Mercury and autism: Foreordained conclusions?

Light blogging today, as I'm in the O.R. (Although there will be one more brief post, which, thanks to the wonders of Movable Type's ability to let me schedule a time when posts are published, will be appearing early this afternoon, while I'm still working.The reason why I'm delaying it will, hopefully, be apparent.)

Light blogging or not, I couldn't resist mentioning a post by Kathleen Seidel in which she's picked up on something that I hadn't noticed but wish I had.

In a long, multi-topic "roundup" sort of post, near the end, she mentions RFK Jr.'s essay Tobacco Science and the Thimerosal Scandal, which he published around the time of his infamous Salon.com article. In this article, he interviewed Mark and David Geier, the father-son tag team mercury wrestlers, who told him that they had a "forthcoming study" that, according to them, would show that autism rates were starting to decline since thimerosal had been removed from nearly all childhood vaccines in early 2003. Quite naturally, Kathleen assumed that the essay was referring to the latest "study" by the Geiers published in that far right wing pseudoscientific rag, The Journal of American Physicians and Surgeons just a couple of weeks ago. RFK Jr.'s article was published to the web in late June, eight months before the Geiers' study appeared.

Kathleen looked at the methods section of the Geier's most recent paper and found something very interesting, namely that the Geiers paper, published in February, had looked at data from the VAERS database through August 31 and the CDDS database through October 4.

Draw your own conclusions.

MIne are the same as Kathleen's: Given that it's likely that their interview with RFK Jr. that was used in his essay very likely occurred several weeks before RFK Jr. published his essay and article, the Geiers had almost certainly decided what their study would show months before they had actually completed it. That was obvious to me by the statistically incorrect way that they "analyzed" their data, and, given this new information, it's even more obvious to me now.

More like this

Of course they had, they decided that before they started.

My guess is that they added more datapoints as long as they fit their predetermined answer and stopped when the numbers started to climb around September/October. (Anybody who can check this?)

Are the Geiers' represented at crank.net ? They sure are qualified.

By Magnus Malmborn (not verified) on 09 Mar 2006 #permalink

I Gant believe it. Where do they get their electricity on Martha's Vineyard anyway? Is it solar? Geothermal? Tidal? I'm trying to remember my bike rides around the island and I just can't recall seeing a nuclear power plant, a fossil fuel generator, or wind turbine anywhere, yet all of those modern conveniences were powered by a mysterious force of some kind. It was Utopian and free from the smog and pollution of the mainland. I would hate to see all of that ruined by offshore wind mills. Some inexperienced flyboy might run into one at night if he isn't properly trained to navigate by instrument. What good is clean renewable energy if birds are too stupid to avoid the blades?

I'm all for removing thimerosal from flu shots but burn all of the mercury laden coal you want Bobby. You can always order the Arctic Char at the Ritz Carlton.

Martha's Vinyard gets it power by undersea cable from the mainland. It was featured in a Jaws movie.

That's right! Well let's get Sheriff Brody and the folks at Woodshole to kill all of the sharks so they don't destroy the cables :-)

If Verstraeten's bosses had planned better, he wouldn't have had to do all those rewrites.

Of course, the reason the Geiers never have to do rewrites is because they make the data fit their hypothesis, rather than changing their hypothesis when new data refuting it come in.

The autism-mercury conspiracy theorists should be cautious about drawing a comparison between the thimerosal-autism "connection" and the tobacco industry's feeble and ham-handed attempts to hold back the tide with junk science.

Am I the only person who sees the autism-mercury proponents playing the part of the tobacco industry's scientists? Check this out:

[1a] The tobacco industry's "research" was overwhelmed by the masses of research data showing a connection between smoking and disease.

[1b]The autism-mercury cabal's "research" is overwhelmed by the masses of data refuting the connection between autism and mercury.

[2a] Government-funded research showed that there was a connection between smoking and disease.

[2b] Government-funded research is showing that there is no connection between mercury and autism.

[3a] The "research" refuting a connection between tobacco and disease was of poor quality and was not taken seriously by real scientists - even at the time.

[3b] The "research" showing a connection between mercury and autism is of poor quality and is not taken seriously by real scientists.

[4a] The "research" refuting a connection between tobacco and disease was aimed at the tobacco consumer (or the legal system), not at the medical or scientific research community (who already knew the truth of the matter).

[4b] The "research" showing a connection between mercury and autism is aimed at parents, who are being advised to consume medical "therapies" (or legal services - such as expert testimony) sold by the researchers. There is no serious effort to "convince" the medical or scientific research community, since the quality of the studies is so poor.

In short, I see the mercury-causes-autism "researchers" and their accomplices playing the part of "Big Tobacco" and its "researchers" in the current drama.

Prometheus

Will one of the statistical wizards here tell me who will win the Derby this year?

Fore Sam,

You might want to ease up on the booze a bit; that last one didn't make any sense.

I think that the point of Orac's posting (and I don't wish to put words in his mouth) is that the Geiers found in the VAERS database only exactly what their clients had put in there. Come on - after over eight years of Bernie Rimland telling parents at the DAN! conferences to report their autistic kids to the VAERS database as vaccine-injured, it would have been a major miracle if Geier and Geier hadn't found a "connection".

Let me make it painfully simple, so that you can understand it through the haze of alcohol:

[1] Starting in 1997 (or earlier) Bernie Rimland tells parents to put reports in the VAERS database connecting their children's autism to thimerosal (also the MMR, which you can also find there, but not so much after about 2000).

[2] In 2003, Geier and Geier look in the VAERS database and find all of those reports. They claim that this supports a connection between autism and thimerosal.

Again, this "proves" (or even supports) that mercury causes autism no more than finding packages under the tree on Christmas morning is "proof" of Santa Claus.

In fact, their "research" seems more like someone wandering into a field of Iowa corn and declaring that it must have come from Mars - despite having film of the farmer planting it in the spring.

If you can't see that, you should probably visit your optician.

Prometheus

Prometheus;
I don't drink, only non-alcoholic beer. I would think anyone who considered themseelves a whiz at statistics would enjoy the challenge of picking some horse races. I had Giacomo last year at 50 to 1 in the Derby so it was certainly worth my while to interpret those stat's. It might be interesting to see how these experts fare.
I never reported my son to VAERS and there are lots of people who never heard of it. Hell, I run into parents of autistic kids all the time who never heard anything about mercury. What would the percentage decrease look like if everyone had made those reports? There are only four thousand something in that big case that Jeff sell is handling but there are over 100,000 kids diagnosed since 1993.

Not mercury;
Disparaging remarks are not answers.

Fore Sam,

I was trying to give you a way out...and you didn't take it.

If you don't know the basic differences between odds-making and statistics (or even probability), then why the heck are you trying to contradict people who do?

Calculating the actual probabilities on a horse race is impossible - there are too many variables that cannot be measured. Although I don't know exactly how the bookmakers come up with the odds, I assume (based on what I've seen from casinos) that their odds guarantee them a tidy profit no matter who wins - on average.

Because the VAERS database isn't verified - or even thoroughly monitored for duplicate entries - it is absolutely worthless for statistical or epidemiological purposes. That the Geiers have attempted to pretend that it is valid epidemiological data just demonstrates that they are either ignorant of the true nature of the database or are cynical opportunists. Perhaps both.

The VAERS database was intended to be a highly sensitive means of picking up unexpected problems with vaccines. It was acknowledged from the outset that most of the "problems" it picked up would be "false alarms". But they didn't anticipate a concerted effort to "stuff" the database, as has been done by autism advocates and opportunists.

Prometheus

Here's what the Geiers had to say on the subject. See how many times they point out "we don't have much time."
http://www.iom.edu/Object.File/Master/19/140/Transcript%202-9-04.pdf

MR. D. GEIER: These are our children. It just
goes on and on -- influenza vaccine --
(Applause)

[...]we have seen many analyses of VAERS
data in this committee, and it usually is accompanied by a
litany of its strengths and weaknesses. That is a part
that is missing in your paper.
I wonder if you could
comment, from your point of view, on the strengths and
weaknesses of using VAERS data for this kind of analysis.
MR. D. GEIER: I would like to comment that the
CDC itself has developed the methodology that we employed
to analyze VAERS. Dr. Robert Chen published specifically
that comparing vaccines administered to similar-age
populations provides accurate qualitative and quantitative
risk-assessment analyses. What we are not reporting is
that VAERS gives you a true incidence rate of the
reactions.
But why should there be a biased reporting in the
way the reactions are? As a matter of fact, we have
actually addressed that in many cases, like with the
thimerosal.

LOL - the Geiers' used Dr. Chen's methodology. Next they'll tell me that their living room lab is a "world-class research facility". No, wait.

Keep trying, guys. You might actually make a coherent, scientifically valid point someday. Or course, when that happens the world will implode and we'll all be sucked into the vortex.

By anonimouse (not verified) on 13 Mar 2006 #permalink

Great reading, keep up the great posts.
Peace, JiggaDigga

By JiggaDigga (not verified) on 06 Apr 2006 #permalink