How Good is Your Doctor? Medicine Doesn’t Know

Medicine runs on an assumption so deeply embedded that we rarely notice it: we assume doctors are competent. Not perfect, not omniscient, but fundamentally capable. Safe enough to trust with our bodies and our lives. But how would you actually know if a doctor was competent? Good? Great? How can you tell the difference between a good doctor and a bad one? What does it even mean to be a ‘great doctor’?

Patients have very few signals to rely on. A diploma on the wall, board certification, a white coat, maybe a recommendation from a friend. Online ratings that often reflect parking logistics and wait times more than clinical skill.

Inside the profession, the signals are not much stronger. To maintain a medical license in the United States, physicians must complete continuing medical education credits and avoid serious disciplinary action. Board certification requires passing periodic multiple-choice exams. These are not meaningless standards, but they are also not particularly good measures of what most people intuitively mean by a ‘good doctor,’ raising a problem hiding in plain sight:

We do not have a clear, shared definition of what makes a doctor good.

The House of God, Samuel Shem’s 1978 darkly comic novel that generations of medical trainees eventually encounter, describes several familiar physician archetypes. The doctor with terrible bedside manner but dazzling clinical brilliance. The doctor beloved by patients but medically mediocre. But the trope that warmth and intelligence exist on opposite ends of a spectrum is mostly fiction. In real life, the physicians most admired by their peers tend to integrate technical skill, judgment, and human connection in ways that are difficult to disentangle.

Ask physicians privately who they would trust to care for their own family members, and there are some patterns that might emerge. Physicians often assume they can recognize excellence in their peers. But that recognition is shaped by context and perspective. We see one another in specific settings – on rounds, in conferences, in meetings – but not necessarily at the bedside when conversations are hard or trust is fragile. A physician admired by colleagues may be dismissive with patients. A doctor who projects confidence may not consistently exercise sound judgment. And a physician struggling with their own limitations may not reliably recognize excellence in others.

Recognition of what makes a good doctor exists – but it is imperfect and difficult to articulate.

In 1964, Supreme Court Justice Potter Stewart famously wrote about pornography:

‘I shall not today attempt further to define the kinds of material I understand to be embraced…But I know it when I see it.’

Defining clinical excellence in medicine is uncomfortably similar.

We may struggle to define it precisely but experienced physicians – and often patients – recognize clinical excellence when we see it: the doctor who makes patients feel heard and at ease, who takes a careful history that reveals what others missed, who can generate possibilities others wouldn’t consider when a problem doesn’t fit the usual patterns, who senses when something isn’t right despite reassuring tests, who explains uncertainty without eroding trust, who makes decisions that are both technically sound and emotionally attuned. But translating that recognition into measurable criteria has proven elusive.

Competence in medicine lives across multiple domains:

· Technical competence: Knowledge, diagnostic reasoning, procedural skill;

· Relational competence: Communication, empathy, the ability to understand what matters to a patient;

· Personalized, evidence-based decision-making: Applying current scientific knowledge to the specific person in front of you while aligning recommendations with their values, circumstances, and goals, under conditions of uncertainty.

We can test knowledge. We can sometimes observe procedures and outcomes. But relational skill and individualized clinical judgment – the parts of medicine patients experience most directly – are extraordinarily difficult to measure. And they are also where clinical excellence most often lives. Excellence is not just knowing the right answer. It is knowing what the right answer is for the specific person in front of you.

And yet, we have almost no systematic way to measure any of that.

When we can’t measure excellence directly, we substitute signals that feel reassuring: credentials, prestigious institutions, job titles, time in clinical care. I have seen this play out in my own career. I have spent much of my career trying to become an outstanding physician – not just competent, but excellent. The kind of doctor that I hope colleagues might trust with someone they love. At the same time, I built a research career. I study health systems, evidence, and how medical decisions actually play out in the real world. More than once, someone has casually assumed that because I am a researcher, I must not enjoy taking care of patients very much -- or that I am probably not very good at it. I remember the first time I heard those sentiments, as well as the offhand certainty of the assumptions. I felt genuinely stunned.

Over time, I came to realize that those reactions were not really about me specifically. They reflected the proxies we rely on to identify clinical excellence, visible signals that are easy to see but are not necessarily accurate. In this case, the proxy was the label ‘researcher’ itself, a visible identity that people used to infer something about clinical ability. The assumption that physician-researchers are somehow not good doctors reflects how easily proxies distort perception. It treats bedside excellence and analytical rigor as separate domains, when in reality they can often reinforce each other.

Even institutional structures sometimes reinforce the same fallacy. At my own university, physicians can be recognized as ‘master clinicians,’ but eligibility depends partly on the current proportion of one’s professional time spent in clinical care (and not necessarily cumulative experience or skill), a crude proxy for clinical excellence. Time spent doing something is not the same as doing it well. What makes this especially striking in my case is that my research training has cultivated skills relevant to clinical decision-making — clinical epidemiology, critical appraisal, and applying evidence to individual patients — skills that improved my ability to deliver personalized, evidence-based care at the bedside. Yet those competencies are largely invisible to the criteria being used. Under current criteria, I would not qualify – not because I don’t practice medicine, but because the metric being used prioritizes visible time over less visible expertise.

All of this points to a central problem: the qualities that make someone potentially a great doctor are often invisible and not measurable in the current systems we use to recognize excellence.

Why does any of this matter? Because modern health care increasingly depends on measurement.

We measure hospital quality. We measure outcomes. We measure productivity. We measure patient satisfaction. We measure costs. We compare physicians to algorithms. We debate whether artificial intelligence might outperform human clinicians. But if we cannot clearly articulate what makes a doctor good, we risk optimizing for proxies that miss the point.

We risk confusing board scores with sound judgment; efficiency with wisdom; friendliness with competence; confidence with accuracy. And, perhaps most importantly, we risk misunderstanding what kind of doctor we actually want.

This is especially visible in the safety-net, where I spend my clinical time. There is, appropriately, enormous focus on measuring disparities: measuring differences in outcomes across race, income, language, and access. Those measurements matter; you cannot improve upon what is unmeasured. But one of the most immediate levers for advancing equity is the quality of the individual care patients receive once they are in front of us – in order words, how good the doctor is – and that is something we measure far less well.

Medicine does track certain aspects of ‘quality.’ We measure adherence to guidelines, complication rates, readmissions, mortality, and patient satisfaction. These metrics are useful for understanding systems, but they are not the same as measuring the quality of how a doctor actually thinks, communicates, and practices. They tell us whether recommended steps happened, but they rarely tell us how well decisions were made – or whether the right decisions were made for each patient. They do not capture whether a physician recognized a subtle diagnosis early, avoided an unnecessary test, adapted treatment to a patient’s life circumstances, or noticed the one detail that changed everything.

In other words, we measure whether boxes were checked — not whether the doctoring was excellent.

That distinction matters enormously for closing disparities. Two patients can have identical insurance, identical diagnoses, and identical demographics, and still receive very different care depending on the judgment, attention, and decisions of the physician in front of them. The most immediate lever that physicians control to improve disparities is not policy or funding. It is the quality of the care we deliver. And yet that is the dimension we have the hardest time defining — and almost no reliable way to measure.

When physicians talk among themselves about colleagues they admire, the conversation rarely centers on exam scores or publication counts. Instead, it sounds more like this:

“She’s incredibly thoughtful.”
“He never misses things.”
“She’s so good with families.”
“He knows when to worry.”
“She’s calm when things get chaotic.”
“I’d trust her with my mother.”

These statements are less about credentials and more about the qualities that are the hardest to measure: judgment, attentiveness, and how someone shows up when decisions are uncertain. Not just what they know, but how they apply their knowledge and skills in real situations to individual people.

That’s what people are sensing when they recognize a great doctor.

And it raises a more personal question: if you needed to choose a doctor for someone you love, what would you look for?

Credentials help. Experience helps. Reputation helps. But much of what matters are qualities that are difficult to observe directly and nearly impossible to measure.

Medicine assumes competence. Patients depend on it. Physicians strive for it. Yet we still lack a clear way to define – let alone quantify – what makes someone excellent.

We know it when we see it.