How to Cite Sources in a Medical Presentation (Without the "(NEJM 2019)" Problem)

The single most common citation in medical slides is the phantom: "(NEJM 2019)". No first author. No volume. No DOI. A learner who wants the paper cannot get to it, and a reviewer who wants to check it assumes you could not either.

A citation that cannot be followed is decoration, not evidence. In a teaching deck it is worse than nothing, because it claims rigor it does not have.

What a Usable Slide Citation Contains

You do not need a full reference list on every slide. You need enough that the trail is followable. Minimum:

• First author + year — "Sanai 2011", not "(Journal 2011)".
• The journal, abbreviated — enough to disambiguate.
• A findable handle on the references slide — DOI or PMID, so a learner gets to the exact paper, not a search.
• The claim tied to the source — which sentence the citation actually supports, not a citation floated at the bottom of a six-bullet slide.

Vancouver Is the Default for a Reason

For medical and surgical content, Vancouver (author–number) is the expected style. It is compact on a slide, it maps cleanly to a numbered references slide, and it is what your audience reads in the literature every day.

Pick one style and hold it across the deck. A lecture that mixes Vancouver, AMA, and freehand parentheses reads as one assembled from other people's slides — because it usually is.

The Flag That Catches the Weak Ones

On a rebuilt deck, citations are surfaced at slide level. A needs-citation flag means a claim is stated without a followable source. A low-confidence flag on a referenced slide means the source as given is too thin to defend.

These flags are where your review time goes. An unflagged citation is one the rebuild preserved as-is; a flagged one is a 30-second decision: tighten it, source it, or soften the claim.

The Honest Limit

A rebuild flags a weak citation. It does not invent a strong one. If the source deck cited a trial wrong, the flag tells you the citation is unreliable — it does not silently swap in a paper it was never given.

That is deliberate. A confident wrong citation in front of faculty is a worse outcome than a flagged uncertain one. The flag is doing its job precisely when it refuses to guess.

The Point

Cite so the trail is followable: first author, year, a DOI or PMID on the references slide, one style throughout. Then let the flags catch the ones that slipped.

A learner who can find your sources trusts the rest of your deck. One who hits a phantom citation stops trusting all of it.