Using the Social Attention and Communication Surveillance Revised (SACS-R) in a Community Based Setting.

Background:

Social Attention and Communication Surveillance (SACS) was designed to prospectively identify children developing Autism Spectrum Disorder (ASD) between 12- to 24-months of age. This tool has 81% positive predictive value and an estimated 83.8% sensitivity and 99.8% specificity (Barbaro & Dissanayake, 2010). Five key behavioural markers identified at 12-, 18-, and 24-months of age as most predictive of an ASD diagnosis (Barbaro & Dissanayake, 2012) were incorporated into a revision of the SACS (SACS-R), where a child is considered as having a high-likelihood of developing ASD if s/he is noted as ‘atypical’ on three out of five key behavioural markers.

Objectives:

The aim in this study was to examine the effectiveness of key SACS-R items in differentiating children with/without autism in a community based setting.

Methods:

The SACS-R was implemented by Maternal and Child Health (MCH) nurses across eight councils in Melbourne, Australia. Children identified as having a high-likelihood of developing ASD on SACS-R were assessed by clinical experts at La Trobe University using the ADOS, ADI-R, and Mullen Scales of Early Learning. Of the 273 children referred and assessed to date, 199 met criteria for ASD. Nurse completed SACS-R data was available for 96 children at 12-months (24 high-likelihood), 159 children at 18-months (76 high-likelihood), and 204 children at 24-months (115 high-likelihood). Fisher’s exact test was used to determine whether the key items at each age point were able to discriminate between children with/without ASD. Binary logistic regression analyses were used to determine whether scoring above the SACS-R cut-off predicted a diagnosis of ASD after controlling for gender and age.

Results:

No significant differences were found at 12-months of age between children with/without ASD on the frequency of atypical responses on the key items, and there was no significant association between scoring above the SACS-R cut-off and receiving a diagnosis of ASD. At 18-months of age, a significantly higher proportion of children with ASD than those without were rated as ‘atypical’ on all five of the key items (i.e. pointing, eye contact, waving, showing, and pretend play), and those scoring above cut-off were 3.5 times more likely to receive an ASD diagnosis. A higher proportion of children with ASD were given ‘atypical’ scores on four out of five of the key items (i.e. pointing, eye contact, waving, and pretend play) at 24 months of age, and those above cut-off were 3.4 times more likely to receive an ASD diagnosis.

Conclusions:

Previous clinical research has identified SACS as the most accurate and sensitive method for the early detection of ASD. The results from this study illustrate that when rated by community based clinicians, SACS-R does differentiate between children with/without ASD at the 18- and 24-month assessments. However, administration by nurses at 12-months did not significantly discriminate children with/without ASD. As these items were previously shown to differentiate between children with/without ASD when rated by a clinical expert (Barbaro & Dissanayake, 2013), additional education is required on the administration and accurate scoring of these items at 12-months for community-based professionals.