Universal Developmental Surveillance for Autism in Infants, Toddlers and Preschoolers: The Social Attention and Communication Study-Revised (SACS-R) and SACS-Preschool

Oral Presentation
Saturday, May 12, 2018: 2:40 PM
Arcadis Zaal (de Doelen ICC Rotterdam)
J. Barbaro, C. Dissanayake and N. Sadka, Olga Tennison Autism Research Centre, La Trobe University, Melbourne, Australia
Background: Barbaro and Dissanayake (2010) utilised developmental surveillance for the early identification of autism within the Social Attention and Communication Study (SACS), where Maternal and Child Health (MCH) nurses were trained to monitor children for autism during routine, community-based health-checks between 12-24 months. The SACS had a positive predictive value (PPV) of 81%, and estimated sensitivity and specificity of 83.8% and 99.8%, respectively. However, all 20,770 children monitored were not followed-up at a later age to identify any false-negatives.

Objectives: The objectives of the current study were to: 1) improve the SACS’ psychometric properties through use of the SACS-Revised (SACS-R) at children’s routine, community-based health-checks from 12-24 months; 2) follow-up all children monitored with the SACS-R at 3.5 years to confirm diagnoses and identify false-negatives; and 3) investigate the psychometric properties of the newly added SACS-Preschool (SACS-PR) check at 3.5 years.

Methods: The SACS-R and SACS-PR contain brief checklists of key social-communication markers of autism at 12-, 18-, 24-, and 42-months-of-age, with 200 MCH nurses trained to utilise these checklists during their routine consultations. A total of 13,838 children were monitored on the SACS-R between 12- to 24-months, with all children identified at ‘high likelihood’ for autism referred for a developmental assessment with the SACS team. Children were assessed by the team every 6 months until 2 years, with gold-standard assessment tools (ADOS-Toddler, Mullen Scales of Early Learning, ADI-R). All children returned at 42 months to confirm diagnoses.

All children monitored by the MCH nurses between 12-24 months were also monitored with the SACS-PR at their 42-month consultation, with all children identified at ‘high likelihood’ for autism referred to the SACS team for an assessment. Furthermore, children identified as ‘low likelihood’ who had any developmental concerns, including autism, or had a community-based diagnosis of autism, were also referred for assessment with the SACS team to determine/confirm diagnoses.

Results: The nurses identified 323 children at ‘high likelihood’ for autism (2.3% referral rate) between 12- to 24-months using the SACS-R. Of these, 234 children have been assessed, with 192 children meeting criteria for ASD (82.4% PPV); the remaining children had developmental and/or language delays. Thus, no “true” false positives (typically developing children) were identified using the SACS-R. 72% of 12-month-olds, 82% of 18-month-olds, and 85% of 24-month-olds met criteria for autism, confirmed at 42-months-of-age. Sensitivity and specificity of the SACS-R was 72.3% and 99.6%, respectively, and Negative Predictive Value (NPV) was 97.7%. Nurses are currently monitoring children at their 42-month-check with the SACS-PR; thus, psychometric properties of the SACS-PR will be presented at the conference. Prevalence rates of autism found in the sample so far is 2.7%.

Conclusions: Developmental surveillance using the SACS-R continues to be the most effective method for the early identification of autism available. Its PPV is higher than the original SACS, with excellent NPV, specificity and sensitivity. These data confirm that developmental surveillance using the SACS-R is the most effective way to identify infants and toddlers with autism, with follow-ups using the SACS-PR useful in identifying ‘false negatives’.