Blog post Part of special issue: Revisiting the Children’s Plan: Towards a new manifesto for tackling early years inequality

Special educational needs policy requires research infrastructure

Ruth Gilbert, Professor of Clinical Epidemiology at UCL Kate Lewis, Research Fellow at UCL Will Farr, Academic researcher at University of Cambridge 24 Jan 2025

Over the past 25 years, there have been successive policy changes to funding, categorisation and implementation of special educational needs (SEN) provision in England. Yet there is limited evidence regarding who actually benefits from SEN provision, and in what way. In this blog post, we summarise lessons from a mixed-methods study funded by the National Institute for Health and Care Research, called Health Outcomes for young People throughout Education (HOPE), which evaluated the impact of SEN provision in primary schools on health and education outcomes.

The context

Parents, children and SEN practitioners who were interviewed and surveyed in the HOPE study reported benefits of SEN provision on participation in school and the mental health of pupils. However, many found that the processes of identification, assessment and access to appropriate provision were distressing, provision was delayed and sometimes fell short of expectations. Further problems with SEN policy are reflected in a recent National Audit Office (NAO) report that pointed to unsustainable rises in funding for SEN; currently over £10 billion per year in England and rising (NAO, 2024). The NAO report recommends that the Department for Education ‘explicitly consider whole-system reform, to improve outcomes for children with SEN and put SEN provision on a financially sustainable footing’. The report also calls for more integration with the health system.

The need for evidence on the effectiveness of SEN provision

The Department for Education has led the way in making high-quality, longitudinal administrative data available for research over the past two decades and provided substantial investment in randomised trials in schools (EEF, 2024). Yet few trials have evaluated the effectiveness of SEN provision in schools.

‘In many comparisons, SEN provision was associated with worse outcomes, which is likely to be due to unmeasured factors, such as children with more severe problems being more likely to be assigned SEN provision.’

The HOPE study assessed the effectiveness of SEN provision on hospital contacts, school absence and attainment by applying appropriate methods for observational studies to the linked ECHILD database of hospital and education records for 14 million children in England. The study compared children with similar health conditions, such as congenital anomalies and neurodisabilities, as proxy measures for groups of children likely to have additional learning needs. Groups of pupils allocated to SEN provision, or not, were made comparable by controlling for measured background characteristics. In some comparisons, SEN provision was found to reduce unauthorised school absence. However, in many comparisons, SEN provision was associated with worse outcomes, which is likely to be due to unmeasured factors, such as children with more severe problems being more likely to be assigned SEN provision. This problem is called ‘confounding’ and could be addressed by plugging key gaps in the ECHILD database, by using experimental study designs and by more systematic and well-documented implementation of policy changes.

What should the Labour government do?

Four ways to develop data and evidence to drive SEN policy change

The government should consider implementing national SEN policy change in a systematic and planned way that maximises opportunities for natural experiments. For example, random, staggered implementation of a change by area, with regular collection of data from all areas to allow comparisons across time.
Research infrastructure is needed to support trials that randomise children, schools or local authorities wherever possible to different SEN interventions or to no intervention. Multiple trials will be needed to find out which types of SEN provision work for whom, when and for how long. For example, as discussed in the NAO report, the effectiveness of preschool SEN provision compared with standard nursery care, and intensive SEN provision in mainstream compared with special schools, should be priorities for trials.
Plug key information gaps in the ECHILD database to improve the validity of studies on effectiveness and equity of SEN provision. First, information on additional needs should be recorded at first identification, focusing on functional needs, not diagnoses. Second, the type of SEN provision actually received should be recorded with information on intensity. For example, is speech and language therapy provided one-to-one or in a group, as a single appointment or with multiple sessions?
More meaningful outcomes need to be measured. The HOPE study measured absence, school attainment and hospital contacts, but these are blunt measures of learning, health and behaviours that SEN provision might affect. Research needs to involve practitioners, parents and children in developing SEN provision and outcome measures relevant to children’s potential, their daily living skills that enable participation in education, and their social and emotional skills, behaviour and mental health.

Change in SEN policy is urgently needed but the Labour government should make sure that change happens within a research and data infrastructure that builds on and continues to generate evidence of what works.

References

Education Endowment Foundation [EEF]. (2024). Evidence and resources. https://educationendowmentfoundation.org.uk/education-evidence

National Audit Office [NAO]. (2024). Support for children and young people with special educational needs. https://www.nao.org.uk/reports/support-for-children-and-young-people-with-special-educational-needs/

Ruth Gilbert

Professor of Clinical Epidemiology at UCL

Ruth Gilbert is a clinical epidemiologist trained in paediatrics. Much of her research uses de-identified, administrative data to address clinical and policy questions related to the health of children and families. Her research uses record linkage to consider health, social and educational trajectories of children from birth to adulthood and to incorporate information from parents and siblings. Ruth co-directs the NIHR (National Institute for Health and Care Research) Children and Families Policy Research which conducts research to inform the Department of Health and Social Care and related government bodies. She also co-directs the Health Data Research UK driver programme on Social and Environmental Determinants of Health. Ruth helped to establish the ECHILD database and leads the NIHR-funded HOPE programme which uses ECHILD to evaluate the impact of special educational needs provision on health outcomes. Most recently, she established a novel database linking de-identified...

Kate Lewis

Research Fellow at UCL

Kate Lewis is a Research Fellow at the UCL Great Ormond Street Institute of Child Health, where her research leverages linked administrative education and health data to study the social determinants of early child health. As part of the HOPE study, Kate is examining variation in, and outcomes of, special education needs provision using biostatistical descriptive and causal methods. She has also contributed to publications on topics such as inequalities in childhood respiratory illness, the association between parental migration history and childhood hospital admissions, and trends in emergency paediatric critical care.

Will Farr

Academic researcher at University of Cambridge

Will Farr is an academic researcher at Cambridge University, Faculty of Education, conducting studies to assess and improve the diagnostic pathways for children with possible autism, digital innovation for diagnosis, and use of AI for individualised therapy. Previously, he was a primary school teacher for 20 years, where he held roles relevant to special educational needs.