Sample policy · Restraint

Restraint and restrictive interventions policy template

Statutory anchor: Regulation 13 (safeguarding service users from abuse and improper treatment, including disproportionate restraint), Health and Social Care Act 2008 (Regulated Activities) Regulations 2014 (SI 2014/2936). This policy also engages the Mental Capacity Act 2005 and the Human Rights Act 1998 (Article 5 ECHR). · primary source

Download the PDF

The PDF version of this template is the same content, formatted for adaptation in your document control system. The disclaimer above is repeated on the PDF cover.

Download PDFTalk to a consultant

Verivius pack version v1.1, 2026-07-19

1. What the regulation says

Service users must be protected from abuse and improper treatment in accordance with this regulation. (Reg 13(1) (the headline duty))

Systems and processes must be established and operated effectively to prevent abuse of service users. (Reg 13(2) (prevention systems))

Systems and processes must be established and operated effectively to investigate, immediately upon becoming aware of, any allegation or evidence of such abuse. (Reg 13(3) (investigation systems))

A service user must not be deprived of their liberty for the purpose of receiving care or treatment without lawful authority. (Reg 13(5) (lawful authority for deprivation of liberty))

ill-treatment (whether of a physical or psychological nature) of a service user, (Reg 13(6)(b) (ill-treatment))

neglect of a service user. (Reg 13(6)(d) (neglect))

The full text is at https://www.legislation.gov.uk/uksi/2014/2936/regulation/13. Where this policy and the regulation diverge, the regulation wins.

2. Plain-English summary

Service users must be protected from abuse and improper treatment. You need effective systems to prevent abuse, and effective systems to investigate any allegation or evidence of abuse as soon as you become aware of it. Care must not be provided in a way that discriminates, uses disproportionate control or restraint, is degrading, or significantly disregards the service user's needs. Service users cannot be deprived of their liberty without lawful authority.

3. Purpose

This policy sets out how the Service prevents restrictive interventions, decides whether an intervention is lawful and necessary, keeps the person safe during its use, and learns from every episode. Its purpose is to make least-restrictive practice visible in care plans, staff decisions, incident records and governance review.

4. Scope

This policy applies to:

Before adoption, the provider identifies which restrictive practices could arise in each service, which roles may authorise or use them, which techniques are prohibited, where emergency clinical help is obtained, and who gives legal or safeguarding advice. The local procedure must state where care plans, capacity decisions, incident records and restraint reviews are kept.

5. The Service's commitment

The Service is committed to the least restrictive practice and to preserving people's human rights. It works to reduce and, wherever possible, eliminate restrictive interventions through positive behaviour support (see the positive behaviour support policy). Restraint is never used as a punishment, for the convenience of staff, or in place of adequate staffing or care planning.

6. Definitions

The Service recognises restraint in all its forms, and distinguishes it from ordinary, agreed support:

7. Roles and responsibilities

The provider records its authorised-role and technique map before this policy is approved. A person must not use a technique merely because they have seen another worker use it.

8. The legal framework: necessity, proportionality and last resort

A restrictive intervention is lawful only where it is:

Where the person lacks capacity for the decision, the Service follows the Mental Capacity Act, makes a best-interests decision, and, where the restriction may amount to a deprivation of liberty, seeks lawful authority (see the consent policy).

Whether a restriction amounts to a deprivation of liberty is determined by a multifactorial assessment following the Supreme Court's AGNI judgment of 2 June 2026. The previous "acid test" is no longer the legal test. The Service follows the current government deprivation-of-liberty guidance and Code of Practice, records the factors considered, and seeks legal advice where the position is uncertain. The deprivation-of-liberty assessment is separate from the decision about consent to care or treatment. The Mental Capacity Act 2005 and Regulation 11 consent requirements continue to apply.

9. Restrictive-intervention workflow

  1. Prevent and de-escalate. Staff follow the person's communication, care and positive-behaviour-support plans, reduce known triggers, offer space or alternatives, and seek senior support early.
  2. Make an immediate risk decision. The authorised worker identifies the harm that is likely without intervention, considers capacity and consent, and decides whether action is necessary, proportionate and the least restrictive available option.
  3. Use only an authorised response. Staff use only approved techniques within their current competence. Prone restraint and techniques that restrict breathing are not part of the Service's planned response and are avoided. Any departure required to respond to an unforeseen emergency is stopped as soon as the immediate danger passes and is escalated for clinical and legal review.
  4. Monitor continuously. One worker monitors breathing, circulation, consciousness, pain, distress and the person's verbal and non-verbal response. The intervention stops immediately if safety is compromised or the risk has passed.
  5. Restore safety and dignity. Staff check for injury or distress, arrange medical help where needed, explain what happens next in an accessible way, and notify the senior person on duty.
  6. Record and report. The episode is entered in the incident register before the end of the shift. Safeguarding, police, duty-of-candour, CQC-notification and deprivation-of-liberty thresholds are considered and the decision is recorded.
  7. Review and reduce. The person and staff are offered separate debriefs. The Registered Manager reviews the episode, updates care and risk plans, assigns improvement actions, and checks whether patterns require a formal restraint-reduction plan.

10. After any use of restraint

11. Training requirement

12. Audit

The Service checks, on a stated cadence, that:

Audit findings are recorded in the provider's audit register. Actions are logged in the improvement-actions register with an owner and due date, and unresolved systemic risks are linked to the risk register.

13. Records and evidence

Every use of restraint must be recorded. The record includes:

The restraint record is cross-linked to the person's care record, the incident register and, where applicable, the safeguarding, risk, training and improvement-actions registers. Records are retained securely for the period set by the provider's retention schedule and applicable records-management guidance.

14. Related policies in this pack

15. Sources and further reading

This template is based on CQC's guidance for providers and managers, the Health and Social Care Act 2008 (Regulated Activities) Regulations 2014, and other topic-specific legislation and guidance listed below. It is a starting point for adaptation, not a substitute for legal, clinical, HR, safeguarding or specialist professional advice.

16. When to seek further advice

Seek specialist advice where the issue involves serious harm, safeguarding, deprivation of liberty, restraint, children, professional misconduct, controlled drugs, radiation, termination of pregnancy, infection outbreak, water safety, employment dismissal, DBS barring referral, or regulatory enforcement.

17. Document control

Version Date Author Changes
v1.0 2026-06-05 Verivius (sample) New template authored to CQC's "what to include" for a restraint policy: a least-restrictive, human-rights policy statement; clear definitions of physical, mechanical, chemical and environmental restraint versus general restriction; the necessity, proportionality and last-resort legal framework; roles authorised by role not by named individual (the CQC red flag); MCA best-interests alignment; implementation guidance on approved techniques, wellbeing monitoring and time limits; post-incident debrief, analysis and care-plan update; and mandatory de-escalation and safe-restraint training from a certified provider, prioritising de-escalation.
v1 2026-06-10 Verivius (sample) Conformed to the Verivius policy standard: anchored to Regulation 13 verbatim quotes and plain-English summary from the guidance manifest; added the standard header, sources and further reading, and when-to-seek-further-advice blocks; updated the deprivation-of-liberty framework to the AGNI judgment (2 June 2026) and CQC statement (8 June 2026), removing the "acid test" as current law and noting that MCA and Regulation 11 consent requirements are unchanged.
v1.1 2026-07-19 Verivius (sample) Added an operational restrictive-intervention workflow, local adoption decisions, fuller evidence fields and register links. Updated the deprivation-of-liberty section to current government guidance and clarified that consent alone does not determine whether a measure is restraint.

This sample policy template was issued by Verivius. It is a template, not a substitute for legal advice or the provider's own policy-development process. Where this template and live law or regulator guidance diverge, the live source wins.

What good looks like here

Written from an ex-CQC inspector's chair, but the point is safe, well-led care your team can stand behind. Each row shows what strong evidence looks like, what thin evidence looks like, and where the expectation comes from.

Restraint is the one intervention where the service deliberately overrides what a person wants, and the harm runs two ways: breathing and circulation can be compromised within minutes, and the fear that follows an episode makes the next one more likely rather than less. For people with a learning disability and for autistic people, restrictive interventions often become the standing answer to a communication need nobody has properly assessed, which is how a service ends up restraining the same person week after week and filing it under behaviour. The restraint record is usually the only surviving account of what the person went through, because the person may not be able to give that account themselves, so a record that omits the monitoring, the less restrictive options tried and the person's own words writes them out of their own story. What shows a service is working on the causes rather than managing the episodes is a restraint count that is genuinely falling, backed by care plans that visibly changed after each one. The deprivation-of-liberty question runs alongside all of this: since the Supreme Court's AGNI judgment of 2 June 2026 it is assessed on a multifactorial, contextual basis, and a measure can restrict liberty whether or not the person resists, so the factors weighed need writing down at the time rather than reconstructing them months later.

  1. Whether each recorded use of restraint actually tests as necessary, proportionate, the least restrictive option and a last resort, or whether the log simply records that restraint happened.

    Strong evidence: The restraint record's fields for the harm the intervention was intended to prevent, the less restrictive options tried (or why they were not safe to try), and the reason the intervention was necessary and proportionate (section 13), tested by the section 12 audit that checks every use met all four.

    Weak evidence: The necessity and proportionality fields read identically on every record, usually some version of "to prevent harm to self and others", and the less-restrictive-options box is blank or filled with "none available". Nothing names the specific harm that was likely that day, or what made it likely, so no one reading the record later can judge whether the intervention was proportionate to the risk actually in front of the worker.

    Law

    Health and Social Care Act 2008 (Regulated Activities) Regulations 2014, reg 13(4)(b) (with reg 13(5) defining restraint); Mental Capacity Act 2005 s.1(6) and s.6(3) for the least-restrictive limb

  2. Whether de-escalation was genuinely tried first and the use of restraint is falling over time, not staying flat or rising unnoticed.

    Strong evidence: The section 12 audit line that de-escalation was tried first and that the use of restraint is reducing over time, read against the de-escalation and less-restrictive options recorded in each restraint record (section 13) and workflow step 1 (section 9).

    Weak evidence: De-escalation appears as a ticked box rather than a note of what was tried, by whom and why it did not work, so the last-resort test cannot be checked after the event. The audit counts restraint episodes but never sets them against the previous period or against one person's own history, so a service that used restraint eleven times last quarter and fourteen times this quarter still reads as "no concerns", and a pattern building around a single person stays invisible.

    CQC guidance

  3. Whether one worker actually monitored breathing, circulation, consciousness, pain and distress throughout the episode, or whether monitoring was only written up afterwards.

    Strong evidence: The 'physical and psychological monitoring during the intervention' field of the restraint record (section 13) and workflow step 4, which requires continuous monitoring and an immediate stop if safety is compromised (section 9).

    Weak evidence: The monitoring field is written up once the person is back in their room, in the same hand and ink as the rest of the record, with a single line such as "breathing normal throughout". There is no note of which worker held the monitoring role while others held the person, nothing on circulation, consciousness, pain or distress separately, and no record of the point at which staff would have stopped.

    Professional guidance

  4. Whether staff used only approved techniques within their current competence, with prone and breathing-restricting holds kept out of the planned response.

    Strong evidence: The training lead's approved-technique map, competence evidence and refresher dates (sections 7 and 11) against the type-of-intervention and staff-involved fields of the restraint record (section 13); workflow step 3 excludes prone restraint and techniques that restrict breathing from the planned response (section 9).

    Weak evidence: The training matrix marks everyone as "restraint trained" with a certificate date, but carries no map of which techniques each person is signed off to use, so competence cannot be matched to the hold recorded on the day. The policy keeps prone restraint and breathing-restricting holds out of the planned response, yet free-text incident notes describe people being "guided to the floor" face down, and nobody has tested those descriptions against the exclusion or escalated them for clinical and legal review.

    Professional guidance

  5. Whether the safeguarding, duty of candour, CQC-notification and deprivation-of-liberty thresholds were each considered and the decision recorded, including the reason where no referral or notification was made.

    Strong evidence: The restraint record's decision fields for safeguarding, police, duty of candour, CQC notification and deprivation of liberty, which must carry reasons where no referral or notification was made (section 13); workflow step 6 requires those thresholds to be considered before the shift ends (section 9).

    Weak evidence: All the threshold boxes are marked "N/A" or left empty, with no reason recorded for the decision not to refer or notify. Safeguarding, duty of candour and CQC notification are treated as one judgement made by one person at the end of the shift, rather than as separate duties owed to different people and bodies, so a local-authority referral is logged as if it discharged the notification too. The deprivation-of-liberty question is skipped entirely on the basis that the person did not resist.

    Verivius default

  6. Whether each episode led to a debrief and an updated care plan and positive behaviour support plan, or whether the record closes with nothing changed to reduce recurrence.

    Strong evidence: The debriefs-offered-or-completed field and the care-plan, risk-assessment, training or staffing changes linked to an improvement action in the restraint record (section 13), and the after-any-use steps requiring plan review and update (section 10).

    Weak evidence: Debriefs are logged as "offered, declined" on every record for months, with no note of what was offered, when, or by whom, and no follow-up. The care plan still carries the review date it had before the episode, and the improvement action, where one exists at all, has no owner and no due date, so the next worker meets the same trigger with the same plan.

    Verivius default

Related

Last verified 2026-07-20

Related Verivius content

Want help adapting this to your service?

A Verivius consultant can read your adapted policy against the live regulation and your service shape. The work fits inside a Mock Inspection engagement or a shorter consulting brief. A 20-minute conversation is the fastest way to find out whether the fit is right.

Get started free

Free to start, no card. A 14-day trial when you subscribe.

Last reviewed 19 July 2026