Mark Geier, David Geier and the VSD

10 Oct

Introduction

One of the many anecdotal lynch pins of the Mercury Militia is the fabled story of what happened when the Geier’s attempted to study the VSD database.

Please bear in mind that to the Militia this story carries a *lot* of weight. It is one of the few supporting crutches left under the hypothesis that since thiomersal was removed from vaccines autism cases have gone down. Educational data has failed them. CDDS data has failed them. The Geier’s paper using VAERS (a non starter if they’d only thought about it) was so bad it couldn’t be published in a proper science journal and so this VSD story is all that’s left.

This story is enshrined in the hallowed pages of Evidence of Harm although the source of the story is unclear. Here’s the Militia version.

What The Geier’s Said

The VSD is the Vaccine Safety Database. This database carries raw data related to vaccine safety. The Geier’s were allowed access to this data, together with their computer expert Vale Kernik who would run the statistical programming tool in the SAS language that the CDC’s VSD uses. SAS is a widely used solution for statistical analysis.

The VSD’s Wikipedia page says that:

Only two outside researchers, Mark Geier and David Geier, have thus far gained access to the raw data. They faced formidable obstacles before being allowed into the CDC computer center, and then resistance from staff and software malfunctions once inside. Nevertheless, they reportedly found highly elevated risks for autism among children in the highest mercury exposure group. The Geiers study on the VSD, “A two-phased population epidemiological study of the safety of thimerosal-containing vaccines: a follow-up analysis” was published in the Medical Science Monitor in 2004 volume 11(4):CR160-CR170.

NB: This page has been edited by a member of the Geier household – against Wikipedia recommendations.

Evidence of Harm (the Kirby book) deals with the same event:

In late July the CDC informed the Geiers that the requested data set had been assembled. After paying a processing fee of $3,200, the Geiers were given two dates in August to come and run their studies. But there was another entirely unexpected wrinkle. Just two days before their appointment, a CDC technician called to make sure they were fluent in the programming language SAS, which is used in the VSD database. The Geiers had never heard of it before. “You must not be epidemiologists,” the technician said, “They all speak SAS.” If that were true it was news to the Geiers……Reluctantly they cancelled the appointment. It took two weeks to find someone who could run SAS…..They got new dates in October 2003… the dad, Vale Krenik, flew in from Texas. The were met by a woman who introduced herself and said she would be their “monitor

Evidence of Harm, p280 – 282.

And then things got very surreal. Their ‘expert’ programmer (who apparently taught himself SAS in two weeks) was stymied by the most dreaded sights for programmers – a command line interface.

How on earth can this be happening?” Mark muttered shaking his head, “Once again they got us.” Silence filled the room. There would be no number crunching today. The men stared at the screen.

Sorry. I’m being facetious. Any ‘expert’ who can’t work in a command line at even a very basic level is _not_ an expert.

The weirdness continued when the CDC monitor who was due to accompany the fearless trio for the duration of their stay popped her head out the door, looked both ways, came back into the room and:

She sat down and took a deep breath. “Don’t tell anyone this,” she said in a low voice, “But I can help you.”….I’m telling you, they know,” she said conspiratorially. “There’s a big problem”…..”The autism numbers are going down,” she said, “We are watching them drop.”

This mystery CDC monitor became known as ‘Mrs Toast’. Over on the EoH Yahoo Group, it was discussed why:

There is a woman who I refer to as Mrs. Toast. She is a CDC staffer who was responsible for monitoring the Geiers when they were instructed to visit the Vaccine Safety Datalink by Congress. When she saw the Geiers datasets, she walked out into the hallway, looked both ways, and came back into the room shutting the door behind her.

The Geiers thought they had epi-evidence. Mrs. Toast told them to look at hers. She told them she was responsible for running weekly autism datasets. She was instructed to run datasets on HMO vaccination adverse outcomes to see what effect removing thimerosal from vaccines was having on the epidemic. She had an affected child and made sure that the Geiers understood that the rates were dropping each and every week.

Author David Kirby had an interview set up, flew down to Atlanta, was in a car on the way to CDC to talk to her, but CDC had found out and they were threatening for end her career if she spoke to him.

When Congressman Dave Weldon found out about her not willing to blow the whistle on CDC’s cover-up he said, “THIS WOMAN IS TOAST!” Which is were I gave her the formal name of Mrs. Toast.

Robert Bloch, EoH Yahoo Group

Hilarious right? Mrs Toast.

And so, off trudge the Geier’s with their expert (the one unable to operate DOS). _Imagine_ their surprise when they get a letter from the CDC that said:

1) The Geier’s had violated the terms of their IRB
2) They asked how to merge datasets in a contradiction of the agreed terms of use of the data
3) They were told they couldn’t and yet they tried to anyway
4) If they had managed to merge the datasets they would have increased the risk of a breach of confidentiality.
5) The research team had attempted to rename data files to make them look like part of the SAS program (by changing the file extension to ‘.sas’)

As a result of this, the Geier’s IRB (Kasier) suspended them from undertaking any more data collection at the VSD.

The Geier’s responded by hotly denying these allegations. They first state that they didn’t violate the protocol but as Kathleen says in her exhaustive look at their reply:

The Geiers here claim to have followed the design of their research protocol, yet simultaneously acknowledge that they were attempting to conduct analyses of information not encompassed by it.

More amusingly, on page two of their reply the Geier’s state:

It is impossible for the datasets given to us by CDC to be merged

And then on page three of the same letter state:

What we were attempting to accomplish was to merge the datasets given to us by CDC to build a record…

And so the situation is now that the Geier’s pet ‘expert’ couldn’t figure out SAS, they had no meaningful results and what they did have was gained under extreme deception to the point their IRB approval was suspended.

And so, they decided to go ahead and publish anyway (well, you would, wouldn’t you?) and thus A two-phased population epidemiological study of the safety of thimerosal-containing vaccines: a follow-up analysis was born in 2005 (hereafter referred to as G05).

Geier, Geier, pants on fire?

G05 made reference to the VSD data that the Geier’s couldn’t collect/collected part of/pick your belief. In this respect it was similar to a paper written by ex-CDC staff member Dr. Thomas Verstraeten which _also_ used VSD data to look at thiomersal and autism in 2000. This paper (hereafter referred to as V00) found a statistically significant correlation between thiomersal and developmental disorders.

Oh no!!!! Doesn’t this back up the Geier’s et al?

Well, it _might_ except that as Verstraeten himself states in a letter to Paediatrics:

The CDC screening study of thimerosal-containing vaccines was perceived at first as a positive study that found an association between thimerosal and some neurodevelopmental outcomes. This was the perception both independent scientists and antivaccine lobbyists had at the conclusion of the first phase of the study. It was foreseen from the very start that any positive outcome would lead to a second phase. The validity of the first-phase results needed urgent validation in view of the large potential public health impact. Did the CDC purposefully select a second phase that would contradict the first phase? Certainly not. The push to urgently perform the second phase at health maintenance organization C came entirely from myself, because I felt that *the first-phase results were too prone to potential biases* to be the basis for important public health decisions.

Because *the findings of the first phase were not replicated in the second phase*, the perception of the study changed from a positive to a neutral study. Surprisingly, however, the study is being interpreted now as negative by many, including the antivaccine lobbyists.

So, in short, the first phase of the study using a small sample size indicated there might be an issue. When the second phase was undertaken with a larger sample size, the issue disappeared. Not uncommon in the slightest. Its standard practice to conduct a small, pilot study to see if there’s any issue to study further before committing large amounts of public money to a full scale study.

But I digress – back to the Geier’s.

They knew about the V00 paper – of course they did, it would be hard not to – and as they wrote G05 then they looked at it again. Remember that the Geier’s had struggled at the CDC VSD headquarters.

As Kathleen once again unearthed, the Geier’s – with a lack of VSD data at their disposal wrote their paper. It had some odd elements to it. Here’s a table of stats from the V00 pilot study:

And here’s a table of stats from G05:

Take a look at the numbers. Aside from one category they’re identical. Further the V00 paper states:

The final number of children thus included in our cohort was 109,993.

And G05 states:

The final number of children thus included in the cohort examined was 109,993

Woah! Spooky! By some miraculous, completely bizarre accidental coincidence, the Geier’s – who had little to no data from their visit to the VSD – have the _exact same cohort numbers and divisional figures_ as a paper written 5 years earlier resulting from a pilot study that showed a now debunked association!! What are the odds of _that_ ? I wish I knew a betting man who could tell me!

And maybe my betting amigo could tell me the odds of those same two papers having over ten more virtually identical passages and/or tables of figures? Maybe the Geier’s should drop the litigation gigs and move to Vegas and live on the strip.

A Different Interpretation

So here’s what _I_ think happened. Just conjecture but persuasive I think.

First of all, this odd SAS programming language. The CDC think its common amongst stat-fans. The Geier’s say its really really rare. Google says there are over four and a half million web related resources for SAS programming. That doesn’t sound pretty rare to me.

A leading SAS expert says:

Millions of people around the world in business, science, government, and education use SAS software to work with data. SAS software runs on many operating systems, including Microsoft Windows, UNIX, OS/2, Mac, MVS, and VMS. Most features of SAS software operate the same way in these different operating systems.

Still not sounding rare. In fact its one of the few apps that runs on Win, Mac and Unix. Not a good indicator of rarity.

And as for how quickly their ‘expert’ was defeated by SAS, SAS author Rick Astor states:

Fortunately, SAS programming is not that hard to learn.

Unless of course you’re a computer expert terrified of command lines. Vale Krenik is quoted and described on this page. His job (and former jobs) is described as:

Business Manager, Strategic Supplier Manager, Global Telecom Manager

It’s true that one of these roles has a techy requirement but absolutely _none_ can be swapped with the title ‘programmer’ or ‘expert’.

I think that the Geier’s needed someone who knew computers and settled on Krenik. When it came to it, Krenik didn’t know what the hell to do with SAS. If you’re reading this Mr Krenik, the three lines of code you need to merge datasets in SAS are available. I think they panicked and tried to grab as much data as they could in a brute force attack and then change the data files appearance to try and make them look like SAS files by renaming them with a ‘.sas’ extension.

I further think that the whole Mrs Toast episode is entirely fictitious. It even reads like a bad John Grisham novel. Bloch states that Kirby had an interview set up with the nameless Mrs Toast and that she cancelled at the last minute. Frankly, I don’t believe a word of it. I wonder who set up and then cancelled this meeting? One of the Geier’s by any chance? Does anybody know?

And then there’s the magically duplicated data. The Geier’s realised their VSD data landgrab had failed utterly and so they copied the data (and hence conclusions) from the V00 paper.

I don’t believe the Geier’s have ever seen VSD data. I don’t believe ‘Mrs Toast’ exists.

Our thanks and appreciation should go to Kathleen for the painstaking research she has assembled on the Geier’s. I know mine do.

Originally posted at Left Brain/Right Brain.

Advertisements

18 Responses to “Mark Geier, David Geier and the VSD”

  1. andrea October 11, 2006 at 02:12 #

    SAS is indeed a very commonly-used statistical program. The analyses are run for a wide variety of disciplines, including many of the sciences. I’ve done SAS coding as part of the statistics work for my MSc in Entomology. It reminds me of BASIC (don’t ask me for the details; I’ve not done programming since the days of punch cards). Even on a Windows machine, it doesn’t have a fancy GUI, and is very much like operating at the DOS level. But hey, that’s part of how it can be formulated to run on Windows, Mac and UNIX platforms.

    It’s not hard to find a SAS technician; any university’s statistics department will be riddled with competant people who can draw up the coding and interpret the results. Of course, I guess it helps to actually be affiliated with a university somewhere.

  2. Joseph October 11, 2006 at 02:33 #

    It’s interesting that the vast conspiracy only manages to produce one ficticious anonymous whistle-blower, who might have been scheduled to be interviewed and might have cancelled, and whose existence no one is really willing or able to confirm.

    Ms. Toast is about as real as the work the Geiers did with the VSD data.

  3. Maggie Rosethorn October 11, 2006 at 02:35 #

    Hi, Kev. SAS is a very common program for statistics. It is also a very OLD program; it came out originally in the 1980’s I believe, and has been updated many times. It has strict command line language requirements, but is a great program for statistical analysis.

    The fact that the Geiers couldn’t find someone who knew SAS meant they didn’t bother looking on the internet; on their website SAS lists a lot of information: class availablility, books, etc. (And the fact that their ‘expert’ taught himself SAS in two weeks tells how easy it is to learn the basics…at least, if you actually learn SAS language…)

    I have to say, from my point of view, the Geiers’ story doesn’t make sense. If the CDC was giving them access to the database, why wouldn’t the CDC have created a program and run the data sets based on what the Geiers wanted to look at? The programming would have been basic and taken a technician a very short time to get the data (although, if the datasets are very large, the RUNNING time might have been long). The whole story just makes no sense.

    Sorry this is so long. I’m actually taking my first SAS programming class starting tomorrow; all my current knowledge is self taught.

  4. Ms. Clark October 11, 2006 at 02:45 #

    Baby boy Geier was in college then wasn’t he, or even in a doctorate program that he might have failed in because all he had was a soft Bachelor of Arts (emphasis on the arts) biology degree not meant for people going on to study the sciences? At any rate, how come baby boy wasn’t able to teach himself SAS at least as well as the Dad Vale K. who flew in from Texas to do his heroic computer whatever.

    I’m guessing that mercury dad Vale knows that there wasn’t a Mrs. Toast with an “affected child” stating that the numbers of autistics were dropping like a rock each week… gimme a break.

  5. Ms. Clark October 11, 2006 at 08:21 #

    Another goodie from the Kirby book, They show the 2004 DDS graph where it looks like the rate of increase of new added 3-5 year olds is slowing down (not the number is decreasing, but the increase is “slowing down”). Then they quote a press-release from SAFE MINDS
    July 19,2004
    “It has long been the scientifically supported belief of Safe Minds that such a trend would be witnessed following the decreased use of thimerosal, and this trend seems to support our research,’ Lyn wrote. “In 2002, CDC said they were waiting to see how much prevalence of autism may have been impacted by the removal of thimerosal from vaccines. I guess now they have their answer.” pg 376

    uhuh. Except now it’s more than two years later since Lyndelle preened about how *right* they (looked like they) were … and turns out they were *wrong*. The numbers of little ones with an autism dx in the California DDS continue to go up. So all they can do is invoke a conspiracy to sneak thimerosal into vaccines, or twitch, clear their throats and quickly change the subject.

    So now a new batch of parents of younger kids will start looking for someone else to sue, I guess.

  6. Broken Link October 11, 2006 at 11:52 #

    The CDDS quarterly numbers for September should be out any day now. Check at

    http://www.dds.ca.gov/FactsStats/quarterly.cfm

    Any money on another quarter of steady increases in the 3-5 year old cohort?

  7. Broken Link October 11, 2006 at 11:58 #

    Oh, and a great post, Kev.

    Also wanted to mention that the AAP also really had doubts about that Med Sci Monitor (G05) study, even without the plagarism issue. To quote from their rebuttal here:

    http://www.aap.org/profed/thimaut-may03.htm

    The AAP author states: “The authors” – referring to the Geiers – ” claim to have performed their own analysis of a Vaccine Safety Datalink (VSD) thimerosal screening study (reference 17 in Geier and Geier), although the raw data needed to perform an independent analysis are not available in the document cited.”

    If you read the whole G05 paper, you will see that they give the numbers of children that they will be studying (in total), and the statistical association of (e.g) tics with vaccines, but they do not give the actual numbers of children with autism or tics or anything else. I guess they didn’t realize that those numbers were important. Or else they never had any numbers.

  8. notmercury October 11, 2006 at 17:28 #

    I think they panicked and tried to grab as much data as they could in a brute force attack and then change the data files appearance to try and make them look like SAS files by renaming them with a ‘.sas’ extension.

    Can you explain this in more detail Kev? How could one tell if a file extension was changed? Did they do it in DOS or just rename the files in Windows?

  9. Ms. Clark October 11, 2006 at 19:33 #

    pg 356 (stuff in parentheses is added by me)

    Lyn’s blood was boiling, but she kept quiet. She knew that Sallie and Barbara (Loe Fisher) must have been ready to explode, too. Then during the question and answer session, a reporter asked McCormick (CDC person) about the removal thimerosal from childhood vaccines and the impact that should have on autism rates if the hypothesis were true.

    “Wouldn’t we see a decline?” the reporter asked.
    “Yes,” McCormick said, “And they looked at this very closely in Denmark. But when teh Danes removed thimerosal from vaccines, autism rates skyrocketed.”
    —-
    (Before someone starts screaming, “DEMARK?? DENMARK!!???” the numbers *did not drop* in any case, even if you think that it’s horrible that the way they added cases from other clinics, the numbers Danish numbers did not drop back to “prethimerosal” levels at the Lyn would have predicted a few years ago)

    —-
    “Lyn saw an opening and took it.
    “As a follow-up to your comment about Denmark,” she said, “recent reports from California show that new autism cases have actually declined for the first time in decades, as we have just seen for the second consecutive quarter this April. Can you please comment on that?”

    ——

    The number of new autisic kids in the DDS hasn’t dropped in California, even after 2 more years.

    How about a press conference where Lyn and friends are called on the carpet instead of the CDC?

  10. Kev October 11, 2006 at 20:03 #

    Maggie – thanks for some interesting back story on SAS. Hope your tutorials go well :o)

    BL – I meant to quote that AAP reference, thanks for saving me some editing.

    NM – hard to tell really. No source mentions what OS the VSD’s machines run. If it was Unix based its highly unlikely there was a GUI whereas with Win there might’ve been, although Andrea’s observations seem to contradict that.

    I think they tried to run the merge program. They knew however that according to CDC governance, no data could be stored by visitors and instead a copy of the completed SAS program was compiled by the CDC and given to researchers on exit. So I think they renamed the merged file from whatever the data filetype for SAS is to SAS program files which by all accounts is ‘.sas’ so that when the CDC were preparing the copy for the Geier’s to take away with them the data files, now renamed, would appear to be simple program files and not be examined.

    However, assuming a command line interface, all file properties are usually displayed – including file modification and creation dates and times. The CDC programmers would know immediately that a SAS program file with a creation date/time the same as the current date was impossible and thus suspicious – which of course, they were right about.

  11. ameliorator October 11, 2006 at 20:11 #

    From Kaiser Permanente’s letter revoking the Geiers’ access to the Vaccine Safety Datalink:

    “External researchers are provided with a copy of their SAS programs at the conclusion of their visit to the RDC. On this occasion, the external researchers attempted renaming data files with the “.sas” extension reserved for program files. This would have allowed the external researchers to remove data files from the RDC, contrary to the rules of the RDC. The violation was detected and only the correct files were ultimately released. The programmer stated that the files were created by mistake.”

    http://www.casewatch.org/fdawarning/rsch/geier.shtml

    If there were a Mrs. Toast, one would think they would have been permitted to waltz right out with those files.

    Oh, and “the files were created by mistake” – I guess the programmer’s fingers were suddenly possessed by the devil and forced to alter the file extensions while he was held in a trance. Paging Bradstreet! This man needs an exorcism, stat!

  12. Kev October 11, 2006 at 20:20 #

    ameliorator – yeah, the ‘mistake’ bit is hilarious. The other really funny bit is the inventor of this little fable thinking that Mrs Toast saying: _don’t tell anyone this but…_ to two published researchers mandated by the Government _to report their findings_ is not stretching credulity to breaking point.

    Potboiler scriptwriters – rest easy in your beds. The Geier’s are not (quite) in your league.

  13. Ms. Clark October 11, 2006 at 21:21 #

    Mrs. Toast “told” the Geiers that the CDC was monitoring the VCD data and watching the numbers of autistic children drop:

    Here’s my quick synopsis of the stuff on pg 282.

    The mystery lady said that the SAS program was so hard to understand that they needed three full time programmers to run the VSD database at the CDC …and so it was ridiculous that the CDC expected the Geiers to know how to run it. Then the Geiers run the numbers that show that supposedly show that autistic kids were far more likely to have had the thimerosal containing vaccine compared to the non-autistic kids … and the woman says, she knows the numbers inside and out. Then:

    “I’m telling you, they know,” she said conspiratorially. “There’s a big problem.”

    Then she claims to be monitoring numbers of autistic children as part of some bioterror monitoring thing called “rapid sampling” and she’s watching the numbers of autistic young children drop since they took out the thimerosal…

    “The autism numbers are going down,” she said, “We are watching them drop.”
    ———————————-

    Uhuh. Why is it that only the Geiers still seem to be saying that the numbers have dropped?

    That’s their story and their stickin’ to it. Where’s Mrs. Toast? She’s not working for the CDC now, as I understand the gossip. What would be holding her back? Oh… she’s being buzzed by black helicopters 24 hours a day! That’s it!

  14. clone3g October 11, 2006 at 22:13 #

    SHE’S TOAST!!!

  15. Robert October 13, 2006 at 00:32 #

    Howdy folks,

    Ok, SAS is good stuff. But there’s a new data crunching programming language in town, the first new data crunching programming language for at least 20 years. As it continues to improve, it will allow you to complete difficult transformation tasks with messy, really messy data with a lot less code than SAS.

    available at http://www.my.opera.com/datahelper

    SAS data crunching technology is good, but is it really different from 20 years ago? No. They’re not trying to make explosive new inventions in productivity. They have a financial chokehold on the pharmaceutical industry, so why should they care?

  16. anonimouse October 13, 2006 at 15:48 #

    This is the part of the Geiers’ story that’s an utter an complete joke. Anyone who’s the least bit competent is statistics understands SAS, or has enough basic enough of another database/statistical language (like SQL) to pick up SAS quickly. Considering that the all of the available literature on the VSD mentions needing SAS knowledge, I am stunned – well, not stunned – that the Geiers went in unprepared.

    So Kev’s right. They panicked. Because they’re incompetent pieces of crap they decided to copy Verstraten’s data right from his paper, even though there’s NO WAY they could’ve gotten that data from the datasets they were provided by the CDC in the first place.

    You see, the Geiers weren’t working with the whole database. They were working with a subset of data (kids who received a certain nummber of DTaP vaccines, I believe) that they asked for. There is no way in heck they could’ve gotten that information from the dataset they were working with.

    They’re frauds. And for anyone to continue to believe them (or anyone who depends on their counsel) is pathetic and stupid.

  17. Kev October 13, 2006 at 21:00 #

    3 – 5 yo cohort still increasing – new Quarterly CDDS figures released.

Trackbacks/Pingbacks

  1. Left Brain/Right Brain » David Kirby: Whats with the scaremongering? - November 17, 2006

    […] This ‘long and unseemly history’ as recounted by David Kirby relates to precisely one incident where the Geier’s tried to get access to VSD data and when they couldn’t, turned to plagiarism. […]

Comments are closed.

%d bloggers like this: