How to Handle Pimping on Rounds (Without Freezing)
How do you handle pimping on rounds?
When an attending asks you a question on rounds, commit to an answer out loud, give the one reason behind it, and name what would change your mind. A committed, reasoned answer that turns out wrong teaches more and lands better than a hedge or silence.
The formula is short to say, harder to practice under a stare: commit, reason, what would change your mind; it tells them where you stand, why, and how new data would move you. Tiny example shape, not about any patient: "I think the answer is X, because Y; if Z appeared, I’d switch to W"; the rest of this post sorts the questions you’ll get into four types, and for broader tactics see surviving your MS3 clerkships.
Silence reads as no thinking under pressure, while a clear stance shows your process and gives them something to teach.
What pimping is, and why attendings still do it
Pimping is the attending quizzing learners on rounds with a rapid series of tough questions. As quoted in Detsky's 2009 JAMA essay, "pimping occurs when an attending physician (the Pimper) poses a series of difficult questions to a resident," and often to students, usually at morning report or attending rounds.
Brancati's 1989 JAMA satire popularized the term and traced it to 17th century London. He cited Koch's Pumpfrage on 19th century rounds, and Flexner documenting it on Osler's rounds at Johns Hopkins.
Brancati predicted it would fade. When Detsky revisited the idea in 2009, it plainly hadn't.
The defense is that, done well, it is Socratic teaching that ties preclinical theory to a specific patient in front of the team.
The critique is sharper. In a 2015 BMJ blog, Suzanne Gordon described attendings asking what trainees can't answer then humiliating them, and quoted a trainee saying, "We're told to respond with confidence and pretend we know stuff we don't know."
You can't opt out on most services. What you can control is the shape of your answer when you know it, when you partly know it, and when you don't.
The four kinds of pimp question
- Recall a fact questions sound like “What are the causes of X?” You’re being asked to list, usually in a standard frame the team expects. Answer shape: “I can give three now in an organized frame, then I’d check a reference after rounds for the rest.” Recall is mostly about reading, and nothing but reading fixes it.
- What else could this be questions sound like “What else is on your differential?” They want your ranking, not a dump. Answer shape: “Top pick is A because of the key feature, second is B given one finding, third is C but less likely because of a missing feature.”
- What would you order next questions sound like “What single test would you get now?” Commit to one test and the decision it informs. Answer shape: “I’d order T, and if it shows R I’d change my plan in this way, but if negative I’d pivot in that way, since a test without a decision attached isn’t the right next step.” For example, a D-dimer is sensitive but not specific, useful to rule out a clot only when probability is low or intermediate, so at high probability it doesn’t move the decision.
- Why did you ask that questions sound like “Why did you ask about that part of the history?” It’s about showing your hypothesis behind the question. Answer shape: “I asked it to test for Diagnosis D, because that question helps separate it from the main alternative in my head.”
Answering "what else could this be?" out loud
Martin Rowe, 55, runs a small joinery firm, usually fit and well. This is an AI patient case written and signed off by a doctor. "I've suddenly gone short of breath with chest pain over the last two days."
Vitals are heart rate 112, respiratory rate 24, BP 120/78, oxygen saturation 91 percent, temperature 37.1 C. He volunteers sudden sharp right-sided chest pain worse on a deep breath, and sudden breathlessness that’s worse on exertion.
Only if you ask, he mentions two small streaks of blood in his spit, a left calf that’s swollen, tight, tender and a bit warm for about a week, a very long-haul flight from Australia five days ago, and keyhole knee surgery three weeks ago.
He’s worried it’s a heart attack, and that coughing blood means something serious.
When the attending asks "what else could this be?", commit first. I think this is pulmonary embolism based on pleuritic pain, tachycardia, low oxygen, and the calf history. Then rank alternatives with the single feature against each.
- ACS, pain is pleuritic, not a heavy crushing central pressure, and it does not radiate to the arm or jaw.
- Pneumonia, no fever and no productive phlegm, and it would usually bring raised inflammatory markers and consolidation.
- Pneumothorax, in this case the chest X-ray is clear and breath sounds are normal.
- Aortic dissection, pain is typically tearing to the back with pulse or blood pressure differences, not this picture.
- Panic, a diagnosis of exclusion that does not explain the low oxygen, the coughed blood, or the calf.
That feature against each is what turns a list into an answer. If you want the quick pattern, read ranking a differential list.
You can also feel the next probe. "What risk factor did you ask about?" is the type-4 question this case sets up. The flight and the recent surgery only appear if you ask.
This is the move you can rehearse on cases where you must name a diagnosis before you see the answer, and then get scored on your reasoning. Try a case like this.
When you don't know the answer on rounds
If you don't know, say so in one sentence, then add how you'd reason from the data you do have or where you'd look it up, and offer to report back. Don't guess or bluff, because, as Gordon pointed out, trainees are sometimes pushed to pretend and bluffing works against safety.
"I don't know the full answer right now, but I'd start from the patient’s history and current labs, check a primary source, and I can bring a one-liner after rounds."
A 2020 Lippincott guide advises you to acknowledge you may not know the entire answer, mention how the information will help you, and say you plan to review the topic outside rounds. For phrasing options, see owning a gap honestly. This post won't repeat it.
How to prepare the night before rounds
Prep your actual patients, not a chapter. The goal is to make tomorrow’s questions about your list, not about trivia, so the conversation stays anchored to concrete data you gathered.
- For each patient you’re following, write your top diagnosis, then two realistic alternatives, and make that list explicit so you can defend it when pushed for reasoning. Next to each, add the single feature that pushes you toward it or away from it. Write it down.
- For every pending test, write what result would change your assessment and plan, and why you asked for it in the first place. If a result won’t change anything, be ready to say that.
- List the risk factors you asked about for this patient, and why they matter to your differential. Tie them to your HPI questions so your reasoning is visible.
- Pick the one topic on this patient you’d be embarrassed to miss, and reread only that. A focused reread beats unfocused scrolling.
This prep is the backbone for structuring your morning presentation, so the team hears a clear problem list with a defendable plan.
Offer to bring a short, relevant reading to teach the team, as Lippincott's guide to pimping suggests.
When pimping becomes mistreatment
Teaching uses questions to probe your reasoning and move you forward, humiliation uses them to make you small. The Lippincott 2020 guide is blunt, pimping should never be used as an opportunity to embarrass a junior member of the team.
Gordon in BMJ 2015 describes a pattern you may recognize, asking questions trainees can't answer, then humiliating them.
When it crosses the line you hear insults or sarcasm, you get singled out repeatedly for ridicule, or questions are used as punishment for small slips like being late.
Write down what happened, when and where, who was present, and the exact words you remember. Notes the same day help, even brief ones.
Talk to your clerkship director. If that feels unsafe, go to your school's ombudsperson or student affairs office.
From what I can tell, reporting routes differ by school, so check yours before you escalate.
Where to practice committing out loud
If committing out loud feels awkward, that is a reps problem. Diagnosica is a low-stakes place to build the habit, since you talk to an AI patient, commit to a diagnosis and a management plan, then see a scorecard after every case.
You can speak or type your history, order investigations, and view results and imaging. Each patient appears as a video avatar that answers back. You commit, then get competency scores and teaching points, on the free tier too. The library has over 130 cases across 18 specialties, and every case is written and signed off by a doctor.
Short on time, use the quick case. It is a two minute mini case where the patient says one line, you get five clues one at a time, and you guess after each. Then you see the answer and a one line teaching point. It keeps your daily streak alive.
Here is the boundary. It will not drill recall facts, it cannot stand in for a real attending, and there is no physical exam practice. It is a simulator. "It doesn't give advice about real patients, and it's not a diagnostic system. Treat every output as educational."
There is a free tier, no card. The Standard plan price sits on diagnosica.com. This habit pays off again in the grind of intern year itself.
Educational use only. Not medical advice. AI-generated; verify clinically against primary sources.
Clinical review pending.
