Despite a national drive to support quality in end-of-life (EoL) care, many dying patients are not identified in a timely manner, with communication and care at this stage often being poor (Healthcare Quality Improvement Partnership, 2019), which may necessitate unnecessary interventions and cause distress (Parliamentary and Health Service Ombudsman, 2015). Do not attempt cardiopulmonary resuscitation (DNACPR) decisions can help promote a dignified death for patients who are thought to be dying (General Medical Council (GMC), 2010). The success rate of cardiopulmonary resuscitation (CPR) on patients with deteriorating progressive illnesses is poor, at typically less than 19% (Nolan et al, 2015), and even less, at 5–10% outside of the acute setting (British Medical Association (BMA) et al, 2016).
While the benefits of advance care planning (ACP) are widely recognised (Dixon et al, 2019), many patients are not given the opportunity to express their EoL choices in advance (The Choice in End of Life Care Programme Board, 2015; Healthcare Quality Improvement Partnership, 2019). The Healthcare Quality Improvement Partnership (2019) identified that, although a DNACPR was in place for 97% of patients who died, a DNACPR discussion with the patient was recorded in only 42% of cases. While some patients are reported to postpone ACP discussions until they are closer to death (Pollock and Wilson, 2015), many individuals want to discuss their future care preferences (The Choice in End of Life Care Programme Board, 2015; Macmillan Cancer Support, 2018) at a time when they feel well enough to process the information but also not too close to being given their initial diagnosis (Cox et al, 2007). DNACPR discussions are often delayed, taking place, on average, at 31 hours prior to death (Royal College of Physicians and Marie Curie Cancer Care, 2014).
Patients who are likely to die soon should have this communicated to them sensitively (Healthcare Quality Improvement Partnership, 2019), respecting their cultural preferences and including families in discussions as appropriate. Patients should be included in decisions regarding their care, and discussions should be timed appropriately to allow for a period for thought, with ongoing opportunities to absorb information and express individual wishes (Leadership Alliance for the Care of Dying People, 2014; National Palliative and End of Life Care Partnership, 2015).
Despite some conflicts within the literature (Pollock and Wilson, 2015), significant evidence generally suggests that patients prefer earlier discussions relating to EoL decisions (Heyland et al, 2006; Cox et al, 2007; Hall et al, 2019). Yet health professionals may, for various reasons, delay such conversations (Miller and Dorman, 2014; Macmillan Cancer Support, 2018). Often, difficult conversations are delayed until they are initiated by patients (Almack et al, 2012; Macmillan Cancer Support, 2018), with health professionals reporting this approach as being easier to manage (Low et al, 2014). However, the initiation of DNACPR discussions by patients is reported to be quite rare (Low et al, 2014; Pollock and Wilson, 2015), and the notion that health professionals feel more comfortable discussing EoL conversations when patients have given cues (Almack, 2012; Granek et al, 2013) contradicts the view that it is the responsibility of the former to initiate conversations about future care (Almack et al, 2012; Royal College of Physicians, 2018).
There is little guidance or training on how health professionals should initiate and manage these conversations (Holland et al, 2013; Low et al, 2014; Kazmierski and King, 2015; Pollock and Wilson, 2015), and research is lacking on how EoL conversations are actually undertaken by health professionals. Saevareid and Balandin (2011) noted that the information given by different physicians varied in terms of the quality and amount. Cox et al (2007) found that oncology staff first acknowledged the incurable nature of the patient's illness, which later helped patients understand disease progression leading to death and allowed them to better understand the meaning of a DNACPR decision. Miller and Dorman (2014) found that DNACPR discussions were had by describing to patients the notion of not being ‘given up on’ and that patients would continue to receive care.
Significant barriers have been reported to these conversations, often due to a fear of causing patient distress (Holland et al, 2013; Low et al, 2014; Miller and Dorman, 2014; Kazmierski and King, 2015), lack of confidence (Royal College of Physicians, 2018) and personal discomfort with discussing death and dying (Granek et al, 2013; Royal College of Physicians, 2018), among other reasons.
The aims of this research were to (1) explore palliative care clinical nurse specialists' (CNSs) experiences of having DNACPR conversations in the community and (2) to determine what language is used when DNACPR decisions are communicated with patients and families.
Methodology
Design
The epistemology (theory of knowledge) within the context of this study considers the social constructivist view that individuals search for reasoning related to their views, behaviours and experiences, while considering the impact of these on interpretation, conclusions or theories reached (Creswell, 2014). This qualitative study uses autoethnography enabling reflection of the researcher's own subjective experiences, while also valuing and comparing others' experiences. The interviews provided rich data, exposing unique experiences and eliciting information that may otherwise have been difficult to obtain (Chang, 2016).
Sampling
Participants were selected using convenience and purposive sampling. An invitation to participate in the study was sent by email to the community CNS team within one organisation (with 15 CNSs), and the proposed research was discussed at the CNS team meeting. With an emphasis on anonymity and the potential benefits of the study, a sample of six participants readily volunteered (consented) to take part. The participants had 5–20 years of experience as CNSs.
Ethics considerations
Ethics approval was granted by the Birmingham City University ethics committee and the governance committee of the organisation where the research was carried out. Potential participants were provided with a participant information sheet, and informed consent was obtained. Since there was a small risk that psychological harm could occur, attempts were made to minimise these, for example, by excluding individuals who had been bereaved in the last year. Patient identity was kept confidential.
Data collection and analysis
Six semi-structured interviews were conducted with experienced palliative care CNSs. Each CNS was interviewed in private consultation rooms away from clinical areas, with interviews ranging from 22 to 49 minutes. The interviews were audio recorded, transcribed and analysed using the inductive thematic analysis approach suggested by Braun and Clarke (2006).
Rigour
In this article, valuable insights from participants' recollections of their experiences and conversations are provided using direct verbatim quotes, and participants' own words are used in some of the sub-theme titles (Smith and Firth, 2011). The author's own experiences (recorded in a reflective diary) are reflected as well, enhancing the credibility of the research (Darawsheh and Stanley, 2014). Each recording was reviewed on the same day. Notes were made and codes were generated, initially by manual colour coding, followed by electronically organising them using NVivo11; these were later developed into themes and sub-themes. Themes were repeatedly refined through the reading of further transcripts referring back to the reflective diary. Participants were allowed to review the interview transcripts and the emerging themes during data analysis.
Findings
Four main themes emerged from the interviews: (1) learning through experiences, (2) feeling responsible to get it right, (3) initiating the conversation and (4) barriers to DNACPR discussions.
Theme 1: Learning through experiences
Sub-theme 1a: Influences on practice
Participants reported that factors such as observation of other health professionals communicating DNACPR decisions with patients throughout their career had influenced their practice. Four participants referred to their technique improving with experience. One participant referred to the notion of ‘trial and error’ when reflecting on their practice while another participant felt that they had ‘stumbled through it’ earlier in their career. The author reflected on how their own practice developed through sharing ideas with colleagues about how DNACPR decisions might be better communicated (to patients).
Sub-theme 1b: Picking up the pieces
Some CNSs reported having to explain existing DNACPR decisions to patients who were previously unaware. The study found that, often, CNSs must manage the aftermath of patients and relatives discovering EoL issues and DNACPR decisions insensitively or inappropriately. Four of the participants in the study referred to occasions where patients (and families) had discovered by chance that a DNACPR decision had already been made.
‘… they had been sent home from hospital on discharge with a sealed envelope, they'd been told to give that sealed envelope to their GP and their GP has then informed them that … there's a do not resuscitate form in there …’
The author's own reflections were similar to those of participants; for example, a patient who was left feeling like he had been ‘thrown on the scrap heap’ following a DNACPR discussion and another patient who overheard health professionals discussing the DNACPR decision before having discussed it with him.
Sub-theme 1c: Making it clear
Two participants referred to the importance of keeping the conversation simple to avoid confusion and using plain and simple language. Two participants said they found it particularly challenging to balance being sensitive while also giving clear information, without distressing or confusing patients:
‘… am I explaining this in a clear way, am I using the rights words, so I'm not confusing people? I'm not wanting to scare people, I'm not wanting to worry them but it's about trying to reassure people that we will still look after them, that we will treat them, even if they've got a DNR in place …’
Theme 2: Feeling responsible to get it right
Sub-theme 2a: Getting it right
All participants conveyed a sense of professional responsibility to communicate DNACPR decisions sensitively to promote a dignified death and maintain a high standard of EoL care. Two participants referred to there being a substantial responsibility and pressure to get it right, with one participant reporting a sense of relief once a DNACPR decision had been communicated. Three participants gave examples of situations whereby DNACPR decisions were in place but were not always upheld across different care settings, for example, patients being resuscitated with a DNACPR in place, DNACPR forms disappearing or distressed family members panicking and calling emergency services at the time of death. One participant felt strongly that initiation of the DNACPR discussion in community settings should not be imposed if the patient is struggling to accept their predicament.
Sub-theme 2b: Rapport
Getting to know the patient and developing a rapport were described as important factors in feeling confident to have the DNACPR conversation by most of the participants.
‘… if you haven't built up a rapport with him or her, or the family, then that makes it uncomfortable and a little bit uneasy …’.
Sub-theme 2c: First visits
Two participants experienced increased unease when it was necessary to have DNACPR conversations during first encounters with patients, with whom they had not yet developed a rapport:
‘… there are times when you have no choice, you come into a house and somebody is clearly dying …’ ‘… those really poorly people that I've never met before, so I prefer very much to have had a chance to get to know the people before having that conversation …’
‘… I find that it's quite hard, thinking oh God these people are meeting me for the first time and here I am having to tell them probably the worst news they've ever had.’
Sub-theme 2d: Gauging responses
Some participants reported that despite having a rapport with the patient, they found it difficult to gauge responses:
‘… I thought I'd totally distressed this person, I've really distressed them and I was very upset about it. I hadn't actually done anything wrong, I brought up the conversation but the person just was never going to be ready to go there and I stopped when I realised that …’.
However, another participant felt that health professionals should expect people to be upset, because these conversations make them realise the severity of their illness.
Theme 3: Initiating the conversation
Most participants spoke about the importance of finding an appropriate time to initiate the discussion. Factors affecting how participants prompted DNACPR conversations included patient acceptance, how well a rapport had been established and finding a time when there were less interruptions.
Participants shared that they would begin DNACPR conversations by initially checking the patient's understanding of their condition and prognosis, inviting patients to talk about their future preferences for care and including death as part of a wider discussion. One respondent referred to testing the water, with others referring to the need to tread carefully and monitor responses. Participants also reported enquiring about patient preferences regarding DNACPR. References were made to chest compressions, dignity and allowing a natural death when describing CPR (Table 1). The majority of participants said they emphasised care continuation, such as continued treatment of infections, deranged bloods and access to emergency services for example, in the event of a fall (Table 1).
| Sub-theme 3a: Prompting DNACPR conversations | ‘… the majority of people erm, I think, find it quite a natural discussion because they are so poorly and they accept that it needs to be discussed’ (P2) |
| Sub-theme 3b: Testing the water | ‘… sometimes you just have to test the water, if you feel they're okay about it, then you have to actually use the words do not resuscitate decision’ (P1) |
| Sub-theme 3c: Asking about the patients' understanding | ‘… ask if he had got a do not resuscitate decision in place …’ (P1) |
| Sub-theme 3d: Asking about resuscitation preferences | ‘… if it became evident that you were becoming more poorly erm, and it appeared that erm … you weren't going to get any better erm, if your heart was to stop beating, and if you were naturally dying, would you want us to try and resuscitate you …’ (P3) |
| Sub-theme 3e: Asking about preferred place of death | ‘Do you feel that you'd like to talk a bit about your care in the future and what might happen if you became more unwell?’ ‘If you became more unwell and you needed treatment, do you feel you'd want to go into hospital?’ (P1) |
| Sub-theme 3f: The act of resuscitation | ‘… would you want people pounding on your chest and would you want to be, you know, hooked up to machinery and be in ITU …’ ‘… Whether you would like to be resuscitated in the event that your heart stopped beating.’ ‘… so it's just in this specific incident, if this happened, where your heart stopped beating, would you want somebody to give you CPR if they were able to and then making them understand that the likelihood of the outcome of that situation would be not great, shall we say, and could lead then to artificial life support, being on a ventilator etc., etc., and in intensive care …’ (P4) |
| Sub-theme 3g: Reference to out of hours and planning ahead | ‘I also explain that if they were to die at home erm, and if it was out of hours or somebody panicked and dialled 999 erm, that if there wasn't a DNR form in place, the Paramedics could be obliged to try and do a resuscitation and that can be very distressing for the family.’ (P3) |
| Sub-theme 3h: Reassurance of care continuation | ‘… only relates to their heart stopping suddenly, a sudden event and it doesn't mean that they're not eligible for other treatments or care’ ‘… if you wanted antibiotics or if you have a mild infection that can be treated, this is purely about the actual, if your heart stops’(P1) |
| Sub-theme 3i: Reference to dignity, futility and natural death | ‘… would you want to be resuscitated or would you feel that it would be futile?’ ‘… it probably would be futile …’ (P2) |
Theme 4: Barriers to DNACPR discussions
The CNSs identified a number of barriers to initiating EoL conversations, such as, talking to younger patients; managing lack of acceptance and understanding among patients and relatives; confidence and training; and the CNSs' own emotions and culture.
Only one participant had received specific training on communicating DNACPR decisions, although all participants had received advanced communication skills training. The conversation was more challenging with patients who were unrealistic about their illness and prognosis or those who were receiving treatments such as chemotherapy. In addition, the participants' own emotions were identified by one participant as a potential barrier to engaging in DNACPR conversations, with four of the participants highlighting DNACPR discussions with younger patients as being particularly challenging.
‘… they won't even acknowledge that they are dying …’ ‘… but we have to respect their wishes, we can't just tell them that you have to believe this because they're not going to believe it if they don't want to, so you can't really have the conversation with those people at all’ ‘… Those four people (young patients) I'm talking about, every one of them died in hospital …’.
Discussion
This study demonstrates that palliative care CNSs often have to manage the aftermath of patients and relatives discover ing DNACPR decisions insensitively or inappropriately. The BMA et al (2016) maintained that concealing information due to discomfort among health professionals is not acceptable, and unplanned discoveries of decisions made may be more likely to cause patient distress. They emphasised that DNACPR decisions do not require consent from patients and relatives as long as the multidisciplinary team agree that CPR is likely to be unsuccessful and inappropriate. However, good practice is to include patients and families in these discussions (BMA et al, 2016). The literature also reflects that it is not unusual for patients to be excluded from DNACPR discussions (Healthcare Quality Improvement Partnership, 2019). Holland et al (2013) identified that hospital-based clinicians involved their patients in less than a quarter of DNACPR discussions. Although only one other small study found patients discovering DNACPR decisions by chance (Kazmierski and King, 2015). Fear of causing patient distress by initiating these conversations is echoed in wider research (Brannstrom and Jaarsma, 2015; Holland et al, 2013; Low et al, 2014; Miller and Dorman, 2014; Kazmierski and King, 2015; Royal College of Physicians, 2018). Yet, it could be argued that the notion of dying is in itself distressing. It may, therefore, be difficult to justify excluding patients from discussions based on concern for their general psychological distress (Berry, 2014), unless, of course, in exceptional cases.
Patient rapport is highlighted as essential when initiating DNACPR conversations, in this and other studies (Holland et al, 2013; Low et al, 2014; Petterson et al, 2014). Health professionals feel greater discomfort where it is necessary to have these conversations during a first encounter with dying patients with whom they have not yet developed rapport. Delayed resuscitation conversations resulting in conversations taking place with unfamiliar professionals is evident in other care settings within the UK (Cohen et al, 2013) and other countries (Garner et al, 2013). Although timing the conversation appropriately is challenging (Cox et al, 2007), earlier initiation of DNACPR discussions with professionals who have an existing patient rapport may be preferable (Hall et al, 2019). Yet, those who have the best rapport with patients may not have the confidence to have these conversations (Kazmierski and King, 2015). Patients receiving bad news generally prefer to receive it from someone with whom they have an existing rapport (Hanratty et al, 2012; Hall et al, 2019), who will allow them sufficient time to understand and ask questions; prior warning about the sensitive nature of the conversation should be issued, so that relatives can accompany the patient where appropriate (Hanratty et al, 2012).
The nurses who participated in this study felt a responsibility ‘to get it right’ and avoid ambiguous words while also finding it challenging to balance the delivery of sensitive but clear information, without distressing or confusing patients. BMA et al (2016) highlighted the importance of communicating DNACPR decisions clearly while simultaneously checking patient understanding. The House of Commons Health Committee (2015) identified that the terms used in the context of palliative and EoL care are often unclear and inconsistent, and Cox et al (2007) found that patients appreciated DNACPR discussions that were ‘straightforward’ and communicated slowly and clearly using ‘plain language’. However, DNACPR conversations can be unclear and rushed (Department of Health and Social Care (DHSC), 2013), with patients finding ambiguous language frustrating (Pollock and Wilson, 2015).
The CNSs in this study alluded to the importance of initially exploring the patient's perception of their illness and acknowledging what the patient wants to know and discuss while demonstrating empathy. These approaches mirror influential and well-recognised strategies for breaking bad news (Baile et al, 2000). The CNSs referred to inviting patients to talk about their future care preferences, and making DNACPR a part of wider discussions has also been recommended in the literature (Hall et al, 2019). The importance of exploring resuscitation preferences was also echoed in the study by Brannstrom and Jaarsma (2015). Further, Taubert (2016) supported this by suggesting that patients' views should be explored by asking them how they imagine their final moments to be, as it is unlikely to include aggressive interventions such as CPR. It is recommended that health professionals adopt their own personal style and frequently check that the patient is comfortable discussing the notion of CPR and understands what is being said (Taubert, 2016).
Offering CPR as a choice to patients may be unfair in situations where it is likely to be unsuccessful (Cox et al, 2007). Consideration should be given to the way patients are asked their resuscitation preferences, given that DNACPR is legally a medical decision (BMA et al, 2016). A significant point arising from the present research was that patients asked the question ‘do you want to be resuscitated?’ may respond in favour of CPR unless the likelihood of poor success with an emphasis on a natural and dignified death is explained. However, phrases such as ‘re-starting the heart’ or ‘getting you breathing again’ could potentially lead patients and relatives to think this is a possibility. Some CNSs alluded to offering DNACPR as a choice while explaining the negative aspects of CPR, and others said they speak to patients in favour of allowing a ‘natural death’. Miller and Dorman (2014) identified that some GPs suggest to patients that CPR would be inappropriate, while others viewed their role as information provision, allowing patients and relatives to decide for themselves (Miller and Dorman, 2014). However, this could be seen as contradicting the notion that DNACPR is legally a medical decision (BMA et al, 2016).
Although electric shocks and injectable drug administration are components of CPR (BMA et al, 2016), the participants in the present study did not mention discussing these aspects of CPR in conversations with patients. This adds to research by Heyland et al (2006), which showed that patients understood only some aspects of CPR and had poor understanding of success rates. The author's own experience has been similar, in feeling hesitation in sharing these discomforting aspects of CPR with patients and their families.
With a focus on dignity, all CNSs mentioned referring to care continuation during DNACPR conversations. Similarly, Miller and Dorman (2014) found that GPs communicated the notion of ‘not being giving up on’ when discussing DNACPR decisions, with examples of continued care at home, pain relief and food. Reassurance of care continuation is essential (BMA et al, 2016), since previous reports have shown that relatives raised concerns when DNACPR decisions led to other aspects of care being withdrawn (DHSC, 2013).
The present study as well as that of Low et al (2014) showed that health providers found it challenging to have DNACPR conversations with patients and families that were unrealistic about the illness and prognosis. These conversations were exceedingly difficult where relatives were less accepting of the patient's deteriorating condition, and one nurse in this study cited an example of family members requesting information to be withheld from the patient. Granek et al (2013) also highlighted this as a barrier to initiating EoL conversations. Cox et al (2007) found that relatives were less accepting of DNACPR discussions than patients, with some relatives preferring to have these discussions closer to death. Brannstrom and Jaarsma (2015) similarly found that speaking to patients and relatives together was challenging for professionals, who felt the presence of the other would hamper open communication. The findings from Cohen et al (2013) further supported this, as the study reported that very few DNACPR conversations took place with both patient and family present.
The CNSs reported that they found broaching EoL conversations with younger patients challenging, and they identified their own emotions in feeling prepared to engage in these conversations as a barrier. Low et al's (2014) findings mirrored these results. Additionally, the CNSs gained their knowledge and skills through experience, role models and advanced communication skills training, unlike junior doctors who reported a lack of role models to enhance confidence in EoL conversations (Royal College of Physicians, 2018). However, specific training on having DNACPR conversations is lacking in both groups and is highlighted as an important factor affecting confidence (Holland et al, 2013; Kazmierski and King, 2015).
Limitations
This was a small-scale research study with a group of health professionals from a single organisation. Therefore, it may not necessarily reflect the experiences of other health professionals. Ideally, further research needs to be conducted with groups of health professionals from multiple organisations.
Conclusion
Community palliative care CNSs emphasised the importance of communicating DNACPR conversations clearly and sensitively with patient and families. Conversations should be initiated by exploring patients' understanding and preferences, with an emphasis on care continuation. Community palliative CNSs have experienced that unplanned discoveries about DNACPR decisions have sometimes caused patient distress. This study may help to improve the way patients receive information, which should be done in a more timely fashion and from someone they have rapport with. Provision of more specific DNACPR education is essential if staff are to communicate key EoL decisions clearly and sensitively to patients and families. For this purpose, a DNACPR workshop was conceptualised and delivered at the author's organisation, incorporating the legal and ethical principles of DNACPR, as well as an interactive demonstration of a simulated DNACPR conversation using actors.