What is already known about this topic?

Autism spectrum disorders (ASD) occur in ∼1 in 63 children/adolescents1 and are characterised by impairments in cognition, communication and/or social responsiveness and interaction. A number of interventions have been developed and studied to target the social skills of children and adolescents with ASD.2 One particularly promising group intervention, namely the Social Skills Training Autism—Frankfurt (SOSTA-FRA),3 targets social motivation, social cognition, self-regulation and peer interaction using cognitive-behavioural treatment techniques provided using multiple methods (ie, social learning, computer-based, behavioural, cognitive methods) within a structured framework. Preliminary data have shown promising treatment effects and acceptability; further study in a controlled setting is necessary.

Methods of the study

The present study represented a six-centre multisite randomised controlled trial in Germany involving children with high functioning ASD between the ages of 8 and 20 years who had IQ above 70 and were diagnosed with either autistic disorder, Asperger's syndrome or atypical autism. The primary study goal was to examine the relative efficacy of SOSTA-FRA together with three parent training sessions (at the beginning, middle and end of treatment) relative to treatment as usual plus parent training. Two hundred and twenty-eight participants were randomised between May 2010 and February 2013 in a 1:1 fashion to either of the treatment arms. The primary outcome was the Social Responsiveness Scale which assesses parent-ratings of child social interaction and reciprocity. Secondary outcomes included the Social Responsiveness Scale Teacher Version and the child-rated Strength and Difficulties Questionnaire. Excellent quality assurance procedures were in place to ensure training of assessors and therapists both of whom were blinded, although treatment fidelity was only modest. Statistically significant but small effects were observed for the SOSTA-FRA treatment condition relative to the control arm at postintervention and 3-month follow-up. There was also a significant effect in favour of the treatment arm in the Strength and Difficulties Questionnaire between baseline and 3-month follow-up assessments but not between baseline and postintervention assays. Teacher ratings did not demonstrate separation between SOSTA-FRA and the control condition.

What does this paper add?

• The study suggests that a relatively brief, structured cognitive behavioural group intervention demonstrates efficacy in improving parent-rated social responsiveness among high-functioning children and adolescents with ASD (d=0.34–0.35).

• The methodological rigour (ie, randomised trial with active comparator), quality assurance procedures (ie, treatment and rater integrity efforts), and multiple investigational sites (n=6) provides evidence for the transportability of this intervention as well as the strength of results.

Limitations

• The sample was predominantly male children and adolescents with high-functioning ASD. Results may not generalise to females with ASD and/or those who are not in the high-functioning range.

• Blinded assessments by external raters of social interaction with peers were not included; the primary outcome was parent-rated social responsiveness, which may be less objective than clinician-ratings.

• Although treatment was manualised and structured, 12 group sessions may not be sufficient to effect significant change as evidenced by modest effect sizes on parent-rated social responsiveness (d=0.34–0.35) and non-significant effects on teacher outcomes (d=0.13–0.23).

What next in research?

In addition to addressing the above limitations (ie, examining the effects of treatment over a longer duration) and extending this research to other non-European countries, dissemination of SOSTA-FRA to frontline clinicians should be considered in future studies given the positive effects observed in the present trial.

Do these results change your practices and why?

Yes and no. These results suggest that SOSTA-FRA demonstrates only modest effects in improving social reciprocity and communication in children and adolescents with high-functioning ASD. However, given the lack of validated and acceptable interventions for this target domain, these data indicate the utility of a cognitive behavioural intervention for improving social functioning in this population.