The nature of the do-not-attempt-cardiopulmonary-resuscitation decision

Focus group participants recognised DNACPR decisions as complex and difficult. Participants described DNACPR decisions as one of the following: signing a DNACPR order, revoking a DNACPR order or not making a decision about a DNACPR order when it might be considered appropriate to make a decision (e.g. in the community or hospital when a patient is very frail).

Senior nursing staff have experience to know whether a patient is likely to survive CPR. They cannot complete the DNACPR form but they and FY2 or FY1 can discuss it and make a note that they have discussed it. If a senior doctor is not available to come down, if the FY2 has discussed it with the senior doctor by phone the FY2 can sign the form but the senior doctor has to sign within 24 hours.

13

I want to give you an instance where this person had a DNR order from the emergency department. The patient came to my ward. I went to see his wife and wife said, they have put a red form in. My husband has got early dementia, he is independent, he has got pneumonia and they have put a red form on this man. That is wrong. I had to reverse that order. He had antibiotics and the patient went home independent, walked out of the hospital.

22

Some consultants, even if you asked them, on a respiratory ward to see a very frail patient, even if you ask them to do a DNACPR sometimes they get distracted or they won’t do it, it won’t be their priority. Whereas others, you know, others it’s the first thing they think about.

19

Participants rarely talked about considering signing a DNACPR order and deciding not to do so, although they did talk about getting on with treatment as their priority for patients in acute situations.

The FY2 or even the registrar automatically just says, yes the patient is for full escalation, and it is not until they get to the next morning and post take round by the consultant who then might make that (make a DNACPR order) decision.

15

You’ve got a patient who is down, they’ve just found this woman, it’s cardiac arrest. I haven’t got time to mess around for 5 minutes waiting for paperwork (a DNACPR order) because that’s 5 minutes I’ve not started resuscitating. So I’ve got to start and when the paperwork turns up I can stop.

4

This sense of urgency and commitment to deal with an acute illness in a timely way formed the context for the discussion about DNACPR for all clinicians working in ambulance services, emergency departments, acute care and critical care. This contrasted with the longer time perspective of general practice. GPs discussed how patients with long-term conditions should be considered for DNACPR orders. The example was given of patients with COPD who have had numerous hospital admissions and in whom treatment is becoming ineffective. The GPs discussed that they were in a position to raise the issue of DNACPR incrementally each time they saw the patient.

When talking about DNACPR orders in the acute care setting, some participants talked, with some concern, about DNACPR orders that specified resuscitation only if the patient had a shockable heart rhythm.

You’ve actually got to get a monitor on and say, okay let’s see if we can do something.

13

A few participants described how a decision not to give CPR might be made but without the paperwork required by local policy.

I’ve come across certainly unfortunate irregularity particularly in our other hospital where it’s apparent from the notes what is right and what’s been decided, but they don’t have the DNAR form. And the form seems to have some sort of biblical quality, if it’s not on that form it doesn’t really count, which to me I don’t think that is correct. The form is there to make it easier to communicate, but the message remains the same. But it doesn’t necessarily specify whether they filled out their local DNAR form . . . If you’ve written clearly, if you’ve said she had capacity and she clearly stated her wishes with the appropriate consultant and it’s handwritten in the notes that to me is as valid as a piece of paper. But I’ve certainly found that if it’s not on the right form it doesn’t get acknowledged.

23

The health professionals involved with do-not-attempt-cardiopulmonary-resuscitation decisions

Many different types of health professionals are involved in DNACPR decisions or have their clinical practice affected by these decisions. Senior nurses on wards commonly initiate the process of a DNACPR order being put in place.

Like last week we had a lady who came to me and said, I’ve thought about CPR and discussed it with my daughter and I’d like to be made not for resuscitation. So then at the earliest opportunity I discussed it with my consultant and then put it in place.

29

Signing a DNACPR order was frequently described as something undertaken by the admitting team on or soon after admission to hospital.

I work in a trust that is actually quite good at identifying patients where it might not be appropriate for this course of treatment. We don’t very often have, the only emergencies we have come in are primary and they often are already in a pretty poor state when they come in, obviously depending on the down time that they had. So our teams are generally pretty good and would either make a decision pretty immediately they come through the door depending on how the ambulance has delivered them, or very quickly within about 1 to 2 hours of treatment whether they would be appropriate, it would be appropriate to have a DNR in place or not.

11

Junior doctors described observing that a patient was deteriorating, and so, in order to avoid having to start CPR inappropriately, going to find a senior clinician to make the DNACPR decision even though that senior clinician might not know the patient.

Not infrequently that it comes to a situation where you see the patient in front of you going down pretty quickly and you know that they’re going to arrest and you need that red form signing quickly otherwise you’ve got an obligation to start doing chest compressions when you know that it’s really not appropriate. And it’s not fair on the patient to start doing things like that. That’s the situation where you ask a consultant . . . that doesn’t know the patient, but you need to get that form signed otherwise we’re going to have to start the process.

34

Many participants commented that junior doctors would seek advice from the critical care team about a patient if their own senior was not available. If the critical care team advised that the patient would not be a candidate for the intensive therapy unit (ITU), the junior doctor was more confident about making a DNACPR decision.

Paramedics were most commonly described as in the position of undertaking CPR in a situation in which there was uncertainty about its appropriateness.

Within the community sector, when we’re looking at commencing CPR, we are taught within our mandatory training that unless we have actually got the signed sheet in front of us to say, DNAR we are to start CPR . . . So if we don’t have that piece of paper right in front of us at that point we start resuscitation. And that’s where we come from. So the discussions have to be obviously before you get to the point where the patient collapses.

20

Revoking a DNACPR order was described, for example, when a patient went to theatre for an operation.

Patients who have palliative care can have palliative procedures but anaesthesia increases the risk involved. One increases the risk of respiratory arrest . . . [but] recovery from an arrest in theatre is better than in-hospital arrest . . . because you get them in time. So you would stop the order for . . . the duration of the surgery and for the immediate recovery period.

12

There are other procedures that increase risk and which create a dilemma for health professionals if it is not well established that a DNACPR order should be revoked.

And also a lot of the intervention we may do apart from anaesthetics [that] increase the risk of arrest . . . the thing that people are worried about is giving him analgesia because you may do something that causes an arrest and then well are you then morally obliged to try and treat that if he arrests because it is iatrogenic rather than the patient’s disease?

12

General practitioners were identified most commonly as the health professionals who would not make a decision about a DNACPR order when it might be considered appropriate.

The GP should have discussed DNACPR at an early stage. The patient had chronic lung disease and home oxygen and attempts at CPR are likely to have been unsuccessful. Even if a DNACPR decision was in place the patient may still have been admitted to hospital but it may have prevented futile attempts at CPR.

7

However, GP participants indicated that GPs are beginning to undertake DNACPR decisions more routinely.

As a GP we are actually doing this proactively with our integrated plans, but that’s for the over 75s. Certainly if we get a patient where the consultant writes and says Mr Smith is now palliative because there is nothing more we can do his ischemic cardiomyopathy that is a conversation we would try and engage in . . . [we say] it doesn’t mean we withdraw treatment it simply is in this event, if your heart stopped would you want us to try and get it going again, bearing in mind particularly for this gentleman, the chance of success are almost nil.

32

For patients in nursing homes, GPs work with senior nurses to ensure that DNACPR orders are put in place where appropriate.

I myself being a GP . . . I look after an EMI [elderly mental infirm] nursing home . . . We have got things in place where as soon as the resident gets admitted the senior nursing staff . . . and one of the next of kin, whoever is around at the time of admission, [discuss] about DNAR straight away. She [senior nurse] makes the provisional decision, obviously the form needs to be signed by myself because I am the primary GP looking after the nursing home. If I don’t agree then I don’t sign for various reasons, but 99% are signed because they are elderly, they have got other comorbidity conditions, they are frail, they have got dementia, that’s the reason they are in the nursing home. So I’m sure my colleagues try to do the same out in the community. And things are changing.

33

Health professional knowledge of the law and policy related to do-not-attempt-cardiopulmonary-resuscitation

Although sufficiently interested in DNACPR to attend a focus group, participants were not always clear about the DNACPR decision-making process in relation to the law and policy. Although the decision not to undertake CPR is established as a clinical decision, not all participants were clear about this.

I have one question, if somebody says that I want CPR . . . and . . . you are making the decision that no you are not for CPR, this is a medical decision. If somebody insists that they want to have CPR, what is the legal situation?

32

Participants thought that the patient had the right to request CPR.

But the patient still has the right to say, I want you to try, and that’s the point.

32

Field notes taken during a GP discussion (GP notes 01) captured uncertainty within the group. The group were discussing a vignette about a man with an aortic aneurysm that was not operable owing to his poor respiratory function from his COPD. Most participants thought that if they discussed DNACPR with the patient they would be giving that patient a choice about whether or not he wanted to be resuscitated. When challenged by one of the GPs, who stated that it was a clinical decision, most of the group disagreed, emphasising respect for the patient’s autonomy and right to choose: ‘It is his and his wife’s decision’. The participants went on to suggest that if the patient would not agree to a DNACPR order, then, were he to develop an acute illness, it would be up to the paramedics or hospital consultant (if the patient was admitted) to make a decision.

This confusion among health professionals is considered further in Duty of care to the patient, where we consider how our focus group participants considered DNACPR in relation to patient autonomy as a core ethical principle underlying current health-care practice.

Health professionals’ distress in relation to do-not-attempt-cardiopulmonary-resuscitation decisions

During the focus groups participants recounted stories related to DNACPR decisions that had caused distress to themselves or their colleagues. However, when we examined these systematically we found that the distress was related to a problem regarding the processes around the clinical decision rather than about making the clinical decision itself, for example where the family disagreed with the clinical decision, had not been involved in the decision or had made a complaint.

One patient who was seen by the registrar and was DNACPR’d and then the family came in and were absolutely livid . . . the DNR order was retracted . . . I ended up seeing the family when I picked this patient up a day or two later. We had a long discussion and at the end of it they understood what it was all about.

12

Things can often go wrong massively, especially if somebody does go into cardiac arrest. And it’s very stressful and if they make a decision and families aren’t involves or discussed it can often cause a lot of issues with the family.

29

You are talking about very time-pressured situations, you have somebody who is critically unwell. In a lot of these situations they haven’t discussed that he is expected to die or that the heart is going to stop very soon. It should all be about this patient’s best interests but equally it’s not the patient who is having a go at you for the next few months as all the complaints go through. And I can easily see that that’s going to figure in your mind, and in everybody’s mind . . . these discussions should have happened at a calmer time and place.

12

Although our participants did not express discomfort with making a DNACPR decision, they described other clinicians who avoided making DNACPR decisions.

[A doctor treating an] elderly patient with bad medical history fails at that point to make a decision, at the moment, because they know when things are going to happen they probably won’t be there.

19

Our participants had volunteered for the focus groups and so they might have been more comfortable with DNACPR decisions or might have not wanted to admit to discomfort with this aspect of their clinical practice. In Professional integrity and responsibility we consider further the distress of health professionals related to DNACPR and how this relates to challenges to professional integrity.

The timing of a do-not-attempt-cardiopulmonary-resuscitation decision

The timing of the DNACPR decision was a theme that ran throughout all focus group discussions. It was the key issue underlying most of the discussion and debate about the process of making a DNACPR decision. The problem of deciding when to consider a DNACPR order was linked to clinical uncertainty of prognosis. Many participants recognised the uncertainty inherent in predicting future health and how rapidly a patient’s clinical condition can change. This uncertainty about prognosis for individual patients is inherent to all clinical practice and can be difficult to discuss with patients.^132,133 Coping with this fundamental uncertainty can be compounded by other uncertainties such as lack of knowledge of the patient, which we will discuss later. Our data suggest that health professionals vary in terms of their recognition and understanding of clinical uncertainty about individual patients. Some participants expressed this uncertainty very clearly.

It’s almost impossible to define when something will be absolutely futile. The best you can do is based on your experience of what is likely to happen with a patient . . . You could say, . . . think it’s very unlikely that this lady is going to pull through, she’s profoundly acidotic, we think part of her bowels just died she’ll only get worse from here, I can’t see her getting better . . . But to say exactly what’s going to happen, you can’t . . . Say we’re not going to take you to theatre to try and repair your bowel, we’re going to make you comfortable. At that point she will almost inevitably die. But to say that’s what would happen anyway you can’t. She may just be dehydrated. She might be. She might pick up.

12

I suppose the only other thing I would want to add is about the temporal nature of this . . . just because we have said right now that’s how we feel [about CPR not being appropriate] and the family understands, things could change in 12 hours or 24 hours. So my anxiety . . . is that you suddenly find that order is still in place 3 days later and the patient is having their breakfast, you know.

16

Clinical uncertainty was sometimes illustrated by stories of patients whom the clinical team considered to be dying but, against expectations, survived and left hospital.

I’m reminded of a patient I saw where the intensivists didn’t want to resuscitate. This was a lady with dreadful asthma who was literally going to stop breathing, and I persuaded him to in fact ventilate her and she went home, no problem.

33

However, these clear expressions of clinical uncertainty were the exception. Most participants talked about uncertainty but without expressing an understanding of how and why it is inherent to clinical practice.

The level of uncertainty about patient prognosis expressed varied across clinical contexts. Participants from palliative care expressed most certainty about imminent death and the need for a DNACPR order.

From a hospice point of view, it is something that needs to be discussed for every single patient that comes in . . . If the CPR form isn’t filled in by the time it gets to within 12 hours [of admission] the nurses are on top of it . . . putting the form in front of doctors and saying, you haven’t made this decision.

9

However, there were participants working in palliative care who talked about potential uncertainty even when a patient was known to have a life-limiting condition.

If we know they’re end of life, and obviously with a cancer trajectory you do, you know the outcome, these long-term condition patient the trajectory isn’t quite so clear, and this chap may have been dipping because of some treatable cause and he may have come out of that and come home. So making a DNAR decision with somebody with a long-term condition I think is more difficult.

20

Some participants working with acutely ill patients expressed considerable confidence about the likely outcome of clinical scenarios.

You should decide only whether or not if at this moment of time when you’re making that decision, do you think that [based on] the information that you have is that going to be a successful resuscitation or not. If the answer is no this person is not going to be able to be successfully resuscitated then I think you should put that order, and that order can be reviewed if you get new information to say that this is an incorrect [decision] . . . you are most likely to make a right decision rather than a wrong decision.

22

Senior nursing staff have experience to know whether a patient is likely to survive CPR.

13

However, other participants were more cautious about making a DNACPR decision in an acute situation and preferred to wait a few days both to see how the clinical condition of the patient developed and to have time to talk to the patient/family when they were less pressured.

I think there’s a lot of pressure to make very quick decisions, and actually sometimes making the decision after a few days can be helpful because everyone can see where things are moving. And I think it’s very hard to make a decision at the front door [of the hospital] because it’s hard to spend enough time in the middle of an acute emergency to have those discussions in an effective and non-directive way.

32

General practitioners see patients over many years during which time the patient’s deterioration may be very gradual. In their discussions GPs talked about how this gradual deterioration could make it difficult to know at what point a DNACPR decision should be considered: when is the patient close enough to death for it to be appropriate to talk about DNACPR (e.g. for patients with dementia or COPD) (GP notes 02 and GP notes 03). Another example given was cancer: a patient can appear fit and well even though they have cancer that is likely to be life limiting. The GPs discussed the difficulty of raising the issue of DNACPR in this situation (GP Notes 03) as it seems to make the terminal nature of the illness certain whereas the apparent health of the patient gives hope that it is not terminal. This problem of deciding when to initiate discussion about a DNACPR order and to sign it, for patients with chronic illness deteriorating gradually over a long time course, was not discussed by participants working in acute care. They talked about DNACPR decisions as if they would be straightforward for GPs to undertake. They suggested that DNACPR orders should be put in place before an acute illness and that, if they were, their problems related to DNACPR would be solved.

This [signing DNACPR order] needs to be done before the acute phase of this episode happens, let’s do it before it’s happened at all; let’s do it months, years before it’s happened but knowing that at some point this person with this many problems is going to have a cardiac arrest and they’re not going to be any better off.

7

The people that are coming in through our doors are acutely unwell and that’s probably not, for some of our patients, that’s not the right time to making the decision . . . it’s a decision hopefully best made or discussed within the community setting

19

Discussions should be initiated early while the patient still has full mental capacity. Often on admission to hospital they lack mental capacity because they are too ill and then the situation is not reviewed as their condition improves.

11

The certainty expressed here, that it is possible to know when patients are sufficiently unwell for CPR to be inappropriate and to know this before an acute event, runs counter to the expressed understanding of clinical uncertainty discussed at the start of this section. These participants want the best for their patients and in their desire for this they are forgetting that uncertainty is intrinsic to clinical practice. It may be that it is possible to improve on current practice in relation to DNACPR but that clinical uncertainty will remain.

The frequency of review of do-not-attempt-cardiopulmonary-resuscitation decisions

The frequency of review of DNACPR decisions is another way in which participants talked about the timing of DNACPR decisions. A number of triggers for review of a DNACPR decision were identified. Change in the patient’s clinical condition was commonly mentioned. For example, patients deteriorating in hospital might have a DNACPR order put in place.

As a night matron we get called to a lot of poorly patients in the night . . . We come across a lot of patients who are you’ve said, not for ITU but you flick to the front of their notes and no DNACPR has been done, and the patient’s deteriorating in front of you. We will actively encourage the doctors to come and do [the DNACPR form] do it now please.

13

Only a few participants mentioned a patient’s recovery prompting the review of a DNACPR order.

The policy is that when they arrive [in hospital] that it’s up to the registrar to re-evaluate that document in light of the current situation. Because if that patient’s gone away and you know, had a heart transplant, kidney transplant, he’s doing marvellously thank you very much.

13

A few participants did talk about their concern that DNACPR orders were not always reviewed when the patient’s condition improved.

An acute event such as a myocardial infarction would also precipitate review of a DNACPR order.

It depends how long before the admission the DNR was issued and how relevant is it to the current scenario. So you know, a patient goes into hospital with an acute event and therefore you have to reassess everything that day when they come in with an acute event.

34

A change in health setting was discussed as precipitating a DNACPR order or reviewing one already in place, for example admission to hospital, hospice or nursing home. A change of ward was also mentioned as a reason for review, although the participants suggested that this was not always done.

In my hospital if you go from one ward to another you’re supposed to have your DNR order reassessed anyway. Not cancelled but there’s an obligation to reassess it. I suspect this is a piece of guidance that we don’t keep to [laughs]. But I mean it’s there for a good reason that something may have changed as the patients move from one ward to another so it should precipitate a review.

15

There was very little discussion about reviewing DNACPR decisions on discharge from hospital except in the GP groups where there was uncertainty as to whether they or the consultant following up a patient was responsible for making or reviewing a DNACPR decision.

A DNACPR order would be reviewed after it had been over-ridden, for example for a patient having an operation in theatre. A few participants mentioned revoking a DNACPR order if a patient had an acute event that could be treated. The example given by a GP was someone having an anaphylactic reaction which precipitated an arrest (GP notes 01).

In order for the family of the patient to have time to accept that their relative was dying, one participant delayed signing a DNACPR order.

You turn up and the family expects you, do something, do something, sometimes, sometimes not . . . Even though it probably isn’t going to work, the family are desperate for attempts to be made, and then when it doesn’t work at least they can say, everything was done, and it helps them accept and cope with the death in the long term. Even though in reality it’s probably not the best for the patient because even if you got them back the family might then say, oh I got that extra 2 days, where although they weren’t with it they could maybe hear and I could hold their hand and they were still alive even though they weren’t really. And that makes perfect sense. I think that it is really important for the family to be able to say goodbye and accept that everything was done.

6

Most participants emphasised the importance of involving the patient, where possible, and their family in the decision to sign a DNACPR order. In acute clinical situations, participants described the difficult timing of this decision-making for the patient, who might not have mental capacity as a result of being severely ill, and for the family in their distress.

And the relatives as well. Especially in her situation where she’s unarousable. I know it’s distressing for the son and it’s not really the best time to have that conversation. It’s one of the worst times you could have that conversation.

19

Reasons for making a do-not-attempt-cardiopulmonary-resuscitation decision

Participants in the focus groups talked about many reasons for DNACPR decisions. We categorised these into reasons related to the patient, the family, the clinician, the organisation and society more generally.

Reasons for making a do-not-attempt-cardiopulmonary-resuscitation decision: society

Participants talked about the use of resources for health care and the role of DNACPR decisions in preventing resources being diverted to futile CPR and subsequent intensive care if the prognosis was poor.

It also helps us with resources. If resus is full we can move someone into a side room, into one of the cubicles where they can have a bit of dignity and a bit of peace and some privacy. Whereas if we’re not aware of it [DNACPR order] we can essentially hold a bed, block a bed, which is unacceptable. Also do unnecessary and perhaps invasive investigations.

29

Senior clinicians in particular were clear that it was their responsibility to make DNACPR decisions on clinical grounds and that this was part of their professional responsibility within the guidance provided by organisations such as the GMC and the Royal Colleges. These clinicians are therefore acting on behalf of society in making these decisions.

Duty of care to the patient

The concept of a duty of care to an individual patient is enshrined in professional codes such as the GMC’s Good Medical Practice¹³⁴ and The Duties of a Doctor Registered with the General Medical Council.¹³⁵ Participants in all focus groups articulated this sense of duty in a range of ways, often expressed as a concern that they were failing in their duty of care by action or inaction around DNACPR decision-making. An intrinsic element of a health professional’s duty of care is to protect their patient from harm. (In the Reasons for making a do-not-attempt-cardiopulmonary-resuscitation decision we have considered harm to family and health-care staff.) Participants talked about the harm of performing CPR on a patient when there was little chance of the patient surviving or when survival was likely to result only in a postponement of death and significant suffering for the patient.

doing CPR on such a patient I think is really brutal and torturing rather than doing anything good.

33

And it’s then to then find when you’ve already started CPR and broken numerous ribs and everything to then find out this poor patient was meant to die with dignity and not like that. That’s horrible as well.

29

The concept of respect for dignity was also seen as important and there was a concern among participants that in attempting CPR on patients with little chance of survival clinicians were denying their patients a dignified death, and thus failing in their duty to respect their patient’s dignity. Respect for dignity is distinct from avoiding physical or psychological harm. Even when death is inevitable, or has actually occurred, there is still an obligation to treat a person with dignity and respect.

If I was this patient’s consultant I would have considered myself failed in my duty to ensure a dignified death.

22

And for me that is a horribly undignified death if when the patient arrests CPR is . . . And everybody’s view is it was always going to be a non-event really. Yes. With ejection fraction of 10%.

27

This is wrong. We see this every day in all NHS hospitals all over the country, and that is the wrong thing to do. You have denied or we have actually . . . I’m very passionate about this, you have denied a good death to this patient by not putting in the DNR order.

22

However, there was also clearly a concern among participants that initiating a DNACPR order, and discussing a DNACPR decision with a patient, could also cause harm. The potential psychological harm to a patient of being told that they would not be considered for resuscitation was clearly a concern to participants, sometimes supported by their own experience.

. . . when she was on her own and quite breathless, she was asked about DNAR and she was told, there’s no point resuscitating you. So she was told that she would not be resuscitated, but with nobody around . . . But because she was explained all this when she was very ill and on her own late at night, it frightened her and it changed how she behaved forever, but she was alive for another year. And when she did come in on her admission where she died she was dreading coming in here because of that DNACPR. It absolutely changed her outlook on coming into hospital.

25

People can react in unexpected ways, such as ‘the 90-year-old’ who was becoming more frail. She was furious when the GP – who had known her for years – raised DNACPR, as was her son who lived in the same house (GP notes 02).

Participants expressed concern that a DNACPR decision generated a perception in patients that they were being abandoned in their hour of need by those who had a duty to care for them.

The patient wants you to do everything for them, they don’t want to be left to die and we need to reassure them on that, that we’re not going to leave them to die.

13

Yes, it’s last hope. And they want every little last bit. They don’t want the decision. They don’t want to say, yes that’s fine, don’t do it. They feel like they’re signing their death warrant.

12

A further potential harm of making a DNACPR decision is that the decision is incorrect and that a patient will die who would have survived resuscitation. Some participants expressed concern that DNACPR decisions were in fact being made precipitately, and indeed questioned whether or not it was necessarily such a bad thing if futile resuscitation attempts were made occasionally. This argument appeared to be based on an assessment that attempting resuscitation inappropriately is less ethically problematic than failing to attempt to resuscitate someone when one should.

Looking at it from the opposite side and they have a number of patients who had had DNARs made on them which have been made very easily. Made the decision just based on what standard of life is. And so although I’m sure we all see where we definitely want them to make DNACPR, there is the opposing side where there’s a number of organisations that are quite concerned at the ease in which DNARs are being made as well.

13

A key problem for clinicians when trying to fulfil their duty of care towards a patient in these situations is the uncertainty that permeates the decision-making process. As discussed earlier, this uncertainty is intrinsic to clinical practice but is compounded by, for example, information not being available. In the emergency or acute care situation, information on prognosis, previous quality of life and likely response to initial treatment may all be difficult to assess, making a decision about what is in the best interests of the patient with regard to CPR much more difficult.

Would we get him back to what he was like before he came in? Because you could possibly do that if he responded to the antibiotic and . . . Well he’s got carers three times a day and a lot of previous medical history of comorbidity, but he might have . . . we’ve just presumed he’s a poorly chap, but he might have quite a nice life generally.

13

Respect for patient autonomy

This is a core ethical principle underpinning the patient–doctor relationship as understood in current health-care practice, with its emphasis on shared decision-making and patient-centred care. It is also the principle that underpins the legal doctrine of informed consent and the realisation of Article 8 of the European Convention on Human Rights: the right to respect for private and family life.¹³⁶ It is, therefore, not surprising that respect for autonomy and what it means in the context of DNACPR decision-making was a dominant theme in our focus group discussions. One conception of autonomy is that a person, in this case a patient, should be free to make decisions about his or her life, including decisions about his or her medical treatment, without interference from others; patient choice should be respected. However, in order to make choices, a patient needs to have the relevant information to inform that choice and a key concern for many participants was that health professionals were not giving patients that relevant information. The lack of open and honest discussion about CPR by clinicians, at all stages of the patient journey, was a recurring theme in our data.

I think you know, in this case for this gentleman he’s been unwell for a long time, and the ideal opportunity is to talk to this gentleman early on in his diagnosis and get some ideas and get him and his family discussing end of life and what his wishes are. So that by the time he comes into hospital extremely poorly . . . it hasn’t been fair on this gentleman or his family that it hasn’t been discussed in hospital, but actually the conversation hasn’t happened before he’s even got there . . . it should have happened much earlier.

28

And what should have been done in this case and may even have been done, is that in the nursing home somebody would have discussed with her at leisure, before a crisis occurred as to what she wanted if she was unable to make a decision.

33

You would imagine that the consultant that’s told him that, or the team that have told him that would have had some forward planning, or I would have thought they would have some forward planning in terms of what to do if and when it did leak or rupture. And if that’s not the case then that’s certainly a conversation that needs to be had.

17

One person we’re not really involving in here is the patient themselves. So I think what we often do, and I’m sure we’re all probably guilty of it, is not actually asking the patient. So we ask the family, we talk to our team, but it’s very difficult to actually sit down with a patient and say, how would you feel if your heart stopped beating, what would you like me to do. And often if you have a frank discussion with the patients they will say, do not . . . do not ever try CPR, I do not want that to happen. So I think the emphasis on . . . so we need to re-educated I guess and actually speak to the patients because a lot of elderly patients find the thought of being resuscitated horrifying.

18

However, the occurrence of a discussion with the patient is not in itself sufficient to enable the patient to make an autonomous decision. The quality of the information provided and the ability of the patient to understand the information and to set it in the context of his or her own experience and values is also required. Several participants identified significant challenges in providing appropriate information in a way that patients would understand, particularly in an acute medical setting.

Yes, you need to have the rapport with the patient and that’s the problem we have. We meet the patient 5 to 10 minutes, they are scared, they are sitting in resus, there’s lots going around them, they might not comprehend everything that’s going on, and then make such a hard decision is a very difficult thing to do I think.

18

So now we have to discuss all the DNACPR decisions with the patient, but it’s very difficult to discuss with the patient, he may not have capacity. He’s barely able to talk and he’s quite distressed, he might actually be much more interested in whether his daughter’s coming in or whether the dog’s been fed or, you know, it can be . . .

8

The need to start the conversation about resuscitation and DNACPR much earlier in the patient journey was raised in several of the focus groups. Participants suggested that GPs may be better placed to discuss this with the patient because they are likely to know the patient better and be able to initiate the discussion at a time and in a setting in which the patient can reflect on the information and make a considered decision. Advance directives were seen as helpful to clinical decision-making in the acute and emergency situation.

I guess that’s part of the point of advance care planning in people who have irreversible conditions, whatever that condition may be and at whatever stage that may be at. Because then you’ve got in black and white what that person wants. Having power of attorney is all well and good but it’s another person’s opinion of what they think that person would want. The advanced care planning is to try and get the individual’s actual wishes.

22

I think in an ideal world if you are having routine follow-up for comorbidities in the community and you’re seeing your GP, that is really in an ideal world where these decisions should be made, at a less vulnerable time, and then that document and that order follows you rather than it being made at vulnerable times.

18

I would probably go one step further than that and say that she lives in a nursing home and that should probably have been addressed in the nursing home, and she should have come to hospital with that decision in place where there was time to have made that decision before it became an acute situation.

16

However, even in these more controlled situations it can be difficult to ensure that decisions are appropriately informed and to take account of the uncertainty of future events. One GP described having such a conversation with an older patient.

. . . although I have said to ’a 92-year-old’, if you collapse in front of me what would you like me to do? To which his reply was, if I have a heart attack don’t bother, but if it’s my appendix I want to be resuscitated.

20

This quote illustrates the importance of seeing CPR as one element in a holistic view of future care: if my heart stops because of a heart attack I would not want resuscitation but I would want it if there was a remedial cause for my cardiac arrest. Advance directives allow for a more nuanced approach to these decisions than a DNACPR form. However, any meaningful discussion with a patient about resuscitation, whether in hospital or in the community, requires time and several participants identified pressure on time as a reason why these conversations were not had or properly recorded.

There’s not the time in the day for GPs to go and have that discussion with people . . . I mean at 88 you don’t know necessarily that she’s in her last year of life, you know, that’s very difficult.

33

Yes. Time constraints, you’re called away and then you forget to go back and document what you have discussed. That does happen. I have to remind myself often enough to go back and record something.

17

Some participants suggested that DNACPR discussions should become a routine part of all hospital admissions to normalise the process and thus reduce patient and clinician anxiety about the subject and again remove it from the acute situation. However, such approaches raise questions about the level of information provided to patients in this context and to what extent this is true respect for patient autonomy.

. . . it should be just on the admission form, you know, name, date birth, do you wish to be resuscitated, yes or no. I think a bit more matter-of-fact rather than aiming this big decision, you know.

18

Adding to that, as a medical student I’ve seen a few departments in district general hospitals they had the medical registrar walked around and actually, like we ask for allergies, it became normal. I found it a bit frightening at the beginning, but then looking back there were a surprising amount of people saying actually, oh I would never want this.

18

Prior discussions and advance directives may help to facilitate autonomous decision-making but respect for autonomy also means respecting a person’s right to change his or her mind. Some participants reflected on the difficulties of predicting what one would want to happen at a future date, giving examples of patients who had changed their mind about an advance directive when faced with the reality of a life-threatening situation.

. . . having said that I’ve been an ITU sister I’ve seen people come in with living wills and then ripped them up and said I’ve changed my mind, oh I feel different now.

18

Accepting the difficulties of facilitating autonomous decision-making there was general consensus among participants in our focus groups that clinicians have a responsibility to have these discussions with patients, reflecting the strong sense that patients should be involved in and informed about all aspects of their care. However, respect for patient autonomy also implies respect for decisions that a patient makes about his or her care and this aspect of autonomy raised significant ethical tensions for participants, namely to what extent does a patient have a right to request CPR (or refuse a DNACPR order) when the clinician considers CPR futile or likely to cause harm to the patient. Some participants were very clear that this was a clinical decision and that the purpose of discussion with the patient or his or her family was to explain the reasons for the decision or simply to inform them of the decision.

And it’s like any other treatment, patients and families can request but they can’t insist and at the end of the day a DNACPR order is our responsibility and our decision. They can express an opinion but it’s ultimately our responsibility.

22

. . . because my understanding is legally that patient cannot demand treatment. They have every right to refuse treatment but my understanding is that actually legally a doctor does not need a patient’s permission to sign a DNACPR. Good practice is to have a discussion with that patient, but ultimately that doctor can say, I understand that but actually I still want to sign it.

21

I think the other thing is that, I don’t know what your approach is, but in some ways you’re not asking a family member to make a decision, you’re saying that this is my professional opinion that this futile or it isn’t, this inhumane it isn’t, and do you understand that I am communicating that to you, and that’s what we’re intending to do. I mean I think the wrong approach is to say, well here’s the scenario, what you want us to do.

16

Others were less sure of both the legal and the moral weight of such a position. Some participants thought that if a patient had capacity and requested resuscitation then a clinician must respect their wishes. One nurse participant referred to a dictatorial attitude exhibited by some doctors in discussing DNACPR with patients and relatives.

The difficulty then comes with if the patient says, I want to be resuscitated, but you feel it’s inappropriate, then that’s difficult and they have to be resuscitated providing they’ve got capacity to make that decision.

18

Participants struggled with their desire to take into account both their patient’s wishes and values and their clinical obligation to minimise harm to their patient. This dilemma highlights the ethical challenge of interpreting and defining the limits of patient autonomy. Does respect for autonomy require that the patient be informed, that the patient is involved/consulted in the decision or that the patient should consent to the decision?

Role of the family

Many patients for whom decisions about resuscitation are made will lack capacity and may be too ill to participate in a discussion with clinicians. In the UK the Mental Capacity Act 2005¹¹⁶ and Adults with Incapacity (Scotland) Act¹³⁷ provide a clear legal framework for decision-making in these situations. This framework, echoed by professional guidance such as the GMC’s guidance on end-of-life decision-making,¹²⁶ emphasises the importance of consulting families about the wishes and values of the patient to inform a best-interest decision. GMC guidance also advises that families should be supported in these situations.

It is important that you and other members of the health care team acknowledge the role and responsibilities of people close to the patient. You should make sure, as far as possible, that their needs for support are met and their feelings respected, although the focus of care must remain on the patient.¹²⁶

Focus group participants recognised the importance of involving family members in decisions about resuscitation, although most talked about involving and informing family members to prevent their distress rather than to seek their views about the patient’s wishes. Even when clinicians were clear about the reason for seeking the views of the family, they found that this was a difficult concept for families to take on board. There was some uncertainty among participants about how to respond when a family member had a lasting power of attorney.

. . . it’s very disturbing and things can often go wrong massively, especially if somebody does go into cardiac arrest. And it’s very stressful and if they make a decision and families aren’t involved or discussed it can often cause a lot of issues with the family.

29

My difficulty I have in these discussions is that when you explain it to the relatives they often tell you what they want and not what they think the patient wants. And trying to distinguish those two in a relative’s head when you’re explaining to them that a family member is very ill and probably going to die, is a very difficult thing to do.

33

This one’s slightly different from a lot of the ones that we see in health care because the health and welfare has lasting power of attorney in there, and that puts an extra, probably a little . . . not a spanner but a little bit more of a confusion on people’s heads about what can we and can’t we do. It’s just a case of all you’ve got to do is communicate that to the son and ask his consent . . .

10

Participants described facing difficult decisions when family members were expecting or demanding resuscitation for the patient when the clinician thought it was either futile or not in the patient’s interest. These situations were often distressing for both clinicians and the patient’s family.

I’ve had lots of examples recently where the patient themselves have wanted the DNAR to be put in place and then the families have put pressure on medical staff and it’s then very difficult isn’t it for medical staff to then say, what do we do, do we go with the family?

13

But then some of the younger patients with cerebral palsy, it can be very difficult, you know for some people who had severe contractures all their lives and bed bound, no cognition and you discuss the CPR and yes we definitely want everything, and you think damn you! . . .

12

We had a similar scenario on neuro quite recently relatively young but had hypoxic brain injuries and we twice had this conversation with [the family]. I was constantly fighting, his sats was below 50 at one point, trying to rectify it, and the family were adamant he was not to be DNAR. And in the end I just had to keep battling away through the night shift to try and keep this man remotely alive. They weren’t actually happy to make that decision; they were very angry and very distressed.

12

Participants described different responses to family pressure to perform CPR or refusal to allow a DNACPR order to be initiated. Some were very clear that the medical interests of their patient were paramount, while others appeared to take into account the impact on family members of seeing their loved ones apparently being left to die.

I have put my foot down and said, this is . . . like you have a duty of care towards your mother, I have a duty of care towards my patient, and I will not do what is wrong for . . . but I offered them a second opinion, that is always, you always . . . and what you must remember is that it’s not us against them, it’s trying to arrive at a consensus.

22

In our case what we did, we . . . we did make him for resus, that was eventually done. Although we thought it’s not going to be successful but the pressure of the family and the patient, we said, OK, that’s fine.

22

. . . even though it probably isn’t going to work, the family are desperate for attempts to be made, and then when it doesn’t work at least they can say, everything was done, and it helps them accept and cope with the death in the long term. Even though in reality it’s probably not the best for the patient because even if you got them back the family might then say, oh I got that extra 2 days, where although they weren’t with it they could maybe hear and I could hold their hand and they were still alive even though they weren’t really sort of thing. And that’s perfect, makes perfect sense, and I think that’s really important for the family to be able to say goodbye and accept that everything was done.

6

Paramedics described sometimes being faced with a relative who was unaware of a patient’s previous advance refusal of CPR. They faced an ethical tension between respecting the patient’s previous wishes, and their interests, and trying to respond sensitively to a family member in extreme distress.

The problem is when I arrived the family didn’t know anything about this when I talked to them and they wanted me to resuscitate. So, on my own, so I started . . . During this time I sort of chatted to the family a little bit more as well . . . in effect we did everything we could and we decided at a certain time to stop. The family were happy, it was like the final thing had been done, you know, they’d seen we’d done our best and we stopped and everybody was sort of happy with that.

26

Patient-centred care

The concept of patient-centred care is central to good clinical practice. This concept recognises the patient as a unique person with his or her own values, life narrative, relationships and physical response to disease. The ethical principle of respect for persons requires recognition of the whole person and the context of an individual case or situation informs and shapes ethical decision-making. Many participants expressed concerns about the tension between DNACPR decision-making and patient-centred care, which manifested in a number of ways.

There was concern that initial decisions for DNACPR were made in the acute situation without sufficient information about the patient’s personal and medical history, leading to assumptions being made based on generalities such as age and comorbidity. Examples were given of these assumptions leading to inappropriate DNACPR decisions.

I would like, when people say, oh this person has got dementia, what degree of dementia because I want to give you an instance where this person had a DNR order put in from emergency department, the patient came to my ward, wife. I went to see his wife and wife said, they have put a red form in. My husband has got early dementia, he is independent, he has got pneumonia and they have put a red form on this man. That is wrong. I had to not only reverse that order, had antibiotics and the patient went home independent, walked out of the hospital.

22

Everybody’s individual aren’t they? . . . a ‘96-year-old lady’ the other day who was telling me she’s just done two parachute jumps for her birthday, and she is . . . and I just . . . I admire that woman so much and I’m thinking . . . basically you know, everybody is individual.

13

And I think there is a duty on the part of the clinical team to ensure that they know who this woman is in the context of her life, and a life that looks like it’s drawing to a close. So to isolate a decision such as if your heart stops beating we will not restart it because it won’t work, has to be done in the context of who she is rather than a cold clinical legal conversation with a patient.

24

A further concern was that resuscitation, and to a certain extent DNACPR decisions, were becoming divorced from a holistic assessment of a patient’s care needs. This could mean that the focus on a DNACPR decision distracted clinicians from the more fundamental need to focus on escalation of care to prevent the need for CPR, or on palliative care when resuscitation was not appropriate. Alternatively, decisions to resuscitate (or failure to make a decision) could be made without consideration of whether other interventions necessary as a consequence of CPR (such as ICU care) would be appropriate or available for the patient.

. . . because we want them to escalate treatment before they make the decision not to resuscitate. Treatment escalation is a priority; making sure this patient doesn’t die, not, what you are going to do when they do die.

13

But we wouldn’t know that on scene, I mean we’ve got abdominal pain and vomiting and that could be anything, and that could be treatable as much as it might not be treatable . . . There’s a lot going on there that we need to do that actually I think actually negates the decision on, well she’s very sick and she’s going to die in the next hour or two, let’s not bother going anywhere, let’s just stay. We don’t get to make that decision.

4

And it wouldn’t . . . I mean just thinking about our own kind of upper GI practice, in this institution it would not be that unusual that a person who had been stented with this diagnosis would not have been referred at that point to palliative care suddenly things move on very quickly and you know, you can see everybody stumbling over themselves around decision-making . . . or the family would say, well we’re not having that form in the notes and, oh okay, well the form is taken away . . . You can see how this all happens, can’t you?

24

And actually what usually happens, the registrars I think will back me up, is that you end up getting them to a point where you’ve got cardiac output back and you’re bagging them, and then you get ITU come down and they say, well I’m not taking the patient, and you’re sitting there going, well now what do I do, you know, and that is the worst position to be in. so if you make it quite . . . so what I often do as a sort of step that says effectively I’m not doing CPR but I’m not putting it that way, I say, it wouldn’t be appropriate with your pulmonary fibrosis to put you on a ventilator.

32

Consequences of a do-not-attempt-cardiopulmonary-resuscitation order on overall patient care

There was real concern expressed by many participants that a DNACPR decision had a range of adverse consequences in terms of patient care. Rather than seeing DNACPR as one decision on a continuum of decisions required to provide optimum care for a patient, the DNACPR decision seemed to categorise a patient as ‘not for active treatment’. The DNACPR label coloured all other interactions between health-care professionals and the patient.

when the nurses hand over to each other they will say, this is patient Joe Bloggs who is a DNA, and then they will talk about him. But the first thing, the very first thing they mention about the patient in their handover, after they have said the name of the patient is the DNA. And I think that’s so wrong in the order of priorities of how a patient should be seen.

25

But you’re assuming then that that patient’s DNARd and not for treatment. A lot of them are DNAR and for active treatment until the point that their heart stops, so you shouldn’t remove . . . necessarily be moving them [into a side room to allow to die].

29

No I think there’s . . . certainly within medical consultant colleagues some of those feel uncomfortable with the idea of having to write a ceiling of care document on a patient because they fear that if they write DNAR nothing will happen to the patient. And a lot of work has had to go into supporting them to make those decisions, but also to I suppose empower the staff to realise that it doesn’t mean palliative care.

15

Most of the trust do a ceiling a care form as a separate from DNAR but in most of the situations within most of the trusts that I’ve worked in the staff, the nursing staff will take a DNACPR as a no-go area and they will not even do the doables, they will probably try to de-escalate from the observations chart, which is from 1-hourly maybe to 4-hourly, or they will pressurise the junior doctors when the consultant has already left to do an end of life discussion with the patients and the consultant was astonished to see the next morning that this was not what was said and why didn’t you treat the treatable cause, which is probably pneumonia or, or maybe even a palliative procedure which could have been done.

32

The data revealed the importance of seeing DNACPR decision-making as one element in a process of assessment and management of the overall care of an individual patient and highlighted a potential danger of treating DNACPR decisions as fundamentally different from other decisions about patient care. Identifying DNACPR decisions as an integral part of holistic patient care means that they are less likely to be avoided or forgotten. However, it also requires recognition of the complexity of these decisions and the need for clinical judgement.

Professional integrity and responsibility

Participants talked about the importance of clinicians taking responsibility for making resuscitation decisions and linking this to their professional integrity. The challenges to professional integrity and the emotional impact of these challenges were recognised, but participants suggested that this was not a reason for avoiding difficult decisions.

Again it’s getting the medics . . . to realise that the buck stops with us regardless of whatever or whoever . . .

13

. . . when I see that someone is just being unreasonable, after having explained everything and is not acting in the best interest of that patient then it becomes my duty if I have not done my duty to make sure that I don’t do the wrong thing for my patient. I will not be browbeaten or basically cowed down to do what is wrong.

22

Professional responsibility included the responsibility for having an honest discussion with the patient and/or their family. Several participants referred to the difficulty of having these conversations and the variation in practice depending on the particular views or the communication skills of the doctors involved. The responsibility to acknowledge the need for training in making these decisions and having the associated conversations was also highlighted.

From my point of view, having sat on many occasions with families and doctors just quietly because I don’t have to do the talking in those situations, it’s not my responsibility, it’s not my decision. And it is very much about the approach of the team, and what the consultants specifically always say. What I think happens is that the doctors who don’t want to have that discussion with the family, they’re the ones that don’t make the decisions about do not resuscitate, they leave it for somebody else to do. And they’re generally the ones that are more uncomfortable with having that discussion. It’s not because they can’t see that it’s the right thing to do, but would prefer to avoid having that discussion. That’s mainly what I’ve observed.

16

But I also would say there is something about individuals recognising an inadequacy in their ability and however much training can we put on actually if those individuals are avoiding having those conversations because they’re difficult. And something about that person identifying their own learning needs as well, but that’s hard.

25

Several participants commented on the importance of having an experienced clinician making these decisions and having the conversations with patients and relatives. There were a range of views on whether or not this doctor needed to be senior and in this context a distinction emerged between ‘informing discussions’ and ‘consenting discussions’. Participants appeared to think that for preliminary discussions, particularly in the community or for patients with established life-limiting diseases, the clinician who had an established professional relationship with the patient and knew them well would be best placed to discuss this with the patient. However, the actual decision as to whether or not an individual patient should have a DNACPR order (except in cases where there was an advance directive refusing CPR) was seen as a clinical decision and one that should, therefore, be made by the doctor responsible for the patient’s care. Participants expressed some concern that this responsibility was not always accepted, with some consultants avoiding making the decision or leaving it to juniors, and GPs not initiating a conversation at an appropriate point in a patient’s illness trajectory.

At the end of the day it’s my responsibility to the patient to relay what information I have and if I think that patient is peri-arrest I have a duty of care to that patient to escalate that to my seniors so that somebody with more experience can help me make a decision.

22

I think from my point of view we have patients who are really quite sick, similar sort of . . . you know, we’ve done investigations and we know they’re not going to be for intensive care. So we . . . I try and highlight those patients to the consultant to say, do we need to think about a resus status here. And depending which consultant is on the ward will depend on what type of discussion we have about that. And I think one of the consultants is saying, well I’m not going to go and have this discussion with the patient, I don’t think it’s appropriate that we have to tell all the patients and all their relatives that this is what we’re doing. So he won’t go and discuss it . . . I’ve had a discussion with him and he told me it’s a medical decision; and if that’s the case then we do need to, you know, if appropriate talk to the patient in an appropriate manner.

29

I think the interesting thing with this scenario . . . these types of scenarios are that if we’re called to nursing homes where is a registered health, or should be, a registered health-care professional on the scene, very often they’re calling us for us to be the decision-makers when actually they should have stepped up to the plate and said actually, OK, fine, our patient has now died, I’ve got the DNACPR, I’ve got the diagnosis, etc., that’s it, OK fine, the patient’s died, we’ll do the rest. But we do get called quite a bit purely to shove the problem to us, for us to sort out.

26

Participants referred to an ethical tension that had been created inadvertently by policies and processes to improve DNACPR decision-making. Rigid policies and requirements for documentation sometimes placed staff in a position of resuscitating a patient when their professional and clinical instinct was not to do so.

I think . . . I think the issues about DNACPR I think to some extent we’ve taken steps backwards because of . . . because of the documentation being much more in some ways clear cut, but also perhaps taking out some of the subtleties. I think there would have been in the past instances where that sort of scenario the staff would have felt comfortable in interpreting what was written without the DNACPR document. And I think people don’t feel now trust in the organisation, so if they’re not covered by a form, they feel obliged to go ahead with something.

28

I mean this is a little bit clearer, we’ve been called, there is something wrong with this patient, we need to deal with . . . the jobs that I dread going to are nursing homes or residential homes. You know you’re going to a cardiac arrest, you turn up and they’ve found them in the morning or what have you, or they’ve just come across them and it’s a case of, right OK, they’ve quite clearly passed away, they’ve been not known necessarily to have been ill, you know, life has taken its natural course. But because we’ve been called we don’t know necessarily when they were last seen, they are warm or what have you, you end up having to do a 20-minute protocol on a patient. And you just think, well you know, the family don’t necessarily want this but there’s no formal DNR in place, no one’s discussed what they need to do with these patients if they are found to have passed away during the night or found to not be breathing. It’s difficult, those are the jobs that I dread.

4

Participants viewed acting with professional integrity and fulfilling their duty of care to their patients in the face of clinical uncertainty and perceived or actual external constraints as major challenges. These challenges generated emotions of anxiety, self-doubt and fear of censure through complaints, professional regulatory bodies or legal proceedings. Participants described a range of fears and anxieties about DNACPR decisions including fear of getting it wrong (do we have enough information to make a decision); fear of harming the patient and failing to do their best for the patient by either upsetting them with conversation or performing inappropriate CPR; fear of upsetting the family; and fear of organisational censure. One striking anxiety or fear that participants identified, which often precipitated action that conflicted with their professional duty of care, was that of litigation or complaints by patients’ families. Few examples of experience of complaints or litigation were given but the threat of it was cited in several focus groups as a reason for not initiating a DNACPR order or resuscitating when it was considered clinically inappropriate.

I think sometimes when families are really angry, I think sometimes doctors back off a little bit because it’s frightening. And so it may be that actually . . . because actually it’s like, oh it’s easier just to say, actually okay we don’t want them to get angry, we don’t want them to make a complaint, rather than actually kind of like tackling the difficult conversations. Because actually at that point there should have been a really good in depth explanation as to why it wasn’t appropriate to resuscitate her and what her wishes were.

24

I think they are taught. I think there is a bit of a fear around making DNACPR decisions and it can be a very grey area, and I think that . . . and it’s affected greatly by the media because what we’ll see is, you know, like a headline in the newspaper, you know, death warrant or death sentence applied to the patient. And you see just this sort of a bit of a biased report, but then, you know, when people are sort of looking to be challenged possibly legally or, you know, for there to be litigation it makes people then afraid of making those decisions. Whereas actually what we need to is keep our focus on the patient and what’s right for the patient at that time.

19

I think the main thing is what is people’s expectations of resuscitation and a fear of medico-legal liability and . . . because I think people perceive it as like oh, would somebody look at me and say I have not done anything. So again that comes with the confidence and that’s where you need kind of more senior people in those types of decision-making.

20

Societal values/responsibilities

Health professionals work within a legal and professional ethical framework that is determined by societal norms, and in recent years there has been a shift in normative emphasis towards respect for patient autonomy and the obligation to involve patients in all decisions about their care. As noted above, autonomous decision-making requires appropriate and accurate information on which to make a decision. A strong theme emerging from our data was health professionals’ concern that patients and their families had received misleading information about CPR from media portrayals, and that societal expectations of the extent to which medicine can save and prolong life create difficulties in discussions around DNACPR.

It’s difficult for patients as well because I mean if you have this conversation about resus with people in clinic they know that Holby City all got resuscitated, and they will survive. Of course that’s what they want.

23

I’m sure you’ve read it – where they’re looking at televised and dramatised where-76% of patients survive. So that tells the public that resuscitation works where we all know it doesn’t in the vast majority of people. But yet, it’s offered to everybody.

10

And actually relatives do get anxious because they think with CPR means it’s a magic thing which just brings everyone back to life, because that’s what they’re shown on Holby City, you know, they do two compressions and then they’re back and normal and functioning and walking and running, you know.

19

The use of society’s health-care resources for the resuscitation of those nearing the end of their life is another concern raised by our participants, as discussed earlier.