STAT AI: An Emergency Medicine Podcast · Ethan Abbott and DJ Apakama

EP12 Dong Yao, MD and Shivam Vedak, MD, MBA Stanford (Part 1)

·39 min·2 clips

After AI triage, sensitivity for critical care rose from 79% to 83% across 174,000 ED visits.

DJ Abhakama and Ethan Abbott start from a New England Journal of Medicine AI piece on triage. It is the kind of topic that becomes concrete fast once it reaches the emergency department. Their focus is on how an AI tool changes the initial assessment in a busy emergency department. The question is direct: what shifts when a machine is part of the first look? The hosts keep returning to transparency. If AI is shaping triage, the patient should know it, and that is treated as a basic expectation rather than an optional extra. They also press on patient experience, even though the study itself does not really explore it. That gap matters because a faster workflow still has to make sense to the person waiting in the room. The conversation turns to race after the study reports a shift in triage patterns tied to AI implementation. That opens a broader question about whether the tool might reduce implicit bias or quietly introduce new bias. The hosts do not pretend that answer is settled. They want more evidence, especially on fairness and on how these systems behave outside one health system. The limits are stated plainly. It is an observational study. It looks at short-term emergency department outcomes, not the longer arc of care after the visit. Even so, the reported effect is practical enough to matter on shift. The tool appears to help nurses triage more accurately and efficiently, especially for the sickest patients and the least sick ones. That can improve flow even when hospital volumes are rising. The episode lands on a clear distinction: augmented intelligence can be a strong partner, but it is not a replacement for the triage nurse. The conversation works best when it stays there, between promise and implementation, where busy clinicians actually have to make these tools work.

As heard by us

A triage AI discussion that stays with workflow, bias, and the people who use it.

DJ Abhakama and Ethan Abbott present AI triage as more than a software story: it is about workflow, trust, and fairness at the bedside. They stay close to a New England Journal of Medicine piece on triage, then keep asking what changes when an AI tool shapes the first look at a…

Read the full review in PlayNext →

Why you'd press play

You get a sober look at AI triage without the usual robot confetti.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from STAT AI: An Emergency Medicine Podcast