Teaching on Rounds: How Senior Residents Teach Without Losing Time

Your actual job Read this first

As an intern, your job was to be right about your patients. As a senior, your job is to make sure the team is right about all of the patients, and to make the interns better than they were on day one. Those are different jobs. The second one is harder, because doing the work yourself is always faster than teaching someone else to do it, today. It is slower every day after that.

An intern who can be trusted to work up a fever alone by week two saves you an hour a day for the rest of the month. That hour is the return on the ten minutes of teaching you invested in week one. Teaching is not a tax on your efficiency. It is how you buy your efficiency.

The one-minute preceptor The only model you need

There are many named frameworks for clinical teaching. You need exactly one, and this is it. Five steps, in order, after a learner presents a patient:

1. Get a commitment. "What do you think is going on?" Do not let them present data and stop. The commitment is where the learning lives, the same reason the presentation guide tells students to end with "I think this is X." You cannot teach someone who has not taken a position, because there is nothing to examine.
2. Probe the reasoning. "What made you land there?" or "What else was on your list, and why not those?" You are not quizzing trivia. You are looking at how they think, so you can find the exact step where it went wrong, or confirm that it went right.
3. Teach one general rule. One. Portable, so it works on the next patient too. "Whenever a heart failure patient bumps their creatinine on diuresis, look at the ins and outs and the exam before you stop the diuretic, because a congested kidney is also an injured kidney."
4. Tell them what they did right. Specifically. Not "good job." Name the behavior: "You asked about medication adherence before anchoring on ischemia. That is exactly the right first question."
5. Correct the error. Also specifically, also aimed at behavior: "You presented the potassium of 3.1 and moved past it. Never present a critical value without a plan attached to it."

The whole thing takes sixty to ninety seconds. The order matters. Commitment before teaching, because otherwise you are answering a question they never asked. Reinforcement before correction, because it is true and because it keeps them talking to you tomorrow.

The one-minute preceptor, worked

Here is the entire model, live, outside a real door. The intern has just presented an overnight admission: a 68 year old woman with cirrhosis, admitted with confusion.

Count what happened. A commitment, the reasoning behind it, one rule the intern will use for the rest of residency, two specific reinforcements, one specific correction, and a plan, with a graded offer of supervision at the end. No slides, no whiteboard, and rounds moved to the next door on schedule.

Feedback that lands Specific, behavioral, timely

Most feedback in residency is worthless, and the interns know it. "You're doing great, maybe read more" is not feedback. It is a pleasantry with a vague insult attached. Feedback that changes behavior has three properties, and needs all three.

Notice the last sentence. Feedback with a follow-up attached is a plan. Feedback without one is a comment. And say the word "feedback" out loud: "Can I give you some feedback on that admission?" Interns often literally cannot tell they are being given feedback unless you label it, and then they report on the year-end survey that they never got any.

Running rounds on time The time budget

Teaching rounds run late for one reason: nobody is holding a clock. As the senior, you are the clock. The attending sets the destination; you set the pace. Budget it like this and say the budget out loud at the start.

Twelve patients with two new admissions comes out near two hours. If your census makes the arithmetic impossible, say so before rounds start and propose the split: "We have sixteen. I'd like to card-flip the eight stable ones in the workroom and walk on the eight active ones." Attendings almost always take that deal when the senior offers it with a plan attached.

Ways to protect the budget without killing teaching:

• Pick your teaching patients before rounds. While you build the run list, mark two patients as today's teaching cases. Everyone else gets efficiency. Teaching on every patient means teaching badly on all of them.
• Use the walk. The ninety seconds between doors is teaching time that costs zero minutes: "Before we go in, what's the first thing you'll look at on his volume exam, and why the JVP over the ankles?"
• Park it. When a genuinely great question surfaces at the wrong time, write it on your card in front of them: "That's a real question and it deserves more than a hallway answer. Parking it for after rounds." Then actually come back to it, because a parking lot you never revisit teaches them to stop asking.
• Hold the presenters to format. Rounds mostly run over because presentations sprawl. Enforce the structure in the presentation guide: tight one-liner, event-driven follow-ups, labs as patterns. Fixing an intern's presentation discipline is itself high-yield teaching, and it pays you back every single morning.

Supervising without taking over Their patients, your net

The fastest way to ruin an intern is to do their job for them. The second fastest is to let them drown. Your position is in between: they own the patient, you own the safety net, and both of you know exactly where the net is.

Practical rules:

• Questions before answers. When an intern brings you a problem, your first sentence is a question: "What do you want to do?" Every time you reflexively hand them the answer, you train them to stop generating plans and start collecting orders. If their plan is safe and reasonable, run it, even when yours is 10 percent better. The 10 percent is the tuition.
• Correct the plan through them, not around them. Do not quietly change their orders or edit their H&P from the workroom. Page them, walk through what you would change and why, and let them place the order and fix the note. An order changed behind an intern's back teaches nothing and tells them you talk about them instead of to them.
• Make escalation rules explicit on day one. "Call me for any transfer, any new pressor requirement or oxygen escalation, any family conflict, any time a nurse is worried even if you aren't, and anything that makes your stomach drop. There is no list of things too small. The intern I trust is the one who calls." Then, this is the part people skip, be pleasant at 3 a.m. Every irritated sigh raises the threshold for the next call, and the next call is the one that matters.
• Grade the supervision to the task. First paracentesis of the year: you scrub in and talk them through it. Fifth: you are in the room, hands behind your back. After that: "Text me when you're set up and when you're done." Say which mode you are in before you start, so nobody is guessing.
• Let the intern run their own show. On rounds, resist answering the attending's questions about the intern's patient. Redirect with your silence, or with "Priya knows this patient better than I do." Then defend her plan in the discussion if it was defensible. A senior who visibly backs the interns gets interns who tell the truth about what is going wrong.

Teaching at night Small doses, high retention

Night float is where interns actually learn to be doctors, because it is the first time nobody is standing next to them. It is also where teaching gets skipped entirely, on the theory that everyone is too tired. Wrong theory. Night teaching just has different rules.

• Teach on the callback, not the call. When the intern calls you about the hypotensive patient, deal with the patient. Thirty minutes later, when the pressure is 110 and everyone has exhaled, call back: "Walk me through what you think happened there. What would have made you call me sooner?" Two minutes, and it lands better than any noon conference, because the adrenaline stamped it into memory.
• Cross-cover is a curriculum. Fever, hypotension, low urine output, chest pain, altered mental status, the drop in sats. Every one of these is a teachable algorithm, and the intern will see all of them by Thursday. Keep it to one rule per event: "Before you give the sleeper, put eyes on the patient. Nobody ever got sued for walking to the bedside."
• Teach the handoff itself. The 6 a.m. signout is a skill, and most interns are never once observed doing it. Watch one, then give one piece of feedback using the structure in the handoff and signout guide: "Your sick patients and your stable ones sounded identical. Flag the two you're actually worried about first, with an if-then for each."
• The 3 a.m. lull is optional teaching, never mandatory. If the list is quiet and the intern is upright, a five minute case from earlier in the night is welcome. If they are horizontal, the teaching point is that sleep is a clinical skill. Protect it.

The habits that make it stick

• Ask, don't tell, as your default. The question mark is the most efficient teaching instrument you own. It costs the same five seconds as the answer and it shows you how they think.
• Admit ignorance out loud, then close the loop. "I don't know, I'll look it up tonight and tell you tomorrow," said by the senior, gives every intern on the team permission to stop bluffing. Then tell them tomorrow. The loop closing is the lesson.
• Teach the student through the intern. When the medical student presents, let the intern run the one-minute preceptor while you watch. Teaching them to teach is the last skill of residency, and you are the only one positioned to teach it.
• Keep a teaching file. One note on your phone. Every time a teaching point works, write it down. By spring you will have thirty reliable ninety-second points and you will never again stand in a hallway with nothing to say.
• End the month with real feedback. Ten minutes, sitting down, specific and behavioral, nothing they have not already heard from you in smaller doses. The end-of-rotation conversation should contain zero surprises. If it does, the surprise is the evidence that you waited too long.