above – and performs regression analysis on them to arrive at the death-by-flu figures the CDC reports. As a matter of fact, the respiratory and circulatory regression number is almost four times larger than the “influenza and pneumonia” one! It’s all spelled out in the link you provided here: http://www.cdc.gov/MMWR/preview/mmwrhtml/mm5933a1.htm?s_cid=mm5933a1_w

    What’s not spelled out in that report though, is what the regression variables used are and how they were calculated. That’s where the likes of PubMed, 50 years of research, etc. comes in. Your accusation of “argument from popularity” only applies to subjective issues. Acknowledging the preponderance of objective, measured, and material evidence that proves the flu vaccine’s efficacy is no more resorting to “argument from popularity” than it is when acknowledging the preponderance of objective, measured, and material evidence that proves evolution and climate change.

    Since neither of us are qualified, or have the resources, to thoroughly evaluate that entire body of evidence, we must trust the process behind the science that resulted in that evidence. However, if you really think your analysis truly challenges that science, then don’t keep it tucked away in a blog. Instead, submit it to a peer-reviewed journal and rock the epidemiology world to its core! I will be the first to congratulate you on your Nobel Prize and be envious as you go down in history as the man who uncovered the decades-long fraud and conspiracy of the flu vaccine that fooled billions. And I’m not just using hyperbole; that’s exactly what happens to the people who make such discoveries.

    I appreciate and encourage your efforts at self-education (I know I’ve enjoyed learning all the stuff I have from our debates), but I would encourage you to apply yourself using the best sources available (a la PubMed) rather than Google searches that can’t discriminate between science and pseudoscience. In this debate, I’ve yet to see you reference a robust study (other than the CDC’s, which isn’t a study per se and which you misinterpreted). Please provide links to the ones you’ve exhaustively analyzed and accurately interpreted to arrive at your conclusions. Otherwise, seek them out to learn what the regression variables are, and what the science behind them is, if you truly want to verify the CDC’s numbers. Without understanding those, you’re just taking shots in the dark.

    02/28/2013 at 11:20 PM
  • Maura Baldwin says:

    Great post! I’ve been doing a similar search with a similar set of questions and you have assembled the information in a highly coherent manner which saved me much time. Not that this is the end of questioning, rather the beginning. And I hope it’s not your Final one on this topic.

    The debate is also of interest and I appreciate the additional resources provided. However, the personal jabs are unprofessional and imply self doubt on the part of the attacker.

    I am interested in “proof of the flu vaccine’s efficacy”. Does this mean the vaccine works in a lab? Has data been gathered to compare people who have had the vaccine, got the flu (which one?) and survived vs. those who had the vaccine and died of the flu vs. those who were not vaccinated, got a flu and died vs. those who were not vaccinated, did not get a flu and lived.

    The question you pose at the very beginning of your post is a good one. If the flu is such a huge killer, perhaps it would behoove the CDC to require medical professionals to perform lab tests to verify the strain of flu present and report this so that the CDC can generate statistics.

    The goal we share, I believe, is to find consistent clear documentation of leading causes of death.


    01/11/2014 at 4:34 PM

Your email address will not be published. Required fields are marked *