Developmental Surveillance Versus Clinical Referral: A Longitudinal Study of Children with Autism Spectrum Disorder (ASD) Diagnosed at 2-Years Via Different Referral Systems

Background: Referral pathways to early ASD diagnosis have received little attention despite their potential influence on the baseline characteristics and developmental presentation. For example, children in the general community who do not receive developmental surveillance may only be identified and diagnosed early if they have more severe symptomology and developmental delays – a thesis consistent with Daniels and Mandell’s (2014) findings, via systematic review, that children with more severe ASD symptomology and cognitive impairments were likely to receive an earlier ASD diagnosis.

Objectives: The primary objective in this longitudinal study was to compare autism symptom severity and cognitive functioning of children who received an early ASD diagnosis via two different referral systems. This was achieved by comparing the baseline characteristics and developmental outcomes at preschool age of children with ASD who were diagnosed at 2-years of age, following two different referral pathways.

Methods: The participants comprised 53 children identified early through developmental surveillance at well-child visits (community group) and diagnosed with ASD (at M age = 25.96; range 23-33 months) and 24 children diagnosed (at M = 25.79, range 21-31 months) following referral to a specialized early assessment clinic by a primary healthcare professional (clinical group). All children were administered the Autism Diagnostic Observation Schedule and the Mullen Scales of Early Learning at 24- and 48-months of age.

Results: Children in the clinical group had parents that were more highly educated, and were more likely to have a relative and/or an older sibling with an ASD diagnosis than those in the community group. There were no differences between the groups in the intensity or type of early intervention received between 24- to 48-months.

A two-way repeated measures ANOVA revealed no significant differences between the community and clinical groups on ASD symptom severity at the 24-month assessment. However, children in the community group showed an improving trajectory in their symptomology such that they had significantly milder autism symptom severity (both social attention and communication deficits and restricted and repetitive behaviour) at 48-months compared with the clinical group, who showed an overall stable trajectory. A further two-way ANOVA revealed that the groups did not differ at either age in cognition, with both groups showing significant improvements in their overall developmental quotient (DQ), which was driven by the subdomain of verbal DQ, with non-verbal DQ being stable between 24 and 48-months.

Conclusions: The expectation that the clinical group may be more severe at baseline with lower developmental scores was not supported. However, the two groups did vary in their autism severity over time with the community group showing a reduction in autism symptoms at outcome despite no group differences in cognition. It is concluded that clinically-referred samples of children with ASD are not representative of community-based children with ASD, signalling that caution is needed when drawing conclusions from these different populations.