Original source: Department for Education — Behavioural and social communication interventions for children and young people with SEMH and autism: a rapid evidence review (ref RR1624, July 2026). Research evidence, not statutory guidance.
This is the newest and most practically focused report in the DfE’s SEND rapid evidence series. Where the earlier reviews concentrated on identification and academic outcomes, this one asks a more pressing question for the classroom: which interventions actually improve behaviour and social communication for autistic pupils and those with SEMH needs? It synthesised 73 systematic reviews — 61 focused on autistic pupils, 12 on SEMH.
The headline: no silver bullet, but real patterns
The review’s conclusion is honest and useful: no single intervention works for everyone. Success depends on matching the intervention, its intensity and its delivery conditions to the individual pupil — plus adequate staff training and reliable outcome monitoring. That is a call to move away from chasing a named programme and towards thoughtful, well-implemented support.
The patterns that emerged are encouraging. Peer-mediated interventions improved social engagement and communication for many autistic pupils, especially in early years and primary settings. Structured behavioural supports — activity schedules, Check In Check Out, Daily Behaviour Report Cards — produced small to moderate improvements in classroom behaviour, particularly for pupils with ADHD or at risk of emotional and behavioural difficulty. Physical activity interventions were linked to reductions in challenging behaviour and better self-regulation, with stronger effects from more intensive, sustained programmes.
A broad family of social-communication approaches also showed positive but variable effects: naturalistic developmental behavioural interventions, play-based approaches, LEGO therapy, structured social skills programmes, video-based interventions and some technology-based tools. These helped social-emotional functioning, emotion recognition and on-task behaviour in the right circumstances.
The caveat that should shape your decisions
The review is unusually frank about what we do not know. Impact is highly contingent on implementation quality, dosage, practitioner training and contextual fit. Interventions vary enormously in intensity, and intensity appears to be related to outcome. Effectiveness is moderated by the child’s age, profile of need, and design features such as whether peers are trained and whether the programme is delivered with fidelity.
Most troubling for a school investing money: maintenance and generalisation are poorly assessed. Most studies measure short-term, observable behaviour, with limited evidence of sustained change, genuine relational inclusion, or broader academic and psychosocial outcomes. In plain terms — a programme may look good on paper and produce short-term results in a study, yet we cannot be confident those gains persist or transfer to your classroom.
The practical reading
For school leaders and specialist staff, three things follow. First, match, do not bolt on: choose interventions on the basis of a pupil’s actual profile and your staff’s capacity to deliver them well, not on the strength of a sales pitch. Second, invest in training and fidelity — the evidence suggests a well-delivered modest intervention beats a poorly delivered ambitious one. Third, build in your own monitoring, because the published evidence rarely tells you whether a programme’s effects last.
Treat the absence of a magic programme as freeing, not frustrating: the reliable ingredients — structure, peer support, physical activity, predictable routines — are things a good school can put in place this term.
The balance of the evidence also matters for resourcing decisions. Of the 73 reviews, 61 concerned autistic pupils and only 12 concerned SEMH, so leaders supporting SEMH-heavy cohorts are working with a thinner base than the headline count suggests. Video modelling illustrates the uncertainty: it showed effectiveness in specific domains, yet its maintenance and generalisation were so poorly assessed that the review could not say whether the gains persist beyond the intervention context.
Practical takeaways for schools
- Match intervention to the pupil’s profile and your staff’s capacity to deliver it well — not to a sales pitch.
- Invest in training and fidelity: a well-delivered modest intervention beats a poorly delivered ambitious one.
- Build in your own outcome monitoring, because published studies rarely establish lasting, generalisable effects.
- Start with the reliable, low-cost ingredients: structure, peer support, physical activity and predictable routines.
More from the DfE SEND evidence series
This is part of a series summarising the DfE’s rapid evidence reviews on identifying and supporting SEND in mainstream settings:
-
– Identifying and supporting SEMH needs in schools
– Identifying and supporting cognition and learning needs
– Identifying and supporting speech, language and communication needs
– Identifying and supporting sensory and physical needs
– Cross-cutting themes in supporting SEND
– Supporting stammering, speech and language in the early years
– Identifying and supporting autistic children and young people
To talk to an adviser about our training, contact us.