Referral Patterns and Early Signs of Possible ASD Among Children Referred for Generic Developmental Early Intervention

Background:  Early identification of autism spectrum disorders (ASD) is essential to providing intervention during a time when neural plasticity allows for optimal social learning. Few studies have looked at the early signs of ASD in high-risk community samples, such as children referred to generic developmental early intervention (EI) programs.

Objectives:  The objective of this study was to describe a developmentally high-risk, community-based cohort of children referred to EI with regard to their demographic characteristics, reasons for referral, source of referral, as well as the prevalence of early signs of ASD in this population. An additional aim was to assess the implications of these findings for EI intake processes and service delivery for children showing early signs of ASD.

Methods: A retrospective chart review was conducted using EI electronic records from York Region, an urban regional municipality in the Greater Toronto Area, Ontario, Canada. All referrals and intake interviews conducted between January 2014 and April 2014 were screened for inclusion. Children with conditions that would automatically generate a referral were excluded (those referred through prematurity pathways, ASD-sibling monitoring pathways and those with established diagnoses that generate referrals, such as Trisomy 21). Data were extracted from the referral form and from the intake interview conducted by York Region EI staff. Descriptive statistics were calculated for referral data and early signs of ASD. ANOVA was used to evaluate the differences in the referral ages based on the referral source.

Results:  During the four month period, there were 759 referrals to York Region EI, of which 81 were excluded due to being automatically generated referrals. The remaining 678 referrals were included in the analysis. The mean age at referral to EI was 27 months (95% confidence interval 12 to 47 months). The source of referral was available for 559 charts. Caregivers were the source of referral in the majority of cases (54%, n=302), compared with physicians (33%, n=185) and others (13%, n=72). There was a significant difference in the age at referral between the three sources, with physicians making the earliest referrals (F=3.5, p=0.015). Screening tools were noted in only 20% of referrals. Delayed domains of development were indicated in 126 cases. Speech and language delays were identified in 86% of these cases, and social/emotional delays (which may be indicative of ASD) were identified in 44%. In the intake interviews, the proportion of the sample showing specific early signs of ASD varied from 1-59%.

Conclusions:  Our results were disappointing in both the age at referral to EI and the reported use of developmental screening tools, given recent local efforts to implement routine eighteen month developmental screening. The number of children with potential early signs of ASD varied greatly, and further characterization of ASD risk in this population was limited by the lack of standardization in the intake process. The inclusion of validated, standardized developmental screening tools in the intake process, including screening tools for ASD, would provide better information on the early signs of ASD at a population level.