Intended for healthcare professionals

Observations Ethics Man

When is restraint appropriate?

BMJ 2010; 341 doi: (Published 04 August 2010) Cite this as: BMJ 2010;341:c4147
  1. Daniel K Sokol, honorary senior lecturer in medical ethics, Imperial College London
  1. daniel.sokol{at}

    All practicable and less invasive alternatives, such as persuasion and de-escalation techniques, must first have been exhausted

    I have encountered the issue of restraint three times in recent months. The first came in my involvement with a Ministry of Justice committee examining the use of a new pain compliance technique in young people in secure institutions. The committee was set up after the Observer ran a front page story detailing violent restraint techniques and the tragic deaths resulting from their use.1

    My second encounter with restraint was in an ethics committee meeting. The patient, who had severe learning difficulties, needed dialysis. Without it she would die in weeks or months. The problem was that she became extremely distressed when undergoing dialysis. The clinicians decided that restraint would not be in her best interests.

    Finally I came across a case that appeared in May 2010 before the Court of Protection, which makes decisions in relation to people who lack the capacity to make the decisions themselves. The patient was a 55 year old woman who needed an operation to treat an endometrial adenocarcinoma.2 Despite her fear of needles and hospitals she had agreed to the operation in the past but consistently failed to turn up for treatment. The judge ruled that it would be appropriate, if persuasion failed, for a consultant anaesthetist to go to her home and put some midazolam in her drink and to restrain her while she recovered after the operation.

    For some people restraint is so repugnant a prospect that it should never be used. It is, they say, a violation of a person’s autonomy and an infringement of their dignity. Restraint is indeed a violation of a person’s autonomy of action. The restrained individual cannot move even though he or she wants to. Often it is also a violation of autonomy of will, defined by Raanan Gillon as “the freedom to decide to do things on the basis of one’s deliberations.”3 Sometimes, however, the situation is more complex. Did the 55 year old patient with endometrial cancer want the operation or not? She agreed to it at first, but when the moment of truth came she refused. Which of her incompatible views is the most deliberated: the agreement or the refusal? In “Ulysses contracts,” named after Ulysses’ instructions to his crew not to untie him from the mast come what may, the autonomous patient agrees to be restrained should he or she later refuse treatment. Restraining the patient thus becomes part of respecting autonomy.

    Ulysses contracts are rare. Restraint usually requires the violation of autonomy. It can be violent, and there is, inevitably, a risk of harm. In the “nose distraction” technique, now no longer used in young offenders’ institutions, the “restrainer” strikes the child underneath the nose with the ridge of the hand. It is easy to get wrong. The restrainer could hit too hard or in the wrong place, causing a nose bleed, a broken nose, and psychological harm. This risk of harm must be justified. One justification is to protect others from harm. The Observer article talks of the use of brutal techniques on “unruly children.” Imagine a large, muscular, irate 17 year old, an improvised weapon in hand, fighting viciously with another teenager; “unruly” is too weak a word. It is easy to dismiss the use of restraints as barbaric and undignified, as remnants of the bad old days when it was undoubtedly abused, but such an outright dismissal ignores the realities on the ground. When a boy has his teeth embedded in another boy’s neck, these arguments lose much of their appeal.

    Another consideration is the risk of harm to the restrained individual. The risk of the restraint must be balanced with the risk of not restraining. One of the key roles of the Ministry of Justice committee was to identify and assess all the risks and harms, physical and psychological, of a new restraint technique. This also included an analysis of the potential harms on the restrainer. The nose distraction technique, for example, left the restrainer’s hands perilously close to the young person’s teeth. Only then can a proper evaluation of the pros and cons of restraint be made. In the case of the dialysis patient the risk of non-restraint was an earlier death than if dialysis was imposed. The clinicians, however, deemed that the profound psychological harm of forced treatment, coupled with the practical difficulty of administering dialysis and enforcing adherence between sessions, outweighed the benefit of a prolonged life.

    Whether in a secure institution, a hospital, or in the community the use of restraint should be a last resort. It is often barbaric and undignified. It generally does violate autonomy and cause harm, sometimes very serious harm. All practicable and less invasive alternatives, such as persuasion and de-escalation techniques, must have been exhausted. The restraint must be effective, its benefits must outweigh its harms, and those performing the technique must be adequately trained. Measures to deal with adverse outcomes must also be in place. In the case of the woman with endometrial cancer, for example, the anaesthetist would need to anticipate the possibility of oversedation and breathing difficulties. If appropriate, post-incident counselling should be offered to the restrained and the restrainer. The hard question is not whether restraint is ever appropriate but in what circumstances it is appropriate.


    Cite this as: BMJ 2010;341:c4147


    • Competing interests: DKS is a member of the Ministry of Justice’s Committee on the Use of Restraints in Juvenile Settings. The views expressed here are entirely his own.


    View Abstract