A study links AI scribes to less burnout. The health system behind it says they caused it.

From the archive. First reported by JAMA Network Open, August 2025. Where it stands now: Still running, as of March 2026, Chief Healthcare Executive, Mar. 11, 2026. Why I saved this: Burnout among surveyed clinicians fell from 52.6% to 30.7%, a striking number, but only 22% of MGB's pilot users answered and MGB has an institutional investment in Abridge. It's a reminder to read the study, not just the press release.

Previously

In February and March 2025, Cleveland Clinic picked Ambience after testing five scribes, and UW Health set out to quadruple its Abridge users to 400. Both systems' results were their own pilot data (Part 1).

August 2025 — A peer-reviewed study of ambient scribes and burnout at two health systems has landed, alongside new rollout numbers. Here's what changed since Part 1 (March 2025), what it cost, and what went wrong.

Cost and outcome (as of August 2025)

Burnout at Mass General Brigham (MGB): 52.6% to 30.7% after 84 days, among 192 respondents. (JAMA Network Open)
Conflict of interest: MGB has an institutional investment in Abridge, per the study's disclosures. The study doesn't name the two vendors MGB tested. (JAMA Network Open)
Survey response: 22.0% of MGB's 873 pilot users at 84 days; 11.1% of Emory's 557. (JAMA Network Open)
Cleveland Clinic rollout: more than 4,000 of 6,000 eligible providers within 15 weeks; 1 million encounters. (Cleveland Clinic)
UW Health: 400 providers across 22 specialties. (Cap Times)
Cost per clinician: not disclosed by any of the systems.

The study

Researchers surveyed 1,430 clinicians in ambient documentation pilots at MGB and Atlanta's Emory Healthcare. At MGB, the share meeting the burnout threshold fell from 134 (50.6%) to 78 (29.4%) at 42 days, and from 101 (52.6%) to 59 (30.7%) at 84 days. Changes in intent to leave weren't statistically significant. At Emory, the share saying documentation had a positive impact on well-being rose from 1 (1.6%) to 20 (32.3%) after 60 days.

The authors concluded that use "was associated with reductions in burnout and increases in perceived well-being." They called it a "noncontrolled study of 2 academic medical center systems with a limited number of vendors," and said that, given the response rates, the findings "may represent the experience of more enthusiastic users and a best-case scenario." Fewer than half of MGB respondents used the tool for at least half their visits.

The disclosures note that "Mass General Brigham has an institutional investment in Abridge." The study doesn't name the two vendors MGB tested, so it doesn't say whether Abridge was one of them.

How MGB described it

MGB's release said the scribes "led to significant reductions in physician burnout." Dr. Rebecca Mishuris, its chief medical information officer and a co-senior author, said: "There is literally no other intervention in our field that impacts burnout to this extent." The release also noted the findings likely reflect more enthusiastic users. MGB said its program began in July 2023 with 18 physicians, tested two technologies, and had more than 3,000 providers routinely using the tools as of April 2025.

Cleveland Clinic and UW Health

On Aug. 14, Cleveland Clinic said active users rely on Ambience for 76% of scheduled office visits, and that it cut time writing and reviewing notes by two minutes per appointment and 14 minutes per day. "People are getting their documentation done faster and are spending less time after hours," said Dr. Eric Boose, who led the pilot. It said the pilot involved about 250 physicians across more than 80 specialties. Quantifying return on investment "will require more time and deeper analysis," the clinic said, and the tool "is not for everyone."

The Cap Times reported UW Health reached 400 providers through the first half of the year. Gordon said about 99.5% of patients agree to the technology, and that UW Health plans to expand it to inpatient units in the coming months.

What's changed

In March, the evidence was each system's own pilot data. Now there is a peer-reviewed study, but it's uncontrolled, has low response rates, and was led by a system with an investment in Abridge. The study doesn't name the vendors it tested.

What went wrong

The release outran the study. "Led to" in the release; "associated with" in the paper.
Few answered. At 84 days, 22% of MGB users responded.
Pilot counts don't match. Cleveland now says about 250 pilot physicians; Fierce reported more than 300 clinicians in February.
A conflict sits in the disclosures. MGB has an institutional investment in Abridge. The study doesn't name the vendors it tested.
What we don't know yet: whether the gains last, what the tools cost, and how non-respondents fared.

Our follow-up question for Mass General Brigham

Will you run a controlled study, and will it include clinicians who stopped using the tools?

Sources

You JG, et al., “Ambient Documentation Technology in Clinician Experience of Documentation Burden and Burnout,” JAMA Network Open, Aug. 21, 2025. doi:10.1001/jamanetworkopen.2025.28056. pmc.ncbi.nlm.nih.gov
Mass General Brigham, “Ambient Documentation Technologies Reduce Physician Burnout and Restore 'Joy' in Medicine,” Aug. 21, 2025. news.massgeneralbrigham.org
Cleveland Clinic ConsultQD, “Less Typing, More Talking: How Ambient AI Is Reshaping Clinical Workflow at Cleveland Clinic,” Aug. 14, 2025. consultqd.clevelandclinic.org
Cat Carroll, Special to the Cap Times, “UW Health grows use of AI notetaking to hundreds of providers,” Jul. 18, 2025. captimes.com
Ron Southwick, Chief Healthcare Executive, “Mass General Brigham's strategy is 'enabled by AI' | HIMSS 2026,” Mar. 11, 2026. chiefhealthcareexecutive.com
Earlier in this series: Part 1. /case/ambient-ai-scribes-pilots

Disclosure: [not in copy] Editor placeholder