Interpersonal Conflict between Clinicians in the Delivery of Palliative and End-of-Life Care for Critically Ill Patients: A Secondary Qualitative Analysis

Abstract

Background:

Conflict between clinicians is prevalent within intensive care units (ICUs) and may hinder optimal delivery of care. However, little is known about the sources of interpersonal conflict and how it manifests within the context of palliative and end-of-life care delivery in ICUs.

Objective:

To characterize interpersonal conflict in the delivery of palliative care within ICUs.

Design:

Secondary thematic analysis using a deductive–inductive approach. We analyzed existing qualitative data that conducted semistructured interviews to examine factors associated with variable adoption of specialty palliative care in ICUs.

Settings/Subjects:

In the parent study, 36 participants were recruited from two urban academic medical centers in the United States, including ICU attendings (n = 17), ICU nurses (n = 11), ICU social workers (n = 1), and palliative care providers (n = 7).

Measurements:

Coders applied an existing framework of interpersonal conflict to guide initial coding and analysis, combined with a flexible inductive approach allowing new codes to emerge.

Results:

We characterized three properties of interpersonal conflict: disagreement, interference, and negative emotion. In the context of delivering palliative and end-of-life care for critically ill patients, “disagreement” centered around whether patients were appropriate for palliative care, which care plans should be prioritized, and how care should be delivered. “Interference” involved preventing palliative care consultation or goals-of-care discussions and hindering patient care. “Negative emotion” included occurrences of silencing or scolding, rudeness, anger, regret, ethical conflict, and grief.

Conclusions:

Our findings provide an in-depth understanding of interpersonal conflict within palliative and end-of-life care for critically ill patients. Further study is needed to understand how to prevent and resolve such conflicts.

Introduction

Complex medical decision making occurs frequently in the intensive care unit (ICU), which often necessitates interdisciplinary collaboration and advanced communication skills to provide high-quality patient care. This process of complex care delivery within ICUs is commonly fraught with conflict. Previous literature shows that major sources of conflict within ICUs center around disagreements about goals of treatment, end-of-life decision making, and inappropriate or inadequate communication.^1–9 “Interpersonal conflict” is defined as “a dynamic process that occurs between interdependent parties as they experience negative emotional reactions to perceived disagreements and interference with the attainment of their goals” and is characterized by the concomitant occurrence of disagreement, interference, and negative emotion.¹⁰ This type of conflict may threaten clinicians' sense of ego and proficiency, with negative impacts on their mental health, job satisfaction, and ability to perform at work; it may also compromise patient care and safety and increase medical errors.^8,11,12

While providing palliative and end-of-life care is an essential aspect of caring for critically ill patients,^13,14 delivery of this type of care may be prone to conflict within the ICU and postsurgical settings. Conflict may arise due to the “rescue” culture of the ICU that is primarily oriented to prolonging life, as well as “surgical buy-in,” an implicit contractual relationship between surgeons and patients through which surgeons feel obliged to advocate for patients to survive postoperatively.^5,15–18 The high levels of stress and strong emotions that are often present in high-acuity situations may also contribute to conflict and potential for medical errors.^1,19 Despite this, few studies have characterized interpersonal conflict between clinicians in the delivery of palliative care in the ICU. Qualitative methods are particularly well-suited to advance work in this area because it has not been previously examined in-depth within health care settings, and it may be a sensitive topic for clinicians.

Thus, we applied a deductive–inductive approach to the analysis of existing qualitative interviews among ICU and palliative care clinicians with the goal of describing interpersonal conflict within the context of delivering palliative and end-of-life care to critically ill patients. In this study, we focus on interpersonal conflict between clinicians, including between ICU and palliative care clinicians as well as between ICU team members.

Materials and Methods

Description of parent study and data collection of parent dataset

The goal of the parent study was to inform the development of a theoretical framework for the process by which specialty palliative care is adopted within ICUs.²⁰ Briefly, 36 participants were recruited from two large academic medical centers in New York City with multiple types of ICUs, and consisted of ICU clinicians (including 17 attending physicians, 11 nurses, and 1 social worker), as well as palliative care providers (4 attending physicians, 1 nurse practitioner, 1 social worker, and 1 chaplain); full participant demographics are shown in Table 1.²⁰ Center 1 had a relatively new specialty palliative care service (established within 5 years of study start) with 8–12 core palliative care clinicians throughout the study period; Center 2 had a mature, large palliative care service (in existence for >10 years at the start of the study) with more than 20 core palliative care clinicians in all the years of study.

Table 1.

Participant Characteristics (n = 36)

Age, median (IQR)	35 (33–42.5)
Self-identified gender, n (%)
Female	21 (58.3)
Male	15 (41.7)
Race, n (%)
White	22 (61.1)
Asian	13 (36.1)
Other	1 (2.8)
Ethnicity, n (%)
Hispanic/Latino	0 (0)
Not Hispanic/Latino	36 (100)
Clinician type, n (%)
ICU attending physician	17 (47.2)
ICU nurse	11 (30.6)
ICU social worker	1 (2.8)
Palliative care cliniciana	7 (19.4)
Years in practice, median (IQR)	5.5 (4–11.5)
Type of ICU, n (%)b
Surgical	17 (47.2)
Neurological	6 (20.7)
Medical	6 (20.7)
Center, n (%)
Center 1	16 (44.4)
Center 2	20 (55.6)

Original table from parent study.²⁰

^a Includes four physicians, one nurse practitioner, one social worker, and one chaplain.

^b Only includes the 29 ICU clinicians. “Surgical” comprises surgical, cardiothoracic, and transplant ICUs.

ICU, intensive care unit; IQR, interquartile range.

Both centers had teams that comprised physicians, nurse practitioners, chaplains, social workers, and music therapists. On average, over the study period, Center 1 had ∼1000 consults per year with 32% of consults coming from the ICU, while Center 2 had ∼1500 consults per year with 20% from the ICU. Among the different ICUs in each center, use of palliative care specialists was variable.²⁰ The study was approved by each center's institutional review board (IRB-AAAQ9525 and HS no. 18-00807). One-on-one, semistructured interviews were conducted in-person by two members of the research team (L.D.F. and M.H.) between 2016 and 2020. Interviews were 44 minutes on average, audio-recorded, and transcribed verbatim for analysis.

The interview guide comprised open-ended questions to explore the current state of palliative care delivery in the ICU, identify barriers and facilitators to specialty palliative care adoption, and elicit participants' views on the use of palliative care specialists (Supplementary Data S1 and S2).

Questions focused on the relationship between ICU and palliative care teams, and probes were used to elicit clinicians' positive and negative experiences in delivering palliative and end-of-life care. For example, participants were asked, “How would you describe the current relationship between ICU providers and palliative care providers? Can you give some examples of the interactions between the ICU team and the palliative care team?” “How did/could the palliative care team earn your trust? What happened to make them not trustworthy?” “Do you have any other personal experiences or anecdotes of using the palliative care consultation service/delivering palliative care that you would want to share?”

As interviews progressed, two members of the research team (M.H. and L.D.F.) coded all the interview transcripts using an inductive and iterative process with constant comparison, meeting regularly (every two to three interviews) to reach a consensus on codes and to revise the interview guide as necessary. Themes were refined until theoretical saturation was reached with no new themes emerging. The team performed “member checking” by sending transcripts and the final theoretical model to select participants to check for accuracy and validity.

Definition of terms

We used an existing framework from Barki and Hartwick¹⁰ that has been previously applied to examine interpersonal conflict in health care²¹ to guide our initial coding and analysis. Within this framework, disagreement occurs where there is a divergence of opinions, values, or goals, while interference refers to behaviors that prevent a task from properly being carried out. Negative emotion is the presence of a strong and undesirable affect, which often manifests as fear, anger, and frustration. Interpersonal conflict can be task-related (i.e., regarding what should be done in a task [task content] or how a task should be done [task process]) or nontask-related (i.e., arising from interpersonal relationships). For this study, we defined the “task” as patient care in the context of palliative and end-of-life care delivery.

Secondary data analysis

Using existing data from the parent dataset described above, we conducted a secondary thematic analysis^22–24 using a deductive–inductive approach to characterize instances of interpersonal conflict. In 2021, three investigators (W.T., L.D.F., and M.H.) identified a distinctive subset of codes from the parent dataset that could plausibly be associated with instances of interpersonal conflict, including codes related to team dynamics, communication, conflict, disagreement, barriers, and negative experiences (deductive approach). From these codes, these investigators independently identified quotes relevant to or describing conflict between clinicians in the context of palliative care delivery in the ICU, using a flexible approach where new codes were allowed to emerge (inductive approach) and an iterative and collaborative process to develop themes.²⁴

As themes continued to emerge, additional quotes and codes were chosen for review if deemed necessary. Themes were developed and refined by comparing and combining conceptually similar concepts until a consensus was achieved among all investigators. We followed the Consolidated Criteria for Reporting Qualitative Research (COREQ) guidelines (Supplementary Data S3).

Results

We identified instances of all three properties of interpersonal conflict (disagreement, interference, and negative emotion) in the context of delivering palliative and end-of-life care to critically ill patients. We did not identify any direct occurrences of nontask-related conflict (i.e., conflict between individuals unrelated to patient care). We identified themes describing the ways in which participants experienced these properties: disagreement about the appropriateness of palliative care in the ICU setting; disagreement about which care plans should be prioritized and how care should be delivered; interference with patient care; and negative emotions around experiences related to palliative care consultations. Representative quotes are included in Table 2.

Table 2.

Key Themes Related to Properties of Interpersonal Conflict within Palliative and End-of-Life Care Delivery for Critically Ill Patients

Disagreement about the appropriateness of palliative care in the ICU setting	There were a lot of trust issues. Pushing too fast. Palliative care believing we needed to be doing one thing. The ICU team thinking we need to do another. The surgical (team) upset that we even got palliative care because for many physicians, the label, as you know, has been labeled as like “you're giving up.” (ICU Attending, P1) A couple of weeks ago I took care of an older guy with cancer who agreed to go through this major operation. It didn't go great but it also didn't go terrible, so it wasn't clear to the surgeon that the guy was dying, but it was clear to the patient and family that he didn't want to go through this anymore. And I think the surgeon felt like taking some more time, like taking a couple more weeks in the ICU, would have helped. The patient and family were like, “No, this is too much, I don't want to do this anymore.” That conflict between patients and doctors can be really intense… They ended up going to the palliative care unit actually, but after I had to basically have a couple of fights with the surgeon on the family's behalf and it was one of the more intense situations in the ICU. (Palliative Care Attending, P25)
Disagreement about which care plans should be prioritized and how care should be delivered	I remember having a debate with a surgical ICU attending who wanted us consulted for goals of care, but every time I went to see the patient, all he wanted to talk about was his pain the whole time, his pain wasn't being treated. So I kept recommending opioids to the SICU attending, and he kept telling me no, he wasn't going to give them because he was trying to extubate the patient and he didn't want to cause respiratory depression. So we just had this ongoing debate and I would keep going and seeing the patient but without being able to get past this pain issue, I wasn't successful in my goals of care discussion because I wasn't able to build the trust of the patient by getting that symptom managed for them. (Palliative Care Attending, P24) A patient was supposed to undergo a major hepatectomy, but the operation was canceled because the patient was considered inoperable. The patient went back to the ICU and was extubated, and I called palliative care because I felt this patient should get to go home and needed to have some idea of hospice care or dying at home and be able to die at home in peace. Instead, the palliative care consult suggested to start the patient—who was perfectly awake and strong and fairly normal until she came to the hospital—to start a morphine drip, right then and there, basically to let her die then. I refused that and later found out when she went to the floor from the ICU that they started a morphine drip from the floor, and four days later she died. I felt like they were pushing her and that was not the reason why I had consulted her for a palliative care consult. I thought she had a really good chance to go home and die with support and hospice care or whatever it needed with the family instead of having to die in the hospital. So that was—it made me very kind of bitter about palliative care. (ICU Attending, P11) This specific patient was on the phone making her own plans of her death… She was probably the most self-aware and most amazing patient I probably ever met… She had such a clarity and comfort with the situation, like, “…There are things I can do to prolong my life, but I want to be comfortable and there is no reason why I can't plan for how and when I'm going to die and how my post-death process can be taken off my family's shoulders.” Lucid, communicative, making these decisions. And a palliative care consultant came in and suggested we sedate the patient… And I was agape, and I said, “She has choices, and she is making choices and her choices are how she wants to live.” …It was the most amazing way of looking at long-term death and for a palliative care consultant to get irate that we were allowing her to experience this pain in this emotional state of horror, it just blew me away. That someone who I think is incredible and strong and was teaching me how death should be really looked at and saying that I'm doing the wrong thing by letting her experience this pain and living though knowing she has days to weeks to months to live… Hearing the differences, it really left such a dramatic impression that someone can have a totally different understanding of this woman. (ICU Attending, P10) One issue that I found disturbing with palliative care is that they have very formalized ways of discussing with families and they do not necessarily appreciate the differences in families, so they use specific phrases that they learn in their books that they should use. My favorite one is when families—say, like, someone religious—they (palliative care) say like, “Oh I really respect your religion.” I don't know what that means. I think it's a meaningless thing that they read in their books that you're supposed to say.” (ICU Attending, P11) The ICUs are very fast paced, so the decision-making was taking too long and we were needing those beds, so we basically just said, “No, we can't wait for…” So that's the disconnect, and that disconnect led to ICU physicians taking control again. (ICU Attending, P16) A barrier to that is that because we're (palliative care) all over the hospital, we ask them (ICU team) to set up a time for the family meeting, and I think because of the work flow of the ICU is such that they have the family meeting whenever the opportunity (arises), like, “Okay the family's here, we're just getting off rounds, let's go have this meeting quickly.” I think our workflows are different, so I take a longer time having the family meeting. If I don't set a time for a family meeting, I'm going to be in a family meeting somewhere else. The MICU's preferred style of having the meetings, it tends to be a little faster. And because of that, when I do get the consult, things have been missed. (Palliative Care Attending, P24)
Interference with patient care	I'm not going to call a palliative care consult if the surgeon and cardiologist does not want them involved. So you are kind of playing a little bit of politics too…you really just have to make sure that all health care providers are ready, especially when you are working with people who have had a very long-term relationship with the patient, such as a cardiologist who's known their patient for 15, 20 years—it's almost disrespectful for me to go against their wishes. When I've only known the patient for a week. (ICU Attending, P12) Like I said, we still have to get permission from the surgeons… that is a huge obstacle for us to come in… sometimes, we get the request that we come in but don't talk about the goals of care, but specifically manage just supporting the patient and the family. But sometimes it's challenging to address the quality of life without addressing the prognosis. So those are the tough situations. (Palliative Care Attending, P9)
Negative emotions around experiences related to palliative care consultations	I have suggested palliative care, but I haven't said “palliative care,” I've just said, “Do you think we need to have a goals of care discussion?” And that attending has said, “Well, what do you mean?” And I have said, “Well maybe we could do comfort measures,” and I've been scolded for that. (ICU Nurse, P8) I recall one attending who got very upset and said, “Never interrupt my rounds again.” That's why we started coming after lunch because once we got totally bitten off and she never gave us a consult again after that. We had times when we were trying to explain to an attending who we were, and he walked away from us while we weren't even done, we haven't even finished talking. And he walked away from us. (Palliative Care Social Worker, P3) “You're fired from the care, I don't want you talking to this patient. You talk to the family, and we'll report you to patient services. Do not come in here again. Do not speak to the patient's family, do not do anything. And if you do, I swear to God I am going to have repercussions.” … The intensivist who was directly involved—to say he was irate didn't quite describe it. He was like beyond levels of anger, like nuclear levels of “get out,” like “how dare you.” (ICU Attending, P10) Whatever the plan is, I think all physicians involved—especially if it's a withdrawal or end-of-life—need to have an agreement. And as much as we have protocols in place, no protocol is perfect… I definitely remember one very difficult case for me, and I think about this all the time. I think we were very heavy-handed, I think the family didn't have enough time with the patient, and I regret it… The palliative care team recommended—I don't exactly remember the doses of the drugs—but it was a benzodiazepine, an amnestic agent, and something for analgesia, and the doses that they suggested were very, very high… In retrospect, I think the drugs that we used were very high doses and I don't think it needed to be done that way. I think it could have been much more gentle and smooth. And personally, I really did feel like what we gave ended the patient's life almost immediately before we turned off the device. And I will never forget that case. My fellow was traumatized, I was very upset… We need all physicians to be comfortable with how this happens… These are experiences that stay with you for a long time, and you never want to feel like they went wrong. (ICU Attending, P12) I think moving to the palliative care unit and enrolling in hospice can be great things when it actually is the choice that they (the patient) or the family is making, but moving somebody who is at the end of life just to decrease a number honestly is kind of disgusting, and it's not really providing anything to the patient and family. I think it feels really gross on our part that we're involved in that, and we sort of hold the power. Somebody cannot move to our unit without us being involved and saying yes to that, so there's often a lot of pressure coming from the ICU, a lot of phone calls and pages and people getting very agitated like, “You need to take this patient right now,” so that feels really bad. (Palliative Care Attending, P25) Unfortunately, I think that where I work, many of the health care providers become very deeply involved in their patients. You know, they spend 10–12 hours in the operating room night after night after night and sometimes when we consult palliative specialists, those physicians are also grieving in a way, and I think that it is something that some people don't acknowledge… I've certainly seen my colleagues grieving for their patients and being unable to let go and having that be a barrier to involving palliative specialists… Sometimes it's difficult for them to acknowledge that at some point, they have reached a point where it's kind of futile, and what we are doing is really causing more pain than we are offering a meaningful solution. (ICU Attending, P21)

MICU, medical intensive care unit; SICU, surgical intensive care unit.

Disagreement about the appropriateness of palliative care in the ICU setting

Participants described ways in which clinicians from different teams, including the ICU, palliative care, or prior primary (e.g., surgery, cardiology) team, disagreed with one another with regard to palliative and end-of-life care delivery. Often, clinicians disagreed about prognosis, which impacted their view on whether a patient was appropriate for a palliative care consultation. This disagreement was often most intense with respect to surgical patients. ICU clinicians were more likely to advocate for palliative care when they perceived the patient to have a guarded or poor prognosis. In contrast, surgical teams often believed that patients would benefit from longer postoperative recovery times in the ICU and thus disagreed with involvement of palliative care. A palliative care social worker described situations in which social workers and nurses believed that patients and family could benefit from palliative care, but were met with “the all-famous phrase that we hear a lot: ‘Oh, we are not there yet’” in response.

Another source of disagreement was the perceived goals of ICU care, in that clinicians were often seen as single-minded in terms of being curative or being palliative. Both palliative care clinicians and ICU nurses described how they thought some ICU attendings operated on the belief that the focus of the ICU is primarily to “treat and get people better as fast as possible,” with the perceived trade-off that “quality of life [while in the ICU] isn't priority.” These palliative care clinicians and ICU nurses disagreed with this view of ICU care, remarking that “people don't always understand that you can have palliative care and critical care at the same time.”

Disagreement about which care plans should be prioritized and how care should be delivered

Even if clinicians agreed to proceed with palliative care consultation for a patient, instances of interpersonal conflict often involved disagreements on which care plans should be prioritized (particularly when goals were perceived to be mutually exclusive), as well as how care should be delivered. Both palliative care and ICU clinicians described multiple instances of disagreement over care plans, which often dealt with administration of opioids for pain control and their adverse effects. Whereas palliative care clinicians tended to prioritize the goal of improving patient symptoms and experience, ICU clinicians prioritized physiologic goals such as blood pressure, spontaneous breathing, and liberation from the ventilator.

In instances of end-of-life care, palliative care clinicians often prioritized pain control, while ICU clinicians who were concerned about hastening death or impairing patient lucidity often felt that recommended dosages of medications were “heavy-handed.” Due to the nature of such disagreements, in many situations, one party's agenda and concerns were often sacrificed at the expense of the other. Because the primary attending was usually an ICU clinician (who might decide not to follow the consultant's recommendations), the ICU team's agenda usually superseded that of the palliative care team.

Furthermore, participants described disagreement related to the specifics of how care should be delivered. Communication was a common source of disagreement; clinicians were often quite exacting in their views of how communication with patients and families should occur, citing issues with how other clinicians presented treatment options, conveyed prognosis, or even provided support. This disagreement occurred not only between ICU and palliative care clinicians, but also between individual members of the ICU team. For example, several ICU nurses described how ICU physicians may get “lost in the trees as opposed to seeing the forest” when discussing prognosis with families, while one ICU attending criticized the “cookie-cutter” phrases used by some palliative care consultants to support families. Another ICU attending more generally described how nuances in word choice were important: “‘Maybe’ might be the split they're looking for, when they've been told by three other services, ‘No, we don't think so.’…A little hope can be dangerous.”

Another aspect of disagreement related to differences in expected time lines for communication and decision making. Participants described ICU workflows as fast-paced, with ICU teams expecting that family meetings and decision making would occur relatively rapidly to accommodate the high demand for patient beds. In contrast, palliative care clinicians preferred to give families more time for the decision-making process. As one ICU attending stated: “This patient is a high-grade acute brain injury patient, he's not going to wake up in 14 days… so don't give that option to the family… what ends up happening is family just disappear, they say, ‘Well we have a week and half now to think about it,’ … And then it's like, well who gave them this option?”

Interference with patient care

In occurrences of interpersonal conflict, participants described observing situations where a variety of clinician types (including ICU residents, ICU attendings, and surgeons) either prevented or were prevented from carrying out a task related to patient care. Disagreement was integral to these instances of interference, which was primarily manifested as prohibition of palliative care consultation or prohibition of having a goals-of-care discussion.

Disagreement over a patient's appropriateness for palliative care sometimes led clinicians to actively prevent the primary team from calling a palliative care consultation. One ICU attending stated, “The surgeons will tell us, ‘You cannot get a palliative care consult.’ It's really that simple.” Palliative care clinicians also cited instances in which the ICU team agreed to consult palliative care but specifically requested that consultants avoid discussing goals of care.

The rationale underlying these requests was to avoid upsetting people who might feel distressed by addressing the patient's potential mortality (including family members or other longtime clinicians such as cardiologists or surgeons). These types of prescriptions clearly interfered with the palliative care clinicians' ability to provide holistic support for patients and families, and often led to challenging situations, particularly if goals of care were broached by family or other team members.

On the contrary, some ICU clinicians described instances in which palliative care clinicians made recommendations without fully understanding the specifics of a disease process or the ICU's preferred workflow and time line, interfering with the ICU team's efforts to move forward with patient care (e.g., transferring patients to rehabilitation centers). In these instances, lack of reliable prognostication schemes and miscommunication between the ICU team, palliative care team, and patients and families complicated prognostic expectations and hindered the delivery of optimal patient care.

Negative emotions around experiences related to palliative care consultations

In the context of disagreement and interference, highly charged examples of negative emotion were described by all clinician types (e.g., social workers, nurses, physicians) in our study. ICU nurses and palliative care clinicians reported instances of being reprimanded and treated with rudeness in response to advocating for palliative care or making recommendations that the other party disagreed with. Both ICU and palliative care clinicians reported experiencing anger, grief, and regret over the way care was delivered, particularly in sensitive end-of-life situations, as well as ethical conflict in struggling with “the fine line between euthanasia and end-of-life care” and with external pressures to meet hospital outcomes.

Participants became quite impassioned while recounting these distressing experiences, even if they had occurred a year or more prior, and often reported persistent feelings of resentment and a decreased willingness to work subsequently with the involved party. For instance, ICU clinicians described no longer wanting to consult palliative care after negative experiences, while palliative care clinicians expressed feeling hurt and upset after being “fired” from patient care.

It should also be noted that perceived hierarchy (e.g., attending to fellow, attending to nurse, primary team to consult team) and power dynamics also played a role in instances of interpersonal conflict, often with clinicians in more subordinate positions being uncomfortable with voicing opposing concerns or ideas. Being reprimanded in these cases and subsequently becoming more reticent often led to negative emotions such as frustration or resentment.

Co-occurrence of disagreement, interference, and negative emotion

Disagreement, interference, and negative emotion often occurred together, as emotions ran high when clinicians disagreed on some element of delivering end-of-life care and subsequently interfered with delivery of care. One palliative care clinician described an experience of the three properties co-occurring:

“One of the most challenging situations…was a case we had a year ago, a patient with a super complicated post-surgical course who was just doing really poorly… We had a family meeting, and the family is like, ‘We want to make him comfort’ and everything. And the surgery team found out and they got so upset. They were really angry with our team and wrote an e-mail to our director and I understand that. It was a complicated situation, but it's hard when a family is so clear on their loved one and he was on max doses of a lot of things, and it just seemed so apparent that he couldn't really pull through the hospitalization. And so we were asked to never come see the family again… And I'll never forget that I went in there just to wish them well… and a different ICU attending saw me in the room and in front of the son said, ‘You're not supposed to be here, never’… and it was just really heartbreaking, and the patient ended up dying two days later.”

Discussion

In this secondary thematic analysis, we applied a deductive–inductive approach to analyze existing qualitative interviews among ICU and palliative care clinicians with the goal of describing interpersonal conflict in the context of palliative and end-of-life care delivery for critically ill patients. We identified the following themes: disagreement about the appropriateness of palliative care in the ICU setting; disagreement about which care plans should be prioritized and how care should be delivered; interference with patient care; and negative emotions around experiences related to palliative care consultations. We did not find additional domains of interpersonal conflict outside of the existing framework specified by Barki and Hartwick in our data.

Within health care, and especially within the ICU setting, interpersonal conflict has been shown to lead to clinician burnout and job strain, and even has the potential to negatively impact patient safety.^{2,8,11,12,25–27} Our findings agree with prior studies on conflict in ICUs that found common sources of conflict to center on inadequate communication, pain control, and end-of-life decision making.^2,6,12,25 Conflict may be occurring due to a lack of mutual respect between clinician teams, miscommunication, or due to clinicians' holding opposing values. However, a notable distinction between these prior studies and our current study lies in their definitions of conflict, which they largely equate with disagreement.

According to the Barki and Hartwick framework used to guide our analysis, true interpersonal conflict occurs when disagreements are inadequately resolved and subsequently escalate into interference and negative emotion. The experiences detailed in our study demonstrate the most salient instances of interpersonal conflict within the context of palliative care delivery and delineate a spectrum of developing conflict, starting from pure disagreement and intensifying into interference and negative emotion. Unresolved disagreements that escalate into interpersonal conflict may obstruct the development of mutual respect required for collaborative working relationships and create a breeding ground for further conflict.^1,10,25,28 Indeed, prior studies have shown that disagreement and conflict are not necessarily harmful; rather, the way the conflict is handled is more important than whether it occurs.^29–31

Although we did not find direct instances of nontask-related conflict in our data (because the majority of, if not all, interactions between ICU and palliative care clinicians take place in the context of the task of “patient care”), it is important to note that inadequately resolved task-related conflict may develop into nontask-related (i.e., relationship) conflict.^10,30 In the parent study that examined factors associated with adoption of specialty palliative care in ICUs, having successful experiences comanaging patients was key to developing the relationship between ICU and palliative care clinicians and achieving “buy-in” for specialty palliative care.²⁰ Our findings suggest that the inverse may also be true, and that interpersonal conflict may make clinicians grow averse to working together subsequently.

Contextualized within the existing literature, our findings suggest that there may be opportunities to mitigate the escalation to interpersonal conflict when disagreements are first identified. By highlighting specific sources of disagreement, our findings may help clinicians take a proactive approach toward addressing issues that may potentially lead to conflict. Disagreement related to the appropriateness of palliative care may potentially be ameliorated by educational initiatives to help dispel misconceptions about palliative care,^32–34 whereas disagreements about care plans and how care should be delivered could potentially be improved through regular proactive communication.^2,31,35–37 In particular, disagreement about medication dosing had the potential to lead to challenging situations, especially at end of life. These disagreements could potentially be improved through targeted discussion regarding medication side effects, as well as agreement on dosing protocols.^13,38,39

Lastly, it should be noted that while pure disagreement might be resolvable by the abovementioned suggestions, instances involving negative emotion may require different strategies, such as involvement of hospital leadership or a trained third-party mediator.^25,28 Other potential solutions to mitigate interpersonal conflict may include training and engagement in collaborative problem-solving, mediation strategies, and other reflective practices.

To our knowledge, this is the first qualitative study to examine interpersonal conflict within palliative and end-of-life care delivery, an area of patient care that may be particularly predisposed to conflict. Our study included the use of semistructured interviews to examine a potentially sensitive topic in an in-depth manner, and a diverse participant sample that allowed us to obtain a plurality of clinician perspectives. Although occurrences of conflict were less common than examples of collaboration and positive experiences in our dataset, these experiences appeared to deeply impact clinicians and their willingness to collaborate on both personal and professional levels.

Our study has several limitations. First, this secondary qualitative analysis used a dataset that was not specifically collected to explore interpersonal conflict. However, because interviews focused heavily on relationships and interactions between palliative care and ICU clinicians, and because participant descriptions of interpersonal conflict were vivid and rich, we believe that our findings are trustworthy. Second, participants with emotionally charged occurrences of interpersonal conflict may have been more likely to share their experiences, and it is possible that less intense examples of interpersonal conflict were not elicited during data collection. Third, no house officers (and only one advanced practice provider) are represented in this study; their perspectives may differ from the other clinician types who participated in this study. Fourth, our participants had a median of 5.5 years in practice, and their perspectives may differ from clinicians with more years of experience. Fifth, our study was conducted at two academic medical centers in New York City, and findings may not generalize to other geographic regions or practice settings. Notably, while programs differed in their maturity level, both would likely be considered mature programs, and conflict may be more prevalent or more likely to occur when programs have had less time to develop interdisciplinary relationships. Lastly, our study focused on descriptions of interpersonal conflict, but we did not ascertain how these experiences may have affected the multiple aspects of clinician well-being or coping mechanisms, patient and family distress, or patient outcomes.

Conclusions

In this qualitative study, we provide an in-depth understanding of how interpersonal conflict occurs during the delivery of palliative and end-of-life care to ICU patients, and the ways in which ICU and palliative care clinicians experience disagreement, interference, and negative emotion within this context. Future studies should focus on understanding how occurrences of interpersonal conflict between clinicians may affect clinicians, patients, and families; how to best manage disagreement to prevent the occurrence of interpersonal conflict; and how to implement collaborative problem-solving strategies or other approaches (e.g., mediation training) at various stages of developing interpersonal conflict to improve the delivery of palliative and end-of-life care for critically ill patients.

Funding Information

This work was supported by the American Federation for Aging Research, the National Institute on Aging, and the National Institute of Nursing Research. The funding sources were not involved in the design and conduct of the study, interpretation of data, preparation of the article, or in the decision to submit the article for publication.

Author Disclosure Statement

M.H. is supported by a Paul B. Beeson Career Development Award K08AG051184 from the National Institute on Aging and the American Federation for Aging Research. K.P.M. is supported by the Comparative and Cost-Effectiveness Research Training for Nurse Scientists (CER2; T32NR014205). There are no additional conflicts of interest to report.

Supplementary Material

Supplementary Data S1

Supplementary Data S2

Supplementary Data S3