Picture an illustrative session: a first-year class of 160 sitting in groups of six around round tables at 9:10 on a Tuesday morning. Nobody is taking dictation. Each student has a single sheet of paper with ten questions on renal physiology, and the room is loud in the way a good session is loud: people arguing about whether a drop in afferent arteriolar resistance raises or lowers filtration, and why.
A session like that looks nothing like a lecture with a few clicker questions added. I think the difference comes down to one change: treating the in-person hour as practice with feedback, and not as delivery of content.
This piece covers what that looks like for first and second year students, and what faculty need to have ready beforehand.
Why the format is worth the effort
A 2018 meta-analysis by Hew and Lo in BMC Medical Education pooled 28 comparative studies of flipped classrooms in health professions education. The authors focused on studies in which pre-recorded videos were provided before the face-to-face class meetings. The authors reported a standardized mean difference of 0.33 in favor of flipped classrooms over traditional ones. They also found the approach worked better when instructors used quizzes at the start of each in-class session. You can read the abstract on PubMed.
I’d suggest taking the finding about opening quizzes seriously. A short, low-stakes readiness check tells students whether they actually did the preparation, and it tells you what to spend the next forty minutes on.
What students do before the session
The pre-work has to be short enough that students will actually do it. I’d suggest keeping it to about 45 minutes and saying so in the syllabus. A typical package is one or two recorded lectures, one assigned reading, and a handful of practice questions with explanations.
Some schools build this from their own recordings. Others license outside material. Faculty who want pre-session video lectures and a question bank in one place can look at an institutional platform for MD and DO programs, which lists video lectures, question banks and flipped classroom use among its features. Whichever route you take, check that the pre-work lines up with the learning objectives for the session, because students notice quickly when it doesn’t.
What faculty prepare
The bulk of the work lands on the faculty side, and it’s front-loaded. A workable checklist for each session:
- A readiness quiz of 8 to 10 items, taken individually and then again as a group
- Two or three application cases that force a decision, not a recall of facts
- A one-page answer key with the common wrong turns written out
- A plan for the last ten minutes, which is where you correct what the room got wrong
The application cases matter most. A good one has students commit to an answer in public, at the same moment, so nobody can drift along with the loudest voice at the table. Team-based learning formalizes this with simultaneous reporting, where every group holds up its answer at once.
What the evidence says about structure
A 2026 systematic review by Ahmad-Naik and colleagues in BMC Medical Education looked at team-based learning in the first two years of pre-registration medical programs. The authors included 37 studies from 15 countries. Sessions were on average about three hours long and moved through distinct phases: preparation, readiness assurance and application. The authors noted limited reporting of the 4S framework for designing the application tasks, and of peer review, and said curriculum governance and faculty development were needed to scale the approach. The review is on PubMed.
Two things stand out to me. First, the phases are not decoration. Skipping the readiness step is the fastest way to end up with a room of students who haven’t prepared and a faculty member who ends up lecturing anyway. Second, faculty development is on the list for a reason. Running a room of 160 students in small groups is a different skill from lecturing to them.
What it looks like when it goes wrong
Three failure patterns are easy to predict.
The first is pre-work that’s too long. Students skim it, the readiness quiz exposes this, and the session turns into a rescue lecture.
The second is cases that are too easy. If every group gets the right answer in thirty seconds, you’ve run a quiz and not a discussion. Write at least one case where reasonable people will disagree.
The third is skipping the debrief. Students will tolerate a lot of group work if they leave knowing what the right answer was and why. They won’t tolerate being left in the dark.
Running your first one
If you’re new to this, don’t convert a whole block at once. Pick one session that already has a clear set of objectives, preferably one where students struggle with application, and flip only that one. Tell students why you’re doing it. Ask them for feedback in the last two minutes, on paper, with two questions: what helped, and what got in the way.
Then change one thing before the next session, such as the length of the pre-work or the difficulty of the cases. Changing one thing at a time makes it easier to tell what helped.
Start with one session and one set of cases this term, and write down what you’d change before you walk out of the room.
