Evidence from previous small trials has suggested the effectiveness of early social communication interventions for autism.

The Preschool Autism Communication Trial (PACT) investigated the efficacy of such an intervention in the largest psychosocial autism trial to date.

To provide a stringent test of a pre-school communication intervention for autism.

152 children with core autism aged 2 years - 4 years 11 months in a 3 site 2 arm single (assessor) blinded randomised controlled trial of the parent-mediated communication-focused intervention added to treatment as usual (TAU) against TAU alone. Primary outcome; severity of autism symptoms (modified social communication algorithm from Autism Diagnostic Observation Schedule-Generic, ADOS-G). Secondary outcomes; blinded measures of parent-child interaction, child language, and adaptation in school.

At 13 month endpoint the treatment resulted in strong improvement in parental synchronous response to child (adjusted between-group effect size 1.22 (95% CI 0.85, 1.59) and child initiations with parent (ES 0.41 (0.08, 0.74) but small effect on autism symptomatology (ADOS-G, ES -0.24 (95% CI -0.59, 0.11) ns). Parents (not blind to allocation) reported strong treatment effects on child language and social adaptation but effects on blinded research assessed language and school adaptation were small.

Addition of the PACT intervention showed clear benefit in improving parent-child dyadic social communication but no substantive benefit over TAU in modifying objectively rated autism symptoms. This attenuation on generalisation from ‘proximal’ intervention effects to wider symptom change in other contexts remains a significant challenge for autism treatment and measurement methodology.