Study reveals inconsistent severity ratings in autism diagnosis among clinicians for toddlers

A new study highlights significant discrepancies among clinicians in rating autism severity levels in young children, raising concerns about the reliability of current diagnostic practices and their implications for treatment planning.

A new study suggests that clinicians do not always agree when assigning DSM-5-TR severity levels to toddlers with autism spectrum disorder, raising fresh questions about how reliably those labels capture a child’s needs.

Researchers looked at 49 children aged 16 to 31 months, with four clinicians reviewing both telehealth and in-person assessments. According to the study, agreement on social communication and interaction severity was only fair, with a kappa score of 0.318, while agreement on restricted and repetitive behaviours was slight at 0.195. The findings point to wide variation in how severity is judged, even when clinicians are using the same diagnostic framework.

That matters because severity ratings are often used to shape communication with families, guide treatment planning and support access to services. The study adds to a broader debate in autism research about whether severity labels reflect core symptoms alone or are also influenced by language level, adaptive skills and other developmental factors. Previous research and clinical guidance have already questioned how well DSM-5 support levels map on to day-to-day functioning.

The Developmental Profile 4, a separate assessment tool often used in developmental evaluations, measures functioning across physical, adaptive behaviour, social-emotional, cognitive and communication domains. Its broad scope reflects the same reality highlighted by the study: autism assessment in very young children rarely rests on one measure alone. For parents and caregivers, the practical implication is that a single severity label may not tell the full story, and clinicians may need clearer guidance to make those ratings more consistent.