‘Do not resuscitate’ decisions in continuing care psychiatric patients: what influences decisions?
Nandini Chakraborty, William J. Creaney



We evaluated the various aspects of ‘do not resuscitate’ (DNR) decisions taken for psychiatric continuing care patients within NHS Ayrshire and Arran. Records were reviewed and nursing staff were asked their opinions about DNR orders in general and the way these were implemented on their wards.


There were 35 DNR orders among 88 continuing care patients in mental health wards for older adults. There were no DNR orders for the 25 continuing care patients in general adult psychiatry wards. Quality of life was the main issue when taking a DNR decision. Medical and nursing staff were involved in all decisions and the family in most. Patients were involved in only two cases. The documentation of the DNR order itself was satisfactory but documentation of the reasons behind the decision was inadequate. Patients with DNR status were perceived by ward staff to have more physical debilitation and more dependence on others. Local guidelines were being followed in most aspects, but these needed to be reviewed, as suggested within the resuscitation policy itself.


Decisions not to resuscitate may often be difficult to reach in psychiatric patients. Wards follow heterogeneous policies despite a resuscitation policy existing within the trust. Documentation needs to be improved and medical and nursing staff must reach a consensus regarding what constitutes quality of life and the appropriate time for a DNR decision.

A ‘do not resuscitate’ (DNR) decision entails complex medical, legal and ethical issues (Levin & Levin, 1980; Cotler, 2000; Cowper, 2000; Murphy, 2002; Thomas, 2002; Vetsch et al, 2002; Berger, 2003; Hartley, 2004). The effectiveness of cardiopulmonary resuscitation (CPR) varies according to the nature of the underlying clinical condition (Dautzenberg et al, 1993). Other issues which influence a DNR decision include the patients’ wishes, the nature of the underlying illness, the relatives’ wishes and expected future quality of life. The joint statement issued by the British Medical Association, the Resuscitation Council (UK) and the Royal College of Nursing (2001) states that a ‘do not resuscitate’ decision would be considered appropriate if CPR is not expected to restart a patient’s heart and breathing, if there are no benefits to be gained from restarting the patient’s heart and breathing and if the expected benefits are outweighed by the burdens.

Most studies on DNR decisions considered physically ill patients in whom CPR would have probably been unsuccessful (Beach & Morrison, 2002; Jackson et al, 2004; Hemphill et al, 2004; Tokuda et al, 2004). A number have explored ethical issues in the paediatric population (Klopfenstein et al, 2001; Da Costa et al, 2002). There are few studies of DNR decisions in those with psychiatric illness, where quality of life issues and the patient’s capacity/incapacity to consent could raise complex ethical dilemmas.

The aims of the current audit were: to look for reasons behind DNR decisions taken in psychiatric continuing care patients within NHS Ayrshire and Arran; to determine who were involved in these decisions; to examine the detail and precision of documentation; to determine how patients with a DNR status vary from those without; and to examine whether the local guidelines (resuscitation policy) were being followed.


Clinical and nursing case records of all continuing care patients in the psychiatric wards of NHS Ayrshire and Arran between July and September 2004 were reviewed and socio-demographic and clinical data were recorded in a pro forma. For patients with a DNR decision, the date of the decision, the reason behind it, who was involved in the decision and whether documentation was clear and complete were also recorded. Missing data were sought from ward staff. In addition, nursing staff were asked for their opinions regarding DNR decisions in general and the way these were implemented in their respective wards. This added a qualitative aspect to the study.

The local DNR guidelines of NHS Ayrshire and Arran are based on those issued by the British Medical Association, the Resuscitation Council (UK) and the Royal College of Nursing (2001).


Patients with DNR order

The proportion of patients with a DNR order varied among different wards (from 0 to 100%). All DNR orders were for the elderly; there were no DNR orders for general adult psychiatry patients. There were 35 DNR orders in total, which constituted 40% of continuing care patients in old age wards and 31% of all continuing care patients.

The socio-demographic and clinical characteristics of the elderly mental health patients in continuing care are presented in Table 1. Those with and those without a DNR order did not vary significantly on any measurable characteristic. However, patients with a DNR status were perceived by ward staff to have more physical debilitation and more dependence on others, suggesting a poorer quality of life. Quality of life was the main factor in deciding on a DNR order, although it seemed to be a subjective judgement with staff varying in their opinions of what constituted quality of life.

View this table:
Table 1.

Characteristics of elderly mental health continuing care patients with and without a do not resuscitate (DNR) order

Involvement in decision-making

When the patient had a living contactable relative, the relatives were always involved in the decision-making. Records show detailed discussions between the staff and family before the final decision was taken. In no case was a DNR decision taken against the wishes of the family. In 6 patients (17%) the family was not involved in the decision. In 2 of these cases, the records showed no living next of kin; in the other 4, no relative could be contacted with the details available. None of these patients had had any visitors for several years. Only 2 patients had made living wills where they had stated that they wished to have DNR orders in place when their condition had deteriorated. They were against resuscitation. No patients had capacity to consent at the time the decision was taken because of advanced dementia. Nursing staff were involved in all decisions and medical staff relied on them for an account of the patient’s quality of life, before beginning discussions with the family.


Documentation regarding who was involved in decision-making was precise and complete in both medical and nursing notes. Documentation regarding the reason behind DNR decisions was for the most part not clear; 21 (60%) of the DNR documentation recorded no clear reason why the decision was being taken. However, most of these case records did have clinical notes where the deterioration of the patient was recorded and from which the reason behind the DNR order could be gleaned. In 14 cases (40%) the one reason given was the advanced state of illness which contributed to a poor quality of life and made the success of CPR unlikely.

Nursing staff in all elderly mental health continuing care wards maintain a list of patients with DNR orders. We did not come across any patient who had a DNR order documented in the clinical records and yet was missing from the list, or vice versa. The wards of NHS Ayrshire and Arran Health use blue forms to record DNR decisions. These are easily detected within the notes and make the documentation of the order unambiguous. However, no blue forms were completed for 11 patients (31%). There were 2 blue forms each for 2 patients, with no evidence of the decision being reversed between the dates shown on these forms. In cases where there were no blue forms, the decision regarding DNR was clearly recorded in the notes and staff were clear who was for resuscitation and who was not.

Observance of guidelines

Guidelines were followed when a DNR decision was taken. The patients’ wishes (where expressed), the families’ wishes, the expected quality of life and the prognosis of the patients’ conditions which influenced the probable outcome of CPR were all taken into consideration. However, the resuscitation policy also states that guidelines need to be reviewed and audited annually and there was no documentation of this happening.

Continuing care patients within wards of NHS Ayrshire and Arran all have an annual review during which a DNR order, if in existence, is always reviewed and documented to be still in effect. We found this documentation in all relevant cases. Owing to the nature of illnesses among these patients, there was no reason to reverse a DNR order once in place.


The study revealed different practices on the different continuing care psychiatric wards despite a resuscitation policy which is supposed to provide the local guidelines for all wards. Particular ward policies are dependent on the opinions of the supervising consultant. Significantly there were no patients in continuing care wards of general adult psychiatry with DNR orders. All the patients with a DNR order belonged to the old age psychiatry wards.

Consultants on general adult wards and those on old age psychiatry wards differed in their opinions about whether DNR orders should be implemented at all. Moreover, consultants on old age psychiatry wards differed in their opinion about when the DNR decision should be considered. Some preferred to raise the subject with the family when the patient was considered to have a poor quality of life but would probably live for years in that condition. Others preferred to raise the sensitive issue only when a patient was clearly approaching the end of their life.

Different attitudes towards DNR orders have been found in other studies (Granja et al, 2001; Kelly et al, 2002) and have been influenced by medical specialty and years of experience and training. We were unable to compare differences in attitudes towards DNR orders within psychiatric sub-specialties. Positive findings from this study about the practice surrounding DNR orders within psychiatric continuing care wards were: the involvement of the nurses, family and patient (where possible) in discussion before a doctor took a DNR decision and the very clear and unambiguous documentation of the order itself. Other studies have reported dissatisfaction with nursing involvement (Castledine, 2004) and ambiguous documentation of the DNR order, sometimes leading to CPR being attempted in patients who were not for resuscitation (Skerrit & Pitt, 1997; Becker et al, 2003). However, we found scope for improvement in the documentation regarding the reason behind the order and in accordance with Skerritt & Pitt (1997). Documentation of who was involved in the decision was satisfactory.

Quality of life remained undefined by the resuscitation policy and staff differed in their opinion about quality of life. Studies reveal that physicians tend to underestimate quality of life in their patients (Junod Perron et al, 2002). This needs further discussion, involving various health professionals, to arrive at a consensus and thus uniform decision-making. Although a patient with a DNR order was always found to be incapable of making a decision because of the advanced stage of illness, there was no evidence of current cognitive abilities being measured by standard tests. Documentation of such an evaluation might add objectivity to the assessment of the patient’s condition at the time a DNR decision is being taken. It is noteworthy that only 2 patients had made their wish to have a DNR order known before significant cognitive deterioration. It may be worthwhile to consider DNR orders as a subject for discussion with a patient diagnosed with a progressively deteriorating condition such as dementia, as early as possible after diagnosis. This would give patients more chance to be involved in DNR decisions while they are still capable.

Another interesting revelation was the difference in understanding of what was conveyed by a DNR order among the nursing staff on the ward. Many nursing staff connected DNR not only with CPR but also with the intensity of medical intervention for any medical condition.


This study reveals that DNR is a difficult decision to take, especially in patients who are incapable of/unable to make their wishes known. Wards follow heterogeneous policies; documentation of reasons behind a DNR decision is inadequate and medical and nursing staff need to reach a consensus regarding what constitutes quality of life and the appropriate time for a DNR decision. An annual review of the resuscitation guidelines within the trust, involving various health professionals, is recommended. This remains a sensitive issue, which has to be handled carefully, but ethical debate will probably always be important.


We thank the consultants on continuing care wards for permission to look into case records. We also thank nursing staff for their help during data collection and offering their honest opinions.


View Abstract