How will DSM-5 affect autism diagnosis? A systematic literature review and meta-analysis.

The authors pooled every paper that compared DSM-IV and DSM-5 autism rules. They wanted one clear number: how many kids lose the diagnosis when you switch manuals.

The team looked at kids already labeled with any pervasive developmental disorder. They ran a meta-analysis to average the drop-off across studies.

DSM-5 wiped out about one in three autism diagnoses. The biggest hit landed on PDD-NOS; seven out of ten of those kids no longer qualified.

Kids with classic autistic disorder mostly kept the label. The stricter repetitive-behavior rule drove most of the drop.

Heald et al. (2020) updated the same question six years later. Their larger pool trimmed the loss to about one in five, so the real-world gap may be smaller than first feared.

Yaylaci et al. (2017) tracked 150 clinic files and saw a 19 % drop, lining up with the newer meta-analysis and showing the 31 % figure may be an early high bound.

Maddox et al. (2015) looked positive at first glance: DSM-5 fit the data better. Tighter rules improve accuracy, even though they disqualify more kids. The two papers don’t clash; one counts heads, the other counts validity.

Rispoli et al. (2011) and Greaves-Lord et al. (2013) explain why PDD-NOS crashes hardest: most of those kids lack repetitive behaviors, so they fail the new criterion set.

Expect service gaps. One-third may be the worst-case, but even a one-fifth drop means some kids lose funding. Re-screen every PDD-NOS case you carry. Document repetitive behaviors clearly. If the child doesn’t meet ASD under DSM-5, check Social Communication Disorder or other codes so treatment can continue.