Case and Commentary
Aug 2018

Should Physicians New to a Case Counsel Patients and Their Families to Change Course at the End of Life?

Shyoko Honiden, MD, MSc and Jennifer Possick, MD
AMA J Ethics. 2018;20(8):E699-707. doi: 10.1001/amajethics.2018.699.


Although new cancer therapies have changed the prognosis for some patients with advanced malignancies, the potential benefit for an individual patient remains difficult to predict. This uncertainty has impacted goals-of-care discussions for oncology patients during critical illness. Physicians need to have transparent discussions about end-of-life care options that explore different perspectives and acknowledge uncertainty. Considering a case of a new physician’s objections to an established care plan that prioritizes comfort measures, we review physician practice variation, clinical momentum, and possible moral objections. We explore how to approach such conflict and discuss whether and when it is appropriate for physicians new to a case to challenge established goals of care.


Dr. T is a medical intensivist who has been an attending physician on the intensive care unit (ICU) service for the past week and will hand off her patients’ care to another intensivist, Dr. B. While signing over her patients, Dr. T took great care to relay each patient’s case details, gathered in part from meeting time with each patient’s family.

John is a 65-year-old man with stage IV non-small cell lung cancer. Previously healthy and active, he was diagnosed 6 months ago and underwent several rounds of palliative chemotherapy and radiation. Despite aggressive treatment, his disease and symptoms continued to progress; due to his previously very high level of functioning, however, he was considered a good candidate for a new clinical trial with a novel regimen that has some positive preliminary results in refractory Stage IV disease. Unfortunately, right before starting this treatment, John became critically ill. Two weeks ago, he was admitted to the ICU with acute respiratory distress syndrome (ARDS) and septic shock from pneumonia requiring multiple medications to stabilize his blood pressure.

Despite the long hospital course, within 2 weeks, John had stabilized. He was on minimal ventilator settings, had weaned off vasopressors, and no longer required antibiotics. However, John remained intermittently delirious, and any time a nurse would try to lighten his sedation he became severely agitated, attempting to pull at lines and his breathing tube. Although spontaneous breathing trials had been attempted daily, John continued to fail them due to severe agitation and an increased respiratory rate.

Although now medically stable overall, John was nearing a point of requiring a tracheostomy due to the prolonged duration of intubation. In light of his advanced disease and protracted hospital course, Dr. T and John’s wife, Lisa, had discussed goals of care on multiple occasions. Because John is currently incapacitated, he has been unable to participate in any decision making. Lisa, having never discussed John’s wishes with him in the past and feeling overwhelmed by the situation, turned to Dr. T for guidance. Considering the combination of widely metastatic disease, severe deconditioning, and inability to wean from the ventilator, Dr. T reasoned that John’s chance of achieving any kind of meaningful recovery was extremely unlikely. Importantly, it seemed unlikely that he would ultimately become well enough to enroll in the clinical trial. Instead, she thought in John’s case that more interventions were likely to inflict harm and increase complications. Dr. T recommended against the tracheostomy in favor of transition to hospice care. After giving it much thought and discussing it with her children, Lisa agreed with Dr. T’s recommendation and decided on terminal extubation with transition to comfort measures only. To give time for family to arrive from out of town, the plan was for John to be extubated Saturday, the day that Dr. B would be taking over John’s care.

As Dr. B listened to Dr. T review the case, he became increasingly uneasy with the plan that he was signed out to implement. Dr. B did not feel comfortable with terminal extubation for John. He felt that his respiratory prognosis remained uncertain and that transition to tracheostomy might facilitate decrease in sedation and improvement in mental status. He could not rule out the possibility that John might eventually become a candidate for clinical trial entry. Dr. B offered, “John has an advanced malignancy, but two weeks ago, was deemed a good candidate for a clinical trial with promising preliminarily results—why not keep going and try to give him the best shot we can?” Although he realized the chances of John making a recovery to baseline were small, he had seen some dramatic recoveries by similar patients, so he did not deem terminal extubation appropriate at this time. Dr. B was inclined to advocate for tracheostomy in hopes that John could recover enough to allow potentially life-saving therapy.

Dr. T, a colleague for whom he had great respect, had established a relationship with the family, knew them well, and had developed a comfort care-oriented plan that the family felt comfortable with and seemed ready to implement. Tomorrow he would come to work and be expected to take over where Dr. T had left off. Perhaps he should leave things the way they were, but now that John was under his care, Dr. B wondered whether he should help John pursue a different, more aggressive acute care-oriented plan.


This case highlights how rapidly evolving cancer therapeutics challenge our understanding of “advanced-stage” disease and lend new nuance to end-of-life (EOL) decision making. According to the National Cancer Institute, the overall 5-year survival rate for those diagnosed with stage IV non-small cell lung cancer (NSCLC) remains around 5%.1 However, thanks to mutation-specific agents and immunotherapies, long-term survivorship has become a possibility for some patients.2  Conversations about prognosis, critical illness, and EOL are understandably influenced by these advances and can yield divergent recommendations from different physicians. Here we will focus on physician-centered and systems-based forces that impact EOL decision making and explore whether it is ever appropriate for physicians assuming care at a critical juncture to counsel patients (and families) to revisit decisions.

Physician Practice Variation as a Force in Decision Making

Surveys suggest that most patients prefer to die without aggressive life-sustaining therapies,3 yet roughly 25% of Medicare beneficiaries die in the hospital.4 Although this issue is complex and reflects many factors, some studies have highlighted the influence of physician practice variation. A recent study of nearly 200 000 patients with metastatic cancer found that the single most influential factor in determining whether a patient died in hospice care was the lead physician’s prior referral frequency to hospice.5 Another study of nearly 22 000 Medicare patients with advanced NSCLC found that 43% received chemotherapy within 30 days of death and that after adjusting for other patient and physician characteristics, physician practice in a smaller independent office was a predictor of more aggressive care.6 What motivates physicians to pursue or reject aggressive care is not known but could include factors such as personal beliefs, personality, knowledge deficits, and cognitive biases.

Practice variation as a potential problem in decision making. Such heterogeneity in physician practice raises a question about whether practice variation is inherently bad. The notion of a second opinion arises from patients’ recognition that different clinicians approach cases differently and that they can seek such opinions to either elicit other perspectives or confirm a prior decision. This process has value independent of the outcome. Most patients with cancer seeking a second opinion at crucial junctures in care report that the process is helpful and reassuring, regardless of whether such consultations yield a divergent diagnosis or alternative therapies.7,8 Within established parameters of best practice, nuanced recommendations are expected and encouraged in the pursuit of personalized, patient-centered care, and informed decision making is supported by the clinician community. As an example, the American Cancer Society provides detailed patient resources on this topic and encourages patients to solicit a variety of perspectives to inform their decision making.9

In the shared decision-making model, respect for patient autonomy is tempered by physician expertise and judgment, and decisions are neither solely vested with the patient nor paternalistically with any one clinician. Confronting uncertainty and acknowledging differing opinions are important aspects of effective communication in this model.10 Rather than undermining clinician credibility, such actions are more likely to foster trust and respect provided a clear recommendation is conveyed. For example, family members who were surveyed after the patient’s death reported that they would have wanted more communication regarding prognostic uncertainty, including knowledge that death was possible or probable, to help inform decision making.11 Withholding divergent opinions, particularly at critical decision points, can undermine effective partnerships. Thus, physician practice variation is not inherently bad—but recommendations must be conveyed carefully and effectively in a dialogue including many perspectives. That medicine is both an art and a science is widely accepted.

Exploring physician variation in the clinical vignette. In the present case, Dr. T and Dr. B have a difference of opinion regarding John’s prognosis. Dr. T viewed John’s chance of achieving ventilator liberation and eventually receiving further treatment as infinitesimally small, while the risks of ongoing harm and suffering were great. Dr. B similarly acknowledged that chances were slim but framed the prognosis differently. He saw John as “stable” with the possibility of further improvement and, given his good premorbid functional status, wanted to give him the benefit of the doubt. The vignette also suggests that Dr. B’s recollection of dramatic recoveries by other patients is an instance of confirmation bias, which is a tendency to interpret or recall information in a way that confirms one’s own hypothesis. The pertinent question here is how likely these recoveries are, how relevant these anecdotal cases are to John’s current situation, and what the risks and benefits of further aggressive treatment might be.

If Dr. T did not acknowledge prognostic uncertainty when Lisa solicited an opinion from him to guide her decisions, Dr. B’s discomfort might be justified. Assuming Dr. T was transparent about prognostic uncertainty, his recommendation for conservative management seems reasonable, given that John had entered a phase of chronic critical illness with low likelihood of a good outcome (a perspective shared by Dr. B). Importantly, Lisa was given time to consider this information and discuss it with her children. With all in agreement, she elected comfort measures and the family was given sufficient time to execute that decision in a meaningful way (ie, waiting for family to arrive from out of town). Ultimately, if risks, benefits, and alternatives were reasonably discussed and patient values and goals were elicited, Dr. T’s recommendation sufficiently balances respect for autonomy with an informed physician perspective, thereby adhering to the principles of shared decision making.

Clinical Momentum as an Underappreciated Force

Clinical momentum is a systems-level force that can propagate unwanted aggressive care during critical illness or at EOL.12 Akin to a biologic cascade like hemostasis, an initial clinical circumstance prompts therapeutic actions that in turn propagate more interventions, even when clinical circumstances have changed. Frequent hand-offs of care in the critical care setting make it even more challenging for clinicians to be cognizant of the power of clinical momentum. Physicians covering a complex and critically ill patient over a weekend or during a 12-hour night shift, for example, might find it comparatively more difficult to appreciate a larger, longer context of a patient’s care than physicians taking care of that patient over a longer duration. They might be more likely to make decisions—even significant ones—based on impressions formed in narrower timeframes.

As in John’s case, tracheostomy frequently signifies a pivotal turning point in the management of persistent respiratory failure, as it represents a commitment to a plan of chronic ventilatory support or long-term weaning.12 And yet, for patients and families, this distinction might seem artificial and difficult to distinguish from the myriad of critical decisions favoring aggressive care that have already been made, such as vasopressor support and mechanical ventilation. Families might acquiesce to a series of smaller decisions that, in aggregate, are not aligned with patient preferences and fail to appreciate that certain choices, like tracheostomy, contribute to the momentum of aggressive care. Indeed, clinical momentum is often unrecognized in the moment and hinders patients, families, and clinicians from pausing to consider alternatives and long-term outcomes.12

In the present vignette, there is no apparent conflict between Lisa and Dr. T about transitioning John to comfort measures. In fact, the shift from aggressive to comfort-focused care signifies a willingness to disrupt preexisting clinical momentum and can motivate appropriate realignment with patient goals, given the prognostic information available. However, one might wonder whether Dr. B is unconsciously influenced by clinical momentum. Dr. B perceives John to be stable (because he is now on minimal ventilator settings and weaned off vasopressors) after a rocky ICU course with severe ARDS and septic shock. He hopes that with a tracheostomy in place, John might become less delirious and agitated, weaned from the ventilator, and ultimately a candidate for future chemotherapy. But how likely is this?

Acute respiratory failure is a leading diagnosis among patients with cancer admitted to the ICU and a significant source of mortality.13,14 In a cohort of 5 000 cancer patients requiring ventilator support for more than 21 days, median 1-year survival was 14.3%.15 Subgroup analysis of those with lung cancer revealed a dismal 1-year survival rate of 6.6%.15 One could postulate that, among patients requiring ventilator support for several weeks like John, achieving a sufficient performance status to receive investigational therapy would be unusual. Such discouraging data support the hypothesis that Dr. B might have succumbed to clinical momentum in advocating for ongoing aggressive measures in a cancer patient with respiratory failure. In upholding the principles of beneficence and maleficence, it is important for physicians to be able to navigate a medical environment primed to propagate clinical momentum and to avoid imposing unwanted aggressive care upon patients and their families.

Dr. B’s Potential Moral Objections

Dr. B’s discomfort with the current plan of care might be broken down into concerns regarding respect for autonomy, beneficence, and, finally, personal objection. Is Dr. B concerned that Lisa was inappropriately counseled by Dr. T? If there is evidence that information Dr. T provided about the prognosis was incorrect or that communication was ineffective or coercive, Dr. B is obligated to revisit the goals-of-care discussion, framing all the facts of John’s case in an objective manner to ensure that Lisa would be able to make decisions based on her representation of John’s best interests. However, if no such concern exists, placing a grieving family at the center of an intellectual conflict is unnecessarily destructive. Professional disagreements about inappropriate care must be distinguished from moral objections to valid care decisions. Our obligation as physicians, first and foremost, is to help patients and families through effective peer-to-peer communication during patient care transfers. Dr. T. could frame the present plan of care by outlining a summary of John’s overall course, his evidence-based prognosis, and his proxy-represented wishes based on this information, while Dr. B could raise his concerns at the time of transfer of care to facilitate a collegial discussion.

A good next question is whether Dr. B is required to execute a plan that he does not fully endorse. Does his objection rise to the level of moral objection—such as being asked to terminate a pregnancy in the face of a religious or spiritual objection? Dr. B’s objection is unlikely to reach that threshold. Physicians sometimes support decisions made by patients that they do not agree with—for example, a fully informed decision made by a patient to forgo intubation even when there is significant chance of benefit and meaningful recovery. This decision might not be one that the physician would personally elect, but that does not preclude another reasonable person from suggesting or selecting an alternate course, and thus patient autonomy should be respected. While not often talked about in the context of another clinician’s decision, the professional autonomy of a clinician should be accorded respect much like the patient’s personal autonomy, especially in view of the reality of physician practice variation previously explored.

There is very little in the literature about how to resolve treatment conflict among clinicians. Some concepts can be borrowed from futility disputes, although such disputes typically pertain to conflict between physicians and families. And while futile care is an extreme scenario not illustrated in this case, resolving concerns related to potentially inappropriate care can draw upon similar interprofessional communication strategies for conflict resolution.16,17 While no single approach can be recommended, effective communication and trust building are necessary for treatment conflict resolution. If Dr. B cannot accept Dr. T’s plan as outlined, an open discussion between the two physicians is an excellent start. The involvement of another impartial physician, much like a second medical opinion sought in futility disputes between a clinician and patient, might be helpful. If deemed necessary to reinvolve family, both physicians should be present to facilitate discussion in a way that does not jeopardize the existing trust and the relationship between Dr. T and the family and avoids undue emotional distress. If the family remains comfortable with the original decision, but Dr. B remains troubled, reassigning John and his family to another ICU team who can execute the plan could be the best course of action.

Fostering Dialogue about EOL Care

In summary, physician practice variation is common and reflects prognostic uncertainty, particularly in fields of evolving therapeutic options. In the shared decision-making model, respect for patient (or surrogate) autonomy and well-informed and well-communicated physician judgment are both considered in aligning difficult care decisions with patients’ values and preferences. Verbalizing prognostic uncertainty or differences in opinion could strengthen trust between physicians, patients, and families by acknowledging that such gray areas exist. We should embrace and invite differing perspectives from our peers and encourage dialogue about critical decisions at transfers of care, neither yielding to inappropriate clinical momentum nor unnecessarily derailing care plans that are thoughtful, supported by available data, and appropriately communicated to patients and families.


  1. National Cancer Institute. Browse the SEER cancer statistics review (CSR) 1975-2014: table 15.14: non-small cell cancer of the lung and bronchus (invasive). Accessed February 14th, 2018.

  2. Rosell R, Neninger E, Nicolson M, et al. Pathway targeted immunotherapy: rationale and evidence of durable clinical responses with a novel, EGF-directed agent for advanced NSCLC. J Thorac Oncol. 2016;11(11):1954-1961.
  3. Dartmouth Atlas of Health Care. End of life care. Accessed February 14th, 2018.

  4. Teno JM, Gozalo PL, Bynum JPW, et al. Change in end-of-life care for Medicare beneficiaries: site of death, place of care, and health care transitions in 2000, 2005, and 2009. JAMA. 2013;309(5):470-477.
  5. Obermeyer Z, Powers BW, Makar M, Keating NL, Cutler DM. Physician characteristics strongly predict patient enrollment in hospice. Health Aff (Millwood). 2015;34(6):993-1000.
  6. Green JB, Shapiro MF, Ettner SL, Malin J, Ang A, Wong MD. Physician variation in lung cancer treatment at the end of life. Am J Manag Care. 2017;23(4):216-223.
  7. Ruetters D, Keinki C, Schroth S, Liebl P, Huebner J. Is there evidence for a better health care for cancer patients after a second opinion? A systematic review. J Cancer Res Clin Oncol. 2016;142(7):1521-1528.
  8. Hillen MA, Medendorp NM, Daams JG, Smets EMA. Patient-driven second opinions in oncology: a systematic review. Oncologist. 2017;22(10):1197-1211.
  9. American Cancer Society. Seeking a second opinion. Accessed February 14, 2018.

  10. Quill TE, Brody H. Physician recommendations and patient autonomy: finding a balance between physician power and patient choice. Ann Intern Med. 1996;125(9):763-769.
  11. Krawczyk M, Gallagher R. Communicating prognostic uncertainty in potential end-of-life contexts: experiences of family members. BMC Palliat Care. 2016;15:59. doi:10.1186/s12904-016-0133-4.

  12. Kruser JM, Cox CE, Schwarze ML. Clinical momentum in the intensive care unit. A latent contributor to unwanted care. Ann Am Thorac Soc. 0014;14(3):426-431.
  13. Azoulay E, Schellongowski P, Darmon M, et al. The intensive care medicine research agenda on critically ill oncology and hematology patients. Intensive Care Med. 2017;43(9):1366-1382.
  14. Aygencel G, Turkoglu M, Turkoz Sucak G, Benekli M. Prognostic factors in critically ill cancer patients admitted to the intensive care unit. J Crit Care. 2014;29(4):618-626.
  15. Shih CY, Hung MC, Lu HM, Chen L, Huang SJ, Wang JD. Incidence, life expectancy and prognostic factors in cancer patients under prolonged mechanical ventilation: a nationwide analysis of 5,138 cases during 1998-2007. Crit Care. 2013;17(4):R144. doi:10.1186/cc12823.

  16. McCabe MS, Storm C. When doctors and patients disagree about medical futility. J Oncol Pract. 2008;4(4):207-209.
  17. Kon AA, Shepard EK, Sederstrom NO, et al. Defining futile and potentially inappropriate interventions: a policy statement from the Society of Critical Care Medicine Ethics Committee. Crit Care Med. 2016;44(9):1769-1774.


AMA J Ethics. 2018;20(8):E699-707.



Conflict of Interest Disclosure

The author(s) had no conflicts of interest to disclose.

The people and events in this case are fictional. Resemblance to real events or to names of people, living or dead, is entirely coincidental. The viewpoints expressed in this article are those of the author(s) and do not necessarily reflect the views and policies of the AMA.