On a recent Tuesday afternoon, I precepted four of our second-year internal medicine residents in clinic. Everyone in medicine knows what I am writing about. For those outside the field, it means I sat in a room and heard a brief presentation on each of the 24 patients the residents saw that afternoon. Some patients required a quick, “Sounds good.” Some patients got a long, considered discussion between me and the resident, a deep dive into their past medical records, and a visit from me with the resident.
It is a privilege to work with our residents. They are smart, kind, hard-working, and dedicated to caring for their patients. Whenever I work with them – precepting their clinics, staffing our urgent care, or, until last July, attending on the inpatient service – I come away invigorated, with renewed hope for the future of medicine.
All that said, I can be a curmudgeon. While embracing the opportunity to teach, I have to avoid extrapolating every plan I disagree with into a commentary about medicine in 2026.
This Tuesday afternoon I recognized five things that I often harp on. Since I am all about “making it count twice,” I thought I’d try to generalize these points into rules. If you’re still in training (aren’t we all?), consider this remote precepting. If you are teaching, maybe this helps. If you are neither, what would you add?
Do something
Often, the best advice for doctors is, “Don’t just do something, stand there.” Except in rare(ish) emergent situations, the best advice can be to stop and think before acting. However, in primary care, it is too easy for trainees to defer management to specialists.
Make management decisions!
You learn by taking responsibility. It is also what makes medicine interesting and (I dare say) fun. Think about whether something needs to be done and do it. Often, what needs to be done is to stop interventions or evaluations. If you are unsure, ask for help.
Before you do order a test, ask yourself three questions
What question am I trying to answer?
How likely am I to find something?
Will the results change my management?
When I am not telling people to take responsibility and do something, I am trying to stop them from ordering tests. I constantly make fun of our residents when they try to order an echocardiogram. I say, “You people don’t think you really know someone without an echo!” (I hope I come off as a good-natured scold and not a jerk.)
The answers to the questions are usually:
What question am I trying to answer? I don’t know.
How likely am I to find something? Not very.
Will the results change my management? Well, if it shows something (what that something might be, I don’t know), then maybe.
A diagnostic test is meant to answer a question. When we think about a test’s utility, we consider pretest probability, test characteristics, and how the results will change the probability of disease. If there is no question, there is no diagnostic utility.
Forget the guidelines
Guidelines can help, but you ultimately need to think about the patient in front of you. A COPD guideline might suggest a second-line inhaler regimen, but there is so much that affects the decision of which inhaler to use for an individual. Is there bronchodilator response on PFTs? Is the patient adherent? (If they are not, why not?) Does the patient want to change therapy? What medications have been used already? Are there reversible factors that are affecting his breathing?
Don’t lose track of the big picture
You need to know that evaluating gross hematuria includes pre- and post-contrast CT and cystoscopy. But, if the patient with gross hematuria also has non-operable pancreatic cancer, is on anticoagulation for a DVT, and their most recent CT showed only BPH, then the work-up should be obvious (see point 2, question 3).
The patient’s chief concern is not the primary problem
Osler supposedly said, “Listen to the patient; he is telling you the diagnosis.” You need to listen to patients. In 2026, the most helpful diagnostic data still comes from the history. I ask, “What do you think is causing your symptoms?” almost every day.
But what a patient tells you needs interpretation and reframing to become useful data. Your job is to represent the problem appropriately. Listen closely, but don’t be hemmed in by how symptoms are presented to you. A patient might complain of a pinched nerve. A detailed history might reveal that she is complaining of groin pain. Dizziness may prove to be vertigo or light-headedness. Fatigue might really mean weakness, dyspnea, or sleepiness.
Photo by Vitaly Gariev on Unsplash
I'd highlight the last one: the patient's chief concern is not the primary problem. Like all great truths, I think the opposite is also true. The patient's chief concern is the most important thing. When I am unsure where to go next, I return to the chief complaint and use that to guide my next move. Also, if a patient comes in for a sore toe and you ignore it because they also had shortness of breath, long after their reaction to someone's perfume has worn off, they're going to be pissed to have to pay a second copay to get their gout diagnosis.
As a long time “preceptor “ of FP residents this piece rings so true. “Baseline “ bloodwork, knee jerk consultations, unhelpful testing, and “do something “ are a large part of the guidance I provide. This piece should be required reading for all docs who do this work.