Relationships between personal attitudes about death and communication with terminally ill patients: How oncology clinicians grapple with mortality

Abstract

Objective

Clinician discomfort with death may affect care of patients but has not been well-studied. This study explores oncology clinicians’ attitudes surrounding their own death and how these attitudes both affect and are affected by their care of dying patients and their communication with them.

Methods

Qualitative interviews with physicians (n = 25), nurse practitioners (n = 7), and physician assistants (n = 1) in medical or hematologic oncology clinical practices about communication styles, care of terminally ill patients, and personal perspectives about mortality.

Results

Clinicians described three communication styles used with patients about death and dying: direct, indirect, or selectively direct. Most reported an acceptance of their mortality that was “conditional,” meaning that that they could not fully know how they would respond if actually terminally ill. For many clinicians, caring for dying patients affected their outlook on life and death, and their own perspectives on life and death affected their approach to caring for dying patients.

Conclusion

An awareness of personal mortality may help clinicians to discuss death more openly with patients and to provide better care.

Practice implications

Efforts to promote self-awareness and communication training are key to facilitating clear communication with and compassionate care of terminally ill patients.

1. Introduction

Most patients with advanced cancer want honest, sensitive communication about prognosis and end-of-life issues [1]. Having these conversations can help patients and their families prepare for death, make informed decisions, and avoid aggressive medical care near death [2–4]. However, open discussions about dying often do not occur [2]. Instead, these discussions are deliberately indirect or ambiguous [3,5–7], or they take place late in the course of the disease [8].

It is clear that many barriers to the delivery of clear communication about death and dying exist [9–12]. Oncologists struggle with finding a balance between providing realistic information and nurturing hope [13,14]. Frequently they provide overoptimistic estimates of survival [15] while failing to recognize or correct their patients’ misconceptions about intent of treatment and prognosis [16,17]. Patients, meanwhile, often overestimate their life expectancy [17,18], misunderstand the goals of palliative chemotherapy even when told it is not curative [16,19], and prefer less information as their illness progresses [20,21]. Often there is a discrepancy between what physicians think patients want to hear and what patients actually want [3,22], as well as what physicians believe they say and what patients understand [23].

Despite suggestions that clinician discomfort with death affects communication, care, and the physician–patient relationship, this phenomenon has not been well-studied [9,11,24,25]. Over thirty-five years ago, Schulz and Aderman found that dying patients of physicians who personally fear death have longer terminal hospital stays [26], and it has been suggested that those physicians who have not accepted their own mortality are more likely to focus solely on biomedical issues rather than dying patients’ emotional concerns [24]. Frequent exposure to death and suffering can evoke intense emotions [27–29] and lead to death anxiety, disengagement from patients, burnout, and nihilism [30,31]. However, such exposure also can yield several positive outcomes, such as greater personal fulfillment and enhanced meaning in life [31,32].

These studies highlight a unidirectional impact of caring for dying patients on clinicians’ well-being, but to our knowledge little is known about the reciprocity of this relationship. One qualitative study has detailed how palliative care physicians’ own spirituality impacts their practice and how their practice impacts their spirituality [33]. While spirituality can be an important component of one’s attitude toward death, our goal was to explore more broadly how oncology clinicians’ perspectives about their own life and death affect and are affected by their care of and communication with dying patients.

2. Methods

2.1. Study design

We conducted and analyzed semi-structured interviews with oncologists, oncology nurse practitioners (NPs), and oncology physician assistants (PAs) to explore how their personal attitudes about death affect and are affected by their care of and communication with dying patients. Data analysis was based on grounded theory [34,35] and followed an editing analytic approach [36].

2.2. Study participants

Oncologists, NPs, and PAs in clinical practices of medical or hematologic oncology were recruited from multiple private and academic oncology clinics in the Rochester, New York, area (two hospital-based oncology clinics, two private community-based oncology clinics, and one cancer center at a large academic medical center). After obtaining consent from the University of Rochester Research Subjects Review Board, clinicians not involved in development of the study were contacted through email by physician co-investigators between October and December 2014. If the clinician indicated interest, the first author followed up by email to describe the details of the study and schedule an interview. Prospective participants met with the first author to verify understanding of the study and give verbal consent to being interviewed and audio-recorded. In accordance with recommendations for sample size in qualitative studies [34], our aim was to conduct at least 20–30 interviews. Sampling was done sequentially in batches (by email) until sufficient responses were received. In total, we contacted a convenience sample of 45 of 51 oncologists, 13 of 23 NPs, and 2 of 9 PAs in the Rochester area. Of these clinicians invited to participate, 25 oncologists, 7 NPs, and 1 PA ultimately were interviewed, resulting in an overall participation rate of 56% for oncologists and 53% for NPs and PAs. The most common reasons for non-participation were non-response to emails (n = 20) and difficulty scheduling (n = 5). Data saturation was determined post hoc during the process of formal data analysis.

2.3. Data collection

A semi-structured interview guide was developed by the first author based on a comprehensive literature review of topics including end-of-life communication [1,37–41]; death anxiety [42,43] particularly as experienced by clinicians [25,26,44–46]; and clinicians’ experiences with, reactions to, and strategies for coping with patient death [27–29,47–55], with a particular interest in proposed links between caring for dying patients and personal attitudes about mortality. The literature review included qualitative and quantitative studies, reviews, and commentaries, which encompassed a variety of physicians (in oncology, palliative care, hospital medicine, intensive care, and other specialties) at different stages of their careers (from internship through senior levels), and non-physician health and spiritual care providers who were caring for seriously ill patients. To ensure the appropriateness of questions in relation to the purpose of the study, interview questions were reviewed and revised by other members of the research team and clinicians not involved in the study.

The first author conducted all individual interviews between October 2014 and January 2015. Interview questions explored how participants approached conversations about death and dying, how caring for dying patients affected their views on life and death, what thoughts they had about their own mortality, and how their views on death affected patient care and communication. (See Appendix A for a list of questions.) Mean interview length was 35 min (range 20–60 min). All interviews were audio-recorded and subsequently transcribed and de-identified. Participants who requested to review their quotes prior to publication were given this opportunity. Demographic data were collected through a survey administered immediately preceding the interview.

2.4. Analysis

An editing style of analysis as described by Miller and Crabtree [36] was conducted by a multi-disciplinary team consisting of a medical student with a background in literature (RR), a medical student with a background in psychology (KR), and an oncologist–researcher (MT). Initially, all transcripts were read by the first author to determine the spectrum of views represented, with initial notes jotted in the margins and potentially significant phrases marked. Next, five transcripts representing a broad range of responses were selected, read line by line, and openly coded by all team members to develop categories of information. From these transcripts a preliminary coding scheme was developed, applied to an additional five randomly-selected transcripts, and further revised. Axial and selective coding was employed to reassemble, compare, and connect categories. This coding scheme was used by all team members for the remaining transcripts, in which no new codes were identified. The first interviews used in the development of the coding scheme were also re-coded. By the point of coding 20 transcripts, no new themes emerged, and data saturation was considered to have been reached. The remaining transcripts served to provide additional nuance and representative quotations. Differences in coding were resolved by consensus. All data were entered into the qualitative data analysis software program Atlas.ti 7.0 for organization and storage, and quotes were extracted for further analysis.

3. Results

Characteristics of the 25 oncologists and 8 oncology NPs and PAs interviewed are presented in Table 1.

Table 1.

Characteristics of oncologists, nurse practitioners, and physician assistants.

Characteristic	Oncologists (n = 25)	Nurse practitioners and physician assistants (n = 8)
Age, mean (SD), years	46.6 (11.9)	40.9 (10.6)
Time in practice, mean (SD), years	13.7 (12.3)	11.0 (8.4)
Sex, no. (%)
Female	6 (24.0)	8 (100.0)
Male	19 (76.0)	0 (0.0)
Race/ethnicity, no. (%)
White	15 (60.0)	8 (100.0)
Asian	8 (32.0)	0 (0.0)
Other	2 (8.0)	0 (0.0)

Below we present clinicians’ accounts of their communication styles when discussing death and dying with patients, then their attitudes about their own mortality, and finally ways in which their own perspectives about death affect and are affected by their communication with patients. Contextual factors that affect how this communication occurs are considered at the end.

3.1. Approaches to conversations about death and dying

Participants described three main styles of communication with patients about death and dying: indirect, selectively direct, or direct (Table 2). Some clinicians described using more than one approach depending on context, even with the same patient. Those who used an indirect approach used vague language, cushioned words, or withheld information from patients, often in order to preserve hope and to avoid terrifying patients with too much information.

Table 2.

Quotations illustrating perspectives and styles of communication about death and dying.

Indirect communication style Many times that’s [an optimistic outlook is] a defense mechanism that is valuable for people and we have to be careful about how much we break that down... The more you spell out to people the unpleasantness or the potential risks of the journey, especially if you don’t emphasize the positives, people may not be prepared to take that road or may choose not to take that road. (Physician 3) I think sometimes we kind of string patients along, you know, we say oh, we’re gonna do this treatment and this treatment and I think sometimes we are a little more optimistic than we probably should be. Again, because you want to give them hope. (NP/PA 4)

Selectively direct communication style So I really try to get a sense of where everyone is and how much information they want. Information can be damaging and you can’t take that back so I think we have to be really careful when we talk in those ways . . . So I try to proceed cautiously. I always err on the side of being more vague than more direct. (Physician 8)

Direct communication style I guess in some respects some people might say the terminology that I use is very direct all the time, although it’s not in your face direct. So I don’t say you’re not gonna make it. I say this is not a cancer that we can cure, meaning there is nothing that we can take off the shelf and give to you that’s going to make the cancer go away and have it never come back. (Physician 2) I’ve found over time that being as direct as possible all the time is the best thing both for me and my patients. And so I’ve gone from a position of, in my thinking and taking care of people, doing a lot of internal editing of what I say to just saying what I think. And so when I see someone and I’m worried about them either being very sick or dying, I say I’m worried about you and then I say why, which is that I’m worried that you might be dying or that you’re very sick from this and that your life is short. (Physician 12)

Perspectives on the d-words e.g.”death,” “die,” or “dying” To tell people okay, you have stage four cancer, that means you’re gonna die no matter what, I don’t think that’s very valuable because it doesn’t help a person live. It doesn’t really help them make decisions on a day to day basis just to say it that bluntly. (Physician 3) You can say we have tried this option and now it’s not working. Eventually with or without treatment the cancer is going to progress and claim your life. It’s still direct but you have to have some degree of empathy planned into it. You cannot just say you’re dying from cancer. (Physician 15) One of my concerns is that medical people, like everybody else in our society, are uncomfortable with the “d” words and so I try in my conversations with patients to use the die, death, dying, something along the way. And you know, pass away is.. . it’s a euphemism. (Physician 17)

The “selectively direct” group used a combination of direct and indirect language with patients. They preferred being direct only in specific circumstances, including when patients asked direct questions, when medically urgent situations arose, and when patients appeared to be in persistent denial.

Finally, clinicians with a direct communication style strove to bring patients to a realistic understanding of their disease in order to help them make informed decisions and settle their affairs, and to avoid blindsiding the patient later on.

Of note, some who reported communicating directly with patients provided examples that others (including the authors) considered indirect: ambiguous expressions and euphemisms about death. Because many perceived d-words (e.g. “death,” “die,” “dying”) as blunt and cruel, they opted to convey the message in softer ways. Others, meanwhile, reported using specific sentence constructions that enabled them to use d-words in caring ways.

Regardless of the communication style they used, many clinicians felt that being honest, exploring patients’ understanding, and asking them how much they want to know were important. Many also acknowledged the difficulty inherent in accurately predicting prognosis.

Further, clinicians in each category of communication style generally described a full spectrum of views regarding their acceptance level of death, how their personal attitudes about death intersect with caring for dying patients, and contextual factors, as described in the following sections.

3.2. Acceptance of death

While all clinicians recognized death’s inevitability, they fell into three groups with regard to the extent to which they accepted and felt at peace with their mortality: full acceptance, conditional acceptance, and non-acceptance (Table 3). The handful of clinicians who reported full acceptance of their mortality reported feelings of peace, lack of worry, and a carpe diem attitude. They denied obsessing about death.

Table 3.

Quotations illustrating clinicians’ level of acceptance about their own mortality.

Full acceptance For me it’s pretty straightforward that someday you are going to die. It’s pretty.. . it’s a given. So the key part is how do you make the best of your life here. So, carpe diem is the advice which is pretty standard advice. (Physician 21)

Conditional acceptance I think I’m pretty accepting that you have to go at some point. I think right now aging is a more difficult thing to come to terms with right now . . . Right now the death issue is there but it’s there in the background. But I think if you get a diagnosis directly that’s when it’s going to be.. . it’s going to really loom large. That’s when it’s going to be, how are you gonna cope with it. Right now it’s there but you try to, you can prepare for it but it’s like in the back burner. (Physician 13) A lot of it for me is just wanting to have a feeling that I lived a life that I can be proud of. Like, I think peace will come at the time of death if I have really done everything I can to make my own life matter to other people, especially my family and my patients . . . I feel like I’ve given it my best so far. So that I guess makes me feel peace, you know, that I will have peace at the time [of death]. (Physician 23) Like everybody is going to die, what are your wants and stuff. I don’t think that death is around the corner, so I don’t have anxiety about it. I think if I was a patient sitting in that chair and it was looking me in the eye, I would have anxiety about it. (NP/PA 8)

Non-acceptance If I were to die now I would have tried to live the last [several] years as great as possible, but I’m not ready to die right now . . . That’s my only fear right now. I feel like if I can get another like 20, 30 years, then I’ll be able to cope with it better. (Physician 6) In the sense that I think I have a baseline anxiety in general that probably reflects that, of not having that level of acceptance... To think that my life could be cut short and miss out, so to speak, not that I should be greedy in that sense, but that’s always sad to think that, as we see with our patients when they die and they will miss their children’s children and all those life experiences. (Physician 18) I’m just worried about, you know, that the kids are okay. And will I feel like I’ve led a good enough life and accomplished enough. I worry about having existential issues that I didn’t do enough. That I didn’t get different things done or that I wasn’t a good enough person, things like that. (Physician 20)

The second group, those with a conditional acceptance of their mortality, expressed an acceptance that was conceptual and intellectual but not experiential. Most clinicians fell into this category. While they currently felt comfortable with their mortality, they acknowledged not knowing themselves fully enough to anticipate how they would react when faced with serious illness in the future. At this point in their lives, they occupied themselves with more immediate concerns such as career or family, and worries about death, when present, tended to be at the back of their minds.

Finally, a small number of participants acknowledged that they had not accepted their mortality; they described discomfort and/or avoidance in thinking about their own death, anxiety resulting from witnessing deaths in their professional and sometimes personal lives, and active worry about at least one aspect of their own death. Not at peace with their own death, they yearned to live much longer.

The most common specific worry about death for those with conditional acceptance or non-acceptance was what would happen to their loved ones. Other worries related to pain, timing, legacy, loss of memory, and having lived a “good enough” life. While many identified specific worries about death, most denied feeling afraid of death.

3.3. How personal attitudes about life and death intersect with caring for dying patients

Many clinicians reported feeling that they coped well with patient death and were able to describe how personal attitudes about death affected how they communicated with dying patients and vice versa (Table 4). Most described positive or neutral features, with a few describing negative features.

Table 4.

Quotations illustrating the relationship between personal attitudes about death and care of dying patients.

Positive effect of patient death on personal perspectives In fact, I’ve achieved more in my out-of-medicine life in the last three years as opposed to any time before in my life. And that is simply because the way those deaths affected me was that life is short; anybody can die anytime so each and every day that you have is extremely valuable... So with that I started living life more and instead of getting more depressed. So it actually had a very positive effect on me. (Physician 9) That’s what I tell my patients, to have a big bowl of ice cream. Like this is what I mean, like just live your life to the fullest extent because you never know what happens. So that really changed my perspective. For holidays I want to see all the family members. You never know what will happen in the next few years. (Physician 16)

No effect of patient death on personal perspectives So, at this point in my career does a patient’s death make me sort of.. .[think about my mortality]? I would have to say no. At least I don’t feel that. Maybe somewhere deep inside it does but I don’t feel that that’s part of the stress. (Physician 3) Not anything more than, you know, everyone is gonna die. That’s one thing that is universally true, right? (Physician 19)

Negative impact of patient death on personal perspectives So anything can happen any day. And you can’t even predict . . . I see it all the time, every day. At home. Here at work. And I think about it all the time. (Physician 20)

Facilitative role of personal perspectives on communication with dying patients Everybody is mortal; you can die at any age. You need to discuss it clearly. And I think me accepting death, helps me discuss death with a patient. (Physician 15) I think that because I don’t fear death terribly, I’m not terribly anxious about it, I think it makes it easier for me to discuss it with my patients. I think it really also allows me to really feel at peace and comfortable about the whole spectrum of care for my patients, so from diagnosis through the end . . . Not having that terrible fear of death kind of allows me to do that where it might be harder for other people who don’t want to face that. (Physician 23)

No effect of personal perspectives on communication with dying patients I don’t think there is much relationship in my situation because just as I said earlier, I just try to be neutral. You know. Give the same information out to everybody. (Physician 19)

Those whose perspectives on life and death were positively influenced by patient death described reflecting on their own mortality, maintaining a broader outlook on life, and living each day more fully. Helping patients through the dying process was rewarding, but they also needed to pursue other sources of fulfillment outside of medicine.

Participants who reported no influence noted rarely thinking or “obsessing” about death; they described “balanced” and, in some cases, “distanced” relationships with patients. Patient death neither affected how they lived their lives nor generated much worry about their own deaths.

The few participants who noted negative effects of caring for dying patients reported feeling overwhelmed and anxious.

The reciprocal effect – the effect of clinicians’ views about life and death on communication with dying patients – was described as either facilitative or unrelated. Many in the facilitative group spoke affirmatively of the interaction between their own outlook on life and death and their care of dying patients. In particular, acceptance of their own mortality helped them speak openly and comfortably with their patients.

On the other hand, a smaller group of clinicians believed that their perspective on life and death was unrelated to how they communicated with patients about death and dying. These participants’ views on death were similarly not affected by patient death.

While the study was not designed for quantitative comparison between groups, we nonetheless sought to identify patterns that might warrant further exploration. In that regard, we did not identify consistent patterns for most clinicians between acceptance of their own death (conditional acceptance or non-acceptance) and self-reported communication styles (e.g. direct, selectively direct, or indirect).

3.4. Other considerations

Personal belief systems, personal loss, cultural backgrounds, and self-reflection seemed to affect clinicians’ perspectives on death and their care of dying patients (Table 5).

Table 5.

Quotations illustrating contextual factors from clinicians’ personal lives.

Belief systems Well, I believe that there is another life. I believe in heaven. I believe that we will be reunited with our loved ones. I think that’s how I cope with it, knowing that this isn’t all there is and having faith. (NP/PA 3) I am not a religious person in any way whatsoever. I’m not sure there is any afterlife. I’m sort of. .. death comes eventually; it’s part of life. I don’t think there is any greater meaning to life. We’re just here to do the best we can and help other people the best we can and. .. and that’s done. (Physician 25) I want to believe there is more to it than you just die and you’re in the ground and you decompose and then that’s it. I’m not sure if I really do believe more than that. (Physician 5) My faith has helped me a little bit but sometimes I still struggle with if there is a God, why do these people suffer like this. (Physician 18)

Personal loss My personal life experience probably significantly colors my view on death and dying, you know, and having had a very close relative die very suddenly... I don’t find it particularly difficult to be upfront and honest with people particularly about these things because I believe so strongly that people would make different decisions if given all the information in many respects. (Physician 2)

Cultural influence In [non-Western country] people understand that you are born and you die. And when it comes time to die most people understand that and they want to make the end – they want to keep some control of the end and they don’t believe that every disease has a cure. Especially in the [U.S], a lot of people think the only reason that you die is because you didn’t try hard enough. (Physician 7)

Self-reflection When I am mindful of it [death] in a way that extends beyond how am I gonna get through the day, it causes one to be attentive to relationships . . . I find that when I’m attentive to death I try to leave the office early. That’s how it impacts me. You know, make sure that you have time to do things that are important. (Physician 12) And I find that if you think about death and dying when you are living and healthy and well, it kind of focuses your mind. It kind of sets up my course. What am I supposed to do? Why am I here? (Physician 14)

Most clinicians either had a secure belief system (whether self-described as religious, spiritual or otherwise) or did not believe in anything. Both groups felt content with and confident about their belief stance, which helped them to cope with mortality both personally and professionally. However, a minority felt uncertain about what to believe; some of these clinicians also neither felt as at peace with their own death as they would like nor coped as well with patient death as they felt they should.

About half of the participants acknowledged that personal loss of close family members or friends had affected them, particularly in the way in which they dealt with patients and families. Meanwhile, a small number of clinicians felt their own cultural backgrounds, particularly from non-Western societies, influenced how they viewed disease and death.

Finally, self-reflection and mindfulness played an important role for many clinicians. A commitment to an examined life and an attention to death enhanced their approach to life.

4. Discussion and conclusion

4.1. Discussion

This study’s main findings are that most oncology clinicians express a conditional acceptance of their own death, and for many, there is a reciprocal relationship between their attitudes toward their own death and their care of dying patients. In particular, caring for dying patients affected clinicians’ outlook on life and death, and their own perspectives affected their approach to caring for dying patients.

In the 1970s anthropologist Ernest Becker argued that humans’ unique condition of being conscious of our mortality creates a ubiquitous fear of death; most human action attempts to avoid or repress awareness of the inevitability of death in order to make life livable [43]. In the same time period, Feifel and Branscomb found that while at a conscious level, most people, whether ill or healthy, report not being afraid of death, below their level of awareness they feel averse to death [42].

Our results provide further nuance to Becker’s argument. Oncology clinicians, because of their continuous exposure to patient death, have a greater awareness of mortality than the general population. While few clinicians acknowledged personal fear of death, many expressed a conditional acceptance of their mortality – that is, a current state of peace that they admitted might change in the future. This view may be considered adaptive, enabling clinicians to care for dying patients without crippling fear while still maintaining a grounded perspective.

Past studies have indicated that for clinicians, death anxiety decreases with age, clinical experience, and religiosity [45,56,57], but our results suggest that several other factors also influence acceptance, including personal loss, mindfulness, and cultural background. In fact, the group of clinicians in our study who did not accept their mortality was diverse in age, clinical experience, and religiosity. Helping clinicians identify their discomfort with death is critical in terms of providing support and training.

Previous unidirectional studies have outlined how caring for dying patients can enhance clinicians’ sense of meaning and fulfillment [31,32], whereas others have emphasized the potential for existential distress and burnout [30,31]. Meanwhile, some studies have focused on how death anxiety [26,58,59] or spirituality [33,60] affects care of patients. Here we illustrate a broader array of factors, such as acceptance, self-reflection, and a positive outlook on life, that contribute to the relationship between clinicians’ perspectives on death and care of dying patients.

We identified two patterns for how clinicians’ attitudes affect and are affected by care of dying patients. One is positive and facilitative, whereby clinicians’ perspectives on life and death were enhanced and open conversations with patients about death were promoted. The other is one of detachment, in which there was no clear pattern between personal attitudes about death and professional responsibilities. We view the pattern of detachment as a missed opportunity, because often what clinicians bring to the patient encounter has to do with their own life experiences rather than learned techniques or skills [61]. Additionally, we did not observe a general pattern between clinicians’ self-reported communication styles and acceptance of their own death. A clinician could believe that their conditional acceptance of death helped them to speak about death openly with patients, while they still approached the topic indirectly.

We also noticed a disconnect between some clinicians’ self-reported communicative approach and the examples they presented, as some clinicians who espoused a direct communication style gave examples that instead comprised indirect language. Few clinicians reported talking explicitly about death and dying. Most, even those with a direct communication style, felt that phrasing such as “you are going to die from this” was too blunt to use in conversations with patients. Meanwhile, honesty was regarded as integral, even by those preferring an indirect style who felt that such an approach was more compassionate. However, there is a risk that patients may misconstrue a message delivered indirectly [3]. Taken together, these phenomena likely contribute to the discrepancy between what oncologists think they say and what patients understand [23] about how long they might have to live and the likelihood of cure.

An openness to discussing death, facilitated by a self-awareness of one’s attitude towards one’s own mortality, is an important ingredient in clear, effective communication at the end of life. However, the findings from this study suggest that it is not enough. While caring clinicians want to mitigate the effect of a potentially harsh message [6], they need to achieve an appropriate middle ground that avoids misleading the patient and be comfortable with discussing the uncertainty in predicting prognosis.

With increased public consciousness on confronting mortality [62], there is an urgent need to help health professionals communicate about death and dying using unambiguous, jargon-free language [3,63] that communicates commitment and compassion and that comfortably incorporates “d-words” [24,64–66]. One approach may be to say, “I worry that you might die from this sooner than either of us had wished” [67].

This study has several limitations. Our data was solely comprised of interviews, yet participants may not have been completely honest about or even conscious of their actual attitudes or behaviors. Also, we included clinicians of a single specialty from one geographical location. There may be differences in communication styles and views on death depending upon medical specialty and geographic region.

4.2. Conclusion

This qualitative study explores and describes oncology clinicians’ perspectives about death and identifies reciprocal interactions between personal attitudes about death and their care of dying patients. While many other factors contribute to why oncologists have difficulty discussing end-of-life issues with patients [9–12], a recognition of one’s mortality and “awareness of a shared mortality” [68] can allow clinicians to discuss death more comfortably with their patients while enhancing empathy and the clinician–patient relationship.

4.3. Practice implications

Communication with terminally ill patients remains problematic, in part due to clinicians’ unexamined attitudes about death –their patients’ and their own. Communication training has tended to emphasize behaviors rather than deeply-held attitudes and values, and our study suggests that, with regard to communication with patients at the end of life, both attention to communication and self-awareness of deeply-held attitudes are fundamental. Such training should help clinicians to become more aware of the schism between their intended (usually direct) communication and their actual (often indirect) communication behaviors. Training methods might utilize feedback based on interactions with standardized patients, role play, or review of recorded clinical encounters in addition to staff support groups, as described by Shayne and Quill, during which team members who regularly care for cancer patients have the chance to share experiences and dedicate time to self-reflection [69]. Such awareness and training can promote the kind of open, compassionate communication that patients desire.

I confirm all personal identifiers have been removed or disguised so the persons described are not identifiable and cannot be identified through the details of the story.

Appendix A. Questions from the semi-structured interview guide

Communication about death and dying

• What is a typical day like for you with regard to patient care in this facility?

• What issues are most important and most difficult to talk about with patients with advanced cancer who are dying?

• How do you decide when to be direct and use words such as “die” or “kill” versus being more indirect and using words like “shortened life expectancy” or “may not make it”?

• Have you ever been hesitant to talk about death and dying with a patient?

• Have you ever found yourself being more optimistic or pessimistic than you knew deep down you should be?

• Are existential issues ever brought up during these conversations, for example the patient asks what the meaning of his/her life was, or wishes he/she had done this or that instead?

Coping and caring for dying patients

• What experiences have shaped the way that you care for patients who are dying (for example, a single experience long ago, or cumulative experience over time)?

• How does patient death affect you?

• Has patient death ever made you think about your own mortality? In what way(s)? How has this affected your views about your own life?

• What has most helped you to “come to grips” with, or cope, while caring for dying patients?

Personal attitudes about life and death

• We all talk about ways in which patients cope with the inevitability of death. How about you?

• What do you believe has had the biggest influence on your present attitude about death?

• What thoughts or emotions come to mind when you think about your own death, for example peace, fear, curiosity, unresolved matters? Do you have any uncomfortable feelings about your own death such as fear or anxiety?

• In what ways have you/haven’t you come to terms with your own mortality?

• How does your perspective about death influence how you discuss information about death and dying with your patients?