Research assessing the effectiveness of treatments in autism spectrum disorder (ASD) is limited by a lack of scalable and quantifiable autism-specific treatment response measures. Assessing treatment success would be benefited by the development of response measures that can be administered ‘blindly’ and are sensitive enough to capture change over short periods of time, flexible enough to be used across studies and are standardized in order to be comparable across sites.

Catherine Lord and her colleagues will test the proof of concept of the Brief Observation of Social Communication – Change (BOSCC), a new 12-minute observation assessment of social communication behaviors in children with autism. The BOSCC is designed to be sensitive to changes in social communication behavior, easily used by naïve, minimally trained examiners and coded relatively quickly by non-expert raters. In addition, the BOSCC will incorporate automatic methods of acoustic signal processing to further quantify changes in social communication.

To assess the utility and accessibility of BOSCC, Lord and her team will recruit 120 children with autism currently participating in short-term, evidence-based social communication treatments at four university-based autism centers in the New York metropolitan area:

Weill Cornell Medical College, New York University Langone Medical Center, Icahn School of Medicine at Mount Sinai and Albert Einstein College of Medicine. An additional 80 children from the same centers, but who are not participating in a specific treatment, will also be recruited. Both groups will receive a BOSCC, Autism Diagnostic Observation Schedule (ADOS) and parent-report measures of behavior and a clinician-rated global impression at the start and completion of treatment or an equivalent period.

Signal processing and machine-learning techniques will be used to develop automatic measures of speech and language characteristics that may yield more sensitive or supplemental measures to behavior coding. BOSCC codes of the brief observation will be compared with alternative measures of change, including BOSCC codes of ADOS videos, as well as the automated acoustic measures. By testing the feasibility of the new BOSCC scale at four different sites and on ASD individuals undergoing different treatments, the team will establish the validity, accessibility, psychometric properties and sensitivity to change of BOSCC for providing scalable response measures of behavioral change in 200 children from early school age to mid-adolescence.