This blog post covers a big JAMA analysis of AI “scribe” technology. These AI scribes listen in on doctor–patient conversations and draft visit summaries.
The study looks at how AI scribes impact clinicians’ documentation time in several major health systems. It also examines who actually adopts the tool and what the time savings could mean for patient care and clinician well-being.
Study scope and participants
Researchers from Mass General Brigham (MGB) and UCSF analyzed 8,581 clinicians across five health systems: MGB, Yale New Haven, Emory, UC Davis Health, and UCSF. Out of these, 1,809 used AI scribes, while the other 6,772 did not.
- Documentation time dropped by an average of 16 minutes per eight-hour workday when clinicians used AI scribes.
- Primary care clinicians saw the biggest benefit, saving about 27 minutes a day.
- Even after these reductions, many clinicians still spent more than 2.5 hours on documentation daily.
- How often clinicians used scribes mattered: those using them for most of their visits saw the biggest drops in documentation time, but only about a third used them that frequently.
The Mass General Brigham program started piloting AI scribes in July 2023. A systemwide rollout began last April.
Now, about 4,500 of MGB’s roughly 7,780 clinicians use the technology. That’s a big jump, but not everyone’s on board yet.
Time savings translate into more patient contact
The average time saved per clinician was modest. Still, the study points out that this freed-up time often turned into practical gains—like seeing about one extra patient every two weeks.
Some clinicians used the extra time to handle patient portal messages. So, the workflow shifted, rather than just shrinking the workload.
Realizing these gains depends on how much clinicians actually use the scribe and how carefully they review and approve notes. Scribes need oversight, which can eat into the total time saved.
Burnout and workload relief: what the data say
Earlier research, including a JAMA Network Open study, saw a 21 percent drop in self-reported physician burnout with AI scribes. The current analysis, though, says the modest documentation reductions probably don’t fully explain the improvement.
Burnout is complicated. AI scribes help with documentation, but they’re not a cure-all.
Rollout status at Mass General Brigham and takeaways
The study makes it clear that benefits vary by specialty and how much clinicians use scribes. Clinicians still need to review and approve scribed notes, which takes time.
MGB’s fast rollout across a big system, but with uneven adoption, shows that AI scribes can be a helpful support tool—especially for primary care. Still, they work best when paired with clear workflows and ongoing training.
Practical implications for health systems
- Targeted implementation—focus on specialties and settings where the documentation burden feels heaviest, like primary care or complex visits.
- Workflow integration—set up processes that let people review notes quickly and accurately, without forcing them through pointless extra steps.
- Monitoring and governance—keep an eye on how folks use the system, how much time it saves, and whether it actually helps with burnout.
- Quality and safety—make sure someone reviews and fixes scribed notes right away, so documentation stays accurate and patients stay safe.
Here is the source article for this story: AI scribes: Study involving Mass General Brigham showed modest benefit for doctors