How to improve ‘do not resuscitate’ decisions in England

This review has highlighted some variations in how ‘do not attempt cardiopulmonary resuscitation’ decisions are made across NHS hospitals. By describing the literature and giving examples where things have gone well and less well in the past it begins to surface promising areas for improvement. These include the designing and implementing of structured forms to record decisions, talking sensitively about the decisions with patients and their families early and letting other health professionals know what has been decided.

A ‘do not attempt resuscitation’ decision allows resuscitation to be withheld following a cardiac arrest. This is usually done if there is little or no chance of success, if the risks outweigh the benefits, or if a person requests not to receive resuscitation. Lack of clarity can lead to confusion and conflicts within a fast-moving medical emergency. This can occasionally lead to complaints later.

The main part of this study reviewed the worldwide literature to look for insights into the processes, barriers and facilitators to ideal care. Other data collected in this study showed variation and inconsistency in current practice.

Why was this study needed?

In the year 2014/15, the UK National Cardiac Arrest Audit recorded 15,779 cardiac arrests taking place in 169 UK hospitals. Despite national guidelines on how decisions not to resuscitate, are made and recorded, evidence suggests there is wide variation in NHS organisations policies, leading to potential conflict, confusion or and complaint between patients, families and healthcare workers looking after them.

This study primarily aimed to review the literature on the topic and then where possible describe how improvements could be applied to England. By understanding better why conflict and complaints arise, the researchers hoped to propose solutions to aid resuscitation decision making, policy and practice.

What did this study do?

The review looked for studies describing interventions to improve resuscitation decision making processes. The researchers read through and summarised the papers where barriers and facilitators to making effective decisions were qualitatively described.

Other parts of the study recorded summary incidents and complaints relating to resuscitation decisions from a sample of NHS trusts. The researchers also collected data on the nature and frequency of issues about resuscitation decisions from a range of patient, government and NHS watchdog organisations. They also undertook focus groups where practitioner opinions on resuscitation decisions were discussed.

The triangulation of evidence from many sources and perspectives, by personal testimony, official policy and real-life practice, increases our confidence that the overarching conclusions are useful.

What did it find?

The review identified 37 studies describing the following processes:

  • Structured communication, two studies
  • Specialist teams, eight studies
  • Documentation, ten studies
  • Nursing home and community interventions, seven studies
  • Physician education, six studies
  • Patient/surrogate education, three studies and
  • Legislation, one study.

Among these 37 studies, over half from US, there were eight trials and 27 uncontrolled before-after studies. Although these provide less certain evidence they were useful in generating theories and provided useful pointers for those considering service improvement.

Good practice was made easier by:

  • A whole-team approach that included the patient, if able, or their family
  • A professional supporting body supporting the process
  • Early decision-making. One that is regularly and appropriately reviewed at a time and in a place where discussion is easy
  • Plain language.

Good practice, decision-making and discussion was made harder by:

  • Inconsistent timing of decisions, pressure of time and decisions being made in inappropriate places.
  • The fact that wishes may fluctuate over time and that patients may wish to reconsider their choice when faced with death was thought to make standard processes harder to implement.

The review also found that there were communication challenges between the decision makers, and patients and relatives. The issues identified included a misunderstanding of professional language, a need for specialist training and skill development, the need to clearly record the reasoning for a decision and to complete documentation and make it immediately available.

Around 1,500 do not attempt resuscitation incidents were reported by staff to the NHS each year, with 1 in 3 reporting harms. Reports of incidents or complaints were rare, at less than 5 in 1000 cases. Problems with communication with patients and or relatives, documentation and handover of decisions were the main sources of complaints.

What does current guidance say on this issue?

The British Medical Association, the UK Resuscitation Council and the Royal College of Nursing published guidance in 2014. The guidance identifies the ethical and legal principles around resuscitation decisions.

Most healthcare organisations have a policy in the absence of a valid recorded do not attempt resuscitation decision, that resuscitation will be attempted if a person dies or had a sudden cardiac arrest.

The UK Resuscitation Council has standardised forms and guidance notes for how health professionals can best record and communicate the decisions.

What are the implications?

This study shows that there are a range of ideas for improving how ‘do not attempt resuscitation’ decisions are made in NHS trusts. These hold potential for reducing the variations that were observed in practice.

To help, health professionals could use standardised forms to record their decisions which could make sharing of decisions across NHS settings easier. Alongside this standardised recording, patients and families could easily become involved in discussions earlier, sensitively, and as encouraged to make the decision as part of the overall treatment plan.

Priority areas of research identified by the study were the possible negative impact of these decisions on overall care, and ways public education and communication around the success of resuscitation could be improved.


Citation and Funding

Perkins GD, Griffiths F, Slowther A-M, et al. Do-not-attempt-cardiopulmonary-resuscitation decisions: an evidence synthesis. Health Services and Delivery Research. 2016;4(11).

This project was funded by the National Institute for Health Research Health Services and Delivery Research programme (project number 12/5001/55).



Joint statement. Decisions relating to cardiopulmonary resuscitation (3rd edition). London:  Resuscitation Council (UK); 2014.

Resuscitation Council (UK) and ICNARC. Key statistics from the National Cardiac Arrest Audit 2014/15. London: Resuscitation Council (UK) and Intensive Care National Audit & Research Centre (ICNARC); 2015.

NHS Choices. First aid – CPR. London: Department of Health; updated 2016.

Resuscitation Council (UK). CPR – FAQ’s. London: Resuscitation Council (UK); 2014-16.

Resuscitation Council (UK). DNACPR model forms and recommended standards for recording decisions. London: Resuscitation Council (UK); 2015.

Produced by the University of Southampton and Bazian on behalf of NIHR through the NIHR Dissemination Centre

Cardiopulmonary resuscitation (CPR) means emergency, life-saving procedures to maintain blood circulation where the heart and breathing have stopped. CPR can range from basic first aid to advanced techniques, to restore heartbeats and breathing.

NHS Choices provides information about carrying out hands-only CPR and CPR with rescue breaths. Basic instructions can also be given over phone after calling emergency services.


Expert commentary

This evidence synthesis illuminates why ‘do not attempt resuscitation’ decisions often cause problems or conflict. Most importantly it recommends solutions to some of the issues identified. The evidence, and hence the recommendations, are clear.

Firstly, do not attempt resuscitation decisions need to be considered and discussed within a broader framework of the overall care and treatment provided; talking about ‘do not attempt resuscitation’ decisions in isolation does not work. Second, talking about (and making explicit to other professionals) the broader considerations of care and treatment should support greater consistency in approach, as well as portability of decisions between health-care settings. Thirdly, experienced clinicians need to lead on this, and recognise the potential complexity/uncertainty of decision-making and the importance of devoting time to do it well. It’s not the routine completion of a form which matters, but the early, open and sensitive discussions about all care and treatment. If experienced clinicians do this well and early, they will alleviate subsequent distress, conflict and complaints.

Dr Fliss Murtagh, Reader and Consultant in Palliative Medicine, Kings College London