Disagreements, including those between residents and attending physicians, are common in medicine. In this Ethics Rounds article, we present a case in which an intern and attending disagree about discharging the patient; the attending recommends that the patient be hospitalized longer without providing evidence to support his recommendation. Commentators address different aspects of the case. The first group, including a resident, focus on the intern’s potential moral distress and the importance of providing trainees with communication and conflict resolution skills to address inevitable conflicts. The second commentator, a hospitalist and residency program director, highlights the difference between residents’ decision ownership and attending physicians’ responsibilities and the way in which attending physicians’ responsibilities for patients can conflict with their roles as teachers. She also highlights a number of ways training programs can support both trainees and attending physicians in addressing conflict, including cultivating a learning environment in which questioning is encouraged and celebrated. The third commentator, a hospitalist, notes the importance of shared decision-making with patients and their parents when decisions involve risk and uncertainty. Family-centered rounds can facilitate shared decision-making.
In graduate medical education, disagreements between residents and attending physicians are common and should be expected. Disagreements may arise in any aspect of clinical care, including diagnosis, management, and disposition. Disagreements also extend beyond training and into clinical practice, in which there may be disagreements between physicians rotating off and on service, the primary and consulting teams, or the medical team and other providers involved in the patient’s care. Variation in practice, which may serve as the basis for disagreements, has multiple causes: physicians differ in their training, knowledge base, clinical experiences, tolerance of risk, and approaches to uncertainty.1,2 In training, acknowledging disagreements can be difficult and should be encouraged. Addressing disagreements can stimulate learning by identifying knowledge gaps and proposing clinical questions for further investigation. In this Ethics Rounds article, commentators discuss a case in which the intern and attending disagree about discharge criteria.
A 9-year-old previously healthy and fully immunized boy presented to the emergency department (ED) of an academic children’s hospital with 6 days of fever (this case occurred before the coronavirus disease 2019 pandemic). His parents reported that he had fevers up to 39°C 1 to 2 times per day, which was accompanied by fatigue, decreased appetite, and a rash. He did not have conjunctival injection, swollen or cracked lips, cervical lymphadenopathy, or hand and foot swelling. The differential diagnosis included incomplete Kawasaki disease, streptococcal pharyngitis, Epstein-Barr virus infection, and juvenile idiopathic arthritis. The diagnostic evaluation, including a complete blood count, complete metabolic panel, urinalysis, chest radiograph, and abdominal ultrasound, was only remarkable for mild leukocytosis. The patient was admitted to the hospital medicine service.
A viral polymerase chain reaction test of a nasopharyngeal sample returned positive for adenovirus on day 2 of hospitalization. An echocardiogram, completed to rule out incomplete Kawasaki disease, revealed normal cardiac anatomy and function. On hospital day 3, the patient’s overall condition was improving; his fever curve was decreasing (but he had not yet been afebrile for 24 hours), his appetite was increasing, and he was meeting his maintenance fluid requirement orally. When the intern prerounded, the child’s mother expressed the desire to be discharged and follow-up with the patient’s primary care provider. When the intern presented her assessment and plan to discharge the patient outside of the patient’s room on rounds, the hospitalist, who had been in practice for 2 years, disagreed. The hospitalist directed the intern to not discharge the patient until he had been afebrile for at least 24 hours. When asked by the supervising resident for the justification for this decision, the attending expressed a concern that an adenovirus infection did not adequately account for the patient’s symptoms and that the patient would decompensate after discharge and require readmission. The attending did not identify any evidence to support his decision.
Drs Timothy Crisci, a Medicine and Pediatrics Resident, and Zeynep N. Inanc Salih, a Neonatologist and Bioethicist, Comment
Unfortunately, resolving disagreements by appealing to the literature begins to break down when there is no clear evidence supporting a single plan of action. If there is no evidence-based guidance, doctors instead rely on clinical experience, reasoning, and judgement. As a form of apprenticeship, graduate medical education relies on the assumption that attending physicians, as a product of being involved in a greater number of cases,3 have more clinical experience and more refined clinical reasoning. Does this mean that the attending is always right, even when there is no scientific evidence to back up their decision and they must rely solely on clinical judgement? How should residents communicate their disagreements with their attending physicians? Moreover, how can residents manage the moral distress associated with enacting care management plans with which they disagree while simultaneously trying to practice and develop autonomy?
Moral distress can be described as the feeling when one cannot conduct an action that one believes is the right thing to do. In medicine, it is not surprising that feelings of moral distress are common because clinicians may not agree on the best practice. It is especially prominent among trainees because of the existence of professional hierarchies. Thus, who makes the decisions and whose opinion counts are at the root of many of the ethical issues trainees experience.4 Ultimately, moral distress may lead to physician burnout.5
Resident autonomy decreases moral distress by allowing residents moral agency to follow their professional values and has been shown to be associated with improved resident well-being and decreased burnout6 ; however, there can be a perception gap between residents and attending physicians in terms of what autonomy looks like. In a survey examining resident-faculty interactions, Biondi et al7 found that both residents and faculty agreed that faculty provided too much direction. In the same study, the authors also found that when asked if “[m]y attendings allow my plan to be followed even if they prefer an equivalent alternative” on a 5-point Likert scale from 1 (“very seldom”) to 5 (“very frequently”), residents’ average response was 2.68, whereas attending physicians’ average response was 3.45 (P < .001).7 Williams and Deci8 would describe this type of direction as controlled and suggest that it can lead to increased anxiety, decreased interest in learning, and lower self-esteem.
Feelings of moral distress often stem from breakdowns in communication.9 In this case, the attending and intern failed to communicate how they weighed risks and benefits for this patient differently; the attending believed the risk of discharging the family home before complete symptom resolution was greater than the cost of an additional night’s stay in the hospital, whereas the intern believed that the risk of the patient decompensating at home and potential hospital readmission was less than an additional night’s stay in the hospital and more consistent with the family’s wishes. It appears they were both intending to do what was best for this patient, albeit from different perspectives. With appropriate training in conflict resolution, the intern could have recognized that both she and the attending agreed on the desired outcome, for the child to be healthy and for the family to return home safely, but were approaching the decision with differing clinical reasoning.
In this case, it is important that the supervising resident asks the attending to share his clinical reasoning, both for education as well as for discussion with the family, who are likely primed to expect being discharged from the hospital; however, the case implies that the attending’s initial explanation was unsatisfactory. Given the hierarchical structure of medicine, the intern may feel that she has only one plausible choice: accept the attending’s decision, share the information with the family, and carry out the attending’s management plan. This choice likely has a moral cost for the intern, who now must do something with which she disagrees. Unfortunately, this is not that uncommon. Chiu et al10 conducted a survey revealing that 32% of pediatric surgical trainees felt compelled to participate in their staff’s management plan, despite moral opposition.
Given the frequency of these encounters, the known relation of moral distress and burnout, and the generational focus of wellness in medical education, residents and attending physicians should be trained to expect situations in which attending physicians and learners do not agree and should be provided with skills to convert these situations from morally distressing to opportunities for education. Interestingly, pediatric residents receive little, if any, training in conflict resolution, and there are limited published frameworks for incorporating conflict management into graduate medical training. The Accreditation Council for Graduate Medical Education (ACGME) Common Program Requirements for pediatric residencies, in particular, include respect for others and effective communication with other providers as core competencies for residency promotion; however, there is no explicit discussion of conflict management, and it is not mentioned in the Pediatric Milestone Project, which is an ACGME-produced guideline for providing feedback to pediatric residents. In contrast, conflict management is explicitly included in the milestones of other residency programs, including anesthesiology, emergency medicine, and surgical critical care.11
In this case, the intern needs an expanded vocabulary to better understand the attending’s reasoning. When the attending is unable to cite literature or guidelines to support his decision, a follow-up question could be the following: “What do you think is the likelihood of there being a missed diagnosis or that the child will acutely decompensate if discharged today?” This type of modified Bayesian question acknowledges that the intern and attending could be approaching the situation from different perspectives. Simultaneously, it provides the attending with a more structured framework to share his reasoning while also creating an opportunity to teach about risk assessment and tolerance. In his response, the attending should focus on sharing evidence and experience, acknowledging the difference between evidence-based medicine and personal management style.
Ultimately, this case highlights the moral distress the intern experienced while deciding what was best for her patient as a result of a breakdown in communication with the attending. To help with both the communication gap and the intern’s moral distress, she and the supervising resident should further delineate the attending’s reasoning while acknowledging that both they and the attending have the same goal, to care for their patient, and the attending should acknowledge the gray area of practice, in which personal practice style can be used when guidance from evidence-based medicine is inconclusive. Furthermore, in the background, the residency program and the ACGME as a whole should make a concerted effort to provide conflict management education to help foster a respectful educational culture to promote teaching, nourish resident autonomy, and minimize the moral distress created by intrateam conflicts.
Dr Ndidi Unaka, a Hospitalist and Associate Residency Program Director, Comments
In medical education, it is imperative that trainees develop a sense of “decision ownership,” defined as the “cognitive-affective phenomenon in which a medical practitioner develops a sense of responsibility over decisions about care for a particular patient and personal investment in this decision-making process.”12 Decision ownership by trainees is a motivating driver for learning. Both trainees and attending physicians experience the tension associated with the mismatch of their respective clinical approaches. Trainees recognize their role as a learner and their limited exposure and experience as a clinician. Trainees also wrestle with how to cultivate their own maturation to independent practice and navigate their disagreements with attending physicians, who are their teachers and guides along the way.
Likewise, attending physicians grapple with the immense pressure of being ultimately responsible for their patients. Although responsibility is related to ownership, the concept is distinct. Responsibility is “the capacity to accept one’s own actions, to have ownership over them and consequently to answer for them and bear their consequences.”13 The pressure experienced by attending physicians can be particularly intense early on as one transitions from trainee to attending. Clinical decisions that were seemingly less anxiety provoking as a trainee become higher stakes as attending physicians think through a variety of scenarios and all that could go wrong with their patients. Attending physicians’ experience and knowledge base may influence the circumstances in which they allow trainee autonomy and how they approach disagreements about clinical decisions.14 Attending physicians who are less experienced may be more conservative in their clinical approach; this is compounded with juggling one’s own practice formation and the responsibility to teach trainees, which may be overwhelming.
Additionally, we must acknowledge the developmental process of each physician in embracing the role of uncertainty and ambiguity in clinical medicine, which is one of the core professionalism milestones outlined by the American Board of Pediatrics.15 A physician’s thresholds for specific management decisions are influenced by the individual’s propensity and tolerance for risk and uncertainty. Moreover, an attending physician’s aptitude and comfort with disclosing and openly discussing their own uncertainty may influence interactions with trainees and overall team dynamics.
Attending physicians must strike a balance between their roles as teachers and the physicians who ultimately assume responsibility for their patients’ care. As teachers, they share information with trainees and provide them graded responsibilities based on the trainees’ skills and abilities. As physicians, they have fiduciary obligations to their patients to put their patients’ interests above their own and to promote their patients’ welfare. As such, they should provide trainees autonomy to implement plans that are likely to have comparable clinical outcomes. If they believe that a trainee’s plan will produce a worse outcome, they should explain their reasoning and be open to revising their conclusion. Unilateral decision-making by the attending physician with minimal discussion may breed trainee discomfort, discontent, and disillusionment.
Trainees must also balance their roles as learners with their duties to their patients. Although the patient in the case is not at increased risk of harm, such scenarios occur. Attending physicians are by no means immune to human error; the role of the medical team and redundancy within the health care system is paramount for promoting patient safety. When a trainee disagrees with an attending physician, the culture surrounding speaking up within a program or institution will influence the trainee’s next steps. An environment that cultivates psychological safety is especially important for trainees in situations in which disagreements with attending physicians involve clinical decisions that may result in significant harm for the patient. When a trainee disagrees with an attending physician’s clinical decision, the trainee should assess (1) whether the decision will place the patient in imminent danger and (2) whether the decision is contrary to reasonable and/or evidence-based practice. Ideally, the trainee should feel empowered to express concerns to the attending physician. If concerns linger or if the trainee does not feel empowered to speak up, the trainee should immediately escalate concerns to training program leadership and use institutional safety processes and mechanisms to ensure the safe delivery of patient care.
Trainees and attending physicians, supported by training programs, should employ several strategies aimed at facilitating learning and open communication when disagreements are uncovered. First, training programs, in partnership with attending physicians, must cultivate a learning environment in which intellectual curiosity and questioning by trainees are not admonished or merely tolerated but rather are encouraged and celebrated. Trainees should frame their differing perspective or clinical approach as a question to the attending physician. Trainee questions should prompt meaningful dialogue regarding the evidence (or lack thereof) to support or refute a clinical decision, the role of uncertainty and its influence on medical decision-making, and, when relevant, the role of shared decision-making with patients and families. Second, attending physicians should openly share effective frameworks and approaches to making challenging clinical decisions. Attending physicians should be transparent and model how to talk through their medical decisions and uncertainty. Trainees require guidance from attending physicians regarding the multiple levers (clinical, health systems, and societal) that influence medical decision-making. Third, training program should provide trainees opportunities for further discussions about clinical decisions. Case-based morning reports or other formal learning sessions are beneficial, especially if other attending physicians are active participants in the discussion. Finally, training programs should encourage trainees to provide feedback to attending physicians regarding their role as an educator, even in times of disagreement, in the form of written, anonymous evaluations.
Dr Jehanna Peerzada, a Hospitalist and Bioethicist, Comments
This situation presents opportunities for both the learners and the educators involved. The intern and attending have conflicting ideas about the appropriate medical management of their shared patient. The attending is not certain about the patient’s diagnosis. For this reason, he prefers a longer trajectory of improvement before discharge. The intern disagrees. In a teaching institution, attending physicians bear ultimate responsibility for the patients they share with trainees. Practically speaking, therefore, the attending’s decision trumps the others. However, the trainees may respond to this situation in several ways that could benefit all of those involved.
First, learning how to handle disagreement with a colleague is a skill that is crucial in the practice of medicine. It involves finding evidence for one’s position in the medical literature and learning how to discuss this evidence with fellow physicians who possess a range of temperaments, levels of experience, insight into their own decision-making process, and openness to being questioned. Hanging in the balance at the end of these discussions is our ability as physicians to fulfill duties of beneficence and nonmaleficence to our patients. If a change of plan materializes after a disagreement, communicating this change to a patient or parent in a way that does not undermine colleagues or promote distrust in the profession is also an important skill.
In the above clinical scenario, the trainees should search for a clinical study that describes the natural history of adenovirus infection and then ask the attending to reconsider his decision in light of evidence that this virus can in fact cause prolonged fevers.16 The language and tone they use in the hierarchical setting of academic medicine will be different from the approach they take later in their careers. But both settings require humility, open-mindedness, and the ability to find quality evidence and use it persuasively and diplomatically.
This clinical situation offers an additional opportunity, which is to learn the art of shared decision-making. The ethical basis of shared decision-making is readily apparent in adult medicine. The model encourages physicians to form a partnership with patients to make evidence-based treatment decisions that support patient autonomy.17 Most pediatric patients are not autonomous decision-makers, although their participation in decision-making can and should gradually increase as they mature.18 Parents or legal guardians make decisions on their behalf and presumably in a manner that protects their interests; parental preference is not the primary driver of medical decisions made for children. Nonetheless, proponents of shared decision-making in pediatrics call for physician and parent collaboration in clinical circumstances in which there is more than one medically reasonable option with similar risk and benefit profiles.19
The attending’s decision in this case appears to be motivated by uncertainty about the diagnosis and a lower level of risk tolerance, which is understandable for a less experienced clinician.20 The mother may prefer discharge because the patient seems improved, and she and her son are exhausted from the hospital stay. Discharge may in fact not be in the patient’s best interests if the differential diagnosis includes a condition that could lead to acute decompensation later at home. In that instance, his parents might not be able to seek medical attention in a timely manner, and discharge could be unsafe for the patient. Barring that circumstance, the real harm in discharging this patient is the burden of returning to the ED: driving back, waiting to be seen, having an intravenous catheter reinserted, and being readmitted to the hospital. Depending on the nature of his concerns, the attending’s decision to delay discharge may be a value judgement that the benefit of feeling rested at home is not worth the burden of returning to the ED, no matter how unlikely it may be. But the magnitude and balance of those benefits and burdens are person specific and best determined by the mother and patient.
To inspire this sort of discussion on rounds that take place away from the bedside, the trainees could present the mother’s preference for discharge as an issue in the same way they might bring up aspects of care that are not strictly medical but nonetheless relevant, such as barriers to follow-up care. They should ask the attending physician how to respond to her request. Faced with the idea of overriding the mother’s preference and needing to offer an argument to the team for doing so, the attending might be forced to examine and clarify his reasoning. At the very least, he would have the chance to explain his approach to incorporating parental wishes into medical decision-making.
An even better approach, although one that requires excellent communication skills on the part of all participants,21 is to conduct rounds at the bedside. In fact, a major advantage of family-centered rounds is the ability to incorporate shared decision-making in real time as treatment plans are formed.22 In this case, the team would encourage the mother to share her thoughts about the patient’s illness. She would have the chance to advocate for discharge with all team members present, ideally sparking an honest and insightful conversation about the risks and benefits of this decision. This might not leave time for a literature search about the natural history of adenovirus, nor would an intern necessarily feel comfortable disagreeing with an attending’s decision in front of a patient and parent. Done skillfully, however, all participants on rounds would leave the conversation at least with a clear understanding of the decision and with the feeling that their perspectives were carefully considered.
Outcome of the Case
After the team’s discussion on rounds, the child was monitored in the hospital for an additional 24 hours, during which he continued to improve clinically. He had no additional fevers and was discharged from the hospital the following day with the diagnosis of an adenovirus infection.
Dr Armand H. Matheny Antommaria Comments
Disagreements between residents and attending physicians are common and should be anticipated. Although attending physicians should support residents’ autonomy, they must also promote patients’ welfare. In cases of conflict, residents and attending physicians should recognize that they have a shared interest in promoting patients’ welfare and should seek a principled resolution through appeal to evidence. If the predominant issue relates to tolerance of risk and approach to uncertainty, it is particularly important for the medical team to participate in shared decision-making with the patient and the patient’s parents because their tolerance and approach may differ.
Drs Crisci and Salih drafted the initial case, commentary, and outcome; Drs Unaka and Peerzada drafted the initial commentaries; Dr Antommaria drafted the abstract, introduction, and concluding comments; and all authors reviewed and revised the manuscript and approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.