Intended for healthcare professionals


When and how to discuss “do not resuscitate” decisions with patients

BMJ 2015; 350 doi: (Published 20 May 2015) Cite this as: BMJ 2015;350:h2640
  1. Zac Etheridge, specialist registrar in geriatric medicine1,
  2. Emma Gatland, barrister2
  1. 1Medicine for Older People, Stoke Mandeville Hospital, Aylesbury HP21 8AL, UK
  2. 2Harcourt Chambers, London, UK
  1. Correspondence to: Z Etheridge zac.etheridge{at}
  • Accepted 30 March 2015

A legal case last year resulted in changes to the law on patient involvement in resuscitation decisions. Zac Etheridge and Emma Gatland set out the implications for doctors and hospitals

Decisions about cardiopulmonary resuscitation are an often difficult yet vital part of medical practice. Such decisions aim to prevent undignified interventions at the end of life and ensure that the patient’s last hours or days are spent as peacefully as possible. In October 2014, the Resuscitation Council (UK), the British Medical Association (BMA), and the Royal College of Nursing (RCN) reviewed their guidance for doctors making decisions about do not attempt cardiopulmonary resuscitation (DNACPR) orders.1 The revised guidance follows a Court of Appeal decision in June 2014 in the case of Tracey v Cambridge University Hospitals NHS Foundation Trust and another. The decision significantly changes pre-existing practice on how and when DNACPR orders should be discussed with patients and others, and may have further legal implications.2

The case

The Tracey case was a claim for judicial review brought by the widower of Janet Tracey, who died in Addenbrooke’s Hospital in March 2011, aged 63. Tracey had terminal lung cancer diagnosed in February 2011 and was given a prognosis of nine months. She was offered chemotherapy and the chance to participate in clinical trials, which she accepted. About two weeks later, she was in a car crash and sustained a high cervical spine injury. She subsequently developed pneumonia, which together with her advanced lung cancer meant that she required ventilation in the intensive care unit. After two unsuccessful attempts to wean her from the ventilator, doctors placed a DNACPR notice in her notes without consulting or informing her. This was despite the fact that she had until that point expressed a strong desire to be involved in decisions about her care. In discussion with one of Tracey’s daughters, one of the doctors said that Tracey would be allowed to “slip away.” Her daughter did not fully understand the implications of the words and was “horrified” when she discovered that a DNACPR order had been placed in her mother’s notes. Trust between Tracey’s family and her doctors broke down. Tracey found out about the notice and believed that her family had agreed to it. The DNACPR notice was subsequently cancelled, but it had caused Tracey and her family considerable distress in the final days of her life.

Mr Tracey brought a judicial review claim, initially in the High Court3 and then in the Court of Appeal, seeking a declaration that his wife’s right to a private and family life under article 8 of the European Convention of Human Rights had not been respected. His lawyers claimed that this arose due to the failure of doctors to involve Tracey in the decision making process that led to the DNACPR order.

The Court of Appeal ruled unanimously that the decision whether to perform CPR on Tracey affected her private life and hence engaged article 8 of the European convention. It went on to find that by not involving her in the decision about whether to resuscitate, her right to respect for her private life had been breached.

Lord Dyson, giving the lead judgment, acknowledged that although it is ultimately a medical decision whether to attempt resuscitation, and patients are not able to demand such treatments, there must be a “presumption of involvement” of the patient, and there must be “particularly convincing justification” not to consult the patient.

Changes to guidance

The Court of Appeal decision resulted in two important amendments to do not resuscitate guidelines.1 Firstly, patient distress is no longer sufficient justification for not discussing do not resuscitate decisions with patients. The Court of Appeal held that it must be more than that. Doctors must discuss a DNACPR order unless they consider it is likely to cause the patient “physical or psychological harm.”

Secondly, it is no longer the case that doctors do not have to discuss do not resuscitate orders when a clinical decision is made that CPR would be futile. The Court of Appeal found that futility is not a valid argument for not discussing a DNACPR notice with a patient. The fact that a treating clinician considers that CPR will not work means that the patient cannot require him or her to provide it. It does not, however, mean that the patient is not entitled to know that the clinical decision has been taken. Nor does it mean that the patient should not be given the opportunity to seek a second opinion.

Dyson also stated that a DNACPR decision needed to be made in accordance with a “clear and accessible policy.” Patients’ rights to be consulted about DNACPR decisions would be undermined if they were unaware of the criteria used in reaching a decision about resuscitation.

Below we discuss the practical implications of the judgment.

What discussions need to be undertaken before issuing a DNACPR?

Patients with the capacity to participate in discussions about resuscitation must be consulted before an order is issued. The only exceptions are if the treating clinician considers that the discussion is likely to cause physical or psychological harm to the patient or the patient indicates that he or she does not wish to discuss resuscitation. In both cases this should be clearly documented in the medical notes.

If the patient does not have capacity to be involved in the decision making process, resuscitation should be discussed with any legal proxy—for example, someone with lasting power of attorney for health and welfare. If there is no legal proxy, it should be discussed with the patient’s family, friends, or others with an interest in the patient’s welfare. In the absence of close friends or family, or if they do not wish to be involved in decision making, an independent mental capacity advocate should be appointed. If the decision regarding resuscitation is urgent, the decision can be made and an advocate appointed as soon as is practicable.

Following discussions, the decision whether to issue a DNACPR order ultimately lies with the treating clinician. However, if the patient or others disagree with the decision, a second opinion should be sought.

How should the issues around resuscitation be discussed, and what additional information should be provided?

When discussing DNACPR orders it is important to state clearly that “resuscitation will not be attempted in the event of cardiorespiratory arrest” (or equivalent in plain language). Colloquial words or phrases such as “slip away” should be avoided. In addition, patients and their families should be given further information, such as a leaflet, about the decision and its ramifications after the discussion. NHS trusts must have an easily accessible policy on DNACPR decisions and literature on what CPR entails.

Is a second opinion required regarding DNACPR orders?

The case of Tracey did not deal directly with whether doctors are legally obliged to offer a second opinion if a patient disagrees with their treatment decisions. However, two of the Lord Justices stated that they thought that doctors are obliged to do so. Part of the basis for requiring discussion of a DNACPR decision is so that a patient can ask for a second opinion, if desired. GMC guidelines also state that a doctor should offer to arrange a second opinion if a patient wishes to receive a form of treatment that the first doctor has not offered.4 Accordingly, if a patient disagrees with a DNACPR decision, a second opinion should be arranged. The exception to this would be if the decision is made by a multidisciplinary team, all of whom agree that a DNACPR order is appropriate, as occurred in the Tracey case.

What are the further ramifications of the judgment?

The Resuscitation Council has issued a preliminary statement setting out what it considers to be the potential consequences of the Tracey decision.5 One consequence might be that article 8 may be breached if a clinician does not discuss resuscitation with any patient who is at risk of cardiopulmonary arrest, regardless of whether the clinician is considering making a DNACPR order. Some patients may not wish to receive resuscitation, and if these wishes are not ascertained a patient may be put through unwanted CPR.

The Court of Appeal endorsed previous case law that states that “it is not for others to say that a life which the patient would regard as worthwhile is not worth living.”6 Many DNACPR forms contain as one of the options for not attempting resuscitation a statement such as: “Successful CPR is likely to be followed by a length and quality of life which it would not be in the best interests of the patient to sustain.” This box should be ticked only if the patient has stated the belief that, following resuscitation, his or her quality of life would be so poor that it would not be worth living.

A further important issue that arises from the judgment, although it is not specifically addressed within it, are advanced decisions regarding the level of care that a patient may receive. Doctors often decide that patients should have only ward based care and not be escalated to intensive or high dependency care. The Court of Appeal determined that since DNACPR orders are made in advance and potentially deny a patient lifesaving treatment, they engage article 8. By analogy, decisions regarding escalation of care and made in advance may similarly engage article 8 and hence need discussing with the patient.

A further notable point is that if cardiorespiratory arrest is not predicted or reasonably foreseeable in the current circumstances or treatment episode, it is not necessary to initiate discussion about CPR with patients. This does not change existing practice but is made explicit in the new guidance.

Summary points

  • Do not resuscitate orders (DNACPR) must be discussed with patients or their proxies unless they have indicated they do not wish to be involved or the discussion is likely to cause physical or psychological harm

  • The wording should be very clear, and the discussion should be clearly documented in the medical notes

  • Doctors should probably discuss resuscitation with any patient at clear risk of cardiorespiratory arrest, regardless of whether DNACPR is being considered

  • They may also need to discuss advance decisions regarding ceilings of care

  • NHS Trusts must provide an easily accessible DNACPR policy


Cite this as: BMJ 2015;350:h2640


  • Contributors and sources: Both authors contributed to this work equally. ZE works in geriatric medicine and has an interest in medicolegal affairs. EG practises in family and civil law. ZE is guarantor.

  • Competing interests: We have read and understood BMJ policy on declaration of interests and have no relevant interests to declare.

  • Provenance and peer review: Not commissioned; externally peer reviewed.


View Abstract