A nurse’s phenomenological enquiry into doctors’ end-of-life decision making

Introduction

How people die remains in the memory of those who live on.

(Saunders, 1989)

While structure exists in the UK to support EoLC, around 47% of the population in England die in hospital (National End of Life Care Intelligence Network (NEoLCIN), 2015), despite only 7% indicating this as their preference (Shucksmith et al., 2013). Both the Francis (2013) and One Chance to Get it Right (Leadership Alliance for Care of Dying People, 2014) reports identify that multi-professional teams struggle to provide consistent quality care and choice near the end of life.

Evidence suggests difficulties for doctors in prognostication (Glare et al., 2008) and recognition of approaching death, compounded by poor communication among patients, relatives and health professionals (Leadership Alliance for Care of Dying People, 2014). The EoLC Programme (2010) and General Medical Council (2010) suggest inconsistency in clinical decision making (CDM) (defined as ‘a choice between alternatives leading to an outcome’; Dowie, 1993: 8) near the end of life adversely influences patient choice regarding preferred place of death, and that minimising investigations or ‘aggressive’ medical management at this point may enable more people to die in a familiar environment.

The first author (the researcher), an experienced hospital palliative care clinical nurse specialist, used to acting autonomously and objectively, found one particular patient drew her more closely into the care of a terminally ill oncology patient and, ultimately, into the decision-making process in the last hours of a young person’s life. Reflection led to recognition of a perceived disparity between doctors’ and nurses’ approaches to CDM at the end of life and between doctors’ and nurses’ interpretation of ‘aggressive’ clinical management within an acute oncology setting, as well as what a ‘good decision’ is for patients thought to be in the last days of life.

In order to explore this perceived disparity the following research question was proposed: ‘What influences trainee oncology doctors’ decision making, around clinical intervention, in acute oncology patients, near the end of life?’

Literature search

A literature search to identify publications on decision making for oncology inpatients in the last days/weeks of life, conducted via CINAHL, PsychInfo, Embase, Ebsco host and Medline (via Ovid), resulted in 14 studies being included. The low number reflects the dearth of evidence around EoLC practice discussed by Smith et al. (2015).

Key words and derivatives related to the research question informed the search strategy: doctor, nurse, identification, oncology patient, palliative care patient, terminally ill, prognosis, qualitative research, decision making.

As Cresswell (1994) suggests, literature was used ‘sparingly’ to maintain an exploratory approach to enquiry. Nevertheless, a number of themes were identified:

Differences in decision-making behaviour between professional groups: whereas nurses tend to rely on a ‘felt sense’, doctors rely more on measurable parameters (Dalgaard, 2010; Dee and Endacott, 2011; Salantera et al., 2003).

OPEN IN VIEWER

Figure 1.

Standing’s revised cognitive continuum of clinical judgement and decision making in nursing – nine modes of practice.

While recognising the risk of nurse/doctor stereotyping and polarisation, findings from the literature resonated with the first author’s reflection on experience regarding paradigm differences in CDM behaviour between nurses and doctors. However, it was not evident what influenced doctors’ CDM in the last weeks of a patient’s life. This was identified as a gap in knowledge and resulted in the question: ‘What influences trainee oncology doctors’ decision making, around clinical intervention, in acute oncology patients, near the end of life?’

Subsequently, the following aim and objectives for an empirical study were identified.

Methodology

To reach the study aim and objectives, the researcher accepted the phenomenological concepts of a ‘tacit’ or unknown lifeworld (Gadamer, 1995, cited in Dahlberg et al., 2008: 35) and a ‘still mute’ experience to be made visible to understanding (Merleau-Ponty, 1968, cited in Dahlberg et al., 2008: 39).

A method of enquiry that brings the tacit world into the open, making it tangible and open to interpretation was necessary.

Epistemological interpretations by Streubert and Carpenter (2011) and Van Manen (1990) support phenomenological enquiry as the most effective approach to elicit understanding of the topic being explored.

In this study, the methodological philosophy of hermeneutic phenomenology, informed by Dahlberg et al. (2008) and Van Manen (1990), facilitated an intersubjective approach, ensuring a deep and rich understanding of the phenomenon.

Van Manen (1990) states that the chosen methodological approach must be suitable to answer the research question and that method should follow methodological principles. Rolfe (2006: 309) suggests that ‘responsibility for appraising research lies with the reader rather than the writer’. In accepting this premise, the writer assumes responsibility for providing sufficient information for the reader to be able to decide on the quality of the research.

It is proposed that the ‘human science’ approach of Van Manen (1990: 11) to hermeneutic phenomenology assures the reader of the credibility of the research by the application of the following principles: expectation that the researcher be ‘self-critical’, ‘systematic’, ‘explicit’ and ‘intersubjective’. This approach enables the reader to believe the findings are a true reflection of participants’ experiences. The scientific approach is further explicated by Dahlberg et al. (2008: 96), who describe the need for reflection at all stages of the research process, ‘openness’, and a purposeful leaving aside of expectations and assumptions in order to approach the research question naively and objectively. Dahlberg et al. (2008: 129) term this ‘restraining’ of preconceived ideas and beliefs, both known and hidden, as ‘bridling’. In terms of credibility, bridling is argued to limit bias when interpreting findings.

Adopting these suggestions conferred credibility on the chosen method.

The first author ensured a ‘bridled’ and reflexive approach throughout the research process by means of a reflective diary, and supervision sessions with the second author throughout the interview process and data analysis. Furthermore, a ‘holding in mind’ of the four lifeworld existentials (Van Manen, 1990: 101–105) – ‘temporality’, ‘spatiality’, ‘corporeality’ and ‘relationality’ – grounded the research in the fundamental principles of phenomenology, conferring a sense of reliability and truthfulness to the interpretation.

Methods

The appropriate research ethics committee and NHS research and development department granted ethics approval in April and May 2014, respectively.

Dahlberg et al. (2008: 104) propose that the researcher should be ‘finely attuned’ to the participant and ‘immersed’ in the phenomenon. Therefore, a method enabling intersubjective engagement between participant and researcher, or a ‘conversation’, which holds the meaning or ‘essence’ of the phenomenon (Van Manen, 1990: 98) is preferable.

Consequently an in-depth interview lasting 60–90 minutes was conducted with each participant, digitally recorded and transcribed by the researcher.

Non-verbal signs were included in transcriptions, adding clarity and meaning to the participants’ words (Streubert and Carpenter, 2011). To minimise bias, an initial question was utilised: ‘I am interested in your experiences of making clinical decisions for patients in the last weeks of life. Can you tell me about this?’

Both to provide structure to the interview and facilitate bridling of early presuppositions, the researcher held in mind the lifeworld existentials of Van Manen (1990). During interview, the researcher paraphrased responses as a form of triangulation, thus precluding the need for member checking, while ensuring the participants’ tacit view of the phenomenon.

After four interviews and concurrent reflection with the second author, it was clear that, despite the small number of participants, data saturation had been reached, with recurrent themes or data units emerging.

In adherence to the principles of hermeneutic phenomenology and the concept of ‘the whole > the part > the whole’ (Dahlberg et al., 2008: 281), the ‘detailed or line by line approach’ to analysis, suggested by Van Manen (1990: 93), was adopted:

Initial reading and rereading of the transcriptions ensured deep understanding of the ‘whole’ phenomenon in context.

To return to the ‘whole’ of the phenomenon: findings were presented as a written description of the essences and meaning of the phenomenon (Dahlberg et al., 2008; Van Manen, 1990).

Results

Participants ‘Alex’, ‘Bobbie’, ‘Chris’ and ‘Dale’ are referred to in the masculine to preserve anonymity. The rich interview data, following analysis, enabled identification of common transcending themes that allowed naming of the essences of the phenomenon itself rather than the individual’s experience.

Through interpretation of the data, three essences were identified as illuminating the fundamental meaning of the phenomenon:

empathy, the unacknowledged influence;

Discussion

The first author’s preconceived view, substantiated by the literature review, was that doctors' and nurses' decision-making behaviour is paradigmatically different.

The findings, however, appear to demonstrate a new concept, not currently reflected in accepted CDM theory: that empathy, although a beneficent approach, may affect doctors’ CDM, adding to the difficulty in making decisions about clinical care of oncology patients near the end of life. Findings from a further literature review suggest the ability to empathise is perceived as a good and helpful characteristic (Selph et al., 2008) and ‘fundamental to effective patient care’ (West, 2012:243). Arguably, the findings of this study suggest the impact of empathy on CDM may adversely affect patient choice re EoLC. Furthermore, the concept ‘empathy’ is not represented in current conceptualisations of CCT by Hamm (1988) and Standing (2008) and is worthy of further investigation.

The essence dichotomy and conflict demonstrates that both internalised professional conflict and external inter-team conflict exist and are perceived negatively. Evidence from the literature illustrates that conflict in clinical practice is distracting and impedes task execution (Greer et al., 2012; Janss et al., 2012). Interpretation of role, learning team rules and placing themselves and their beliefs within this context appeared to cause most conflict for these four doctors. Participants did not describe conflict between nurses and doctors as a common problem. Contrary to the literature, comments generally suggested a perception of nurse–doctor interactions as usually positive and helpful. ‘Dale’ finds that understanding the patient’s perspective for him ‘hinges around the nursing team’. Comments on conflict between participants and nurses, raised in interview, represented differing interpretations of protocols rather than fundamental differences of opinion.

All participants acknowledged difficulty in making decisions when patient wishes were not documented, particularly if patients were perceived as too unwell to include in conversations around escalation and limits of care. However, despite this acknowledgement, an internal conflict exists about identifying the appropriate time to discuss EoLC choices. This reflects issues raised earlier around inconsistency and poor prognostication adversely affecting place of death (EoLC Programme, 2010) and appears worthy of further investigation.

The third essence, subverted emotion uncovered through reflection, was not apparent in the literature relating to senior oncology doctors and was interpreted as an element of the tacit or unknown lifeworld this research brought to light.

The interview process invited reflection on past experiences, appearing to fulfil a need for reflection in the participants. Each one, even ‘Dale’, who barely acknowledged the issue of emotion, appeared to have reached a new understanding of previous experiences and what they had brought to those decisions.

Taylor (2010: 29) suggests reflection may enable professionals to examine their protective behaviours or shields, utilised to limit the impact of the ‘relative tragedies and uncertainties of daily practice’. The unexpected degree of emotion felt and expressed by participants during interview demonstrated that participating in a research interview allowed an ‘unburdening’, begging the question of what support is available for senior oncology doctors in practice.

Further literature highlighted the concept of a hidden curriculum (Hafferty, 1998; Lindberg, 2009; Pitkala and Mantyranta, 2003) relating to doctors’ socialisation into a role or team. It is proposed that, although this concept is addressed for undergraduate doctors, the role development of trainees may involve a unique hidden curriculum that promotes subverted emotion as the norm.

Presenting this research to oncology doctors (trainees and consultants) from the participating organisation in September 2015 provided powerful validation of the findings. Doctors recognised the essences as a true representation of their world, with consensus among the audience that the findings, although previously unacknowledged or ‘subverted’, were true.

The authors acknowledges the following limitations:

Participants self-selected for interview, potentially introducing a non-response bias (Bowling, 2009).

Future work should take account of these factors.

Conclusion

In conclusion, the authors propose that the chosen methodology and method – hermeneutic phenomenology, underpinned by the human science approach exemplified by Van Manen (1990) and Dahlberg et al. (2008) – ensured the aim and objectives were met. The findings illuminate the phenomenon of oncology doctors’ experience of decision making for patients in the last days/weeks of life.

Findings, expressed as essences of the phenomenon, demonstrate that in contrast to findings from literature and the first author’s own pre-reflective perceptions, doctors appear to bring more than an analytical approach to the act of CDM. Factors of empathy and their own world views impact and may cause conflict both internally and as part of the wider team structure. Expected behaviours associated with the role, including those that may be a result of the ‘hidden curriculum’, may impact negatively on outcomes for patients, who can feel powerless at this critical point of their lives, and create an emotional burden for doctors. These factors appear worthy of further enquiry.

The presentation of these findings may raise awareness of the impact of this complex process for both patients and doctors, and ultimately enable a more caring and empathic approach between professionals.

Further research into the impact of empathy on oncology doctors’ decision making with patients and further enquiry into the role of nurses in relation to decision making could influence practice and, ultimately, improve patient choice at the end of life.

Key points for policy, practice and/or research

The experience of CDM for oncology doctors is complex and can cause distress for them.