Sacking Dr. Nick

Yesterday’s post on the CDDS numbers drew a response from JB Handley. Cool. However it seemed that he entirely missed the point of the post, either through misunderstanding or wilful choice. I’ll discuss this aspect of his commentary later, but for now, lets try and remove all ambiguity from the point of the CDDS post.

Poor Dr. Nick.I fondly imagined that using a well known quack would demonstrate the usefullness of CDDS for autism epidemiology as subtly as the dialogue in an Arnie movie demonstrates Arnie enjoys big guns. Sadly it seems that that optimism was misplaced as both commenter ‘666sigma’ and Brad showed they didn’t get it.

In one sentence, the previous post was made to illustrate the inability of CDDS data to show anything meaningful regarding the epidemiology of autism. Using CDDS I managed to ‘prove’ that the fastest growing cohort on CDDS was the 62 – 99 year olds.

I did this as for the last few years Brad, David Kirby and Rick Rollens have made a series of suggestions, proclamations and predictions regarding this data. All revolve around the central hypothesis of thiomersal containing vaccines causing autism. Lets look at what they’ve said.

In an interview with FAIR Autism Media, David Kirby said:

It’s now 2005. Mercury started to be removed from vaccines roughly in 2001, we don’t know exactly when as the FDA won’t tell us, but kids entering the system now, four year olds for example in California entering the Dept of Developmental Services [CDDS] were born in 2001. So those kids theoretically get less mercury on average than kids born in 2000. So we should see fewer cases entering the system this year than we did last year.

So in 2005 David Kirby says CDDS should show a dropoff in numbers starting in 2005. 2005 came and went and guess what? No dropoff. He then claimed he’d never said that and instead said in an email to blogger Citizen Cain that:

if the total number of 3-5 year olds in the California DDS system has not declined by 2007, that would deal a severe blow to the autism-thimerosal hypothesis….total cases among 3-5 year olds, not changes in the rate of increase is the right measure.

No explanation was given for this tacking on of two extra years.

And so, 2006 came and went and here we are in 2007 and guess what? Still no decline.

No explanation has been offered by David Kirby as to this. He is now claiming, in debates about vaccines, that enviornmental mercury from forest fires, crematoriums and China is partialy to blame.

Lets be clear on this – he might be right. But a) No data has been offered to support this hypthesis either and b) this has absolutely no bearing on the vaccine hypothesis.

David Kirby put forward a hypothesis that thiomersal containing vaccines cause autism and that using CDDS data we would see a decline in 2005. We didn’t. He then claimed he meant 2007. Again, we didn’t.

And what of Brad? IN an interview with the Times Union in July 2005, Brad said:

Late 2006 should be the first time that rates go down,” said Handley. “If they don’t, our. hypothesis will need to be reexamined.

Just to reiterate, rates did not go down. They are still rising according to CDDS data.

And so we come to the crux of this issue. Do the mercury militia want to carry on using CDDS or not? If they do, then they need to explain why all of their predictions have so far failed to come true.

If they do not then they need to contradict Rick Rollens, who launched another epidemic-supporting missive yesterday. He says that there must be an epidemic because 78% of the CDDS autism population are under 18. Rollens claims this works out to approximately 14,000 ‘missing’ adult autistic people.

Mr Rollens reached these figures by using CDDS data. Amusingly, Brad quoted Mr Rollens in his comments yesterday and thus exposed the poor logic he utilises.

Which is true? Is there an epidemic of autism as shown by CDDS numbers? Or does thiomersal cause autism?

In the past, the ‘epidemic’ hypothesis supported the thiomersal hypothesis but as we have now passed three separate deadlines for a drop in numbers and as a drop in numbers fails to materialise (as evidenced by Mr Rollens) then these two hypothesis have now become competing rather than complimentary. You can’t have an ongoing epidemic and still say thiomersal causes autism. As I’ve discussed before, at an ACIP meeting in 2002 it was revealed that the total thiomersal containing vaccine (TCV) population was 1.9% as of Feb 2002.

So, on one hand we have virtually no TCV’s and on the other we have a steadily increasing amount of autistic people in CDDS across all cohorts.

The combination of this extreme paucity of TCV’s and Mr Rollens continued insistence we have an autism epidemic are thiomersal hypothesis killers.

But lets turn this whole thing around a minute. The point of yesterdays post was to illustrate that CDDS is not a good source of epidemiology. Why do I say that?

First we should note that Brad accused me of only saying this _after_ my post. As I said at the start of this one, I would’ve thought Dr. Nick guest blogging this would’ve been a bit of a giveaway, as would the title ‘manipulating CDDS’ but still….

I first recall mentioning the shortcomings of CDDS in Feb 2006. I can also recall Do’C, Joseph, Diva and Jon posting at various times between then and now about how poorly the mercury militia were using CDDS figures. And in fact, I can remember Brad being part of those debates too. He must’ve forgotten that.

So why are so many of us saying that CDDS is a poor source of data for autism epidemiology?

Well, mainly because thats what CDDS says themselves. There is a PDF on the front of the site with a warning against doing exactly what Rollens, Kirby and Handley have done/are doing. In April 2005 Autism Diva wrote to CDDS for clarification as to why this was. The following points emerged (bulleted – read the whole entry on Diva’s site for detail – quotes from a CDDS email to Diva)

1) Numbers reported do not represent all persons with developmental disabilities in the State of California
2) The numbers can not be used to report the incidence of autism
3) The number does not represent only individuals newly added (or new intakes) to the DDS system.
4) New intakes may be one component of the net value.
5) The difference in the number of persons reported with a given characteristic from one quarter to the next is simply a net value of the above factors
6) It is inaccurate to represent the change in the numbers reported from one quarter to the next as “new intakes”
7) In some cases the developmental disability recorded on an individual’s CDER may be changed to another type of developmental disability

Surely it must be clear to even the most opportunistic of armchair epidemiologists that using this data is _a bad idea_ ?

Diva has also provided further clarification as to the makeup of those autistic people in the CDDS system. It has long been an article of faith to the mercury militia that CDDS figures are good because they are made up _only_ of ‘full spectrum’ autism (classic, low functioning, whatever) and they use this to exclude Aspergers, High functioning etc. Here’s David Kirby:

…the Golden State, however, is said to operate the gold standard of autism epidemiology, having always tracked “full-blown” autism only, as defined by the DSM-IV manual. In other words, children with milder forms of the disorder, such as PDD and Apserger Syndrome, need not apply for services.

It seems that that just isn’t so:

The following was published in the Journal of Autism and Developmental Disabilities. It is a letter written by Rita Eagle PhD of the California Dept. of Developmental Services (DDS). (Journal of Autism and Developmental Disorders, Vol. 34, No. 1, February 2004)…..The California Department of Developmental Services (April 2003) reports that the proportion of “higher functioning” children among the autism population has substantially increased and is on the rise, with a steady decline in the proportion of persons with autism who also have MR……To many clinicians, it appears that more and more children who, in th epast, would never have been referred to the regional centers–for example, bright but anxious and slightly socially inept kids with average or better IQs and children who, in the past, had been or would have been diagnosed as ADHD, OCD, ODD, anxiety disorder, learning disabilities, psychotic, and so forth—are now being diagnosed wit high-functioning autism and/or Asperger syndrome and referred to the regional centers for services.

So not only are the mercury militia using bad source data, they are presenting it to back up a scenario that simply doesn’t exist.

But hang on here – what am I saying? Surely if I’m stating that CDDS data is no good for autism epidemiology then I must be giving more weight back to the thiomersal hypothesis? If the Rollens/CDDS supported epidemic now undermines the CDDS supported thiomersal hypothesis (which it does) then surely by undermining the epidemic I’m re-establishing the thiomersal hypothesis? Haven’t I and lots of others been using CDDS data to show that autism rates are still increasing even after 5 years plus of TCV removal?

No. What we were doing was saying, ‘OK, if you want to use this data then use it but be aware that we can show that the same data shows the serious flaws in the thiomersal theory.’ Personally, I don’t believe CDDS data shows anything one way or the other. That was the point of yesterdays post – if you can use CDDS data to show that the heaviest area of growth in the ‘epidemic’ is the 62 – 99 cohort then I think we can pretty much say anything.

However, what I _know_ is that its now been five years since thiomersal was removed from vaccines. If it was even a partly contributing causative agent to a rise in autism then we should’ve seen a massive fall in autism. This fall would’ve been absolutely unmissable.