What medicine owes A former patient came back to see me after several months with another doctor at a practice where I had once worked. The patient told me that the other physician had made disparaging comments about me. The words stung. Someone had questioned my judgment and character, and I felt it. The patient had a different problem. Two doctors now seemed to be at odds, and the person seeking care had been left to decide whom to trust. Before we could deal with what had brought the patient back, we had to address doubts about the care that came before. I had not heard the other conversation. There might have been a concern about treatment worth discussing, an unjustified personal attack, or something the patient’s account did not capture. I needed to keep those possibilities open. I also needed to stop my own hurt from taking over the visit. The obligation went beyond being polite. It included explaining my reasoning, reviewing concerns about the care, and deciding whether a separate conversation or formal review was warranted. It also meant refusing to turn the patient into a witness for my defense. Whatever happened between professionals, the patient should not have to repair it. Who carries the work when an institution stops short of meeting a need? That question follows us well beyond the consulting room. A dispatcher must recover from a terrible call while answering the next one. A family must organize care after an intensive care admission. A patient must decide whether a confident recommendation deserves trust. Each person does work the institution may not count as its own. These situations demand different responses. A staffing shortfall cannot be addressed in the same way as sexual assault and institutional betrayal. The connection lies in where we look for responsibility: beyond the official account, toward what happened to the person relying on it. Medicine records completion. Consent signed. Referral placed. Training finished. Discharge summary sent. These records matter; without them, care would be less reliable. But they can also let us stop asking whether the person received what the procedure was meant to provide. A referral does not establish that a patient can obtain an appointment. A course does not establish that a physician has time to listen. A signature does not establish that a research participant could reasonably decline. The action may have been performed correctly while the person’s need remains unmet. When we miss this distinction, we can defend the procedure and miss the failure of care. As a psychiatrist, I am particularly concerned with how readily suffering becomes a judgment about the person who experiences it. Distress may be called weakness, disruption, noncompliance, or manipulation before anyone has examined what happened. A diagnosis can correct that error by making an illness recognizable and opening a path to treatment. It can also become another stopping point if naming the condition ends our curiosity about the person’s circumstances. The same caution applies in the other direction. Explaining distress through social conditions does not establish that treatment is unnecessary. An understandable fear can become disabling. A harmful institution can contribute to an illness that still needs clinical care. No one should have to wait for a workplace, a military organization, or society to improve before receiving help. The task is to care for the person while remaining honest about what lies beyond that person’s control. Care requires time to understand a problem, access to someone qualified to help, and choices a person can actually use. It also requires someone to respond when the plan fails. Who follows up an abnormal result? Who helps when a referral goes nowhere? Staffing, budgets, and rules assign these responsibilities, whether their authors acknowledge that work or leave it to chance. This argument places demands on clinicians, but it does not ask them to become endlessly available or absorb unlimited distress. A system that requires exceptional personal sacrifice for routine care has made its reliability depend on something it cannot reasonably guarantee. Patients deserve more dependable arrangements, and clinicians deserve working lives in which they can practice their professional commitments without making exhaustion a condition of belonging. Nor does the argument impose an unlimited duty on institutions to meet every request. Resources are finite. Some proposed treatments will cause more burden than benefit. Some services cannot be provided safely in a particular setting. Research can fail, and a study can become ethically or scientifically indefensible. A responsible institution must be able to say no. It must also explain the refusal, identify the consequences, and remain answerable to the people affected. Research can tell us whether an intervention helped in a particular setting. It cannot decide what an institution owes. Those decisions require judgment, and the people affected deserve to know the reasons behind them. The examples introduced with an invitation to imagine or suppose are hypothetical. The opening account of a patient carrying criticism between physicians is personal. It brings me back to an uncomfortable duty: before defending a profession, a process, or ourselves, notice what the person seeking care has been left to carry.