How to structure a grand rounds presentation

Case-based or topic-based

Most grand rounds follow one of two formats, and many departments expect one or the other.

• Case-based: the talk opens with one or more real patients, de-identified, and uses the case to raise a clinical question. The rest of the talk reviews the evidence on that question and returns to the case at the end. This suits diagnostic puzzles, rare presentations and decisions where guidelines are unclear.
• Topic-based: the talk reviews a subject, such as a new guideline, a change in practice or the state of research on a condition, without a single case at its center. Visiting speakers and faculty often use this format, and it suits audiences that span several subspecialties.

A hybrid is common: a short case to frame the question, followed by a topic review. Whichever you choose, decide early on the one clinical question the talk answers. A grand rounds without a question tends to become a textbook chapter read aloud.

A standard structure

1. Title slide: the talk title, your name, role and department, and the date.
2. Disclosures: relevant financial relationships, or a statement that you have none. Accredited continuing education activities generally require this, and many departments ask for it regardless.
3. Learning objectives: two to four, written with observable verbs: “Describe…”, “Compare…”, “Apply…”. Continuing education credit often depends on them.
4. Case presentation: history, examination, investigations and course, told in time order and cut to the details that matter to the question.
5. The clinical question: one slide that states what the case raises.
6. Background: the condition, its epidemiology and the mechanism, pitched at the audience's level.
7. Literature review and evidence: the key studies, with design, population and results.
8. Guidelines: what current society guidelines recommend and how strong the recommendations are.
9. Discussion and controversy: where the evidence is thin or conflicting, and how reasonable clinicians differ.
10. Back to the case: how the evidence applies to this patient and what happened.
11. Take-home points: three to five, matched to the objectives.
12. References: full citations, usually on one or two closing slides.

Timing and slide count

A grand rounds slot is often about 45 to 60 minutes including questions, but check with the organizer: some departments run 30-minute sessions. For a 60-minute slot, a typical allocation is:

• 5 minutes: introduction, disclosures and objectives.
• 10 minutes: the case.
• 20 to 25 minutes: background, evidence and guidelines.
• 5 to 10 minutes: discussion, controversy and the return to the case.
• 10 to 15 minutes: questions.

Plan for no more than one slide a minute of speaking time, fewer if your slides carry study tables or figures. Thirty to forty-five main slides is usually plenty for a full hour, with backup slides held after the references.

Pitch it to the room

Grand rounds audiences are mixed: attending physicians, residents, students, nurses, pharmacists and sometimes allied health professionals or visitors from other departments. A talk pitched only at subspecialists loses most of the room.

• Define every abbreviation the first time it appears, and avoid ones local to your service.
• Give one slide of mechanism or anatomy before the evidence, so non-specialists can follow.
• State why the question matters to people outside your specialty: who else sees these patients, and when.
• Keep the advanced detail in the backup slides, ready for the subspecialists' questions.

Patient privacy and de-identification

Case-based talks put real patients on a screen in front of colleagues, and sometimes on a recording that circulates beyond the room. Treat de-identification as a requirement, not a courtesy.

• Remove direct identifiers: names, initials, medical record numbers, dates of birth, exact admission and procedure dates, addresses and the names of referring clinicians. In the US, the HIPAA Privacy Rule's Safe Harbor method lists 18 identifiers to remove; your institution's policy may go further.
• Use relative time (“day 3 of admission”, “six months later”) instead of calendar dates.
• Check images. Radiology and pathology images can carry names, numbers and dates burned into the pixels or stored in file metadata. Crop or mask them, and export clean copies.
• Avoid photographs that show a face or distinctive features unless the patient has given written consent for that use under your institution's process.
• Consider whether the case is identifiable by its rarity or by local news coverage, even without identifiers. If it is, seek consent or change non-essential details and say so.
• Follow your institution's rules on recording and sharing slides after the session.

If a colleague in the room could recognize the patient, the case is not yet de-identified.

Designing slides for evidence

Evidence slides are where grand rounds talks most often become unreadable. A few habits help.

• One study per slide, with a title that states its finding in a sentence rather than the trial's name alone.
• A consistent study table: design, population and sample size, intervention and comparator, primary outcome, effect size with confidence interval, and follow-up. Use the same columns for every study so the audience learns to read them once.
• Forest plots, simplified. Each study is a square sized by its weight, with a horizontal line for its confidence interval. A vertical line marks no effect, and a diamond at the bottom shows the pooled estimate. Label which side favors which arm, enlarge the text and highlight the row you are discussing.
• Show absolute as well as relative effects where the paper reports them, so the audience can judge the size of a benefit.
• Cite each study on its slide (first author, journal, year) and give the full reference at the end.
• When summarizing a guideline, show the recommendation's strength and the quality of the evidence as the society grades them, with the society's name and the year.

Handling questions

• Repeat or paraphrase each question so the whole room, and any recording, hears it.
• Answer briefly, then point to the evidence or the backup slide that supports the answer.
• Distinguish what the evidence shows from your own practice or opinion, and say which is which.
• If you do not know, say so, and offer to follow up. Senior colleagues respect that more than a guess.
• When a question turns into a debate between attendings, let it run briefly; that discussion is part of what grand rounds is for.

Common mistakes

• A case presentation that runs for twenty minutes and includes every laboratory value.
• No clear clinical question, so the evidence section has nothing to answer.
• Screenshots of journal tables too small to read from the back of the room.
• Identifiers left on images, in file names or in slide notes.
• Outdated guidelines, or a single study presented as settled practice.
• Overrunning and leaving no time for discussion.
• Take-home points that do not match the learning objectives.

Grand rounds checklist

• Is there one clinical question, stated on its own slide?
• Are disclosures and learning objectives included in the form your department requires?
• Is the case free of identifiers, including on images and in metadata?
• Does every evidence slide cite its study and show the effect size with its confidence interval?
• Are the guidelines current, with their grading shown?
• Does the talk return to the case before the take-home points?
• Do the take-home points match the objectives?
• Does the timing leave 10 to 15 minutes for questions?