Archive | Epidemiology RSS feed for this section

Kirby launches torpedo at Verstraeten, sinks Geier

8 Oct

The thimerosal/autism study by Thomas Verstraeten is one of the big targets for those with the vaccines/mercury cause autism agenda. For what it’s worth, Autism’s False Prophets goes into the history of the Verstraeten study and clearly explains the history of that study.  Not surprisingly, the answer is somewhat different than you might find in, say, Evidence of Harm.

In his recent briefing on Capital Hill,  David Kirby took another jab at the Verstraeten study. He tried to assert that (a) the NIEHS claimed that the Vaccine Safety Datalink was unusable for autism studies and that (b) the CDC agreed. He was incorrect, and, luckily, a staffer caught Kirby at it.

Mr. Kirby is trying to explain his actions in a blog post in which he posts an open letter to that congressional staffer.

Let’s consider something here: the congressional staffer, an M.D., knew enough about the subject to catch David Kirby misquoting the NIEHS. I wouldn’t have been quick enough on my feet to catch the misquote.  Now, David Kirby wants to educate this gentleman. Frankly, the information should be flowing the other way. If Mr. Kirby had shown himself open to such education, say when EpiWonk made it abundantly clear (twice) what Mr. Kirby’s mistakes were, perhaps it would be worth the staffer’s time to discuss this with Mr. Kirby. That said, let’s take a look at Mr. Kirby’s letter.

In regards to Mr. Kirby’s misquotes, he has recently “clarified” his position.  He is writing to the Doctor who corrected him in his briefing here:

As you rightly pointed out (and as I concurred that day) I omitted an important detail in regards to Dr. Gerberdings’s letter to the Committee. I regret that, and never meant to mislead people in the room.

It was a rather artless sin of omission.

I think the lesson for me here is that, when you try to cram a two hour presentation into 25 minutes, it is wise to not include very complicated and, as you put it, “somewhat arcane” details that are difficult to explain in such a short period of time. In retrospect, I probably should have focused solely on the NIEHS report itself, and left the Gerberding letter out of the presentation entirely.

Mr. Kiby iscorrect, it is a confusing situation.  There are two documents–an NIEHS report and Dr. Gerberding’s response for the CDC. But, does that excuse misquoting the head of the CDC in his legislative briefing?

Here’s what David Kirby in his capital hill briefing “quoted” the NIEHS report as saying:

NIH: “We identified several areas of weakness that were judged to reduce the usefulness of the VSD for addressing the potential association between exposure to thimerosal and risk of ASD.”

That isn’t in either the NIEHS report or Dr. Gerberding’s response.  Here’s what Dr. Gerberding actually agreed to:

The panel identified several serious problems that were judged to reduce the usefulness of an ecologic study design using the VSD to address the potential association between thimerosal and the risk of AD/ASD.

Emphasis is mine.  But, we’ve already discussed that: Dr. Gerberding didn’t claim that the VSD has reduced usefulness in addressing the thimerosal/autism question. It made a claim that the ecological studies using the VSD had limitations. But, the recipient of Mr. Kirby’s letter would know that.

Back to Mr. Kirby’s open letter: David Kirby is now presenting his own interpretation of the NIEHS report, in place of Dr. Gerberding’s.

As I interpret things, the panel concluded that the database itself suffered from several weaknesses and limitations, which in turn reduced its usefulness for studies of autism risks from thimerosal (ie, Verstraeten) AND ALSO reduced the feasibility of future studies (ie, ecological ones) that are based on data collected within the VSD.

As EpiWonk aptly pointed out, Mr Kirby’s assertion is not the case. The NIEHS panel suggested a number of possible studies on autism using the VSD.  From the NIEHS report:

An alternate future study design that was viewed positively among panel members was a study of a high risk population, defined, in this instance, as siblings of individuals diagnosed with AD/ASD. A sibling cohort from the VSD would allow comparison of AD/ASD risk in siblings as a function of their thimerosal exposure through vaccination and the sample size would lend itself to supplemental data collection. A related study design based on sib-pairs or sets could be used to address discordant ASD/AD status in relation to thimerosal exposures. Another possibility that generated support by the panel was an expansion of the VSD study published by Verstraten et al (2004). The availability of several additional years of VSD data was seen as an opportunity to provide a more powerful test of any potential association between thimerosal and AD/ASD and would enable reconsideration of some aspects of the original study design (e.g., exclusion criteria). A related idea was to conduct a VSD retrospective cohort study using California-based MCOs linked with the California DDS, which would improve the diagnostic data and provide more complete ascertainment. For each of these designs, the ability to link medical records from mothers with those of their children was deemed critical.

As this reader interprets things, NIEHS seems to find that there is quite a bit of value in the VSD for studying autism, including an expansion of the Verstraeten study.

EpiWonk made the point first, but how can the NIEHS say that Verstraeten study design is not a good and that future use of the VSD is not useful, while at the same time suggest expanding Verstraeten?

The bottom line is that there are limitations to using the VSD alone in ecological studies of autism. One can overcome these limitations by going to chart reviews and other methods–as used in Verstraeten et al. and, more importantly, by VSD studies ongoing at CDC (one of which looks at autism).  As noted by Dr. Gerberding:

The VSD currently has a number of priority studies underway to address a range of important immunization safety questions, none of which utilize an ecologic study design. Instead, these current studies, including one study evaluating associations between thimerosal-containing
vaccines and autism, all evaluate individual-level data. This typically involves the review of individual medical charts to confirm the vaccines each individual received as well as the outcomes being studied. Studies using individual rather than group data provide stronger scientific evidence.

Mr. Kirby seems to be neglecting the fact that the CDC’s ongoing study (and the Verstraeten study) is not soley dependent on the VSD for the data.  He seems to be arguing that since the VSD, as a single data source, has limitations, the CDC can’t use it for any study. It’s like saying,

But, let’s take a closer look at what this says….and what Mr. Kirby is saying: The VSD on it’s own is not a good source of data to look at the thimerosal/autism question.

Now, anyone remember all the consternation that has been created by the fact that the VSD is not open to just any outside researcher?  Why should the VSD be opened to, say, Mark and David Geier?  Could they do the individual level data collection needed to make a VSD study valuable?

Apparently not. Recall this study by the Heather Young and the Geiers: Thimerosal exposure in infants and neurodevelopmental disorders: An assessment of computerized medical records in the Vaccine Safety Datalink

This was a study paid for by the petitioners in the Omnibus proceding.   It, on it’s own, was bad enough that EpiWonk disassembled it.  Twice.

The recent Heather Young/Geier paper didn’t look at individual level data.  Any future study by the Geiers almost certainly wouldn’t as well.  Given the argument by the NIEHS, Dr. Gerberding…and David Kirby, the above study and any proposed study by the Geiers on the VSD would be useless.

Some how I doubt Mr. Kirby will make statements confirming that. But, I can’t see how he could hold any other opinion, given the arguments he, himself, has made.

Vaccines on the Hill III

26 Sep

Somehow I never thought there would be a “Vaccines on the Hill II”, much less III. That said, a question from Lisa (from about.autism.com) got me thinking and, well, I’d rather do this a post than a response.

I admit, this isn’t directly related to her comment, who commented on how David Kirby makes a point of stating he is not “anti-vaccine”.

Instead this is about frustrations with Mr. Kirby. As an example, let’s discuss how Mr. Kirby “quoted” a response that the CDC made to an NEIHS report in his congressional briefing. Yes, “quoted” is in quotes for a good reason.

On his presentation, page six, Mr. Kirby “quotes” (there’s those quotation marks again!) the NIEHS report:

NIH: “We identified several areas of weakness that were judged to reduce the usefulness of the VSD for addressing the potential association between exposure to thimerosal and risk of ASD.”

With the response from Dr. Gerberding at CDC of:

Gerberding General Response: CDC CONCURS

What was the real quote?

The panel identified several serious problems that were judged to reduce the usefulness of an ecologic study design using the VSD to address the potential association between thimerosal and the risk of AD/ASD.

Emphasis mine.

Yep. Mr. Kirby left out the fact that the NIEHS was specifically talking about ecological studies.

Makes a BIG difference in how that phrase is interpreted. This was a major part of two epiwonk blog posts, here and here. Mr. Kirby’s original blog post on this was retracted, so Mr. Kirby is well aware of the importance of the fact that the NIEHS limited the statement to ecological studies.

By the way, the real CDC response?

CDC Response: CDC concurs with this conclusion and does not plan to use VSD for ecological studies.

They did most certainly not concur with the statement that Mr Kirby “quoted”. Instead, they see the limitation for ecological studies. There is strength in using the VSD. They don’t see it as valuable for discussing the thimerosal/autism question, as we’ve discussed before.

Here’s the NEIHS report, and here, the CDC response.

Mr. Kirby’s “quote” of the NIH was incorrect. This isn’t incorrect in the way Dan Olmsted thinks that “has” vs. “have” is an important difference. No, the quote by Mr. Kirby completely changed the very meaning of the statement that NIEHS made and implied the CDC concurred with.

It sounds like Mr Kirby was caught red-handed trying it too, by a staffer who obviously came in very well informed. The bright side is that the legislature got an idea of Mr. Kirby’s tactics. The down side, they may not realize that the entire autism community is not represented by Mr. Kirby and his tactics.

This misinformation effort has already had an effect. Mr. Kirby’s original treatment of the CDC response made people think that the CDC position is that the Verstraten study was flawed. As epiwonk makes very clear, the opposite is true. The NIEHS panel suggested expanding the Verstraten study (which was not ecological) with additional years.

And people wonder why I get frustrated with Mr. Kirby.

How the Hidden Horde were hidden

1 Aug

One of the (many) controversies within the autism community is the question of the hidden horde. The basic argument is:

1) Autism diagnosis have ‘increased’ massively in recent years. Prevalence now stands in the UK at approx1 in 100 and approx 1 in 150 in the US.

2) Something(s) must have caused this large scale increase.

This is where the division point is. Devotee’s of the ‘autism is vaccines/TV/mobile phones/whatever’ ideas say that the increase is not in diagnosis but in autism itself. That there really is a massive increase since the early 90’s in the amount of autistic people. They call this ‘the epidemic’.

People like me think that there may be a small ‘real’ increase but it is very small and what we are seeing is the effect of (to quote an authority on the subject):

The shift in how we view autism….is part of a broader set of shifts taking place in society.

…..

Doctors now have a more heightened awareness of autism and are diagnosing it with more frequency, and public schools….which first started using the category of autism during the 1991 – 1992 school year are reporting it more often….Epidemiologists are also counting it better.

…..

Still, these rates may not be proof of an epidemic. Why? Because the old rates were either inaccurate….or based on different definitions of autism than the ones we use now.

Interestingly, in the Autism Omnibus hearings, the families are now arguing (after years of ‘epidemic’ talk) that the amount of children allegedly poisoned by vaccines is so small as to be undetectable. Hardly a hallmark of an epidemic.

One of the arguments used by people who believe there has been an epidemic of autism is to say ‘if there has _not_ been an epidemic, then where are all the adult autistics?’ meaning that if the rate of autism has always been 1 in 100 or 1 in 150 then there should be an equal number of adult autistic people to children.

Its a logical thought but it doesn’t take into account one crucial fact; as far as I know, *no* epidemiological study has tried to count the adult autistic population in any country. So we have no real idea how many adult autistic people there are.

We have some clues – such as the 2004 Scottish audit that revealed that 45% of local authorities in Scotland considered adult prevalence grossly underestimated. For example, Perth and Kinross commented:

Figures for adults reflect the national findings that the numbers known to services/diagnosed represent a significant underestimate of those individuals likely to be affected. For example day centre managers locally consider a number of people to be on the spectrum who have had no formal diagnosis.

And this year, the UK Gvmt announced the would be undertaking the first ever audit of autistic adults in England.

But we do have the odd clue thrown to us now and again that shows where the so-called Hidden Horde might be. As the Scottish Audit suggests, they live amongst us, unrecognised or wrongly diagnosed.

Two recent studies from the Netherlands have shed a bit more light on what may be happening with adults.

In ‘Autistic Spectrum Disorders in Adults‘, the abstract states:

The expression of impairments in social interaction, communication, imagination and mental flexibility changes during development into adulthood.

Autism spectrum disorders in adults may mimic, or be overshadowed by, other psychiatric disorders.

Almost a direct agreement with the Perth and Kinross statement from the Scottish Audit.

The second paper ‘Recognition of autism spectrum disorders in adults‘ has an Abstract worth quoting in full;

Autism spectrum disorder was diagnosed in three adults. The first patient, a married man aged 41, was referred to a psychiatrist with ‘impending burn-out’. The second was a 32-year-old male student with schizophrenia and a depressive disorder who was referred to a centre for autism because a friend of his mother’s knew someone with Asperger’s syndrome. The third patient was a 25-year-old woman with a ‘fixation on food’ who was referred by her general practitioner to a psychiatrist for evaluation of longstanding use of antidepressant medication. Autism used to be thought of as a condition of childhood. Only recently has the diagnosis and treatment of autism spectrum disorders become the focus of attention in adult psychiatry. It is made all the more difficult as during development into adulthood, the expression of disorders of reciprocal social interaction, communication, imagination and repetitive stereotypical thinking and actions, change.

This again shows how autism can be ‘masked’ and how diagnostic tests suitable for children may not be suitable for adults. It also touched on another key issue – that only recently have adults begun to be looked at. It also thirdly touched on another issue – comorbidity. All these three people had ‘other’ psychiatric issues. I have no idea if their diagnosticians considered their autism to be comorbid or if their other diagnoses were considered comorbid to their autism. In the end it doesn’t really matter, except in one important regard: By failing to help these people properly when they were children, did their other psychiatric issues grow so pronounced that their autism was ‘eclipsed’ until a suitable diagnostic test was undertaken? If that is the case then we need to be very aware that there is indeed a large population of adults who have not got a full and proper diagnosis and thus are missing out on help they need and deserve.

Autism Speaks refuses to talk: 1.5 million autistic people? How do you figure?

12 Jul

Not long ago Michelle Dawson, an autistic adult and autism researcher in Montreal, Canada, wrote that she had tried to get Autism Speaks to explain how they arrrived at their well known number of autistic people in the US (or just autistic children in the US, depending on which of their statements you are reading). She phoned the Autism Speaks offices not too long ago and eventually was handed off to a man named Michael. This apparently was Autism Speaks epidemiology “expert” of sorts, Michael Rosanoff, a young man in a sort of low level position at Autism Speaks and without much training in epidemiology it seems.

You can read some of what transpired in Michelle’s conversation
on her QT board And on her blog

I like this paragraph in particular from that blog

On the other hand, if Autism Speaks applies their advertised 1 in 150 prevalence figure to the entire US population (which cracked 300 million in 2006), then the result is a total of 2 million autistics in the US, of whom about 500,000 are children and 1.5 million are adults. But this would mean that there has been a high stable rate of autism. This is a scientifically sound position, but one that Autism Speaks and autism advocacy in general has rejected.

jypsy also contacted Autism Speaks through a form on their website and was sent to check out a page on the Centers for Disease Control and Prevention’s website that explained the 1 in 150 number, which was not jypsy’s question. Hers was the same as mine, “How did you get the 1.5 million autistics in the United States number?”. jypsy was told to check back with Autism Speaks if she didn’t get her answer on the CDC website. She checked back with them reiterating that she wanted to know how they got the 1.5 million number and they didn’t respond.

So I emailed Michael Rosanoff (mrosanoff@autismspeaks.org) and the AS general email address (contactus@autismspeaks.org) my question, which is below Mr. Rosanoff’s response here from Monday, July 7, 2008:

Ms. Clark,

Thank you for the e-mail. I will be glad to answer your questions. Would you be available for a brief phone call? If so, please provide me with some dates/times you are available as well as a number at which you can be reached. I look forward to speaking with you.

Best,

Michael

—–Original Message—–
From: [Ms. Clark]
Sent: Friday, July 04, 2008 5:57 PM
To: contactus; Michael Rosanoff
Subject: Question about autism epidemiology

Dear Autism Speaks and Michael Rosanoff,

I am gathering information for an article I want to write about
autism prevalence in the United States. Autism Speaks uses the CDC’s
latest estimate of autism prevalence, that is 1 in 150.

Autism Speaks also uses a figure for the total number of people with
autism in the United States, 1.5 million.

1) Can you show me how you arrived at the 1.5 million number?

2) Which census numbers were you using when you arrived at 1.5 million
and which prevalence numbers were you using?

3) Is it possible that the number of autistic people in the US is
significantly higher that 1.5 million? That is, if you divide the
currently population of the United States by 150 would you arrive at
1.5 million?

4) Can you tell me how you see the 1.5 million breaking down by age
groups, in other words, about how many of the 1.5 million are about
age 3-15 and about how many are 16-21? Also, how many are over 21,
and how many are over 50? If you can break out the ages using
different age groups that would be fine, but basically, how many
youngsters and how many adults?

5) Please indicate if you are using numbers for the whole spectrum:
autistic disorder, pdd-nos and Asperger’s disorder, or if you believe
that a set of numbers only applies to people with “classic autism”.

Thank you very much in advance.

[Ms. Clark]

I sent my question to Mr. Rosanoff on Friday, July 4, and he responded on Monday which was nice and prompt. I answered him on Tuesday saying that he could call me Thursday and gave him a time slot. I then told him if he confirmed that time was good I’d send him my phone number.
He didn’t respond.
Then I found out that it was likely that Mr. Rosanoff would not be responding, that it was likely he was just trying to give the impression that he wanted to respond… since really, he could have responded in the email to me instead of asking if we could talk on the phone. The whole thing about needing to talk on the phone struck me as a little bizarre anyway. Why couldn’t he just answer the questions, he being Autism Speaks epidemiology expert? Shouldn’t he be able to crank out the answers in his sleep? For that matter why isn’t there a page or two on Autism Speaks devoted to explaining how many autistic adults they are trying to serve as the big mondo world encompassing AUTISM (b)org?

So after getting no response about my offered time slot for Michael Rosanoff to call me, I sent Mr. Rosanoff and a few others at the now curiously silent Autism “Speaks” (including to media@autismspeaks.org) semi-cranky email saying this shouldn’t be a so difficult you guys, and asking them just to answer my questions, please, thank you. And I have had no response from any of those I emailed.

So I encourage any of you to also email Autism Speaks or to call them and ask them, how do you get your 1.5 million number. About how many are adults and how many are children? You could also ask them if they have a vague idea of how many of the total have “autistic disorder” and how many have another of the ASDs/PDDs.

If you’d like to send them a letter you could address it to:

Autism Speaks
2 Park Avenue
11th Floor
New York, NY 10016

This is their general phone number (212) 252-8584 and their fax: (212) 252-8676, if you prefer faxing the question. I suppose they answer the phone from 9-5 Monday through Friday, Eastern Time excepting holidays. I encourage all to use your “indoor voice” when calling and be polite as possible.

This is not a trivial question. If Autism Speaks wants to be BIG AUTISM* the borg (We are Autism Speaks. Resistance is futile. You will be assimilated.) of autism organizations then they should be able to give basic facts about the community they are claiming to serve when they are out pounding on people’s emotions to get them to fork over cash to support Autism Speaks.

*”Big Autism” is an apt term that was coined by the blogger Prometheus.

It isn’t RhoGaM either

30 Jun

If it’s a vaccine, or similar to a vaccine, especially if it contains thimerosal you gotta figure someone, somewhere has blamed it as a cause of autism. Such is the case of anti-D immune globulins, like RhoGaM. RH-negative mothers are given these shots while pregnant to protect their RH-positive babies.

If you are thinking like I was at first, “RhoGaM…where I have heard this recently?”, probably the most famous case involving RhoGaM is that of the Sykes family in what became Sykes v. Bayer. This was blogged extensively at neurodiversity.com, including the recent dismissal of the suit. On the way, the Plaintiffs subpoenaed Ms. Seidel (the more than a mere mother who runs neurodiversity.com). Said subpoena was quashed, and Mr. Shoemaker, the attorney who subpoenaed Ms. Seidel was sanctioned for what was deemed a misuse of his powers as an officer of the court.

But, I digress…Back to RhoGaM and similar products.

Earlier this month, a study by Lisa Croen (and other people familiar in the epidemiology of autism) was published:

Maternal Rh D status, anti-D immune globulin exposure during pregnancy, and risk of autism spectrum disorders.
Lisa A. Croen, PhD; Marilyn Matevia, MA; Cathleen K. Yoshida, MS; Judith K. Grether, PhD

It’s worth reading, and, I would bet money, is soon to be attacked. Why? Take a look at the last line of the abstract:

“These data support previous findings that prenatal exposure to thimerosal-containing anti-D immune globulins does not increase the risk of autism. that prenatal exposure to thimerosal-containing anti-D immune globulins does not increase the risk of autism.”

One could speculate whether the parties in Sykes v. Bayer were aware that this paper was in press, and whether that had anything to do with the dismissal. Without any more information, it would be just that: speculation.

Something that rises above speculation, but is below the level of a conclusion of the paper is this question: does this mean that this level of thimerosal exposure to pregnant women (at least at the times similar to that of RhoGaM injections) is off the hook in general? Or, to put it more simply, does this tell us that thimerosal containing flu shots given to pregnant woman are likely not a contributor to autism risk?

This is an interesting question in some circles. In the shifting sands of the desert that is the thimerosal-caused-autism science, the thimerosal containing flu shot is gaining prominence. Ignore the fact that a minority of pregnant women get the flu shot. Somehow, these flu shots are supposed to be taking up the slack left behind by the phase out of thimerosal from the pediatric schedule. (I can hear people now, “oooh, he’s perpetuating the myth that thimersosal is gone. There are trace amounts left. Trace amounts!” Another of the sand-dunes of the desert.)

OK, for the ultra involved, I bet some people fixated on the statement: “These data support previous findings…”. Some people will be angry with the thought that Croen et al. seem to be ignorant of “Neurodevelopmental disorders, maternal Rh-negativity, and Rho(D) immune globulins: a multi-center assessment.” By Geier, Mumper, Gladfelter, Coleman and Geier. Yep, our good friends the Geiers. I believe that is also our new friend, Dr. Mumper as well.

But, rest assured, Croen and company are aware of Dr. Geier’s team and included their work in their analysis.

Similarly, Geier and Geier found that the frequency of maternal Rh-negative status among 53 consecutive non-Jewish Caucasian patients with ASD referred to their genetic clinic was significantly higher than the frequency among 926 non-Jewish Caucasian pregnant women who presented at their clinic for prenatal genetic care (28.3% vs 14.4%, P ” .01), and all ASD patients with Rh-negative mothers received RhIg during pregnancy. Given the authors’ belief that thimerosal-containing vaccines cause autism, it is likely that the ASD patients who seek out their clinical services are skewed toward higher perceived mercury exposure. For that reason, these study findings may be biased and should be viewed with caution.

And, yes, they also note the Holmes “baby haircut“study as well. But that study has been through the wringer so many times, it’s best to move along.

As they indicate, Croen et al. are not the first to look for and fail to find evidence of a link between Rh status and immune globulins. Miles and Takahashi’s 2007 paper, Lack of association between Rh status, Rh immune globulin in pregnancy and autism found, well, a lack of association.

Of course this isn’t going to be the end of the discussion. The Miles and Takahashi study was met in short order by a comment by Bernard, Blaxill and Redwood. I’d bet the cash in my wallet that a response to the Croen et al. study is in the works.

And, unfortunately, that is the problem we as a community have to deal with. As noted (with some sadness, I am sure) in the Omnibus testimony of Dr. Mumper, the question of thimerosal and autism is a closed book to the scientific community. There are a few studies still ongoing, but indications are that they will also show no effect.

It will be hard enough to keep the interest level high amongst legislators as the science quashes the guilt factor of vaccine induce autism. It will be even more difficult if we are perceived to be clinging to the failed theories and ignoring good science.

Autism isn’t “a novel form of mercury poisoning“. Not from vaccines, not from immune globulins, not from plumes of mercury from China. Let’s stop moving goalposts and patting ourselves on the back for being flexible. Instead, let’s accept that falsifiable hypotheses have had their chance and been falsified. That’s learning.

Quag-Geier

14 Jun

I propose that any researcher or scientist who unwittingly gets into a quagmire with the Geier’s should be referred to as being ‘quag-geier-ed’. Its a handy way of referring to people who’ve (possibly accidently) stumbled into a great big pile of shit.

I therefore nominate Professor Heather Young as the inaugural QuagGeier. She has published a paper with the Geier’s which alleged to find:

associations between neurodevelopmental disorders (NDs) and exposure to mercury (Hg) from Thimerosal-containing vaccines (TCVs) by examining the automated Vaccine Safety Datalink (VSD)

One would imagine that any serious researcher who valued her career would be reluctant to associate with the Geier’s who have odd ideas about what is valid research but maybe Professor Young simply didn’t know.

Anyway, I forwarded the paper itself onto Epi Wonk, a blogger who:

…has a Ph.D. in epidemiology from an Ivy League university. Before that I got a bachelor’s degree from a different Ivy League college, a master’s degree in developmental psychology, and a master’s degree in medical sociology from another Ivy League University. I worked for more than 30 years as an epidemiology professor in medical academia and schools of public health, and in the senior biomedical research service at the Centers for Disease for Disease Control and Prevention (CDC). During my career I have been the editor of two epidemiology journals and one more general biomedical journal. I am now retired.

So very, very bad was the quality of this paper that Epi Wonk took three (and possibly a couple more in the future) posts to tackle the numerous issues with it. I plan to recap them here but here is Epi’s take on the paper itself.

So in part I, Epi found the following:

dubious “imputing” or imputation lies at the bottom of the author’s little trick…..

Imputastion is simply – using known data to ‘guess’ at unknown data. Epi gives an example:

…let’s say a researcher has a file of data on children and 8% are missing data values on parent’s household income, 4% are missing data values on gestational age at birth, and 1% are missing data values on birth weight. She decides to use an imputation procedure to impute values for parental income, gestational age, and birthweight where they were missing. Perfectly fine, legitimate, and scientifically valid under most circumstances.

However, when we are dealing with something like autism….

She examines the data and sees that in certain cohorts in her study population the distribution of autism isn’t quite what she would like. So she “imputes” autism cases into the data set. Except that she’s not imputing a value on a variable for an existing study participant. She’s adding imaginary autism cases into the analysis. This isn’t imputation — it’s cooking the data.

Epi was very disturbed about this to the point that xe said:

This is just not done. It’s not valid. It’s not ethical. Adding imaginary cases into a data set borders on scientiific fraud.

Later on in the comment thread that developed, commenter Andrea asked:

Does this mean that Young, Geier and Geier added 45 and 80 cases that were not in the original data sets, that they MADE UP those 125 cases just to add imaginary data points to make the stats results look more like what they wanted?!

Thats exactly what happened.

In Part II of xes detailed look at this paper, Epi concluded that when it come to controls and particularly controlling for any confounding variables:

….there’s no attempt to control for, or adjust for, the confounding effect of birth cohort. Just one look at Figure 1 (or a basic knowledge about trends in autism) tells you the regression coefficients (slopes) are being driven by increases in autism risk over time. Given the increase in frequency of autism (and other neurodevelopmental disabilities) during time time period, you could do an ecological regression analysis of almost any factor that varied over time and you would find an an association with autism. I would bet that you could enter number of sushi bars per capita into an ecological regression and you’d find an association with autism rates.

This is not a good or valid paper. It seems, based on the expert analysis of a professor of epidemiology that this paper is fundementally flawed.

A quiet concession

3 Jun

While much has been made of a certain “concession” lately, it is worthwhile noting that the past month has seen a significant shift in the mindset of people pushing the autism “epidemic”.  In vaccine court it was the petitioners (plaintiffs) who opened with the admission that if there is any group of people with autism as a result of vaccine injury, the number would be so small as to be invisible to epidemiological studies.  (This didn’t stop them from closing with the latest Geier paper claiming that they can see the effect of thimerosal on the autism rates.)

But, this hasn’t been such a sudden shift.  As far back as last November, David Kirby noted:

“Finally, to all those who are going to post comments about the autism rates in California not coming down, following the removal of thimerosal from most vaccines: You are right. The most likely explanation is that thimerosal was not responsible for the autism epidemic.”

Baby steps….baby steps…

That’s my way of saying to those who would bravely rise up in their seats and proclaim, “stop the hate speech!” that the above is progress.  Sure, we still have to continue to endure the “epidemic” language for now.  Why he can’t just admit it in full, I am not sure.  Oddly, Mr Kirby doesn’t put forth a valid suggestion of what caused this “epidemic” of non-thimerosal-induced autism.  He does try a last-ditch effort to prop up the possibility of a thimerosal-induced-autism-epidemic with claims of immigration and pushing the goalposts back to 2011.  

But, I felt I had to acknowledge the fact that even Mr. Kirby is stating that the most likely answer is that it is not thimerosal before I go on.  Because, if one looks at the data just a little farther, one can find a likely answer for why there is a larger number of young people identified with autism than older people.  It isn’t news to us, and it isn’t news to Mr. Kirby and the rest: one likely reason is substitution.  I don’t know which they would try to avoid more, admitting that substitution is a big effect or an MMR jab, but let’s look a the data.  Just a peak. 

Consider  Autism as a function of age in the CDDS (California Department of Developmental Services) data.  Keep in mind, this is service data–i.e. this is a count of who is receiving services from the CDDS, not a count of who has autism in California.  This is not epidemiological grade data.   But, since it has been used so much, let’s move forward, keeping the caveats in mind. (click on the graph if you want to see it larger and more clear)

 

So, Autism count vs. age (data in black), what do we see?  Of course, we see a big peak in the younger ages.  This is the “Tidle Wave” or “Tsunami” or “Insert Other Hate Speech Term Here” that people like to use to make this scary.  But, as you have noticed, I have included the data for “Mild Mental Retardation” as well, in red.   What do we see there?  A dramatic dip in individuals identified with mild mental retardation in the younger ages.  Does anyone believe that something happened in California to reduce the number of people with intellectual disabilities?  If so, they aren’t very vocal about it.

If I were reading this for the first time, I would be suspicious of someone using only one mental retardation category.  Believe me, all mental retardation categories show much lower numbers in the younger age brackets.

This is a clear indication of “administrative substitution”.  [edit–see comments below] Joseph has done a lot of work on the CDDS data in the past and notes (below) that it is very difficult to draw conclusions about substitution from the publicly available data.  However, this is similar to what people have reported in the literature especially using special education data: people with autism have very likely been mislabeled, their autism label was “substituted” for a different label.  They weren’t “missed” as the favorite straw-man arguement would have it, but mis-labeled.

One person in the AutismOne media discussion (as reported by AutismNewsBeat) kept saying, “you ain’t seen nothin’ yet”, implying that the future is going to be very different from the past.  I have to agree, in the future, more adults with autism will be correctly identified.  They won’t be in hospitals with schizophrenia diagnoses, for example.  This will be a good thing.

Frankly, I’d like to see that change come sooner.  I’d like to see adults properly identified and appropriately served now.  We have to move away from baby steps and take bigger steps away from the “epidemic” and acknowledge that unidentified/misidentifed adults with autism are almost certainly out there.

Instead, we have people asking to “Green Our Vaccines“.  The “Green Our Vaccines” idea is a very cute way to mask the typical “epidemic” speech.  But it is the same old story: autism is vaccine injury and new.  Note that almost all of the groups sponsoring the “Green our Vaccines” rally are autism-related.   Sorry to divert to that topic somewhat, but I can’t help thinking that all that effort is being put into that rally by people who actively supress the existance of adults with autism.  I can’t help thinking that the present and the future would be better if they were acting to help everyone with autism, now.

(noted, this has been edited to more correctly reflect Joseph’s opinions on substitution.  Also, I missed another Green Our Vaccines link.)

New blog worth following

20 May

I don’t usually make recommendations about new blogs. Not because I’m above all that – course I’m not – but mostly because so many people have their set ideas about what makes a good blog and they don’t need me pushing my opinions on them.

But this is a little different. Its the first blog I’ve seen that concentrates on epidemiology and is written by someone who:

…has a Ph.D. in epidemiology from an Ivy League university. Before that I got a bachelor’s degree from a different Ivy League college, a master’s degree in developmental psychology, and a master’s degree in medical sociology from another Ivy League University. I worked for more than 30 years as an epidemiology professor in medical academia and schools of public health, and in the senior biomedical research service at the Centers for Disease for Disease Control and Prevention (CDC). During my career I have been the editor of two epidemiology journals and one more general biomedical journal.

Thats some pretty impressive credentials.

I wouldn’t (and I doubt Epi would either) claim that the blog is about autism or vaccines, or autism related science but the two posts I’ve read that have discussed autism have been clear, concise and easy for non-experts to parse.

So, I hope that Epi will continue to blog tangentially about autism from time to time as there are big issues surrounding autism epidemiology that we could all learn about. But more than that I plan on reading Epi’s blog on a regular basis in order to learn.

That’s not to say I expect to become an epidemiology expert simply by reading a blog! Of course not. But that doesn’t preclude me from being able to hopefully discern from an expert what is important in epidemiological studies and what is not.

I would _love_ to see Epi turn her attention to some of the Geier’s epidemiological studies for example. I think we all might learn a lot from a detailed critique of that particular body of work!

I’d also like to see Epi’s opinion on some of the epidemiological studies being utilised in the Autism Omnibus hearings. As we’ve all seen, the epidemiological basis on the autism/vaccine hypothesis seems to have undergone a substantive revision of late. I’d like to see a professionals take on it.

Best of all though? The name Epi Wonk. It reminds me of Wonko The Sane from ‘So Long and Thanks For All The Fish”. Anyone that sounds that much like a man who believes he’s living in a perpetual lunatic asylum can’t be all bad 😉

So – visit Epi Wonk, see what you think.

Newer Entries →