AI Scribes Boost Physician RVUs but Could Also Raise Healthcare Costs

 Ambient artificial intelligence (AI) scribes may boost physician productivity and earnings, but their widespread adoption could also contribute to higher healthcare spending and other trade-offs, according to recent analyses published in JAMA journals. 

Interest in AI platforms has grown in recent years, reflecting a push to embed the technology into multiple aspects of healthcare, including clinical documentation, prior authorization, and decision-making.

One of the analyses, published on January 9 in JAMA Network Open, examined whether AI scribes are associated with changes in physician productivity.

Ambient AI scribes use audio recordings of patient-clinician conversations to draft clinical notes that providers can review and edit before signing, reducing documentation time.

Researchers from the University of California San Francisco (UCSF) Department of Medicine reviewed electronic health record data from outpatient visits completed by attending physicians at UCSF Health between January 1, 2023, and April 1, 2025, comparing physician productivity before and after the health system began offering commercial AI scribe tools. UCSF Health made the technology available to 2018 physicians in August 2023, followed by a broader rollout in March 2024.The use of the tools was optional and was not tied to any defined productivity requirements.

The final analysis included approximately 1.2 million ambulatory visits involving 1565 physicians. About 45% of those physicians adopted an AI scribe during the period, accounting for roughly 15% of patient encounters.

Among physicians who adopted an AI scribe, weekly productivity — measured by relative value units (RVUs) and encounter volume — rose after adoption. Compared with physicians who never adopted the technology, physicians using AI scribes generated an additional 1.81 weekly RVUs and completed about 0.8 more outpatient visits per week.

The authors estimated that the productivity gain per physician translated into $3044 more in annual Medicare revenue based on the 2025 physician fee schedule.

Notably, the analysis did not find evidence of a rise in billing denials following AI scribe adoption, though the authors said they could not assess whether the higher RVUs stemmed from additional clinical services or more accurate documentation and coding. Productivity gains also appeared to grow over time as physicians became more accustomed to the technology.

AI’s Return on Investment (ROI)

In an accompanying editorial published in JAMA Network Open, researchers from Stanford University School of Medicine, Stanford, California, described the per-physician revenue gains as “modest,” estimating the increases at about 5.8% more RVUs and 2.8% more patient encounters per week. When scaled across organizations, they said, the supplemental revenue could help offset monthly subscription fees, which can run several hundred dollars per clinician.

“We now have quantitative evidence that ambient AI scribes can enhance financial productivity without compromising billing integrity,” the editorial authors, Shreya Shah, MD, and Patricia Garcia, MD, told Medscape Medical News. Still, they cautioned that viewing ambient AI scribes “solely as revenue enablers” may overlook other equally important ROI, including “the indirect gains from reduced burnout and retention risk and the longer-term benefit of building an AI-ready care ecosystem.”

Some health systems are already monitoring these more comprehensive metrics.

Jennifer Goldman, DO, chief medical information officer and chief of primary care at Memorial Healthcare System, Hollywood, Florida, told Medscape Medical News that the organization implemented AI scribes about a year ago for up to 300 physicians across ambulatory, emergency, and inpatient settings.

Since then, Goldman said internal measures show that the technology has contributed to a 6% increase in appointments per day and a 17.5% reduction in documentation time per encounter, shaving roughly 5 minutes off the time physicians spend on each patient’s notes.

“The decrease in cognitive load during the visit is pretty substantial, since you can relax and focus completely on the patient,” she said. After-hours “pajama time” work, another significant source of documentation burden and burnout, has also declined by about 7%.

Goldman said AI scribes are emerging as a “new norm in medical care,” increasingly viewed as a necessity rather than a nice-to-have, similar to how dictation software gained traction as its benefits became better understood. She said the health system considers the technology worth the investment and plans to expand access this year.

“You can’t always use a straight ROI calculation,” Goldman said. “Some of the return is in physician and patient experience — things you can’t always put a price tag on.”

Angela Bohnen, MD, neurosurgeon at Neurosurgery One and chief of surgery at AdventHealth, Littleton, Colorado, described similar benefits. She told Medscape Medical News that it was common to spend up to 3 hours/d completing operative notes and patient documentation, whereas her AI scribe drafts notes in real time that she briefly reviews for accuracy.

That efficiency allows more quality time with patients and captures critical billing information for complex neurosurgical cases, she said.

“The model is essentially self-sustaining,” Bohnen said. “Higher patient volume and lower administrative costs allow more visits with no additional personnel.”

Productivity Gains Raise Spending Concerns

Despite these promising gains, other research suggests that improvements linked to ambient AI scribes could contribute to higher healthcare spending and other unintended consequences as adoption becomes more widespread.

In a recent policy analysis published in JAMA Health Forum, researchers from the Division of Health Policy and Management at the University of Minnesota School of Public Health, Minneapolis, proposed that vendors often market ambient AI scribes as a path toward “higher-intensity billing.” In that scenario, overall spending rises, and the “scribe-adopting organization stands to capture much, if not all, of the additional spending,” study authors said.

For example, an AI scribe’s ability to capture a greater number and severity of diagnoses may be particularly advantageous in value-based payment arrangements, translating into higher reimbursements for participating health organizations.

However, the researchers acknowledged that the additional spending driven by AI-generated documentation may not necessarily improve patient care, instead potentially leading to “simply justifying a more lucrative service code for an otherwise-identical office visit.”

The authors encouraged health systems and policymakers to take a longer view of ROI calculations when implementing AI scribes, emphasizing improvements in clinician turnover and well-being, rather than using efficiency gains to pivot to even higher productivity targets.

https://www.medscape.com/viewarticle/ai-scribes-boost-physician-rvus-could-also-raise-healthcare-2026a10001g4