Application of the Brief Observation of Social Communication Change (BOSCC) to a Short Term Parent Mediated Intervention Trial

Background: The measurement of small yet meaningful social communication change in children with autism spectrum disorder (ASD) is critically important to the development of effective short-term interventions (McConachie et al., 2015). To that end, the Brief Observation of Social Communication Change was developed (BOSCC; Grzadzinski, Lord et al., 2016). The BOSCC is a measure of social communication change for use in children with little to no language coded from parent-child play interactions. It has been found to have high inter-rater, test-retest and convergence with other measures (Gradzinski et al., 2016; Kitzerow et al., 2015).

Objectives: To assess the sensitivity to change of the BOSCC over the course of a short term (10-week; 20 sessions) randomized intervention trial and three months later at follow-up. Additionally, to determine the effect of a number of moderating variables on BOSCC scores at entry.

Methods: Eighty-six parent child dyads were randomized to receive a ten-week parent mediated social communication intervention. A prior analysis found that the group receiving the parent-mediated JASPER intervention spent significantly more time spent jointly engaged and showed higher parental responsivity compared with the psychoeducation control group (Kasari et al., 2015; Shire, Gulsrud & Kasari, 2016). The primary outcome measure for this analysis was the BOSCC, scored by two reliable coders from a ten-minute unstructured parent-child play interaction. Additionally, data from the Autism Diagnostic Observation Schedule (ADOS), Early Social Communication Scales (ESCS), the Mullen Scales of Early Learning (MSEL), the Reynell Developmental Language Scales-Revised (Reynell) and Maternal Behavior Responsiveness Scales (MBRS) as well as demographic information were collected at each time point.

Results: Group differences on BOSCC TOTAL composite or domain scores were not found and so data was collapsed across groups. Change across time was assessed at the item, domain, and composite level using a linear mixed effects model. Overall, there was a significant decrease (signifying improvement) on BOSCC scores for the total composite score (d=.56) as well as both the social communication (SC; d=.56) and restricted and repetitive behaviors domains (RRB; d=.27). These decreases maintained at follow-up. Seven of ten SC domain items evidenced significant decreases from entry to exit (table 1). A median split was conducted for significantly correlated predictors and independent samples T-tests were run comparing the TOTAL composite scores of the median split groups. Significant differences on BOSCC scores at entry were found between the median-split groups for mental age (MA), IJA, composite language score, and parental responsivity (Table 2) with lower BOSCC scores associated with higher abilities.

Conclusions: Overall, the sensitivity of the BOSCC as a measure of social communication change was confirmed. Significant differences were found over the ten-week intervention period which were maintained at follow-up three months later. Additionally, it was found that a number of predictors such as MA, IJA, language, and caregiver responsiveness moderate BOSCC scores at entry. As caregiver responsiveness was found to be related to BOSCC scores, future directions include further exploration of the parental factors that influence the validity of the BOSCC as a measure of children’s social communication change.