Exclusion rates in Wales are rising. Behind every statistic is a child whose educational journey has been interrupted — and a school, family, and community left to pick up the pieces. The financial cost is staggering. The human cost is incalculable.

The Numbers

Research consistently estimates the long-term cost of a single permanent exclusion at around £370,000 when accounting for lost earnings, increased benefit dependency, health costs, and the criminal justice system. In Welsh schools, fixed-term exclusions have been rising year on year, with SEMH (Social, Emotional, and Mental Health) needs cited as the most common reason.

Yet the vast majority of exclusions are not inevitable. They are the endpoint of a journey — a series of unmet needs, missed warning signs, and interventions that either weren't available, weren't right, or weren't given enough time to work.

Why Early Intervention Matters

The evidence for early intervention is overwhelming. Children who receive targeted, appropriate support at the first sign of difficulty are significantly less likely to reach the point of exclusion. The challenge is not the willingness of schools — it is the availability of the right information at the right time.

Most schools currently choose interventions based on what they know, what they can afford, and what happens to be available in their area. Very few have access to outcome data that tells them which interventions are most likely to work for a child with a specific profile of need. That information gap is enormously costly.

The Inclusion Imperative

Inclusion is not just a moral imperative — it is an economic one. A child who remains in mainstream education, properly supported, is more likely to achieve qualifications, gain employment, and contribute to their community. The investment required to provide that support is a fraction of the cost of exclusion and its consequences.

Wales has made inclusion a cornerstone of its education policy. The Additional Learning Needs and Education Tribunal (Wales) Act 2018 strengthened the rights of learners with additional needs. But policy without data-driven delivery is a promise without a plan.

What Needs to Change

Three things would make the biggest difference to exclusion rates in Welsh schools:

1. Better data on what works. Schools need access to intervention outcome data that goes beyond their own limited sample. A network approach — where outcomes from hundreds of interventions across dozens of schools are aggregated and analysed — makes it possible to recommend what is most likely to work for a specific child.

2. Earlier identification. Wellbeing indicators, attendance patterns, and behaviour data often signal emerging need weeks or months before a crisis. Systems that surface these signals early give schools time to act.

3. Joined-up safeguarding and wellbeing. Child protection and child wellbeing are often managed in separate systems by separate people. Integration of these two functions gives a complete picture of a child's situation and supports more coordinated responses.

Ask a school how they choose which interventions to commission, and the honest answer is usually some combination of: "it worked before", "we know the provider", or "it was the only thing available." Almost never is the answer: "the data told us it was most likely to work for this child."

The Problem with How We Currently Choose Interventions

In most other areas of professional practice, evidence-based decision making is the standard. Medicine, engineering, finance — all have developed systematic approaches to evaluating what works and applying those lessons to future decisions. Education has made significant strides with evidence-based teaching practice. But intervention management remains largely intuitive, relationship-driven, and anecdotal.

This is not a criticism of the schools and DSLs who work within this system. They are doing their best with the tools and information available to them. The problem is structural — no infrastructure exists for sharing and analysing intervention outcomes at scale.

What Evidence-Based Intervention Management Looks Like

An evidence-based approach to intervention management does not mean rigid protocols or replacing professional judgement. It means enriching that judgement with better information. Specifically, it means:

Knowing what has worked before — not just in your school, but across a network of schools facing similar challenges with similar children.

Understanding why it worked — which dimensions of need the intervention addressed most effectively, and for which profiles of learner.

Measuring whether it is working now — systematically tracking engagement, progress, and outcomes throughout the intervention, not just at the end.

Feeding that information back into future decisions — so that every intervention makes the next recommendation slightly better.

The Network Effect

The power of this approach grows with scale. A single school running ten interventions per year has limited outcome data to learn from. A network of 85 schools running thousands of interventions generates the kind of dataset that makes genuinely predictive recommendations possible.

This is the logic behind ODISSYS. Every school that joins the network contributes to a growing body of evidence. Every outcome recorded — whether an intervention succeeded or fell short — makes the recommendations to every other school in the network more accurate.

The Practical Difference

Consider two scenarios. In the first, a DSL identifies a Year 8 pupil showing signs of anxiety and social withdrawal. They commission the intervention they have used before, from a provider they have a relationship with. It may or may not be the best fit.

In the second scenario, the same DSL enters the pupil's profile into ODISSYS. The system draws on outcome data from hundreds of similar referrals across Wales and surfaces the interventions with the highest statistical likelihood of a positive outcome for this specific profile of need. The DSL still makes the decision. But they make it with significantly better information.

The difference between these two scenarios is not just better outcomes for one child. It is a fundamentally different — and more professional — approach to one of the most important decisions a school can make.

The role of the Designated Safeguarding Lead has never been more demanding. Increasing caseloads, evolving legislation, growing awareness of mental health and wellbeing needs, and the ever-present pressure of inspection readiness — DSLs are carrying more than ever before. Their technology should be helping, not adding to the burden.

The Gap Between What DSLs Need and What They Have

Most safeguarding platforms were designed a decade ago, built around compliance and record-keeping rather than active case management and analytical insight. They are digital filing cabinets — better than paper, but not by as much as they should be.

The DSLs we speak to consistently identify the same frustrations: fragmented information across multiple systems, time-consuming disclosure recording processes, limited ability to spot patterns or trends across the school, and weak integration between safeguarding and the broader wellbeing picture.

Seven Things a Modern Safeguarding Platform Must Do

1. Make disclosure recording fast and standardised. Every minute a DSL spends wrestling with a form is a minute not spent supporting a child. Recording a concern should take seconds, not minutes, and the process should guide the user through the correct steps automatically.

2. Surface patterns, not just records. Individual disclosures recorded in isolation tell you what happened. Patterns of disclosures — across a pupil, a year group, a family, or a time period — tell you what is happening. A modern platform should make those patterns visible without requiring manual analysis.

3. Connect safeguarding and wellbeing data. A child whose safeguarding record shows a disclosure and whose wellbeing data shows declining attendance, increasing behaviour incidents, and withdrawal from activities is showing a coherent picture of distress. Seeing that picture requires the data to be in one place.

4. Ensure the right people are notified immediately. When a concern is raised, the right colleagues need to know immediately. Automated notifications, clear escalation pathways, and audit trails of who knew what and when are not optional — they are fundamental to safe practice.

5. Support inspection readiness without additional workload. Estyn and Ofsted inspectors scrutinise safeguarding records carefully. A good platform should make it straightforward to demonstrate that concerns were recorded properly, responded to appropriately, and followed up consistently — without DSLs having to compile evidence at short notice.

6. Protect sensitive information appropriately. Not everyone who uses a safeguarding platform should see everything in it. Role-based access controls, audit logging, and data security are non-negotiable when the information relates to vulnerable children.

7. Integrate with intervention management. Safeguarding and wellbeing are not separate concerns — they are two lenses on the same child. A DSL who identifies a safeguarding concern should be able to seamlessly commission an intervention, track its progress, and see its outcome, all within a single system.

Where Most Platforms Fall Short

Of the seven requirements above, most existing safeguarding platforms deliver on perhaps three or four. They record disclosures. They maintain audit trails. Some send notifications. But the integration of safeguarding and wellbeing, the surfacing of patterns, and the connection to intervention management remain largely absent.

This is the gap that ODISSYS was designed to fill. Built by practitioners who experienced these frustrations firsthand, the platform treats safeguarding and intervention management as two parts of the same system — because in the life of a vulnerable child, they are.

Social, Emotional and Mental Health — SEMH — is one of the most significant challenges facing schools in Wales and across the UK today. Understanding what it means, how to identify it, and what to do about it is no longer optional for school leaders. It is essential.

What Does SEMH Mean?

SEMH is a broad term used to describe a wide range of difficulties that affect how children and young people manage their emotions, develop relationships, and engage with learning. It is one of four categories of Special Educational Needs and Disabilities (SEND) defined under the SEND Code of Practice.

Children with SEMH needs may experience conditions including anxiety, depression, attachment disorders, conduct disorders, attention deficit hyperactivity disorder (ADHD), and the effects of trauma or adverse childhood experiences (ACEs). What these conditions share is that they manifest in behaviour — in how a child interacts with their peers, their teachers, and the school environment.

Crucially, SEMH is not about a child being "badly behaved." It is about a child whose internal emotional world is making the external world of school enormously difficult to navigate.

Why is SEMH Increasing?

The prevalence of SEMH needs in schools has risen significantly over recent years. The NHS Mental Health of Children and Young People survey found that one in five children aged 8 to 25 had a probable mental disorder — up from one in nine in 2017. The causes are complex and interconnected: the legacy of the pandemic, social media, rising economic pressures on families, and reduced access to specialist services like CAMHS all play a role.

Waiting times for CAMHS referrals in Wales can exceed 18 months. In that gap, children are in school — and schools are on the front line.

How Does SEMH Affect Schools?

The impact of unmet SEMH needs ripples through the entire school community. For the individual child, it can mean chronic absence, exclusion, academic underachievement, and a trajectory that leads away from education and employment. For the school, it means staff time, behaviour incidents, SEND resource allocation, and — at its most serious — safeguarding concerns.

Exclusion is the most visible endpoint of unmet SEMH need. In Wales, SEMH is consistently cited as the most common reason for both fixed-term and permanent exclusions. The long-term cost of a single permanent exclusion — to the child, their family, and society — is estimated at £370,000.

But exclusion is not inevitable. It is the outcome of a journey — and at every stage of that journey, there are opportunities to intervene.

What Can Schools Do?

Early identification is everything. The earlier a child's SEMH needs are recognised, the wider the range of effective interventions available and the better the likely outcome. Schools that invest in regular wellbeing monitoring — using validated tools like the SDQ (Strengths and Difficulties Questionnaire) or SCWBS (Scottish Children's Wellbeing Survey) — are better placed to spot emerging need before it reaches crisis point.

Whole-school approaches work. SEMH is not a problem that can be solved by one intervention for one child. Schools that create environments of psychological safety, train all staff in trauma-informed practice, and embed wellbeing into their culture consistently see better outcomes than those that treat SEMH as an add-on.

Evidence-based intervention selection matters. Not all interventions are equally effective for all children. A child whose SEMH needs stem from attachment difficulties requires a different approach from one whose difficulties are rooted in anxiety. Schools that can match individual pupils to interventions based on outcome data — rather than availability or familiarity — achieve significantly better results.

Safeguarding and wellbeing must be connected. Children with significant SEMH needs are frequently also children with safeguarding concerns. Managing these two dimensions in separate systems — or worse, by separate people with no shared information — increases risk and reduces the effectiveness of support.

The Role of Data

One of the most significant shifts in SEMH practice over recent years has been the move towards data-driven approaches. Schools that systematically track intervention outcomes — recording not just what they did but whether it worked, and for which children — are building an evidence base that improves every future decision they make.

This is the principle at the heart of ODISSYS. By aggregating outcome data across a network of schools and providers, ODISSYS can surface the interventions most likely to succeed for a specific child profile — making evidence-based SEMH support accessible to every school, regardless of size or resource.

Choosing an intervention provider is one of the most consequential decisions a school makes for its most vulnerable pupils. Yet in most schools, the process is informal, relationship-driven, and rarely based on outcome evidence. Here is how to do it better.

The Problem with How Schools Currently Choose Providers

Ask a DSL or SENCO how they found their current intervention providers and you'll typically hear some version of the same story: a recommendation from a colleague, a provider who approached the school directly, or someone the SENCO had worked with before. These are not unreasonable starting points — professional networks and trusted relationships matter. But they are not sufficient as the basis for decisions that will significantly affect children's lives.

The fundamental question — does this intervention actually work, for children like mine, in circumstances like ours? — is rarely asked. And when it is asked, it is rarely answerable, because the data to answer it simply doesn't exist in most schools.

Eight Questions to Ask Any Intervention Provider

1. What outcomes do you measure, and how? Any reputable provider should be able to tell you precisely what they measure, using which validated tools, at what intervals, and over what time period. Vague claims about "positive impact" or "improved wellbeing" are not sufficient. Look for providers who use standardised assessments — SDQ, SCWBS, NMRQ, or similar — and who can show you aggregate outcome data.

2. What does your outcome data show? Once you know how they measure, ask to see the results. What proportion of children show measurable improvement? Over what timeframe? What does "improvement" look like in practice? A provider who can answer these questions with specific data is a provider who takes evidence seriously.

3. For which children does your intervention work best? No intervention works equally well for all children. Good providers know their strengths — which presenting needs, age ranges, and profiles they are most effective with. If a provider claims their service is equally appropriate for every child with SEMH needs, be sceptical.

4. How do you handle safeguarding? Any provider working with vulnerable children must have robust safeguarding procedures. Ask specifically: what happens when a child makes a disclosure during a session? How is the information recorded? How and when is the DSL notified? What is the audit trail? The answers should be immediate, specific, and confident.

5. How do you communicate progress to the school? A provider who completes sessions and sends a summary report at the end of term is not giving you what you need. Progress information should be accessible in real time, enabling the school to respond to what is happening with a child rather than what happened last month.

6. What is your approach to co-production with the school? The best interventions are not delivered in isolation from the school — they are integrated with it. Ask how the provider works with teaching staff, pastoral teams, and parents. How do they share what they are learning about a child? How does the school's knowledge of the child inform the intervention?

7. Are you registered, insured, and accredited? Check that the provider has appropriate professional registrations, public liability insurance, and DBS-checked staff. For any provision involving direct work with children, these are non-negotiable baseline requirements.

8. Can you provide references from schools similar to ours? Outcome data and accreditations tell you a lot, but hearing directly from a DSL at a school with similar demographics and challenges is invaluable. Any confident provider should be able to provide references.

The Matching Problem

Even armed with answers to all eight questions, schools face a fundamental challenge: the information required to match a specific child to the most appropriate intervention is rarely available at the point of decision. A school might know that Provider A has good outcomes with primary-age children presenting with anxiety, and that Provider B specialises in trauma-informed work with adolescents — but knowing which of those profiles best describes the child in front of them, and which provider's specific approach is most likely to resonate, requires a level of granularity that most schools don't have.

This is the matching problem — and it is the problem that data-driven intervention management is designed to solve.

How ODISSYS Changes the Equation

ODISSYS aggregates outcome data from interventions across a network of schools and providers, making it possible to answer the matching question with statistical confidence rather than professional intuition. When a school refers a child through ODISSYS, the system draws on outcome data from hundreds of similar referrals to surface the providers with the highest likelihood of a positive outcome for that specific child.

This doesn't replace professional judgement — the DSL still makes the decision. But it means that decision is informed by evidence that no individual school could ever generate on its own.

For providers, the logic runs in reverse: ODISSYS identifies the schools and children most likely to benefit from their specific approach, making every engagement more likely to succeed and building a reputation grounded in evidence rather than marketing.