Policy paper

Independent Restraint Review Panel (IRRP): summary report to end 2025

Published 20 July 2026

Applies to England and Wales

1. IRRP 2025 report

Throughout 2025, the Independent Restraint Review Panel made 12 visits to establishments. These included all four Young Offender Institutions housing children and the Secure Training Centre at Oakhill. During those visits we reviewed a total of 111 restraints where Minimising Physical Restraint (MMPR) techniques were used.

We reviewed all incidents where Pain Inducing Techniques (PIT) had been recorded as having been used. We selected the incidents to be reviewed from the master records held locally by establishments. These incidents were selected from a time period spanning broadly back to when we had visited previously and in some instances just before this, meaning that in terms of timing, there is not direct comparability with official statistics.

Across the year, we identified and reviewed 11 incidents when PIT, as defined by MMPR, was recorded as used. In each of these incidents, PIT was used just once. This represents a substantial reduction compared to the previous year when PIT was recorded as having been used 43 times across 25 incidents of restraint. Of the 11 uses of PIT in 2025, we assessed 7 as fully meeting the policy criteria set out in the MMPR manual, that

the application of a pain inducing technique should never be used where a non-painful alternative can safely achieve the same objective

however, the use of a pain inducing technique may be justifiable if that is the only viable and practical way of dealing with a violent incident which poses an immediate risk of serious harm to the young person, other young persons or staff

Where we assessed the criteria for use of PIT as not having been met, our views sometimes aligned with those of the establishment based on their local post incident reviews, but not always.

Our reviews additionally included incidents where pain had been experienced beyond what might result from MMPR techniques. One resulted from the use of a personal safety technique by an officer who reasonably believed that he was otherwise about to suffer serious injury. His actions were assessed as reasonable in the circumstances faced. Another involved application of a wrist hold, which although not intended to do so, clearly caused pain to the young person.

With the Ministry of Justice decision to make Pava available for use in the children’s estate during 2025, IRRP was asked to review independently those instances where Pava was both drawn and deployed. We were not asked to review instances where Pava had been drawn but not then deployed, although potentially these could have been looked at as part of our existing work. We reviewed one such case.

During 2025 we reviewed 6 incidents where Pava was both drawn and deployed. Pava had been deployed 8 times across these six incidents. These incidents exclude one which occurred late in December 2025, during which Pava was both drawn and deployed. It was not reviewed by IRRP until 2026. Our reviews followed local establishment reviews and national Youth Custody Service reviews. We understand that our findings will be looked at as part of the consideration of the Pava evaluation..

What we observed was that when deployed, Pava quickly led to de-escalation as children were either incapacitated or retreated/demonstrated de-escalation to avoid being sprayed. We challenged the justification for the use of Pava in half of the deployments reviewed.

2. The Panel and approach

During 2025, the constitution and approach of the panel continued as in the previous year (see 2024 report). Local Authority Children’s Services representation became stronger, with representatives often also joining focus groups with children as well as the formal panel. On one occasion, the Panel was also joined by a representative of the Independent Monitoring Board. Where the governor wanted this, attendance by specialists such as local healthcare or social worker representation was also facilitated.

3. What the children told us

A subset of the Panel met with children on every visit. These meetings were conducted in confidence in a bid to provide a safe space for the children to open up to us. We met both individuals and with groups, boys and girls and with children of mixed backgrounds, age, custodial experience and ethnicity. Although our focus was to understand their views on and experience of restraint – positive and negative – often the children wanted to broaden the discussion and as a result we were sometimes able to identify wider issues which they found frustrating, as well as their ideas for improvement.

What we heard from the children was fed into each full Panel in anonymised form and relayed to the Youth Custody Service nationally.

Many children told us that they didn’t recall having been told anything about restraint during their induction. As the year went on, after we had raised this apparent gap, newer arrivals had some recollection of it being touched on during their induction although, universally, they were clear that their focus during their early days was on visits, phone calls, canteen, etc. Most children told us that they got to understand how restraint operated through their peers and through what they observed. On most visits we heard children quote the rule book related to restraint, as they understood it. Generally, their understanding was correct.

Although children generally knew who we were and that we wanted to talk with them about restraint, they clearly hadn’t been coached or influenced beforehand in what to tell us. That said, turnover and behaviours locally, meant that those who had been restrained the most, weren’t always available for us to speak with. We have since changed our approach to try to increase access to a wider mix of children.

At one establishment, we identified concerns amongst children that black boys felt themselves to be treated differently. This was followed up locally through focus groups and targeted awareness work with staff. On our next visit, the children we met raised no concerns. Some children told us of concerns about racist comments or behaviours by other children. We didn’t hear this said about staff. When we dug deeper, the children said that sometimes staff didn’t tackle head on what was said by other children. Racist, abusive and threatening comments made towards staff were often heard when we reviewed restraints. Almost universally, staff responded professionally to even the most provocative of behaviours. Where we heard staff responding inappropriately – we did hear occasional swearing and some inappropriate language - we called this out.

Most children told us that they considered restraint when used, generally, to have been justified by the circumstances. Some children felt that staff could and should have been slower to use force and that had they been given more time and been listened to better, it may not have been necessary to use restraint.

Some children went as far as to tell us that, knowing staff would respond quickly should there be a fight, provided a sense of safety. Some children talked about individual staff who they felt were quicker to use restraint than others, or who seemed consistently to be the first responders when alarms were raised. Some felt that previous behaviour influenced the likelihood of restraint and the level of force deployed. Some black boys felt that when they were restrained, staff tended to expect more resistance and, consequently, used more force. Relationships with staff were cited as important to whether restraint was likely in particular circumstances.

In almost every establishment, the children identified individual staff who they felt were doing a good job, describing these staff variously as fair, respectful, caring, doing what they say they will do, knowing what they are doing, etc. New staff were felt generally by children to be less confident during restraint and quicker to intervene when behaviours escalated.

By contrast, children were very reluctant to identify staff who didn’t behave like this, preferring instead to talk in generalities. On the occasions when children were happy to go further, they talked about concerns related to how complaints if made, may sour relationships with staff or about having limited confidence that change would result, should they complain. Some talked about the fear of retribution, should they complain. As the year progressed, it was evident to us that establishments had tried to improve children’s confidence in local complaints procedures but that there remained more to be done. All children we spoke to knew how to make a complaint.

As reported previously, Bernardos continued to be valued by the children, though more as a source of support to address individual needs than as an organisation to which any concerns about restraint might be put. Youth Offending Team (YOT) contacts were generally seen by the children as having limited influence within custody.

When exploring the sorts of situations that were likely to trigger restraint, we were told that fighting, use of weapons, threatening behaviour and refusing a lawful order were the main triggers. Although we saw force being used to prevent self-harm, this was rarely raised by the children – perhaps because of the individual sensitivity.

Triggers which the children questioned the most as justification for the use of force were threatening behaviour and refusing a lawful order. On the first, the children said that venting frustration or challenging things they disagreed with didn’t mean that they were about to resort to violence. Unpacking this, the children often acknowledged the challenge for staff in assessing what might happen next when tensions arise, but many children felt that creating space and time might have allowed them to de-escalate. We saw some good examples of where staff did exactly this.

On the question of refusing a lawful order, some children talked about the power imbalance between staff and children, often describing the best staff as not needing to rely on power to get respect. In a very small number of instances, children talked about individual staff who they felt relied too much on their power status and authority. The overwhelming feedback was that most staff are considered fair, respectful and that they care about the children. Often, and usually without us prompting, the children talked about how difficult the job of a prison officer must be.

On several occasions, children told us of individual restraints about which they had concerns. Sometimes these comments came from children not directly involved but who felt that a restraint had not been justified by the circumstances. We reviewed a number of these and, albeit with the benefit of hindsight, in some instances we also questioned whether the initiation of restraint was justified or potentially avoidable had a different approach been taken.

In one example, the initiating officer had reasonably refused a request by the child but then could potentially have consciously created space for the child to deal with his frustration. Having not created that space, it developed into confrontation and a restraint. In another, a message to the children from the governor to explain the circumstances around a restraint which many of the children had been discussing as unjustified and unfair, had gone a considerable way toward settling concerns.

4. Pava

Pava became available for use within the children’s estate during 2025. IRRP had no direct input to this decision, but we were asked to extend our remit to provide an independent view on each occasion when Pava was drawn from the holster and deployed. Our independent review followed and drew upon the local establishment review and a YCS chaired national review of each deployment.

We had reported last year that speculation was circulating that Pava may be being considered. We had picked this up from staff as, in turn, had children. During 2025, children had picked up that an announcement was expected and once Pava had been introduced, all the children we spoke to knew this to be the case.

As noted above, we reviewed 8 instances of Pava deployment across 6 incidents. A deployment 24 December 2025 was not reviewed during the period covered by this report.

Our reviews were assessed against the policy criteria set for the use of Pava during the 12 month exceptional limited deployment model, focussing particularly on the risks faced and whether there may have been alternatives to Pava as a mode of restraint. Initially we were concerned that all alternatives to Pava had not been considered fully in a number of the cases that we reviewed – around half of the deployments we reviewed were judged either to not meet the policy guidance or we were unable positively to satisfy ourselves from the information available to us that the guidance had been satisfied. We raised this with YCS and understand that this was in turn discussed with operational managers. Since then, there appears to have been a marked reduction in the deployment of Pava.

What we did see was the clear impact of deployment in dangerous and volatile situations, often involving multiple assaults or individuals being subjected to serious assault by multiple assailants. Often staff numbers were far outweighed by the number of children. Weapons were often present and seen to be being used with intent to cause serious harm. Mixing on exercise yards, often involving some pre-planning by children was common to these group assaults. The introduction of new children into an existing group, was often identified as a trigger for the violence. Pava when used, resulted in rapid cessation of violence or provided the opportunity for staff to separate and to protect those fighting.

To reduce the likelihood of violence starting, if Pava is to continue to be available to staff, we suggested that there may be a deterrent benefit in considering the overt availability of a Pava carrier in situations where mixing is to take place and where intelligence would suggest there to be particular risks.

Talking to children it was apparent early on, that they did not believe staff would use Pava, other than in the most serious multi-handed fights where weapons were also being used. Where there was violence involving fists and kicks, children felt the use of Pava not to be justified.

Aftercare for those exposed to Pava - both children and staff – was observed to be good.

5. Placement considerations

We surfaced a small number of instances where the placement of a child in a particular establishment raised questions. Similarly, we identified instances where move on arrangements appeared to be too slow. As an example, IRRP raised concern that, in part because of the small numbers in custody (outside of secure children’s homes, SCH), individual girls were sometimes isolated and their specific needs were not always provided for well.

We questioned assumptions made as part of the girl’s review, that SCHs would accept those girls presenting the most challenging behaviours or whether, because there were no other options available to them, the YCS placement team would be forced to allocate them to Oakhill STC or the Secure school. Allocation decisions should in our view be based on the needs of the child.

We identified individual children whose movement on within the children’s estate appeared to have become bogged down. Some were well over 18 and presenting risk to other children and staff. Individual intervention by YCS senior management successfully unblocked some such cases.

We became aware of a placement of a vulnerable, quite young child outside of the SCH network because, based on previous extreme behaviours, no SCH would accept them. They were instead placed at Oakhill.

The establishment had done a good job in identifying need and providing care for the child but a combination of intervention to prevent self-harm and to protect against violence meant that in one month they were involved in 25% of all restraints at Oakhill. For this child we suggested Director of Children’s Services (DCS) to DCS dialogue in the hope that on release a suitable package would be in place to support them in not re-offending or needing to be returned to custody.

On occasions, there was very limited information available to both the placement team and the receiving establishment. Timely information sharing is a challenge across the system the absence of which hampers effective care and behaviour management.

6. Summary findings

Previously recorded early reticence to engage with IRRP at establishment level has largely passed. Governing Governors and Directors have continued to support IRRP’s work. Whilst our views were sometimes challenged locally, the hoped for learning environment we sought to create within IRRP has been achieved. Feedback on our approach has been positive. YCS nationally, responded to our findings in a timely fashion and in some instances, followed up immediately when issues were identified.

Children engaged with us constructively. On the small number of occasions where we met children on a second visit, they were often able to reflect on what had happened in the interim, sometimes reporting that issues they had raised with us previously, had been addressed.

We were increasingly told that restraint is now explained during induction.

We continued to see improvement in the quality of post incident reports from staff, with attention to not just what happened but also to what led up to the incident and why staff made the decisions that they did. We saw good examples of reflection by staff on whether different approaches might, with hindsight, have led to a different outcome. And, where there may be safeguarding or other questions, we often saw these proactively advanced by staff as issues to be looked at.

Although there is nothing to see when interactions result in no force being used, we saw good examples of staff trying hard to de-escalate emerging tensions. But we also saw incidents where de-escalation opportunities were not tested as we would have hoped or where staff could have created more time and space for children. In some incidents, we questioned the initial decision to initiate the use of force.

Generally, there was good CCTV or body worn camera (BWC) footage available. The short pre-record facility now available on BWCs proved particularly helpful in understanding the immediate antecedents.

We saw good evidence of local post-incident assurance reviews with clear action plans where improvement or learning opportunities were identified.

Across every establishment we saw attention to de-escalation with consideration being given in many instances to whether a prone relocation was avoidable. We still saw too many occasions when testing the opportunity to release or lower the level of holds could more proactively have been tested. We saw too many instances of children under restraint being instructed by multiple voices but not being actively listened too. Too often we heard staff telling children “Work with us” – something which has little meaning to a child being restrained by several adults. And, as previously, we saw instances where it was not clear that the incident manager had created themselves the space to be able to manage effectively.

Where Serious Injury Warning Signs (SIWS) arose (eg. a child stating that they could not breathe), other than in a tiny number of instances, these were responded to quickly and effectively by staff.

We commented last year on space restraints making it extremely difficult for staff in most establishments when relocating or restraining children in their rooms. Whilst cost is clearly a consideration, we must again register this constraint and the challenge/risk that it presents.

We have commented before on examples of outstanding work done by staff in the most challenging of circumstances. We saw the same during 2025. We also saw plenty of examples of children consciously not getting involved in violence when had they done so, staff would have been at risk.

Where PIT was used, we didn’t always hear a clear warning. In some instances, the use was to ensure compliance rather than to prevent immediate serious injury, alternatives were available or we could not see risks that justified the use of a pain inducing technique. But the overall drop in PIT use is to be welcomed.

Our reviews of Pava use represent very early usage. Pava when used has an immediate effect. Since the period covered by this report, usage appears to have reduced significantly.

Colin Allars
Chair Independent Restraint Review Panel