Autism Street

If there was machine that could pick an appropriate bumper sticker to accurately represent the information that underlies common beliefs and scientific knowledge, would the information you hold to be true, generate this sticker?

Hopefully not, and I certainly don’t intend to criticize anyone for the holding of any particular belief in and of itself, but I do think it potentially worthwhile to examine the basis of such beliefs or understanding from time to time.

Suggestion of being lost is not necessarily a suggestion about intelligence or gullibility. It’s possible that being lost could be a result of being misled. As I saw it written by someone I think is a great blogger in the comments of another blog,

“It is a myth that you have to be stupid or gullible in order to be misled.”

Now it’s time to read up, and see if some information relayed by ASU professor Dr. Jim Adams could potentially be a little on the misleading side.
 

ASU Professor Dr. Jim Adams on autism epidemiology

From March 2nd of this year: Article in Tucson Weekly

“There is no doubt that the number of reported cases has dramatically increased, from three or four cases per 10,000 20 years ago to one in 166. And according to the best data we have, which is from California, where there’s a fixed set of diagnostic criteria and a good reporting system, cases have increased every year that they’ve been counting–until last year, when there was a slight decrease,” says Adams.

“That could be a statistical blip, but coming after an unbroken string of increases, it may be significant.”

Were you able to spot the potential problems with this one?

Potential Problem 1 - Here’s a hint: What does the 1 in 166 refer to?
That’s right, autism AND autism spectrum disorders like Asperger’s and PDD-NOS.

Study Questions: Aside from likely increased awareness/recognition and diagnostic substitution, did the autism diagnoses of twenty years ago include Asperger’s and PDD-NOS? When did autism become recognized as a broad spectrum disorder? Is “three or four cases per 10,000 20 years ago to one in 166″ evidence of an epidemic, or comparing apples to oranges?

For a starting point and good blogging about this, I recommend the following articles from Autism - Natural Variation:

No Autism Epidemic: An Update
No Evidence of “Epidemic” from Prevalence Studies
No Autism Epidemic: Summary of the Numerical Evidence

And this one from Autism Diva: 

One point five? 

Potential Problem 2 - Here’s a hint: Was there really a slight decrease in cases last year?
That’s right, it depends on what you mean by “decrease” and ”cases”.

Study Question: What data does California’s DDS report, and was there really a decrease last year?

For a starting point and good blogging about this, I recommend the following articles from Autism - Natural Variation:

CDDS Data 101
One More Time on “New Cases” for David Kirby and Others
Open Letter to Journal of Physicians and Surgeons

From a recent telephone interview with Jim Adams conducted by another blogger (sometime after the recent Dateline piece that featured Jim Adams):

“The incidence in autism is best been measured by, um the state of California. And their latest statistics, as Rick Rollens has steadily reported, the latest statistics seem to show a decrease in the last uh year… and so …will that continue, or is it just a blip? We don’t know. But it’s been increasing steadily for 20 years and now to see a decrease right about the time we thought it would, is um certainly compelling.”

Hey that sounds kind of familiar. In fact it sounds similar to what Adams said back in March. But wait…didn’t California release two new full quarters of data since then? As a matter of fact they did.

Have you spotted the potential problem with this one yet?

Potential Problem 3 - Here’s a hint: Has Rick Rollens reported on CDDS data for Q2 of 2006?
That’s right, apparently he hasn’t. Adams’s statement looks to be based on old data, and a possible misinterpretation of the data at that. Note: It’s possible that the interview was conducted prior to California’s release of the 2006 Q2 data. If that’s the case, perhaps Dr. Adams can clarify that, and whether or not he maintains the same opinion following examination of the Q2 data, as he did when the interview took place.

Study Question: Please look at the following graph. This is a graph of the CDDS 3-5 year-old autism caseload cohort data for the past 4 years (16 consecutive quarters), and includes the data for Q1 and Q2 of 2006.

Source Data

What do you see? Any decreases? Any statistical “blips”?

It’s pretty easy to understand that if you’re looking for a change in administrative incidence due to a hypothesized effect of the removal of Thimerosal-containing vaccines, the clearest picture of any hypothesized effect is likely to be found in this age group first.

Apparently, even Mr. “Evidence Of Harm”, David Kirby himself, has acknowledged this in an e-mail to blogger Citizen Cain.

Slouching Toward Truth– Autism and Mercury

But in an e-mail, he did address the key point, and concede 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.”  He also conceded that total cases among 3-5 year olds, not changes in the rate of increase is the right measure.

For good blogging about why total cases are the right measure, see the following article, also from blogger Citizen Cain:

Questions for David Kirby– Dataphobes Beware

As to what might happen in the next couple of quarters in California is anybody’s guess. Remember, if you think “the mercury in vaccines dunnit,”  it’s got to be a drop in total caseload, not a decrease in caseload growth. Decreases in caseload growth (a leveling off) are a known eventuality.

Other recommended blogging on the subject:

CDDS Data, Flu Vaccines And Likely Predictions