Physician Approaches to Conflict with Families Surrounding End-of-Life Decision-making in the Intensive Care Unit. A Qualitative Study

Abstract

Rationale: Families of critically ill patients are often asked to make difficult decisions to pursue, withhold, or withdraw aggressive care or resuscitative measures, exercising “substituted judgment” from the imagined standpoint of the patient. Conflict may arise between intensive care unit (ICU) physicians and family members regarding the optimal course of care.

Objectives: To characterize how ICU physicians approach and manage conflict with surrogates regarding end-of-life decision-making.

Methods: Semistructured interviews were conducted with 18 critical care physicians from four academically affiliated hospitals. Interview transcripts were analyzed using methods of grounded theory.

Results: Physicians described strategies for engaging families to resolve conflict about end-of-life decision-making and tending to families’ emotional health. Physicians commonly began by gauging family receptiveness to recommendations from the healthcare team. When faced with resistance to recommendations for less aggressive care, approaches ranged from deference to family wishes to various persuasive strategies designed to change families’ minds, and some of those strategies may be counterproductive or harmful. The likelihood of deferring to family in the event of conflict was associated with the perceived sincerity of the family’s “substituted judgment” and the ability to control patient pain and suffering. Physicians reported concern for the family’s emotional needs and made efforts to alleviate the burden on families by assuming decision-making responsibility and expressing nonabandonment and commitment to the patient. Physicians were attentive to repairing damage to their relationship with the family in the aftermath of conflict. Finally, physicians described their own emotional responses to conflict, ranging from frustration and anxiety to satisfaction with successful resolution of conflict.

Conclusions: Critical care physicians described a complex and multilayered approach to physician–family conflict. The reported strategies offer insight into pragmatic approaches to achieving resolution of conflict while attending to both family and physician emotional impact, and they also highlight some potentially unhelpful or harmful behaviors that should be avoided. Further research is needed to evaluate how these strategies are perceived by families and other ICU clinicians and how they affect patient, family, and clinician outcomes.

One in five deaths in the United States occurs in an intensive care unit (ICU) (1, 2). Although most Americans express a desire for less intensive care at the end of life (3), the ICU setting can make it difficult to prioritize this goal. Unfortunately, discussions about whether to pursue full aggressive measures or to limit interventions at the end of life are often not considered until patients are critically ill and no longer able to participate in decision-making (4, 5). Families are thrust into the proxy role and asked to make difficult choices from the imagined standpoint of the patient (“substituted judgment”) at a traumatic time, often with incomplete understanding on which to base their decisions. It is hard to imagine a more difficult scenario: high-stakes choices fraught with emotion, complicated by complex intrafamily dynamics and the need to negotiate these decisions with one or more ICU physicians with whom they have no prior relationship.

Thus, it is no surprise that conflict between ICU physicians and family members is common (6–9), often arising in “family meetings” that occur at critical junctures to discuss a loved one’s care. Disagreements between physicians and families over end-of-life care can lead to significant stress for all parties. Conflict has been independently associated with both provider burnout and family post-traumatic stress disorder (10, 11).

Professional societies have issued guidance to help ICU physicians navigate communication and end-of-life decision-making (12, 13). Yet, little is known about how conflict is managed in real-world settings (14). Hence, we sought to elucidate the ways in which ICU physicians approach communication and decision-making when there is conflict with families, as well as the emotional impact on families and physicians themselves. Gaining insight into physicians’ strategies may inform best practices for improved communication and conflict management with surrogates, provide pragmatic guidance for critical care teams, and identify areas for future exploration and research. Some of the results of this study were previously reported in the form of an abstract (15).

Methods

We recruited academic ICU physicians who provide care in medical ICUs at four hospitals (Table 1). Two investigators (H.M.M. and J.B.M., both of whom are ICU physicians) invited colleagues perceived to have a reflective approach or particular interest in end-of-life decision-making to participate in semistructured interviews. Enrollment was stopped after 18 interviews when thematic saturation was reached.

Table 1.

Structural and process characteristics of participating intensive care units

	Patient and Family Involvement in Rounds	Palliative Care Consultation Available	Nurse Involvement in Family Meetings	Conflict Resolution Training
Tertiary care academic medical center	Family included in rounds at provider’s discretion; bedside rounds rare and not standardized	Yes, same day	Standard	None
Urban safety net hospital	Family included in rounds at provider’s discretion; bedside rounds rare and not standardized	Yes, same day	Standard	None
Veteran Health Administration hospital	Family included in rounds at provider’s discretion; bedside rounds rare and not standardized	Yes, same day	Standard	Online module for attending physicians
Community hospital	No family involvement in rounds; no bedside rounds	No	Encouraged	None

Semistructured in-person interviews were conducted by a single investigator (H.M.M.), loosely following an interview guide (see online supplement) (16). Prespecified topics were informed by a literature review and included approaches to decision-making and communication with families, situations in which conflict arose, strategies for resolving conflict, and sources of provider stress. The interview guide was iteratively revised throughout the interviewing process, allowing exploration of emerging topics from diverse perspectives. Interviews were audio recorded and transcribed.

Following the constant comparative method of grounded theory methodology (16–18), transcripts were repeatedly read and reviewed to achieve immersion, familiarity, and holistic understanding. Key passages representing important concepts were coded with descriptive labels. Codes were then sorted into thematic categories based on conceptual relationships and linkages. Contextual linkages within and between transcripts were examined by arraying the coded data in a case-by-code matrix to define sequences suggestive of communication processes. The study team (H.M.M., J.A.C., and R.S.W.) met frequently to review and discuss transcripts. This included in-depth discussion of individual codes and themes, how particular quotes and excerpts from transcripts fit into the coding framework, and iterative revision of the coding framework until consensus was reached. H.M.M., J.A.C., and R.S.W. each performed close readings and annotations of the first several transcripts; once the coding framework was finalized and agreed on, H.M.M. performed formal coding of all transcripts with supervision from J.A.C. and R.S.W. All procedures were approved by the institutional review board of the Boston University School of Medicine.

Results

Overview

Our diverse sample is described in Table 2. All participants recalled experiences of conflict or disagreement with patient proxies over end-of-life decision-making. In every instance of conflict described, proxies favored the implementation or continuation of aggressive measures, whereas the healthcare team favored a less aggressive course, such as withdrawing or withholding life-sustaining treatments or supportive measures.

Table 2.

Participant characteristics (N = 18)

	n
Male sex	10
Race
White	13
Nonwhite	5
Years of attending experience, no. of subjects
≤5	5
6–10	7
11–20	3
>20	3

As summarized in Figure 1, physicians conceptualized the process of managing conflict and its impact on family and physician emotional well-being. This multidimensional process unfolded over three stages: 1) the period leading up to the initial “decisional” family meeting; 2) the initial decisional meeting itself, in which conflict would often arise; and 3) the period after the initial decisional meeting, in which the conflict continued to be negotiated. Physicians described strategies to anticipate and potentially circumvent conflict, as well as to manage conflict once it arose. Their approach was modulated according to characteristics of the patient’s family (or other proxies), which they carefully gauged. Throughout this process, physicians strived to minimize the family’s distress while also pursuing the course of action they believed was in the patient’s best interest. This often entailed active relationship maintenance and repair. Finally, physicians themselves experienced complex emotions and described how these emotions evolved throughout the process of negotiating conflict with families.

Figure 1.

Physician approaches to managing and mitigating effects of conflict with families. Physicians conceptualized the decision-making process as comprised of three distinct phases: 1) the period before the first decisional family meeting, 2) the initial decisional meeting, and 3) subsequent interactions between provider and family. Components of decision-making are positioned in boxes along this horizontal timeline. Physicians’ accounts of tending to families’ emotional health are also illustrated on this timeline.

Preparing for the Family Meeting: “Sizing Up” the Family

Physicians described several strategies for assessing the family’s opinion, state of mind, and preferences for information delivery and decision-making role in anticipation of end-of-life decision-making. These strategies allowed physicians to routinely align their approach with the family’s preferences and also prepared physicians for potential conflict, in some cases allowing them to avoid it entirely. In the best scenarios, informal encounters with families increased receptivity to physician input and expedited concordance between the healthcare team and the family.

Physicians described “sizing up” the family in advance of formal decision-making conferences (Table 3). One strategy was to solicit input from ICU nurses, who often spent more time with family members and could comment on their emotional state and understanding. Several physicians routinely scheduled early “informational” family meetings in the ICU, before the point at which important decisions needed to be made, to build rapport and gauge family dynamics. Informal interactions with the family at the patient’s bedside also fostered rapport building and provided insight into family members’ psyche and situational understanding. Some providers found “bedside rounds” with family in attendance to be particularly useful to increase transparency and help the family understand their loved one’s condition.

Table 3.

Representative quotes describing strategies for “sizing up” families

Solicit input from nurses:

“[T]alking to nursing staff, who spend a bit more time with the family. Going into a meeting, it’s very helpful to get a better understanding of where the family is at emotionally and what they understand is going on.” (Physician 17)

Early “informational” meetings:

“I have an ‘early and often’ approach to family meetings in the ICU; the earlier you meet with the family and start to set expectations and establish a relationship with the family, the better it goes. Early family meetings help hugely; you get an early assessment of the family dynamic, competing factions, and understandings or interpretations within the family, how big [the] family is, who all needs to be involved in decision-making, and a quick assessment of health literacy and realism.” (Physician 12)

Informal bedside interactions:

“When you’re meeting the family ahead of time, you try to get a feel when you’re talking with them what kind of reaction you might get. Sometimes you can feel it when they first come in, when you first meet them and see them, as to how upset or accepting they are of what the circumstances might be.” (Physician 7)

Bedside rounds:

“My practice is to do bedside rounds in the ICU, and I ask any family members present to stay in the room, listen in, ask questions, and I find that really helpful in preparing for family meetings—what they say at the end.… I invite family to stay or tell them in advance when we’ll be rounding. I try to prepare them for follow-up family meetings during informal interactions at bedside or during rounds.” (Physician 5)

Definition of abbreviation: ICU = intensive care unit.

Initial Approaches during Formal Family Meetings

Completing the narrative

Physicians carefully modulated their information delivery in family meetings. Many acknowledged that the goal of pursuing less aggressive care consciously influenced the way they presented information and options to family. Some presented information in a way that effectively eliminated options:

“When you talk to the family you provide a narrative. Whether you admit it or not, the narrative is geared so that the logical conclusion of the story is what you think is the appropriate outcome. In my experience, the vast majority of the time families end up being reasonable and completing the narrative in the appropriate fashion.” (Physician 13)

To strengthen construction of the “narrative,” physicians found it useful to review with the family the patient’s condition and quality of life in the time preceding their hospitalization.

Offering an opinion

Virtually all physicians stated that they routinely offer families their opinion on the best course of action:

“If a family comes and says to us, ‘We want you to do everything,’ I think the onus is on us to say, ‘We believe that’s not the right thing to do.’” (Physician 18)

A few physicians added the caveat that they withhold their recommendation when families make it clear it is unwelcome:

“When a family only wants information and doesn’t want my opinion, I pretty much give them what they’re asking for and won’t give any of my opinions.” (Physician 6)

Managing Disagreement

Physicians’ approaches to managing conflict with families ranged from relative acquiescence to aggressive persuasion. They recognized the importance of a sincere “substituted judgment” and their obligation to inform families fully of the likely consequences of medical decisions.

Approach 1: acquiescing to families

Some physicians were relatively acquiescent, driven by several underlying factors and concerns. One factor was the ethical principle of patient autonomy:

“Autonomy is still our highest ethical principle.… With the right to make decisions comes the right to make what I think are bad decisions, but they’re not my decisions to make. The best we can do is provide [families] with as much information as we can, find out where they stand, let them know what their options are, but the decision is theirs to make.” (Physician 4)

In a similar vein, physicians acknowledged the subjectivity of their own opinions and the validity of family desires for aggressive care even when it ran counter to their recommendations:

“There’s no such thing as futile care; it’s just that most of us don’t value the life enabled by the care that we’re providing.… [We’re] doing something, just not something that I would want done for me or my family, but it’s the family’s decision.” (Physician 11)

Others were more likely to acquiesce when families appeared to have more conviction in their choices and seemed unlikely to change their minds:

“Those who’ve already made up their mind about aggressive care and want you to be purely informative, there’s nothing you can tell them that’s going to change their decision.” (Physician 1)

Approach 2: persuading families

Some physicians were more likely to persuade families, citing information imbalances between physicians and families that were so great as to be unbridgeable by counseling and information exchange:

“In some cases it seems obvious [to me] what the trajectory of the case is in a way that can’t be obvious to the family.… It’s not appropriate to be a disinterested party that’s merely informing the family of the facts. I think my job is also to advocate for the patient.” (Physician 2)

Physicians who took this approach described several distinct negotiation strategies, which we label as “pushing hard,” “biding time,” and “changing the conversation” (Table 4).

Table 4.

Representative quotes illustrating physician approaches to persuade families

“Pushing hard”:

“When we disagree, I will push hard, but I won’t be disdainful of their decision or disrespectful. I will be explicit.... The best example is anoxic brain injuries. I’ll tell them, ‘There’s really not a good chance that this person’s ever going to get any better, so you’re committing them basically to being in bed, cared for, unable to communicate, recurrent hospital visits as their body starts to waste and they fall apart.... We’re talking about bedsores starting to develop, coming to the hospital over and over again with infections, getting stuck (by needles) over and over again, having tubes placed in them, not being able to communicate, can’t talk, can’t walk, by themselves most of the time.’ I’ll lay it out in stark terms, that’s how hard I’ll push, because I feel like that’s the reality they’re going to encounter, [and] they need to be clear on what they’re deciding for [the patient]. And if they say, ‘Well, we still want a tracheostomy,’ I’m like, ‘Okay, then we’ll go ahead and arrange it.’” (Physician 9)

“An important mentor had a very paternalistic decision-making style, and it seemed to really resonate positively with families and save them a lot of pain and grief and agony.… In a metastatic cancer or post-cardiac arrest case or something where the medical scenario is extreme, the patient’s not going to come off life support or not going to survive, and you’re proceeding to what you believe will be torturing them for days to weeks, I’ll try as best as I can to understand what the patient would’ve wanted, then say, ‘We’re going to need to stop care.’ If there’s pushback, [I’ll] continue to meet periodically [with the family].… I don’t take no for an answer without at least three tries.… I keep emphasizing that I’m very uncomfortable with this, I don’t think it’s in your mother’s best interests. Usually the family will back off in that situation.” (Physician 5)

“Biding time”:

“Sometimes I feel like what the ICU is for is to give people time to acknowledge the mortality of their loved one, and you can have all sorts of economic arguments about that, but I think acknowledging that there’s a difference of opinion, and then sort of agreeing on the time you sit down next, so sort of a time-limited trial. You know, ‘This is what we’ll do, let’s sit down again in three days and see where we are,’ because sometimes I feel in that situation, where family wants everything and for us the writing is on the wall, it sometimes does take a little time.” (Physician 16)

“To try to maintain a good working relationship with that family, I might say, ‘Alright, we’ll keep pushing ahead, but gosh, I’m really worried about things, why don’t we get back together tomorrow or in a couple of days and talk about where things are.’ It’s sort of meeting families where they are. I might not be as explicit on what my thoughts and opinions are initially because of wanting to maintain a good relationship with that family so they know I’m advocating for them.” (Physician 14)

“Changing the conversation”:

“[It can help to] just reframe the conversation and help the family understand that they have done everything, that we have done everything. I think sometimes there can be a switch that flips, instead of saying, ‘We will not do XYZ,’ if you sort of emphasize, ‘You have done everything and we have done everything for your loved one.’” (Physician 17)

“With very few exceptions, I think that when you start to phrase the conversation in terms of suffering, not in medical terms, but just, ‘When I go in the room I see your mother suffering. Do you see that?’ it can be really helpful.” (Physician 13)

Definition of abbreviation: ICU = intensive care unit.

“Pushing hard”

Two physicians described aggressive approaches to persuade families. One described painting an increasingly vivid picture of the potentially undesirable effects of continued aggressive care. Another voiced his recommendation as an imperative. He acknowledged that this might be construed as paternalistic, but he believed it often resonated positively with families.

“Biding time”

Many physicians believed that families need time to come to terms with their loved one’s prognosis and that allowing time often softened the family’s resistance to de-escalating care. These physicians allowed families time to process their loved one’s condition over a series of meetings while carefully modulating their approach to minimize adversarial encounters. Maintaining a good working relationship aided in eventually fulfilling the objective: less aggressive care.

“Changing the conversation”

Physicians described several ways of redirecting conversations to convince families to change their minds. Strategies included highlighting the extent of the therapies that had been undertaken and discussing patient suffering.

Modulating Factors

Physicians identified several factors that impacted the aggressiveness of their efforts to persuade families, most notably the quality of substituted judgment and the perception of patient suffering (Table 5). Physicians were more comfortable accepting a family’s decision if the family made a sincere effort to exercise substituted judgment. Despite coaching on the meaning of substituted judgment and apparent sincerity in attempting to exercise it, physicians perceived that families commonly experienced difficulty separating their own preferences from those of the patient—a circumstance that made physicians more likely to engage in persuasion. Similarly, in cases where patients appeared to be experiencing discomfort (e.g., pain, dyspnea) that could not be adequately controlled, physicians felt a greater impetus to push for de-escalation of care.

Table 5.

Representative quotes demonstrating modulating factors in physician approach

Quality of substituted judgment:

“I think that we tend to push harder when we feel that perhaps the family is, although they may think they’re operating on what the patient’s best interests are, maybe they’re thinking more about what they’d want for themselves than what the patient would want.” (Physician 18)

Patient suffering:

“If I think that we’re prolonging suffering, I push harder. I try to emphasize the fact that what I’m concerned about is that we’re prolonging suffering without a reasonable hope for recovery. I feel like my duty as a physician is to protect patients from things that cause them suffering.” (Physician 12)

Evolving perspectives on prognosis and quality of life:

“If there’s not consensus, I sometime wonder if I’m doing the right thing by being so forceful with my opinions.… We never really know, but we compensate for that, I think, in the meeting by pretending to be more certain than we ever are about the outcome.” (Physician 5)

“We make so many value judgments on impaired quality of life, so I’ve become a little bit more cautious maybe.… Part of it is feeling the responsibility to that life, and to make sure I’m not giving up too soon.… As [my mentor] says, ‘Death is forever.’” (Physician 16)

Physicians described the evolution of their approaches to conflict over their careers. Early in their careers, they were more likely to enter family meetings focused on fulfilling specific goals, such as changes in code status or withdrawal of life support. With experience, they realized the importance of gauging if families were “ready” to make these decisions. Whereas some had gained confidence in their ability to predict outcomes, others noted that critical illness outcomes remained difficult to predict, making them more circumspect about pushing for de-escalation of care. Experience also increased recognition of the subjectivity of “quality of life” and of death’s finality, augmenting circumspection (Table 5).

Tending to Family

Physicians expressed concern for families’ emotional well-being throughout a loved one’s critical illness, including hospitalization, death, and its aftermath (Table 6). Physicians worried that families would be burdened by guilt if they felt responsible for their loved one’s death and observed that this could contribute to families’ reluctance to de-escalate care. Physicians sought opportunities to assume some of the decision-making burden, at either the family’s explicit or implicit request. However, physicians had to tread carefully because misreading the needs of the family could result in conflict and damage to the physician–family relationship.

Table 6.

Representative quotes showing how physicians tend to families’ emotional well-being

Mitigating psychological toll on family:

“I think … about their anxiety, depression, PTSD related to the death of a loved one, and how that can be mitigated. If there’s a tremendous amount of conflict, especially right up front, then I think that can really unfortunately also negatively affect people down the line.… It’s a balance of how much we’re putting the patient through for, say, another day versus trying to also treat the family.” (Physician 14)

Shouldering responsibility:

“If I tell [families], ‘You should do this,’ it takes some of the responsibility off their shoulders, and they find that it’s a little bit easier. [But one family member] in particular did not care for that, she was very offended in some ways, how would I dare put myself in her position.… I misread them.” (Physician 15)

Reassuring families of commitment to care for the patient:

“It’s a conflict, and then you’re in a position of providing all this care that you didn’t think was right, so then I … try to really just ensure that the family [understands] that we’re going to follow your wishes and do everything we can at this point, so that we’re not giving up or phoning it in just because our opinions were different.” (Physician 14)

Definition of abbreviation: PTSD = post-traumatic stress disorder.

Physicians also worried about the impact of conflict with the healthcare team on families’ well-being. This concern made some physicians hesitant to forcefully push their agenda, despite their desire to advocate for the patient. Physicians continually worked to repair any damage to their relationship with families caused by conflict. In the aftermath of conflict, physicians feared that families would be unable to trust that the healthcare team was fully dedicated to the patient’s care.

Physicians’ Emotional Responses

Physicians reflected on their own emotions surrounding decisional family meetings and conflict with families (Table 7). Leading up to decisional meetings, some physicians experienced positive emotions, including hope that any decisions made would result in more clarity moving forward. However, many reported anxiety, particularly when conflict was anticipated. Family meetings were seen as draining, requiring intense focus to avoid missteps that could damage the physician–family relationship, and as disruptive to workflow in a busy ICU. This led several physicians to distance themselves emotionally or to postpone or avoid family meetings.

Table 7.

Physicians’ emotional responses to conflict

Leading up to initial decisional meeting

Emotional distancing: “One thing I do before family meetings is increase the ‘clinical detachment factor’ because some of what comes up in meetings is so raw that I probably put walls up and say, ‘I don’t have feelings.’” (Physician 3)

Anticipatory anxiety: “Before all family meetings, I’m nervous and apprehensive, burdened by the weight of the huge decisions being made.… [S]ometimes I even procrastinate a family meeting because it’s so uncomfortable for me.… I fear that I’ll contribute to the family’s distress or that there’ll be some standoff or disagreement. There’s awareness that what I’m doing in the meeting is going to have really lasting repercussions, feelings, and memories for a lifetime for the individuals in that room. That’s a heavy burden.” (Physician 5)

After family meetings in which there is conflict

Sadness and regret, drained

“Family meetings usually don’t go exactly as planned, and I screw up often and regret things I said and the way the meeting went or how I handled situations.” (Physician 5)

“A lot of times when I come out of a meeting I feel really drained, like exhausted, and sometimes kind of sad and also this kind of tension between having this very emotional response and then having to go to the next thing, if it’s a busy day.... I let myself feel it for a few minutes, then I pack it away, you know. There’s this kind of very pragmatic moment of like, ‘Okay, now I can move on.’” (Physician 9)

Frustration

“At times I feel frustrated; often this happens if I find the family is just not getting it and they’re completely unrealistic in what they’re thinking, and no matter how I’ve presented it to them or had other people try to present it to them, they’re just not getting it.” (Physician 6)

Anxiety

“I get just as anxious when I feel as though I’m trying to convince them to withdraw care. If I go in there and after the conversation, they say, ‘Okay, we should withdraw care,’ I sort of feel as though that’s been my decision and that also is a little anxiety provoking.” (Physician 11)

Satisfaction, relief

“I’m very happy if it’s a successful meeting, which means that you communicate what you want to communicate to the family, so that the family actually understands that the patient is ill, the chance of meaningful recovery is very small; if they understand that, at least I will be very happy.” (Physician 10)

“I often feel relief after a meeting if we’ve been able to come to a decision that I think is the right decision for the patient, meaning I think it’s going to allow them to die with dignity.” (Physician 12)

Physicians experienced a range of emotions after decisional meetings. After meetings that resulted in a withdrawal of care, many reported sadness for the family but also relief of tension. The finality of these decisions provoked anxiety among some physicians. After meetings in which there was unresolved conflict or lack of consensus, physicians commonly experienced deep frustration, sometimes causing them to second guess their own opinion. At times, frustration led physicians to describe families and their beliefs in pejorative terms (e.g., “crazy families”). Finally, even in cases of disagreement, some physicians found satisfaction in smaller “victories,” such as helping a family fully understand their loved one’s condition and the likely outcomes of their choices.

Discussion

Critical care physicians described their complex and strategic approaches to managing conflict with surrogate decision-makers, reflecting on their experiences with introspective openness. They devoted effort to determining how families’ preferences could be both honored and influenced. Physicians’ inclinations to accept the family’s decision (even if it conflicted with their own opinion) versus attempting to persuade families to change their minds varied, as did specific strategies to resolve conflict and the circumstances in which they applied them. Physicians were concerned for families’ emotional well-being and worked to rebuild trust and repair any damage to the physician–family relationship in the wake of conflict. Managing conflict with families raised complex emotions for physicians, including both anticipatory and conflict-provoked anxiety.

It is encouraging that these physicians worked with families to avoid inappropriate prolongation of the dying process even in the face of initial conflict. Their approaches to families are consistent with previously described contributors to lower-intensity care near the end of life: a broader definition of the “dying” patient, concern for harms of commission rather than of omission, high decision-making self-efficacy, and a view of family requests for continued aggressive treatment as part of a normal response to a loved one’s critical illness rather than obdurate resistance to de-escalating care (19, 20). Tailoring information delivery and the decision-making role according to the family’s preference, making recommendations, allowing families time to process and come to terms with overwhelming emotions, and shouldering decision-making burden when appropriate are strategies our participants used to achieve their goals. These strategies are consistent with previous research and recommended approaches to family-centered communication and shared decision-making in the ICU (12, 14, 21, 22).

Experts in the field and professional society guidelines endorse the use of shared decision-making in ICU family meetings (12, 23–26). Because our study was focused on conflict in the ICU, we cannot comment on the degree to which shared decision-making is employed when conflict is absent. In the case of conflict, our study suggests that although physicians do take family preferences into account, their approach to decision-making may be driven by the goal of achieving the outcome that the healthcare team believes to be in the patient’s best interests—often the withholding or withdrawal of life support or other aggressive treatments when meaningful recovery is unlikely. Physicians clearly incorporate elements of shared decision-making when they judge them useful in that pursuit, and they also employ other artful strategies to size up, engage, and shape families’ views in the service of giving a voice to patients who can no longer speak for themselves. Although this is a laudable goal, the potential pitfalls of approaching these encounters with an agenda must be acknowledged.

In particular, some reported tactics for achieving the goal of less aggressive care may be counterproductive. Presenting a “biased” narrative of events and engaging in aggressive tactics to persuade families is manipulative; may heighten conflict; and could engender distrust, psychological trauma, and emotional exhaustion for both families and clinicians. Moreover, physicians’ frustrated perceptions of families as unreasonable or unable to grasp the concept of substituted judgment are likely misplaced. These negative encounters with families may instead reflect shortcomings in physicians’ communication skills or the detrimental impact of burnout (11, 27), a response to chronic emotional and interpersonal work-related stressors marked by exhaustion, cynicism, and inefficacy (28). Recent guidelines call for increased training for ICU clinicians in family-centered communication and shared decision-making (12, 22).

Physicians recognized the emotional toll that conflict can take on everyone involved. The rich descriptions our participants provided may both inform best practices to mitigate this emotional toll and highlight counterproductive strategies that may increase conflict and burnout, such as approaching family meetings with a rigid agenda and “pushing hard” for it. Uniformly and humanely, our participants expressed concern about families’ emotional well-being during and after conflict. One strategy they used to mitigate negative psychological sequelae for families was assuming decisional burden when families expressed a desire for this approach. Yet, assuming decision-making control weighed heavily on some physicians. Participants in our study described personal strategies to reduce the emotional impact of conflict as well as an increasing peace of mind and circumspect approach that come with experience.

When unrecognized, physicians’ emotions can negatively affect both the quality of medical care and their own sense of well-being; it is thus important to increase physician self-awareness to identify and manage these emotions (29). For example, easily implementable interventions such as “death rounds” for periodic debriefing and reflection may help ease ICU clinicians’ emotional burden (30). Our findings also underscore the need to support clinicians engaged in emotionally draining encounters. One effective strategy may be judicious use of palliative care and ethics consultative services, which can improve family emotional outcomes and reduce ICU length of stay and treatment intensity (31). Recent work has begun to quantify key aspects of palliative care specialists’ approach to managing conflict that may help explain the benefit of palliative care involvement and that could be the focus of interventions to improve critical care clinicians’ conflict resolution skills (32).

Our study has limitations. We recruited a small sample of ICU physicians from four hospitals in Boston and cannot assume that our findings are generalizable to other settings. Moreover, we purposely recruited subjects perceived by their colleagues to have an interest in end-of-life decision-making in the ICU. These physicians may have a distinctive approach to end-of-life conversations, thus introducing bias. Our interviewer had a colleague relationship with some participants, which may have influenced responses. Our findings may also have been influenced by our own preconceptions and experiences. We used methods commonly exercised in qualitative research to minimize this impact, including reflexivity and continuing interviews to thematic saturation (33, 34). Finally, the investigator who conducted the interviews was also the primary qualitative data analyst, representing a potential methodological limitation.

Our study provides nuanced descriptions of the complex interpersonal skills, artfulness, and care required to guide families through conflicts surrounding end-of-life decision-making in the ICU. Important goals endorsed by physicians included avoiding inappropriate prolongation of life at the expense of patient suffering and minimizing emotional trauma to families at an extraordinarily difficult time. This study provides insight into pragmatic approaches used by practicing physicians for negotiating disagreements with families, including concrete strategies to operationalize the theoretical approach described in recent statements from professional societies (12, 13). Despite their best efforts and intentions, at times physicians engaged in behaviors that were potentially unhelpful or even harmful to families. Further research is needed to elucidate how the reported strategies relate to patient and clinician outcomes, approaches used by other ICU team members to mitigate conflict, and family perspectives on these issues.