Short report: Autism diagnostic impressions in young children formed by primary care clinicians and through telemedicine expert assessments.

Abstract

Formal autism diagnosis is often critical for children to access early, autism-specific services and supports. However, barriers to traditional in-person evaluations, including long waitlists, delay diagnosis. The goal of the current study was to compare diagnostic impressions (i.e. clinical judgments) made by primary care clinicians and autism experts conducting brief telehealth sessions, with expert diagnosis from in-person gold-standard evaluations. Participants were toddlers (n = 32, age 12-36 months) referred for any developmental concerns by four primary care clinicians from one pediatric practice in the United States. Primary care clinicians indicated their diagnostic classification and families then completed telehealth evaluations and in-person evaluations with one of five autism diagnostic expert clinicians. When primary care clinicians classified a child as having definite autism (n = 11), they were 100% accurate, but only 57% accurate when they indicated a child definitely did not have autism. Experts providing classification after a telehealth evaluation accurately classified 72% of children and were confident in the diagnosis for 55% of cases. In high-confidence cases, telehealth diagnosis matched final diagnosis 88% of the time. These findings indicate that when primary care clinicians believe a toddler is autistic, or when autism experts indicate autism telehealth classification with confidence, the child should begin receiving autism-specific services and supports right away.Lay abstractThere are long waitlists for autism evaluations, which greatly delay the start of interventions that are known to improve children's outcomes. We tested the accuracy of primary care clinicians' impressions of autism versus other developmental delays during well-child visits, and of experts during brief telemedicine visits, and found that more than half of the children were accurately identified through these streamlined methods. These findings support a tiered approach in which children identified through these more efficient methods begin autism intervention immediately; this approach also benefits children with more complex differentials by shortening waitlists for comprehensive evaluations for those who require them prior to treatment entry.