Universal Developmental Surveillance for Autism Spectrum Disorders in Infants and Toddlers Using the Social Attention and Communication Study-Revised (SACS-R)

Background:  Despite increasing knowledge of the early markers of ASD, no screening tools have demonstrated sufficient psychometric properties for universal ASD screening in infants and toddlers. Barbaro and Dissanayake (2010) utilised developmental surveillance for the early identification of ASD within the Social Attention and Communication Study (SACS). The SACS involved training Maternal and Child Health (MCH) nurses to monitor children for ASD at their 12, 18 and 24 month health check-ups.  The SACS has been found to be the most accurate and sensitive method for the early identification of ASD to date, with 81% positive predictive value (PPV), and estimated sensitivity and specificity of 83.8% and 99.8%. However, not all children monitored (n = 20,770) were followed-up to identify any “missed” cases of ASD (false-negatives).

Objectives:  The objectives of the current study were: 1) to improve the psychometric properties of the SACS by using the SACS-Revised (SACS-R) in a low-risk, community-based sample, at children’s routine health consultations from 12-24 months; and 2) to follow-up all children monitored with the SACS-R at 3.5 years to identify any false-negatives and confirm diagnoses made at 24 months.

Methods:  The SACS-R contains brief checklists of key social-communication markers of ASD at 12, 18, and 24 months. 200 MCH nurses in Victoria, Australia, were trained on the SACS-R for use at their 12-24 month routine consultations. All children identified “at risk” for ASD on the SACS-R were referred for developmental assessments every 6 months until 2 years of age, and followed-up at 3.5 years to confirm diagnoses. Children were assessed with the ADOS-Toddler and Mullen Scales of Early Learning at each assessment, with the ADI-R administered at 24 months. All children monitored between 12-24 months were also followed-up with a SACS-preschool checklist at their 3.5 year MCH consultation, with all children “at risk” referred for assessment. Furthermore, all children who have a confirmed diagnosis of ASD in the community, but are “not at risk” on the SACS-R, will be referred for assessment.  

Results:  To date, 11,379 children have been monitored, with 200 children identified “at risk” for ASD thus far (1.7% referral rate). Of these, 139 children have been assessed, with 113 children meeting criteria for ASD (81.3% PPV), with the remaining children having developmental and/or language delays. 78% of 12-month-olds, 85% of 18-month-olds, and 80% of 24-month-olds assessed currently meet criteria, or are showing the early signs, for a diagnosis of ASD. There were no “true” false positives identified, as no child “at risk” on the SACS-R was typically developing.

Conclusions:  Developmental surveillance using the SACS-R has been shown to be the most effective method for the early identification of ASD to date. It is on par with the PPV of the original SACS, and is showing promising results in increasing sensitivity given the ASD prevalence found in the current sample (1%). These preliminary results confirm that developmental surveillance, not screening, is the most useful way of identifying infants and toddlers with ASD. Updated results to May 2015 will be presented at the conference.