Conflict Management by Physicians: A Heavy Hand in Preference-Sensitive Decisions

Conflict implies opposing forces. In the healthcare setting, patients, their family members, and physicians are ideally aligned in promoting patients’ well-being. Therefore, conflicts that arise between physicians and patients or family members regarding patients’ goals of care constitute failures of therapeutic alliance, communication, or the shared decision-making model (1). In the intensive care unit (ICU), these conflicts may be due, in large part, to lack of agreement on information underlying preference-sensitive decision-making, such as whether family members accept the prognostic estimates provided by physicians (2, 3), inadequate understanding of the patient’s medical condition or the risks and benefits of therapies, or physicians’ ability to communicate this information effectively (4).

The heightened emotional state of those involved in critical care decision-making may magnify misunderstandings and communication difficulties. In large part, this is due to predictable changes in cognitive processing when we as human beings are threatened (5), as family members are when facing a loved one’s potential mortality, as well as our tendency to demonstrate optimism bias (2). Physicians must acknowledge that these cognitive distortions are not personal shortcomings of family members but rather their behavioral adaptations to stress. Failure to do so creates opposition between those advocating for the patient, resulting in conflict.

Mitigating such conflict is a key skill that ICU physicians should possess to maintain appropriate focus on care that will maximize patients’ well-being. In addition, managing conflict well is likely to improve family members’ outcomes after involvement in ICU care (6) and decrease the frequency of moral distress and burnout among physicians (7, 8). Such conflict-management skills include recognizing potential or actual conflict, avoiding its development, and resolving conflict if it does occur. Yet, all physicians may not possess these skills.

There is little empirical work describing how ICU physicians understand and approach conflict with families. This foundational step would advance the patient- and family-centeredness of ICU care for two reasons. First, this understanding is necessary to develop evidence-based communication practices aimed at decreasing conflict in the ICU. Second, evaluating the extent to which physician behaviors, such as conflict management, vary among ICU physicians will enable the development of effective interventions resulting in undue variation in care patients receive. There are few identified mechanisms for the well-defined variation in critical care delivery at regional and center levels (9, 10). Physician behavior may be an important contributor to this variation (11) because of differences in the way in which physicians practice critical care, including shared decision-making.

In this issue of the AnnalsATS, Mehter and colleagues (pp. 241–249) provide a first step in producing necessary insight into how ICU physicians approach conflict with families, which also illustrates that there is a wide range of practice behaviors among even this small group of physicians (12). The authors focused primarily on conflicts that arise during goals-of-care discussions, conducting a series of semistructured interviews among 18 critical care physicians selected for their reputation as having an effective approach to or interest in end-of-life care. This group of physicians represents those who practice at four distinct settings within a single geographic area.

The authors propose a conceptual model, derived from their interview data, of ICU physicians’ approaches to conflict management. This model focuses on three main domains of physician behavior: 1) goal-focused actions intended to achieve what they perceive as the optimal patient outcome, 2) supportive actions intended to support the family during the ICU experience, and 3) the experience and management of their own emotions. Within these domains, the authors find patterns of physician behavior that may be helpful or detrimental to patients and their family members. Perceived helpful behaviors include the use of supportive skills, such as acknowledging family members’ emotions and expressing nonabandonment should goals shift away from restorative care. Potentially detrimental behaviors include those that may harm patients and their family members when used inappropriately, such as assuming a forceful role in decision-making (e.g., the use of persuasion), suppressing key information (e.g., eliminating potentially appropriate treatment options or not correcting family members’ false hope), or failing to uphold his or her professional duty to patients by acquiescing to inappropriate requests. However, the range of approaches to dealing with conflict described by these 18 physicians also sheds light on the dramatic influence a single physician may have in shaping decisions that should be highly dependent on an individual patient’s preferences, values, and goals.

Mehter and colleagues work reveals physicians’ emotions matter in how they approach, manage, and respond to potential or actual conflicts with family members that arise during high-stakes decision-making (12). The interviewed physicians expressed anxiety, relief, emotional distancing, and sadness relating to family meetings, conflicts, and poor patient outcomes. They acknowledged that their strategies for approaching families were influenced by these emotions. Although the authors focus on the importance of this finding in informing the study of physician burnout, this also influences physicians’ facilitation of shared decision-making. Physicians are at risk of the same predictable “irrationality” as all humans. Their behavioral attributes are likely to influence how they prognosticate, communicate, and provide recommendations to surrogate decision-makers. Although this work is focused on physician–family conflicts over end-of-life decisions, the findings have broader implications for the degree to which preference-sensitive care is dependent on or determined by physicians’ rather than patients’ goals.

There are clear limitations to the presented work. There is wide variation in the physicians’ approaches to conflict and no evidence that these strategies improve patient or family outcomes. Thus, it is premature to promote the use of the described strategies. Furthermore, physician behavior was not observed. Given that the physicians were selected to participate based on their existing relationships with the investigators and their reputation for being skilled at end-of-life care, there may have been significant bias introduced into how they presented their typical behaviors. The choice to use a single physician as the interviewer and primary data analyst is also limiting from a methodologic standpoint, as multiple analysts are less likely to inadvertently introduce bias into qualitative results. Finally, and most importantly, there were only physician perspectives represented in this study of conflict. Prior work has shown that physicians are less likely to perceive conflict than family members (4, 7). In addition, as acknowledged by the physicians who participated, nurses and other nonphysician ICU clinicians also assist in the avoidance of and response to conflict. Thus, physician perspectives are an incomplete depiction of conflict development and management in a multidisciplinary and family-centered critical care setting.

Nevertheless, this work provides novel, hypothesis-generating insight into physician perspectives of conflicts in the ICU. The authors identify several areas for future research that build on these findings, including developing physician-directed interventions aimed at reducing conflict in the ICU. Furthermore, these findings demonstrate that physicians perceive that they yield a heavy hand over preference-sensitive decisions, providing further evidence that physician behavior may lead to undue variation in critical care delivery. Conflict in the ICU suggests that there are multiple strong forces advocating for a patient. Although the high-stakes and highly emotional setting in which this advocacy occurs can lead to opposition, understanding the behaviors that lead to and mitigate conflict is an important first step toward achieving the goal of providing care that is best matched to patients’ values and preferences.