A Prospective Study of the Concordance of DSM-IV and DSM-5 Diagnostic Criteria for Autism Spectrum Disorder.

O et al. tracked the children who already had an autism diagnosis under the old DSM-IV rules. They asked: if we apply the new DSM-5 checklist, how many still qualify?

Two psychologists rescored every case blind. They noted which kids had classic autism, Asperger's, or PDD-NOS under the old system.

Almost every child with classic autism kept the label under DSM-5. Only a handful slipped through.

But half of the former Asperger's cases and two-thirds of the PDD-NOS group no longer met criteria. Most of these kids were older, had higher IQs, or showed milder social problems.

Zander et al. (2015) looked at toddlers and found the same squeeze. When they used a strict impairment cutoff, two-thirds of 2- to young learners lost the diagnosis. Together the studies show DSM-5 is tough on mild cases across age groups.

Giserman-Kiss et al. (2020) followed preschoolers and saw 88 % keep the label over time. Their sample was mostly classic autism, so the high stability matches O et al.'s finding that severe cases pass the new bar easily.

Meimei et al. (2022) reviewed telehealth screens. Their best tools hit 70–100 % sensitivity, but some dipped as low as 38 %. The same mild, high-IQ kids that DSM-5 misses are the ones brief screens also overlook.

Check the file for older clients, bright girls, or anyone with a previous Asperger's note. They may need extra data—ADOS-2 modules 3–4, SRS-2, or school observations—to qualify under DSM-5. Document social impairment in everyday settings so the child is not denied services simply because the new rules are tighter.