Rate of Autism Diagnosis: A Case of Operational Definitions and Check Sheet Data Collection

A CDC report released on last week reveals that the rate of Autism has increased 23% from 2006-2009, or 1 in 88 children has a diagnosis within the spectrum of Autism. Most are quick to point out that the increase could be due to an increased screening, not necessarily an increase in prevalence of Autism or related disorders. But, at closer inspection, there might be more to the story.

The chart below shows the rate of increase over the last several years ¹:

By most accounts, the rate of increase is staggering, which has led many to question what is truly going on? Is Autism really on the rise? Or, is there something else happening?

Below are the hypotheses presented so far:

• The prevalence of Autism has actually and in truth increased.
• The definition for what constitutes Autism is murky ²:

“To qualify for the diagnosis, it’s a little bit murky,” Carey said. “It’s a pretty wide spectrum, it goes from very severely disabled kids to kids who are unusual, awkward, have social difficulties.”

Carey said he’s skeptical the report demonstrates a real increase in the rate of autism spectrum disorders. He cites financial support from the government as one of several reasons why families might push for their child to receive an autism diagnosis.

“Any time a very rare condition goes to very common, especially in these young kids, I think you have to be wary about what’s going on,” Carey said.

• Because of broader awareness, there is now more testing and screening of Autism. In other words, the prevalence hasn’t increased, just our discovery of the disorder.
• The definition of what constitutes Autism, while murky, has also changed, leading to an increase (but it is artificial). In other words, the Check Sheet and the Operational Definition used has changed mid-way in the data collection.

The last point above is important. If the diagnostic criteria for Autism has changed, then that materially can lead to a possible increase in the rate of Autism. In other words, if the criteria which was once narrower but is now broader, then children under the former criteria might not be diagnosed as Autistic, but under the new definition might be diagnosed as Autistic:

There is no brain-imaging test for autism, let alone a blood test or other rigorously objective diagnostic. Instead, physicians determine whether someone fits the criteria laid out in the American Psychiatric Association’s Diagnostic and Statistical Manual, or DSM.

The manual has undergone significant changes over the years, including in the diagnostic criteria for autism. In its current version, someone must fit at least eight of 16 criteria, including symptoms involving social interaction, communication, and repetitive or restricted behaviors and interests.

The previous version was stricter, describing one diagnostic criterion as “a pervasive lack of responsiveness to other people.” In the current manual, that became “a lack of spontaneous seeking to share …. achievements with other people” and friendships that appear less sophisticated than the norm for a child’s age.

The earlier manual also required “gross deficits in language development” and “peculiar speech patterns” for a diagnosis, while the current one lists difficulty “sustain(ing) a conversation” or “lack of varied . . . social imitative play.”

Morton Ann Gernsbacher, a professor of psychology and autism researcher at the University of Wisconsin, Madison, and others have cited these changes to question the reality of the reported autism increase.

All of this points to Operational Definitions and how important it is to set clear definitions. Indeed, we’ve seen this before – when the diagnostic criteria was changed for Diabetes Type 2, we saw a significant increase in the rate of Diabetes. But, it turns out, according to many Epidemiological studies, that the increase in rate for that time period was mainly due to the change in the diagnostic criteria ³.