“It’s like heart failure. It’s chronic…and it will kill you”: A qualitative analysis of burnout among hospice and palliative care clinicians

Abstract

Context

Although prior surveys have identified rates of self-reported burnout among palliative care clinicians as high as 62%, limited data exist to elucidate the causes, ameliorators, and effects of this phenomenon.

Objectives

We explored burnout among palliative care clinicians, specifically their experiences with burnout, their perceived sources of burnout, and potential individual, interpersonal, organizational, and policy-level solutions to address burnout.

Methods

During the 2014 AAHPM/HPNA Annual Assembly, we conducted 3 focus groups to examine: personal narratives of burnout; how burnout differs within hospice and palliative care; and, strategies to mitigate burnout. Two investigators independently analyzed data using template analysis, an inductive/deductive qualitative analytic technique.

Results

We interviewed 20 palliative care clinicians (14 physicians, 4 advanced practice providers, 2 social workers). Common sources of burnout included: increasing workload, tensions between non-specialists and palliative care specialists, and regulatory issues. We heard grave concerns about the stability of the palliative care workforce, and concerns of providing, high-quality palliative care in light of a distressed, overburdened discipline. Participants proposed anti-burnout solutions including: promoting the provision of generalist palliative care, frequent rotations on-and-off service, and organizational support for self-care. We observed variability in sources of burnout between clinician type and by practice setting, such as role monotony among full-time clinicians.

Conclusion

Our results reinforce and expand upon the severity and potential ramifications of burnout on the palliative care workforce. Future research is needed to confirm our findings and to investigate interventions to address or prevent burnout.

INTRODUCTION

There has been a recent surge of interest regarding burnout among palliative care clinicians. Burnout is a negative psychological state characterized by emotional exhaustion, depersonalization, and a low sense of personal accomplishment.(1) Left unresolved, burnout has been linked to hastened retirement,(2) poor professional conduct,(3) increased risk of medical errors,(4) as well as deleterious effects on personal health and relationships.(5, 6) Given constraints in the palliative care workforce,(7) burnout threatens the strength and sustainability of this discipline.

In 2015, we reported that 62% of clinician members of the American Academy of Hospice and Palliative Medicine (AAHPM) responding to a survey were burned out.(8) This rate, compared to other medical specialties, ranks among the highest.(1) Whereas a recent U.S. survey by Yoon and colleague found burnout rates among palliative care clinicians lower than ours (23%),(9) that study used a validated single-item measure of burnout, compared to our more exhaustive assessment that used the Maslach Burnout Inventory.(10) Nevertheless, although quantifying the prevalence of burnout within palliative care clinicians is important, there remains a need to understand the lived experience of burnout in this population as defined by clinicians themselves, as well as the sources of burnout, its protective factors, and downstream effects. To date, the few qualitative studies of burnout have been restricted to individual organizations, limiting a comprehensive understanding of this phenomenon.(11, 12)

Recognizing the urgency for interventions to address burnout, in 2016 Back and colleagues proposed a conceptual model for resilience promotion among palliative care clinicians.(13) Synthesizing the limited extant literature on burnout in palliative care, the authors posited that burnout and resilience are products of individual and workplace factors. For example, managing unrealistic personal standards for workload and pace (i.e., individual factors) are helpful insofar as they align with organizational structures and policies regarding patient census (i.e., workplace factors). Given that, to date, the majority of evidence regarding burnout in palliative care emanates from quantitative surveys, the literature is without rich, respondent-driven narratives that support or challenge constructs in conceptual models and related interventions.

In this study, we explored burnout experiences from the perspectives of palliative care clinicians representing multiple organizations across the U.S. – how they experience burnout, its sources, as well as their views on strategies to prevent or alleviate this phenomenon.

METHODS

Design

This paper reports on a qualitative follow-up study to our electronic survey that quantified burnout in palliative care clinicians who belong to AAHPM.(8) We used focus group interviews as this methodology allows for individuals to iteratively formulate their own ideas and use the group to produce voices that capture collective lived experiences.(14) We held three focus groups of palliative care clinicians at the 2014 Annual Assembly of the AAHPM. Groups were stratified by clinician type (i.e., physician or non-physician) to capitalize upon shared identity within groups. Our protocol was approved by the Duke University Institutional Review Board (Pro00052330).

Sample

We contacted individuals who had previously participated in our Internet survey to gauge their interest in participating in focus groups. Participants were eligible if they were palliative care clinicians (defined as physicians [MD/DO], physician assistants [PA], nurse practitioners [NP], or social workers [SW]), aged 18 years or older, and in clinical practice at least part time. In this manuscript, we use the term “palliative care clinician” to encompass both hospice and palliative care clinicians. Nevertheless, we delineate differences in findings between hospice and palliative care clinicians when appropriate. Given that we invited clinicians to participate regardless of whether they scored positively for burnout in the preceding survey, this study uses phenomenological methods to elicit participants’ experiences with burnout as a phenomenon that they themselves define, rather than individuals’ experiences with a socially-constructed “diagnosis” of burnout.

Data Collection

We developed an interview guide based on a literature review regarding clinician burnout, as well as insights gained from the aforementioned survey of AAHPM clinician members(8). We engaged participants in discussion regarding three topics: 1) experiences with burnout; 2) sources of burnout; and, 3) strategies to prevent or alleviate burnout (Appendix 1). The moderator used probes to explore responses with greater precision, as appropriate. As a way to spark discussion, we asked participants to independently write down responses to the following introspective “grand tour”(14) question: “What does burnout feel like?” The moderator used these responses as prompts for additional clarification throughout the rest of each group.

We pilot tested the interview guide with two palliative care physicians and one palliative care NP, revising questions as necessary to improve clarity. Each group was moderated by the first author, a health services researcher with extensive training in qualitative methods. An assistant took field notes of verbal and non-verbal data, such as key ideas and remarkable interactions.

Analysis

Focus groups were audio-recorded and transcribed verbatim. We used template analysis, a qualitative approach that combines content analysis and grounded theory.(15) This hybrid analytic approach allows for inductive (i.e., allowing themes to emerge from the data) as well as deductive (i.e., approaching the data with preformed hypotheses) reasoning; this approach is appropriate given that while qualitative studies exist on burnout in other specialties, very limited work exists vis-à-vis palliative care.(11, 12, 16–18) Two investigators (DK and DS) coded transcripts independently, meeting to discuss coding after each transcript. Using the constant comparative technique,(19) data were compared with previously-coded transcripts to ensure stability and relevance of themes. To ensure inter-rater agreement, we held coding meetings until both coders reached consensus on coding rules. Qualitative analysis was performed through a process of compound queries using NVivo10.(20) After each session’s transcript was coded, its corresponding set of field notes was reviewed by the first author to confirm the stability of themes identified. Because we hypothesized that burnout is a multilevel phenomenon, we applied a socioecological framework(21) to data analysis in order to disentangle relationships between emergent themes on four levels: personal, interpersonal, organizational, policy. We invited two participants from each clinician type to review a version of this manuscript’s results section and comment upon our interpretations (a.k.a. “member checking”).(22)

RESULTS

Of 1,357 individuals who participated in the quantitative survey, 383 (28%) expressed interest in participating in focus groups. Of those, we sampled the first 20 individuals who were able to attend one of the three 90-minute focus groups whose times had been predetermined. Two groups comprised physicians, and one group consisted of nurse practitioners, physician assistants, and social workers. Our sample was 65% female; overall, 55% of participants had practiced in palliative care for 0–5 years, 30% for 6–20 years, and 15% for ≥21 years (Table 1).

Table 1.

Focus Group Participant Demographics

Characteristic	n (%) N=20
Clinical role
Physician	14 (70)
Nurse Practitioner	3
Physician Assistant	1
Social Worker	2 (10)
Female gender	13 (65)
Marital status
Married	15 (75)
Single, never married	1 (5)
Partnered	3 (15)
Divorced	1 (5)
Children living in the home, yes	11 (55)
Age in years, median (IQR)	40.5 (35–56)
Practice setting where you work the most
Academic medical center/university	11 (55)
Private practice	2 (10)
Veterans Affairs hospital	1 (5)
Other	6 (30)
Duration of practice in hospice/palliative care, years
0–5	11 (55)
6–10	2 (10)
11–20	4 (20)
21–30	2 (10)
>40	1 (5)
Clinical care setting(s) in which you see patients1
Hospital	19 (95)
Inpatient hospice unit	8 (40)
Residential hospice	2 (10)
Nursing home/assisted living/SNF/LTC	9 (45)
Home visits	9 (45)
Outpatient clinic	8 (40)
Average hours worked per week in clinical hospice/palliative care	12 (60)
11–20	2 (10)
21–30	3 (15)
31–40	2 (10)
41–50	8 (40)
51–60	4 (20)
>60	1 (5)

Key: IQR, interquartile range. SNF, skilled nursing facility. LTC, long-term care.

¹= Totals and percentages may exceed 1 or 100% as participants could choose all responses that applied.

Personal Experiences with Burnout

Figure 1 offers a sample of participants’ verbatim responses to “What does burnout feel like?” Discussion regarding the experience of burnout generally reflected onset, timing, and duration, as well as effects on physical and emotional wellbeing. We heard descriptions of burnout as “insidious” with a gradual onset of affective and somatic symptoms; however, the cognition and description of the insidious nature was typically a retrospective reflection. Typically, participants became aware of burnout after an acutely stressful event (e.g., conflict with a non-palliative care colleague, a particularly traumatizing patient death). The affective and emotional aspects of burnout were often described as feelings of emotional exhaustion, boredom, anxiety, or irritability, whereas somatic aspects of burnout were reflected in participants’ physical experiences (e.g., fatigue, or sleeplessness).

Repeatedly, participants expressed surprise at experiencing burnout earlier than expected: “…I had anticipated that burnout was something that didn’t happen until after you had been doing your work for 15 or 20 years, and was surprised to find myself feeling burned out, and wondering how that happened…” (MD/DO, female) Participants earlier in their careers (often, younger in age, as well) particularly endorsed such sentiments.

Figure 1.

Focus group #1 participants’ responses to “What does ‘burnout’ feel like?”

Participants often used terms that conveyed fear or remorse regarding their experiences of burnout and its residual and sapping effects. For many, burnout was a phenomenon that, even when temporarily addressed, is never fully resolved:

NP: I think it’s like heart failure. I think it’s chronic, and it’s there, and if you don’t take care of it, it will kill you…. And it gets to a point that you can’t reverse it. (Male)

PA: I think that’s a good analogy because it has that cycle of you know, you do better, and then – but do you ever get back up to your top functional level? Um, that’s a great analogy. (Female)

NP: ‘Cause I think the memory will always be with you -- what that burnout felt like…. And you’ve lost that virginity, so to speak. You can’t go back. (Male)

We noted a general ambivalence (across groups) regarding whether burnout was a natural and expected phenomenon among palliative care clinicians that would “ebb and flow” over time, or if it was a “disease” which required aggressive intervention. Notably, participants in focus group #2, which predominately comprised older male physicians, dispassionately described burnout as endemic to and expected of palliative care practice, whereas participants in other groups discussed burnout as an unnatural outcome.

Sources of Burnout

Given the complexity of many palliative care patient cases, participants’ descriptions of the sources of burnout often included multiple themes in Table 2. Often this complexity was also expressed in an ambivalent manner – on one hand, the feeling of satisfaction, when producing benefit for patients and families, was restorative to clinicians and would, “fill up the tank.” Conversely, instances of perceived futility or inefficacy were described as exhausting. In general, participants expressed ambivalence about palliative care practice as a source of burnout. This overarching theme of ambivalence emerged in each focus group, and appeared to influence many other themes. For example, the gratification from providing care during an active death was a deep source of fulfillment; yet, active deaths that included unpredictability and traumatic elements, could spawn difficult emotions and professional fatigue.

Table 2.

Sources of burnout identified from focus group participants

Level	Theme	Exemplars
Individual	Lack of Boundaries	“…early in my career, very few boundaries. Um, a 32-year-old who died of metastatic breast. And, I drove to her funeral three hours away…Got to the funeral, really was important for me to go, really wanted to be there. Anyways, long story short, everyone left and I’m still there, crying at the gravesite. And her friends had to come up to me and say, ‘I guess we have to take care of you.’” (SW, female)
	Not Making a Difference	“Last week, um. It was probably burnout. I’d had two patients who had nothing but nausea and vomiting, intractable that I wasn’t fixing and it was really getting to me ‘cause in this profession, if you can’t make their symptoms better, then what are you there for?…I was feeling really, totally incompetent.” (PA, female)
	Self- Judgment	“I think when I was feeling most burnt out it was because I felt like I was putting in 110% of my energy, but the product was still inadequate. That my patients were not getting the care that I felt they deserved. And that’s what caused me to have a lot of moral distress. For me, there’s moral distress in the equation for burnout.” (MD/DO, male)
Interpersonal	Inter- specialty Conflict	“…But when you’re put off, and put off, and put off, and you know they’re saying other things to the family…and you’re trying to pin down that other provider to have that conversation, and it goes on, and you can see it goes on to the detriment of the family and the patient. Boy, that, that empties my glass really quick.” (MD/DO, female)
Organizational	Role Monotony	“…The doctors rotate every two weeks, at least where we are. And they have an opportunity to really do a whole lot of balancing different things, and um, they’ll often come in at two weeks and say, ‘How do you do this all year round?’ And that’s a good question. And at the same time it sort of irritates me. Like, why isn’t the standard for me, you know – for my discipline? It’s OK that you’re putting me in this work 40 hours a week, 24/7, but you get to do it three times a year in two-week shifts? So can we create something else for the rest of the team members that offers more creativity, more opportunities, more breaks to do other things so that when we come back in maybe we’re a little fresher?” (NP, female)
Policy	Regulatory Issues	“I spend a lot of my time, you know, just jumping through the hoop. And it, anyone who does hospice knows, it’s all regulatory….It’s like, this isn’t why I went into this. I didn’t go into this to worry that my hospice is going to go out of business because we’re providing care to dying patients who aren’t dying fast enough or under the [local coverage determinations]….If anything is going to drive me our of this profession, it’s going to be that.” (MD/DO, female)

Sources of burnout were often framed, implicitly or explicitly, as inherent to the nature and structure of palliative care practice. Themes broadly mapped on to four levels: individual (lack of boundaries, not making a difference, self-judgment), interpersonal (inter-specialty conflict), organizational (monotony and system stress), and policy (regulatory issues) levels (see Table 2). Furthermore, we identified a crosscutting belief that the nature of palliative care itself inherently leaves clinicians susceptible to burnout. One participant explained, “We’re not in a dentist’s office, with all due respect.” (SW, female) Throughout the groups, participants explicitly or implicitly spoke of palliative care as having inherent differences from other specialties (e.g., patient acuity, high-risk decision making, suffering).

At the individual level, participants articulated the struggle to maintain boundaries between their professional work (i.e., providing immersive, empathic palliative care) and their personal self. For example, one participant (SW, female) described an experience earlier in her career, “like it was yesterday,” when she “glommed onto” a young adult patient who eventually died of metastatic cancer and broke down “crying for every single patient I had lost… it was completely overwhelming.” A female physician described how a particularly complex case “drained” her, as well as her team:

“We had a patient in the ICU…he was young…He was in pain. We would call it ‘crying without tears’ ‘cause he would rock in the bed… And I probably took care of him about four or five weeks – my colleagues had stopped rounding on him….[they felt] there was ‘nothing more that [they] could do.’…We were splintered, we didn’t have a whole team that could take care of him. We had all these people — him being in agony and pain, his [situation being misrepresented to his family] and not being able to sit and talk with [his family and] a colleague as a multi-disciplinary group…I’d be up in the middle of the night thinking about him, and – I’m sorry, I’m getting a little emotional – and I would call the nurse practitioner [voice breaks] in palliative care, who, we would just talk, you know? ‘Cause it was so hard to take care of him [crying]. It just emptied all of our tanks…I was the only one who wrote notes on him but yet…it was a really hard one. [sobs] It burnt us all out.”

Participants spoke of close relationships with patients, which affected them long after the patient had died. Across groups, participants struggled to reconcile the balance between overly-protective boundaries and a natural inclination to immerse themselves in the patient and family’s experience. Further, we sensed a tension in the way our participants talked about who bears responsibility and agency for creating and enforcing those boundaries. On one hand, participants talked about needing to make boundaries themselves, yet also about their expectation that organizations need to change the structure of palliative care practice (e.g., role variability) so as to prevent burnout.

Participants described how instances of being unable to “fix” patients’ suffering were internalized as personal failures. When discussing perceived inefficacy, participants frequently mentioned feelings of self-judgment and disappointment with their performance as a clinician vis-à-vis patients, families, and colleagues. For example, one participant described feeling “incompetent” when patients were experiencing intractable symptoms. For others, the demands of managing multiple patients engendered feelings that they were providing impersonal or inadequate care. These perceived failures were often juxtaposed with the reflection that many participants chose to specialize in palliative care because they found fulfillment through alleviating suffering and in their strong bonds with patients and families during the “sacred” period of failing health.

Interpersonal, organizational, and policy-level influences were also presented as sources of burnout. For example, inter-specialty conflict was a significant source of frustration when participants felt there were conflicting priorities between palliative care and referring clinicians, resulting in “the detriment of the family and the patient.”

On an organizational level, role monotony (i.e., one’s constant exposure to witnessing patient/family suffering without opportunities to diversify their professional experiences) led to faster and more severe burnout. This theme was most prevalent among non-physician clinicians, where participants contrasted their experiences with physicians. One participant asked, rhetorically, “It’s OK that you’re putting me in this work 40 hours a week, 24/7, but you get to do it three times a year in two-week shifts?” (NP, female) This theme was also prevalent among full-time hospice clinicians without academic appointments, physicians and non-physicians alike.

Finally, the regulatory and reimbursement-related pressures faced by palliative care and hospice practice emerged as common sources of burnout. Frustrations with Medicare policy (e.g., local coverage determinations, audits) were particularly resonant in group #3, which comprised the largest proportion of hospice clinicians.

Effects of Burnout

We heard descriptions of perceived burnout as pervasively affecting participants’ lives personally and professionally (Table 3). On the individual level, these effects fell into three categories: psychological, physical, and clinical performance. Group discussion focused on how burnout harms patient care, in a variety of ways including impaired empathy, decreased engagement with patients and families, and quality of care. Burnout also had implications for interpersonal communication with colleagues, including unsympathetic interactions with trainees.

Table 3.

Effects of burnout identified by focus group participants

Level	Theme	Exemplar
Individual	Psychological Effects	“…I think it was last year – I think my colleagues were at the New Orleans [AAHPM] conference and I was the guy. Staying at work…It was a coincidence where, [my colleagues] were both out-of-town…for like three weeks in a row. And I had a particularly difficult death where somebody, um, uh, was gasping, um, and I remember his son yelling at me like, ‘This isn’t how it’s supposed to be!’ And uh, so what I found over the period of days after that was I really didn’t share it with people, and then I was having flashbacks of them gasping. And it was making me really fearful of doing other uh, terminal or compassionate extubations…” (MD/DO, female)
	Clinical Performance	“…For me, [burnout is] just a switch from the enthusiasm of practice…it’d be like going to do a visit with a patient and not having any interest. I must do this. This is part of my expectation, but I just don’t want to do this. I’d rather be anywhere else.” (MD/DO, male)
Interpersonal	Interpersonal Relations	NP: …Somebody had died on the hospice unit and I went in there with…an NP student there, and I just remember talking clinically about death and blah blah blah and we were by ourselves. And she actually had a feeling. I was so removed from um, recognizing that this was a person who had died, and this was really an important and sacred moment, and I just kind of blew through that piece. (female) SW: And you said she had a feeling? What was her feeling? How did you know she had a feeling? (female) NP: Well, she was quiet, and a little sad, and she was just in a different place. And I was in my, ‘Yeah, another dead body,’ you know, ‘Here’s what we know, here’s what we do’ place. And I just went, ‘Ooh, that’s not how I want to be, and that’s not how I want to teach.” (female)
Organizational	Leaving the Field	“I think [burnout] should be looked at as a workforce issue. I think that that’s got the attention of the [American] Academy [of Hospice and Palliative Medicine], uh and it should have our attention. ‘Cause we’re going to lose a lot, a lot more people. And we don’t have any to lose…” (MD/DO, male)

Participants described burnout as an invasive feeling, as “…[burnout] attacks my identity of who I am and kind of, how I define myself.” (MD/DO, female) Participants also felt that burnout robbed them of their ability to provide the patient care that they were accustomed to providing. In the groups, quality of care was framed as a function of the clinician’s empathy and ability to immerse oneself in the patient’s total experience. In contrast, burnout gave rise to feelings of apathy and detachment. One participant explained, succinctly, feeling like “I’d rather be anywhere else [than visiting a patient].” (MD/DO, female)

The chronic effects of burnout on organizations’ ability to provide quality palliative care was commented on by all groups. We repeatedly heard personal or witnessed narratives related to “leaving the field” or “getting out while you can.” There was consensus in all groups that burnout is not simply an individual-level phenomenon; rather, if not addressed, burnout impacts the entire organization and discipline. As stated by one female Social Worker, “…when we lose a member of the team, the whole team falls apart.”

Factors Protective against Burnout

Across the groups, participants identified individual-, interpersonal-, organizational-, and policy-level factors they felt were protective against burnout (Table 4). These factors were relatively consistent across groups. Groups referred to a number of self-regulation strategies, such as enforcing work/life boundaries and proactive self-care (e.g., mindfulness, psychotherapy).

Table 4.

Protective factors against burnout identified by focus group participants

Level	Theme	Exemplar
Individual	Self-regulation and protective strategies	“My personal rule is that at four o’clock, I don’t answer a new consult. I finish what I’m doing, and I’m out the door at five. It hurts some days – I really don’t like doing that. But I know that if I don’t do that, I won’t be at work again the next day. And my colleagues, sometimes, some of my colleagues have trouble with that. My medical director, my direct supervisor, pushes me out the door. And I thank God for him.” (NP, male)
	Positive impact	“…A really good family meeting. One where you go and you’ve felt like you’ve really connected, and you walk out and you feel like, ‘Gosh, I just made a real difference.’ And that can fill up a tank…” (MD/DO, female)
Interpersonal	“Safety in Numbers”	“I think there’s safety in numbers. I did my fellowship at [institution], and there’s a huge team of people and it’s completely different when, you’re not the only one who’s not finishing at the end of the day. But to be in a system where you’re the sole person, and it’s you not finishing every day – you take it a lot more personal, and it’s a lot harder. Or you’re on weekends on call, and you’re the only one there that weekend, it’s a lot harder. Um, and having someone else take call is even, is important also, ‘cause I was in that situation where I was the sole call person, now I’m just a preferred call person. [group laughs]” (MD/DO, female)
Organizational	Culture change	“…It’s something we’re working on, to change the culture so that the palliative care service doesn’t provide the palliative care. The hospital provides the palliative care, and the palliative care service comes and aids, or directs…” (NP, female)
Policy	AAHPM’s role	“I’m just concerned. You know, I didn’t, I agree, I didn’t realize that burnout was such an issue. And you know, I think it sounds like it’s going to have to be addressed. I don’t think we can hide it. You know, I think we need to be – the Academy [of Hospice and Palliative Medicine] needs to be proactive…I think we need to go beyond these [focus group] sessions and there has to be some concrete plan to address it…” (MD/DO, male)
	Regulatory	“I think that there’s certain structural components that force us to do things a certain way. So for instance, I’m listed as the attending at the hospice house so that I can bill. If I’m not listed as the attending, I can’t bill, and if I’m listed as the attending, then none of the other nurse practitioners can bill. So I’m stuck at the hospice house. Period….So, we’re almost chained to our roles, so we don’t have that flexibility. I don’t know if y’all could fix that?” (NP, male)

Other individual-level protective factors included reflection on the positive impacts of the care they provided, for example, or in their interactions with patients’ family members. Interpersonal support and the notion of “safety in numbers” (i.e., the size and strength of one’s professional network) was also important; participants often commented that colleagues were important sources of support, both formally (e.g., Balint groups) and informally. Participants also noted that they tried to seek support from colleagues, rather than their own families at home, as a means of buttressing a separation between their work and personal lives. Participants spoke of the need to maintain boundaries, and expressed the tension between a sense of obligation and a cognitive awareness of the risk for burnout.

Groups also elaborated protective factors at organizational and policy levels. For example, it was important to remind themselves that other providers in the care team could be equally competent, and that, “I’m certainly not the only one” (SW, female) capable of providing quality care. Participants discussed the potential for changes in organizational culture such as the dissemination of primary palliative care so as to defray excessive burden on specialty palliative care teams.

Reflecting on the influence of groups such as AAHPM to address burnout, participants believed that AAHPM could have a trickle-down effect on the organizational culture and structure of palliative care, for example, by “normalizing those sorts of policies.” (SW, female) Changes to the administrative and regulatory structure of palliative care were necessary, for example, so that providers were not “chained to our roles” (NP, male), again relating to the theme of role monotony.

DISCUSSION

To our knowledge, this is the first multi-institutional qualitative study of burnout among palliative care clinicians. We documented sources of burnout on four interrelated levels: individual (lack of boundaries, feelings of “not making a difference,” and self-judgment), interpersonal (inter-specialty conflict), organizational (role monotony and system stress), and policy (regulatory issues) levels. Participants also noted a variety of potential solutions to burnout, also corresponding to the aforementioned four-level hierarchy: individual (self-regulation strategies, reaffirming one’s positive impact), interpersonal (relying on colleagues for “safety in numbers”), organizational (culture change to reduce burnout), and policy (appeals for AAHPM’s involvement to promote resilience, and regulatory changes such as reimbursement policies).

Our study provides empirical support for the key feature of Back and colleagues’ conceptual framework of resilience promotion – that burnout and resilience are not exclusively individual-level phenomena; rather, the role of organizational and policy-level influences also must be considered.(13) Participants frequently spoke of the need for organizations to adopt policies that might reduce burnout, such as the promotion of primary palliative care, such that the rapid growth of palliative care does not overtax the already constrained palliative care workforce. Similarly, we identified that the clinical responsibilities of clinicians in the field are unique among roles, resulting in differential sources of, and experiences with, burnout. For example, participants in full-time clinical practice, such as community-based hospice clinicians, lamented the inflexibility of their roles, comparing themselves to academic counterparts who could diversify their workload with administrative or teaching responsibilities.

Clinicians reported downstream effects of burnout that have important personal and professional consequences. We heard concerns regarding psychological effects, clinical performance, exacerbations of interpersonal conflict, and a concern regarding individual professional viability. Further, we heard that the burden of burnout is shared by the entire palliative care team. Our previous survey demonstrated that a leading reason palliative care clinicians report possibly leaving the field in the future is “burnout”, followed by “retirement.”(8) Evidence regarding the “contagiousness” of stress and burnout are growing, highlighting that the suffering of an individual clinician does not occur in isolation.(23) These findings reinforce the notion that resilience interventions in palliative care should be situated within organizations as a way to address systemic issues but also to reinforce the notion that resilience is just as a natural component of clinical practice as is patient care.(13)

Participants in our sample described the central role that boundaries, or lack thereof, contributed to burnout. This important theme was also documented by another recent study of palliative care clinicians at a single institution.(11) Many clinicians find themselves fervently attracted to palliative care due to a powerful personal or professional experience with a loved one or patient with serious illness. Yet, this passion, if not paired with the discipline of setting and enforcing boundaries, may present challenges to professional sustainability.

Several limitations merit consideration. Given the non-probability convenience sampling used in this study, our findings may not generalize to the overall population of palliative care clinicians. Furthermore, our participants were attendees of the 2014 AAHPM Annual Assembly leading to potential selection bias; notably, the perspectives of clinicians who were unable to attend are missing from our analysis. These individuals may arguably have different experiences than those of clinicians who have the means and flexibility to attend the Annual Assembly. Additionally, it may be that participants in our sample may have higher burnout, another form of selection bias. Further, given that we did not restrict our sample to clinicians with a positive screening for burnout, these results should be interpreted as clinicians’ experiences with their own construction of burnout, rather than one’s experiences with a “diagnosis” of burnout. Although ideally we would have allowed ourselves the opportunity to conduct additional groups, we were constrained by logistical limitations at the Annual Assembly such as room availability. Lastly, the use of focus groups (versus individual interviews) may have affected the type of data generated. Given group dynamics, it is conceivable that focus group participants have less time to build rapport with the moderator and with fellow participants and may therefore be less inclined to disclose intimate or potentially taboo information. Although this is a concern in any focus group, the amount of rich, moving, and deeply personal data offered by many participants in our groups dampens these concerns.

Our work advances our understanding of burnout experiences in palliative care beyond prior studies. First, many prior studies have focused on prevalence and severity, often measuring current burnout through closed-ended surveys that typically constrict analyses to a predetermined definition of burnout, which may vary across individuals and specialties.(1, 18, 24, 25) Further, nuance may be lost with use of dichotomous constructs (i.e., either one has burnout or does not) as optimal characterizations of burnout occur along a continuum, better explored using phenomenological methods as we did. Lastly, prior qualitative studies have investigated burnout experiences among clinicians within a single organization,(11, 12) thus limiting generalizability, or used cursory methods that precluded rigorous examination.(16)

Although palliative care practice is inherently taxing, burnout threatens the future of a sustainable workforce to meet the needs of our society’s frailest individuals. Our findings suggest that stressors experienced by palliative care clinicians represents one of the greatest threats to the sustainability of our field. The remarkably high prevalence of burnout directly endangers the delivery of compassionate care to all with serious illness, owing to the existing palliative care clinician workforce shortage(7) and the predicted looming loss of half the workforce over the next decade.(8, 26, 27) Future research is needed to identify feasible interventions to improve resilience and address burnout, as well as to explore the presence of burnout among palliative care clinicians who were not included in this study (e.g., registered nurses).

Appendix 1. Focus group interview guide

1. Let’s find out a little bit more about everyone here. How about we go around and state your names, where you live, what your roles are in hospice and palliative care (HPC), and for how long you have been practicing in HPC.

2. What does the term “burnout” mean to you? How do you define it?

   1. Self-reflection activity

      1. Using these markers and the large Post-It sheets on the wall, please take a few minutes to respond to these two questions:

         1. “What does burnout feel like?”
         2. “What gets you through burnout?”
   2. Is burnout an acute, discrete event for you, or is it more of a chronic feeling?
3. Think back over all of the years that you’ve been in HPC. Please tell us an example of a stressful event when you felt burnt-out.

   1. What about that experience made you feel burnt-out?
   2. What happened to help you get through the burnout?
4. From what or where do you find support to help you get through feelings of burnout?

   1. How open are you about your feelings when you’re feeling burnt-out?
   2. Who do you talk to when you feel burnt-out?
   3. What do they tell you?

5. Let’s suppose that you are mentoring a medical student who is interested in a career in HPC. What would you tell him/her to protect him or herself from burnout?

6. How do you think that burnout in HPC relates to burnout in different medical specialties? How is burnout in HPC different than, or similar to burnout in other specialties?

   1. How prevalent do you suspect that burnout is among your HPC colleagues?

7. We’ve spent a lot of time this [morning/afternoon] discussing burnout, what it is, and how it affects you and your colleagues. Now, I want to give you all the opportunity to brainstorm as a group about strategies to try and mitigate burnout. I’ll invite you all to think about strategies on any of three different levels: individual-level strategies, local organization/agency-level strategies, and specialty-wide strategies.

8. Of all the things we’ve discussed today, what to you is the most important?

   1. What questions would you have asked if you were me? What have I missed?