Pathways to Autism Spectrum Disorder Diagnosis through a Multi-Stage Screening Process

Background: For Autism Spectrum Disorder (ASD), evidence suggests that early screening improves detection rates. Our team implemented a multi-stage screening protocol for ASD in Early Intervention settings based on the “clinical reasoning model.” According to this protocol, referral decisions were based on screening results and the expressed clinical concern from parents or Early Interventionists. That is, children were promoted to successive stages if screening results were positive or clinical concern was reported subsequent to screening.

Objectives:  To assess the utility of a posterioriconcern-based referrals for ASD within Early Intervention settings and conduct a basic cost/benefit analysis of this multi-stage screening process.

Methods:  1022 children between the ages of 14-33 months participated in a case finding process for ASD involving stage 1: parent-report questionnaires, stage 2: structured observations, and stage 3: full diagnostic assessments. Quantitative methods were used to assess the effectiveness of different diagnostic pathways. Qualitative interviews of providers (N=20) were conducted to understand factors that influence their decision making processes.

Results: Among children with positive screening results, parents and/or Early Interventionists were concerned 64% of the time. The most efficient pathway to diagnosis was a positive screening score at stage 1 along with reported concern [5.6 assessment hours/diagnosis]. The least efficient pathways were those in which there were positive screening scores but no reported concerns [11 assessment hours/diagnosis]. Notably, pathways in which the stage 1 screening was negative but there was at least some reported concern fell in the middle [7.5 assessment hours/diagnosis]. Qualitative results suggest EI interventionists employed dynamic and distinct strategies to move families through the three-stage screening process when clinical concerns, whether the parents or their own, are present.

Conclusions:  Results demonstrate the value of allowing referrals based on clinical concern subsequent to screening. Although, it is unclear the degree to which screening results influenced these concerns (concerns were coded after stage 1 screening results and are therefore not independent). Results also highlight that Early Interventionists find value in screening that go beyond the psychometrics of the binary result to include instruments' ability to enhance conversations and to collect clinically meaningful information about observed behavior within natural settings. These data show how important screening is and how crucial it is that we do not solely rely on parent concern in drawing conclusions about the value of ASD screening. If screening does not occur until parents are concerned, opportunities will be missed to diagnose children and therefore limit their chance to receive high intensity services at a younger age.