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

EP13 Dong Yao, MD and Shivam Vedak, MD, MBA Stanford (Part 2)

·36 min·2 clips

Could race, ethnicity, sex, or primary language subtly influence who gets an ECG or troponin?

DJ Abhakama and Ethan Abbott keep part two centered on charting. They bring in Shivam Vidak and Dong Yao from Stanford University, then ask a practical question about whether AI can actually make documentation easier for physicians or whether it simply shifts the work around. The discussion stays grounded throughout. The answer, they suggest, depends on phenotype, meaning the kind of user, the way that person thinks, and the way they approach documentation, so the same tool can feel helpful to one clinician and irritating to another. Perfectionists tend to get less out of it. For people who care deeply about the shape of every note, the episode suggests AI is unlikely to feel finished, while people who see documentation as a completed output may get value sooner. Ambient dictation changes the tradeoffs. The hosts describe a move toward accepting a good-enough note if it means getting home in time to have dinner with your kids, and they frame that shift as a real change in how people approach documentation. Even so, that shift is not automatic. The conversation keeps returning to trust, because the current problem is not only whether the tools can do the job, but whether clinicians trust them enough to use them without constant second-guessing. Timing matters too. They describe the current moment as the first wave of ambient dictation and AI documentation, which leaves the jury out on day-to-day usefulness even as the long-term upside still looks significant. No one is pretending the work is done. The episode keeps its feet on the floor by weighing the promise of faster charting against the reality of review burden, imperfect output, and the extra attention a clinician still has to give the note. That gives the discussion a clinical feel. It is less interested in hype than in how the tool fits into real emergency medicine routines, from documentation habits to the pressure of leaving on time. By the end, the show makes a restrained case that AI documentation may eventually help in a meaningful way, but only if trust can keep pace with the technology.

As heard by us

AI charting looks useful here when clinicians value workable notes, trust the system, and accept that the verdict is still early.

Episode 13, part two, stays with a focused question: what happens when AI enters documentation, and how do clinicians handle notes that are good enough rather than perfect?

Read the full review in PlayNext →

Why you'd press play

If documentation is on your mind, this is a sober checkup on AI charting in emergency medicine.

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