A large-scale clinical trial is testing whether a therapy that helps autistic children communicate in a clinic can be extended to work at home and school. The core problem is that autistic children often learn new skills in one setting but fail to use them elsewhere—a difficulty called poor generalisation. No previous autism intervention has improved overall symptom severity across different contexts. This trial, called PACT-G, builds on an earlier clinic-based treatment that improved parent-child communication, adding targeted strategies to help those gains transfer into everyday life. The researchers will randomise 244 children aged 2–11 into either the enhanced therapy plus usual care, or usual care alone, measuring autism symptoms, language, and real-world functioning at home and school. If successful, the intervention could give families and teachers practical tools to support autistic children’s communication across all parts of their lives, rather than only in a therapist’s office. The study also includes a mechanism analysis to understand *why* generalisation fails, which could guide future therapy design.
View original technical description
Background: The evidence base for early intervention in autism shows that behaviours proximal to the intervention delivered (e.g. dyadic interaction measures) are amenable to change, but it has been difficult to generalise treatment gains successfully acquired from one context into another, and no studies to date have demonstrated improvement in general autism symptom severity (difficulty with the generalisation of acquired skills is a key problem in autism). Aim: This proposal tests an intervention designed systematically to promote generalisation of previously demonstrated clinic-assessed treatment gains into home and school contexts. It includes a detailed mediation analysis building on our previous work and a mechanism study that will enable for the first time a detailed approach to understanding the difficulties of generalising down a known mediation pathway into other contexts. Hypotheses: 1) The intervention will show the added efficacy and cost-effectiveness of preschool and school-age autism outcomes in home, school and research settings compared to treatment as usual; 2) There will be an increase in the generalisation of acquired communication across contexts and persons, shown by mediation and the mechanistic study. Design: Three site two parallel group randomised controlled trial of the experimental treatment plus treatment as usual (TAU) versus TAU alone. Initial pilot first stage with pre-specified progression criteria. Population: Children 2-11 years meeting criteria for core autism on gold-standard measures. Interventions: The experimental intervention builds on our clinic-based MRC Preschool Autism Communication Treatment model (PACT), delivered with the primary caregiver using methods that gave maximal intervention effect on child social communication, plus an additional series of targeted theory and evidence based strategies designed to enhance the generalisation of this effect into naturalistic home and education contexts (details below). The control intervention will be treatment as usual. Primary outcome: Autism symptom outcome, researcher assessed in standardised setting. Secondary outcomes: Autism symptoms, child interaction with parent or teacher, language and reported functional outcomes in home and school settings. Outcomes measured at baseline and 11 month endpoint in all settings with interim interaction measurements (6 months and 8 months) to test mediation. Sample: 244 (122 intervention/122 TAU; 82/site). Primary Analysis will test for between-group change in primary outcome using analysis of covariance plus planned subgroup analysis by age-group stratifier and test of moderation. Mechanism analysis will use regression models to test for mediation of parent-child interaction on primary outcome. Duration: 42 months, with 6 month start-up, 6 months pilot stage with progression criteria, 24 months for the main trial, and 6 months analysis and write up. 244 cases will be recruited over months 9-26 (4.8/month/site). This is very feasible given that our previous PACT trial with the same inclusion criteria recruited 2.6/month, and recruitment here will be from three times the population pool due to expanded age criteria (<15% of available cases during this period from population estimates).
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know