Research

The evidence for AI scribes just got a major upgrade. A study published this week in the Journal of the American Medical Association — the largest of its kind — tracked ambient documentation across five U.S. hospitals over two years, comparing more than 1,800 clinicians using AI scribes against nearly 7,000 controls at the same institutions.

0.5

Extra patient visits per weekThe average productivity gain from AI scribe use, per the JAMA study — modest but measurable, and most pronounced for primary care physicians and advanced practice providers who used scribes in at least half their encounters.

The study, spearheaded by UCSF Health and Mass General Brigham, found meaningful reductions in EHR charting time among frequent scribe users. But its senior author, Dr. Rebecca Mishuris of Mass General Brigham, pointed to something less expected: "Modest reductions in documentation time are unlikely to fully account for changes in burnout," she said, suggesting the tools are altering how clinicians engage with care in ways that go beyond simple time savings. Female clinicians and primary care providers showed the strongest improvements — a notable finding for a profession where burnout falls unevenly across specialties and demographics.

Public Trust

Healthcare AI adoption is accelerating inside hospital walls. Outside them, something different is happening. A national survey of 1,007 adults commissioned by The Ohio State University Wexner Medical Center and published April 7 found that public openness to AI in healthcare has dropped significantly — from 52% in 2024 to just 42% today.

42%

of Americans are open to AI being used in their careDown from 52% in 2024, per the Ohio State Wexner Medical Center national poll — a 10-point drop as clinical AI deployment has simultaneously accelerated.

The numbers land at an awkward moment. Hospitals are rolling out AI faster than at any point in history, yet the patients those tools are meant to serve are growing more skeptical, not less. The trust gap isn't a technical problem — it's a communication and accountability one. If health systems cannot explain what AI is doing, why it is doing it, and who is responsible when it gets something wrong, public confidence will continue to erode. The survey is a clear signal that deployment speed and public trust are not moving in the same direction — and that the industry needs to treat patient education as an operational priority, not an afterthought.

Health IT

Electronic health records have a reputation problem. Clinicians describe them as outdated, clunky, and designed around billing workflows rather than patient care. Most AI improvements to EHRs so far have been bolt-ons — tools attached to legacy codebases that were never designed to accommodate them. This week, Greenway Health introduced Novare, which it calls the first EHR built from the ground up with agentic AI at its core rather than layered on top afterward.

160

Skilled nursing facilities going all-in on AI this weekCreative Solutions in Healthcare announced a company-wide rollout of ExaCare AI across all 160 of its SNF locations — now processing over 1,500 patient referrals daily, automating a process previously slowed by paperwork and calls.

Where legacy systems require clinicians to adapt to the software, Novare is designed to adapt to the clinician — anticipating documentation needs, automating revenue cycle tasks in real time, and eliminating the manual workarounds that have made EHR use synonymous with physician frustration. The launch comes in the same week that the EU committed €20 million to accelerate AI deployment in cardiovascular care, after a Joint Research Centre report confirmed that AI can already perform coronary artery calcium scoring — a leading predictor of heart attack risk — as accurately as a specialist radiologist. Taken together, the week's news draws a clear line from infrastructure investment to clinical impact: the scaffolding is going up, and the applications are arriving faster than most predicted.

Sources