Early evidence shows AI scribes reduce burnout, but without financial improvement
Add Axios as your preferred source to
see more of our stories on Google.

Illustration: Natalie Peeples/Axios
AI-powered medical transcription tools are alleviating clinician burnout but not yet saving practices money or making care more efficient, according to a report from the Peterson Health Technology Institute.
Why it matters: AI scribes that can securely transcribe medical notes for providers have exploded in popularity, attracting hundreds of millions of dollars in investments.
State of play: There are now about 60 companies offering technology that listens to and transcribes a patient appointment using AI.
- The promise of so-called ambient scribes to alleviate clinician stress is driving health systems to pilot these tools, the Peterson institute says.
- As providers increasingly adopt AI scribes, the institute convened a task force of developers, health systems, investors and academics last year to discuss the outlook for the technology.
What they found: Mass General Brigham found a 40% reduction in reported burnout during a six-week survey pilot of an AI scribe, the institute's report says.
- MultiCare, a health system in Washington state, found a 63% reduction in clinician-reported burnout due to AI scribe use.
- Anecdotes and survey evidence from other providers also show that AI scribes improve clinician and patient experience.
Yes, but: It's too early to tell if the AI tools are making health providers more efficient. A study of 112 physicians using an ambient scribe at Atrium Health showed that the technology did not reach that goal overall.
- Task force members also haven't seen significant financial improvements from their scribe technologies.
- "As interest grows beyond a subset of clinicians and costs remain significant, health systems will increasingly need to demonstrate a clear return on investment, particularly at current prices," the report notes.
Go deeper: Health systems embrace AI scribes as business booms (🔒 Axios Pro)
