Background: Psychosocial treatments are the mainstay of management of autism in the UK but there is a notable lack of a systematic evidence base for their effectiveness. Randomised controlled trial (RCT) studies in this area have been rare but are essential because of the developmental heterogeneity of the disorder. We aimed to test a new theoretically based social communication intervention targeting parental communication in a randomised design against routine care alone.
Methods: The intervention was given in addition to existing care and involved regular monthly therapist contact for 6 months with a further 6 months of 2-monthly consolidation sessions. It aimed to educate parents and train them in adapted communication tailored to their child's individual competencies. Twenty-eight children with autism were randomised between this treatment and routine care alone, stratified for age and baseline severity. Outcome was measured at 12 months from commencement of intervention, using standardised instruments.
Results: All cases studied met full Autism Diagnostic Interview (ADI) criteria for classical autism. Treatment and controls had similar routine care during the study period and there were no study dropouts after treatment had started. The active treatment group showed significant improvement compared with controls on the primary outcome measure--Autism Diagnostic Observation Schedule (ADOS) total score, particularly in reciprocal social interaction--and on secondary measures of expressive language, communicative initiation and parent-child interaction. Suggestive but non-significant results were found in Vineland Adaptive Behaviour Scales (Communication Sub-domain) and ADOS stereotyped and restricted behaviour domain.
Conclusions: A Randomised Treatment Trial design of this kind in classical autism is feasible and acceptable to patients. This pilot study suggests significant additional treatment benefits following a targeted (but relatively non-intensive) dyadic social communication treatment, when compared with routine care. The study needs replication on larger and independent samples. It should encourage further RCT designs in this area.