The Anatomy Of A Doctor Visit

Follow on

Working as I do in an academic medical center, I’m frequently called upon to teach medical students. I do most of my teaching now with the third-year medical students when they rotate through our outpatient clinic. Usually, I’ll send them into an exam room on their own to see my patients (after requesting my patients’ permission to do so, which they nearly always grant) and then have them present the patient’s history to me in front of the patient. This model is efficient and accomplishes much: the medical students have the opportunity to sharpen their history-gathering and history-presenting skills, patients have the opportunity to clarify the history the students provide me as well as to make a contribution to the education of tomorrow’s doctors, I hear a (hopefully) concise and accurate story that I can pick through to arrive at accurate diagnoses of my patients’ problems, and I can evaluate and provide feedback on students’ clinical and personal skills having watched them in action.

I typically give a small talk to the medical students who rotate with me on the thought process I use in tackling patients’ problems in my clinic, and I thought readers might find it interesting to know how doctors (at least, this one) typically approach patient visits in an outpatient clinic.


The moment I enter an exam room, before my patient even has a chance to speak, I’ve already started my assessment. This first part of the process happens quickly and is largely unconscious. In a flash I take in the picture I’m seeing: is my patient sitting comfortably reading a book or fidgeting in a strained posture? Does he stand to shake my hand when I enter the room or avoid all eye contact? What are his first words and how are they spoken? No right or wrong answers exist to these questions, and I’m careful to not infer too much from them. But I do always wonder about what the answers might imply: am I seeing anxiety, depression, impatience, anger? It matters a great deal which, if any, are present, both to the process of gathering a history and to the correct assessment of a patient’s problems.

Each and every time a patient comes to see me, I remind my medical students, they have an agenda and I have an agenda. (By “agenda” I mean simply a list of things we each want to discuss.) They may have only one thing they want to discuss or they may have twenty. Our agendas may overlap (I may also want to talk about their arthritis and cholesterol, for example) but often they’re quite distinct. (Obviously, I don’t yet know what new complaints they’ve brought with them.) What I do know patients bring with them in variably insignificant, small, and large quantities is anxiety—not only about what terrible malady I might end up telling them they have but also about being prevented from communicating their concerns. In today’s world, doctors tend to demonstrate non-verbally (and sometimes even verbally) their sense of urgency to be elsewhere. Many of us seem (and are) constantly distracted by the next thing we have to do, the end result often being that we’re not only half-listening to what our patients tell us but also are rushing through the visit, failing to ask clarifying questions that just might reveal the key to our patients’ problems. (The only remedy I’ve found for the impulse to rush through a patient visit lies in direct primary care, where physicians reduce their panel sizes significantly so they have enough time to spend with each patient and don’t feel that they have to rush on to the next.)

To combat this problem (before I transitioned into a direct primary care practice), I deliberately spend a moment making small talk, deliberately slowing myself down and communicating by my action that I’m going to focus all my attention on my patients and listen to what they have to say. Then I frequently (though not always) do the following: I ask them to list for me everything they want to talk about, requesting that they avoid going into any detail as they do so. (That will come later, I reassure them.) After each item (“my ankle hurts”) I prompt them to continue by asking, “Anything else?” When they finally pause to consider if they have anything else left that concerns them and at last answer, “That’s about it…” I always respond, “Are you absolutely sure?” Usually they pause again, conduct a mental inventory or consult a list they’ve written out beforehand (something I heartily endorse), and tell me they’re done.

At this point, I’ve accomplished two important things (usually in under two minutes): I’ve learned the universe of their concerns and I’ve made them feel heard. When patients complain that their doctors rarely spend enough time with them what they’re really irritated by isn’t the time not spent but how little listening their provider seemed to do. In my experience, it’s possible to spend only five minutes with a patient (depending, of course, on the reason(s) they’ve come in) but still have them come away completely satisfied with the visit quality. (Accomplishing this feat in so little time isn’t the goal, of course, but the fact that it happens proves the point.)

Having heard and understood the full extent of a patient’s concerns, I’m now free to combine their agenda with mine and arrange them into one large list, prioritizing them using my clinical judgment according to their potential seriousness. Due to anxiety, for example, a patient may have mentioned “chest pain” as their last complaint, hoping whether consciously or unconsciously to downplay its significance. Luckily, because they weren’t given any time to go into detail before I gathered the entire list of their complaints, I can make their last complaint mentioned the first complaint discussed, ensuring I have enough time to gather those all-important details to the degree I require. If I’d failed to obtain an exhaustive list from them at the beginning, the complaint of chest pain might only surface at the end of the visit (something that used to happen to me all the time). I’d then be forced to spend an additional twenty minutes eliciting the details around this most important complaint, making me late for my next patient and preventing me from spending the appropriate amount of time and appropriate amount of focus on them.

Sometimes, of course, as I mentally glance over my combined list, I find no medically compelling reason to prioritize any one problem over another. In that case, I turn to the patient and ask them which complaints they want to discuss first.


Once I’ve had my patient list their complaints, we start discussing the details of each. The medical history, I teach medical students, remains the most powerful diagnostic tool we have. Most students are skeptical of this for a long time, thinking diagnoses are mostly made with technology. Certainly, I acknowledge when I hear them say this, technology has dramatically increased our ability to make many diagnoses. But with it has come an unintended consequence: an increased likelihood that providers will exhibit lazy thinking. Why bother to consider the diagnostic possibilities past a certain level of detail, after all, if you can simply order a test to get the right answer?

The answer, it turns out, is that if you apply technology with insufficient forethought, not only will you order a plethora of unnecessary tests on the way to the diagnosis, you may very well miss the diagnosis entirely. As just one example, consider pain—something that often has a functional cause rather than an anatomic cause, rendering the all-too-often mindlessly ordered CT scan useless. Unfortunately, I’ve lost count of the number of times I’ve had medical students suggest ordering abdominal CT scans without knowledge of a single qualifying detail of a patient’s abdominal pain. This thought error may be forgivable in medical students (it’s my job, after all, to train them), but I’ve seen many medical residents commit the same kind of error. I’ve come to believe it doesn’t happen just because ordering a CT scan is easier than thinking or even that residents are terribly rushed (which they are), but rather because many newly minted doctors simply haven’t yet learned to trust the data the medical history offers them. It’s a trust that seems only to develop with experience.

Nevertheless, each fact a health care provider gathers in the taking of a medical history is, in fact, a test itself. When I ask a 55-year-old man complaining of chest pain if he gets it with exertion and he says yes, in the right clinical setting that positive result carries as much prognostic value as a stress test. With each subsequent question, I adjust up or down the likelihood of the various diagnoses I’m considering until I arrive at a final “pretest” probability for each. The term “pretest” probability is used to indicate the probability of a disease being present that’s been calculated from the history alone, before more traditional tests are ordered. But it’s a misleading term in one sense because the testing has long since already begun—with the first question I asked.

Interestingly, the amazing advances in technology we’ve enjoyed in the last fifty years have added far more to our ability to treat than to diagnose. This is why training medical students to take a thorough medical history will always remain relevant: no matter how much better our diagnostic tests become, we only know to order them in the first place because the history we gather first leads us to consider the diagnosis they’re able to make. Medicine, it turns out, isn’t a science at all, but rather the art of applying science to symptoms in such a way that yields us a diagnosis.