Hospitals are discarding what I call “digital gold.”
You’re reading the web version of STAT’s AI Prognosis newsletter, our subscriber‑only guide to how artificial intelligence is being used in health care and medicine. If you’re interested, you can sign up to have it delivered to your inbox every Wednesday.
Last call: If you haven’t already, please complete your 2025 health AI bingo card and/or share 2026 predictions before tonight. It’s worth your time—there are real insights and lively debate in the comments.
As a thank‑you, here’s a little surprise: a video of a sasquatch playing saxophone in the woods. Note: this isn’t a rickroll, but you’ll only learn that once you’ve filled out the bingo card. There’s an honor system involved.
Two epidemiologists chasing the trail of missing transcripts
Katherine Goodman and Daniel Morgan, both epidemiologists at the University of Maryland, are exploring a question that matters: are there moments in clinical decision‑making where what patients say or how they phrase things isn’t fully picked up by clinicians? Are there subtle differences in patient descriptions of symptoms that could help doctors detect conditions earlier and intervene sooner?
To study this, researchers need the patient’s own words. Historically, this would have been tough, but today ambient AI scribes make it routine to record conversations during patient visits, opening new avenues for understanding communication gaps in care.
This article is exclusive to STAT+ subscribers
To read the rest and access in‑depth analysis, newsletters, premium events, and alerts, please subscribe to STAT+. Sign in if you already have an account.
If you’re not yet a member, you can view all plans and options on the STAT+ page and start your trial or subscription today.
Note: If you’re sharing this content externally, please keep in mind that full access is limited to STAT+ subscribers, and some sections may require a login.
