New criteria may reduce autism diagnoses
NEW YORK (Reuters Health) - The number of U.S. kids diagnosed with autism has been on the rise, but that trend could turn around with new diagnostic criteria coming into effect, researchers say.
By applying the new symptom checklist to 6,577 children who already met the old definitions for autism and related disorders, the study team found about 19 percent of the kids would not get an autism diagnosis today.
"Parents have no reason to be concerned about the findings of this study which is not about re-diagnosing people or taking away their diagnoses," Dr. Brian King told Reuters Health in an email.
King is director at the Seattle Children's Autism Center and director of child and adolescent psychiatry at the University of Washington and Seattle Children's Hospital. He was not involved in the new study, but he was involved in drafting the new diagnosis requirements.
The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) contains the symptom checklist for every psychiatric disorder recognized by the U.S. healthcare system.
Experts have suggested that it sets a higher threshold for diagnosing Autism Spectrum Disorder than the previous version, DSM-4-TR. Small studies attempting to test that theory have not been conclusive, however.
For the new report, the researchers looked at a larger national sample of kids, led by Matthew Maenner of the National Center on Birth Defects and Developmental Disabilities at the Centers for Disease Control and Prevention in Atlanta.
"This is a population-based study that's really looking at how children are being evaluated and what symptoms are being described in everyday community settings," Maenner told Reuters Health.
The new DSM-5 doesn't separate the so-called autism spectrum into separate conditions such as autistic disorder and Asperger disorder. The previous subtypes are simply classified as Autism Spectrum Disorder (ASD).
The new manual uses seven diagnostic criteria rather than 12 in the previous edition to define ASD. In addition, the new version takes historical behavior into consideration along with current behavior.
Maenner and colleagues applied the new criteria to 6,577 eight-year old children who were diagnosed with autism and related disorders from 2006 to 2008.
About 81 percent of those children would still meet the criteria for having ASD. This translates to a rate in the population of about 10 per 1,000 people having ASD instead of 11.3 per 1,000 as is currently estimated.
The results were published in JAMA Psychiatry.
Maenner said that children who had been given a special education classification of autism were more likely to meet the new criteria.
"Children that had an intellectual disability were slightly more likely to meet the new criteria than children without intellectual disability; and children with a history of developmental regression, which is either a loss or plateau of skills, were also more likely to meet the new criteria," Maenner said.
He noted that many of the kids who would not be diagnosed under the DSM-5 only missed by one criterion.
Parents may be concerned about how the new diagnostic changes might affect their kids, but Maenner says that from a parent's perspective, the evaluation for ASD isn't necessarily going to change, at least in terms of the signs that will be looked for in developmental screening.
"The things that would indicate early developmental delays are going to be the same, and you'd want to follow up if there are any concerns," Maenner said.
He acknowledged that looking at kids diagnosed by one set of rules in the past doesn't give a perfect picture of what will happen going forward.
"We looked backwards in time, from the most recent data we have, but the diagnostic practices could change, so the impact on service eligibility might actually be smaller than what we're reporting in terms of the impact on prevalence," Maenner said.
When physicians and health professionals adapt to the new criteria, Maenner added, they might be looking for the symptoms that would substantiate a diagnosis - they might pay more attention to symptoms that receive greater emphasis in the DSM-5.
King also said that looking backward at old records could be a problem when you're interested in new criteria.
For example, some of the kids who would not be diagnosed with ASD under the new DSM-5 might have met the new criteria, but the symptoms weren't recorded. After the threshold for diagnosis has been met, there's no reason for a clinician to "over-report" still more symptoms, he explained.
In addition, finding that the numbers are different with a second look (any second look) is to be expected to some degree, he said.
In that light, "finding that the boundaries around the overall spectrum are within 10 to 20 percent with this particular method could actually be viewed as pretty good," King said.
SOURCE: bit.ly/LLN3qy JAMA Psychiatry, online January 22, 2014.