During one 28-hour call shift, I took care of a critically ill man in the cardiac intensive care unit. He had been sick for years, and had been getting worse over several months. It was a relatively quiet night, so I was able to spend a lot of time in his room. I watched him become increasingly confused and sleepy, his skin become mottled with net-like patches, and his blood pressure fall lower and lower despite adding multiple medications to keep it up. All of these were signs that he was close to death, and there was little chance he would get better.
His wife stayed at his bedside until late in the evening. She and I talked about his condition. She said he had been through “so much” during multiple moves back and forth between the hospital and a nursing home. He had told her that he was tired and didn’t want to suffer anymore.
Given that my patient could stay awake for only a few seconds at a time, his wife would need to make medical decisions for him if there were any important changes. After she left to get a few hours of rest, his condition continued to worsen. By the morning, I was sure he was near death and that escalating medical interventions would both be futile and not match his wishes.
During morning rounds, I described to my co-residents and the attending physician — my supervisor — the patient’s condition, my view of what was happening, and the conversation I had with his wife the night before. But after the attending examined the patient and reviewed his vital signs, he told me to start more aggressive interventions: more medication to support the man’s blood pressure, a catheter in an artery to track his blood pressure, and high doses of steroids.
I was surprised at how different his instructions were with what I thought we should do. I framed my disagreement as a question: “Do you think these interventions are within the patient’s goals of care?” I also suggested that we speak with the patient’s wife to see what she wanted us to do.
My feeble objection didn’t get very far, and I dropped the issue. Minutes later, I was inserting a catheter into an artery in the patient’s wrist and ordering the new medications, just as my attending had told me to do.
But when the man’s wife returned to the hospital an hour later, she was dismayed. Hadn’t she clearly explained what she wanted for her husband? She did not want us to pursue any of these interventions, as I had believed but had been unable to make my attending understand. She asked that we withdraw all care, and her husband died later that morning.
The relationship between attending physicians and residents is complex. Attendings are legally responsible for their patients’ outcomes, so the stakes are high for them if anything goes wrong. Attendings are also supposed to provide residents with the practical knowledge they need to eventually practice independently. Residents, in turn, handle more mundane tasks and execute medical plans on behalf of attending physicians.
The seniority of attendings leads to a strict hierarchy in which residents must defer to attendings on all major clinical decisions.
Disagreements between residents and attendings happen all the time. Given the hierarchical leadership structure, the attending always has the right to overrule a resident’s decision. But virtually every resident can rattle off a list of times he or she thinks an attending made the wrong decision.
Attendings have more experience than residents — sometimes decades more — but that doesn’t mean residents can’t occasionally be right when there’s a disagreement. A resident might be familiar with new research or guidelines because he or she has had more recent education on a topic. Attendings who are specialists in a field like cardiology or oncology may be less comfortable with conditions outside their expertise.
Perhaps the most common reason that residents can be right is because they typically spend much more time with patients and their families than attendings. That means they may have a stronger understanding of important psychosocial factors that affect patients’ clinical care.
In theory, a resident can file a complaint with the residency program director when there is a disagreement with an attending. In practice, this rarely happens. When decisions need to be made in a matter of minutes or hours, it isn’t practical to wait for a meeting with the program director to discuss the differing opinions. What’s more, vigorously arguing about a decision can strain a relationship, and when attendings have influence over a resident’s fellowship placement, the career risk from arguing with an attending can be significant.
If discreet questions don’t work, most residents defer to their attendings, even when they disagree with the decision. Residents must then carry out actions they do not think are in their patients’ best interest, which presents an ethical predicament. In the case of my intensive-care patient, I inserted the arterial catheter, causing him pain, even though I did not think it was the right thing to do. In other cases, residents may give dangerous medications they do not think are appropriate or may even arrange for a patient to undergo a surgery they disagree with. This can be both stressful and demoralizing to residents, not to mention potentially harmful for patients.
Disagreements between residents and attendings can also create confusion and undermine a resident’s credibility. Overnight, as I cared for my dying patient in the intensive care unit, a nurse had asked if I should place a catheter to monitor his blood pressure. I explained that I didn’t think it would change his course or be within his goals of care. The following morning, I had to explain to the nurse why I had reversed my stance, adding tension to our relationship.
In many cases, residents and attendings could reach an agreement if they had a frank discussion. But these discussions often don’t happen. Residents can be too timid to advocate strongly for their stance, and attendings are sometimes domineering and unwilling to entertain alternative perspectives.
There is a better way. Some attending physicians are masterful at encouraging their residents to challenge them. I once had an attending outline two possible treatment approaches for a complex patient and then ask each resident his or her view on which option would be better. It was only after hearing the opinion of each team member that he voiced his own opinion, explaining his thinking by drawing from what a couple of residents had said. He did not make the decision final until we had considered every resident’s concerns.
Another attending repeatedly drew me into conversations about patients I was not even caring for to ask my opinion on their cases. In doing so, she showed that she respected my clinical reasoning. Later on, when she made a decision I disagreed with, I felt comfortable raising my concerns because of the trust she had built up over our two weeks working together. We had an open conversation and recognized that there was a better solution that combined elements of both our approaches.
By encouraging dissenting opinions, attending physicians can foster richer conversations about important decisions. In some cases, these conversations will sway an attending toward the resident’s perspective. In others, the resident may gain a better understanding of the attending’s rationale and change his or her mind. In either case, both physicians will feel more comfortable about the decision, and the patient will benefit from a more thoughtful decision making process.
Alex Harding, MD, is a second-year resident in internal medicine at Massachusetts General Hospital in Boston.