AI-enabled clinical information tools have moved into everyday practice for hundreds of thousands of U.S. clinicians, but physicians are not adopting them evenly. A new analysis of one of the most widely used platforms shows medicine-based specialties turning to the technology far more than procedural and imaging-focused fields, relative to their share of the workforce.
The cross-sectional study, published Oct. 15 in JAMA, examined about 271 million queries that 952,008 clinicians with verified National Provider Identifiers — including 556,200 physicians — submitted to OpenEvidence between January 2024 and June 2026.
To compare specialties, researchers calculated a “use ratio”: the percentage of all physician queries submitted by a group divided by that group’s percentage of the U.S. physician workforce. A ratio of 1 means a specialty’s query volume is proportionate to its size.
Here are the specialties with the highest use ratios:
- Oncology and hematology: 2.00
- Internal medicine: 1.88
- Family medicine: 1.80
Here are the specialties with the lowest use ratios:
- Ophthalmology: 0.23
- Radiology: 0.24
- Orthopedics: 0.24
- Anesthesiology: 0.37
The gap held when researchers looked at how heavily active users relied on the tool. Family physicians submitted 51.7 queries per active user-month and oncologists and hematologists 49.7, compared with 19.9 for ophthalmologists and 18.5 for orthopedic surgeons. Across all physicians, the average was 37.1.
The same specialties sat at the bottom of the topic rankings. Anesthesiology accounted for 1.1% of physician queries, ophthalmology 1.2% and radiology 1.3%. Cardiology was the most common topic at 8.6%, followed by infectious disease at 7.5% and oncology and hematology at 7.4%.
Many physicians used the platform to answer questions outside their own field. Emergency physicians directed 88% of their queries to topics beyond emergency medicine, while oncologists and hematologists did so 35% of the time. Surgeons directed 70% of their queries to nonsurgical topics.
Career stage was an even sharper divide than specialty. Physicians within a decade of graduation had a use ratio of 2.54, and residents and fellows 1.84. Physicians 20 or more years past graduation had a ratio of 0.31.
Across specialties, physicians mostly used the platform for treatment questions rather than diagnosis. Clinical decision-making made up 51.9% of physician queries, with treatment and management alone at 33.1%, medication information at 14.1% and diagnosis at 13.9%. Administrative tasks accounted for 3%. The authors said those patterns differ from the diagnostic tasks emphasized in current AI evaluations.
The findings reflect a single platform and likely undercount total clinician AI use, the authors noted. Three of the study’s authors disclosed roles at OpenEvidence, which provided the data. The company, which raised $250 million at a $15 billion valuation in September, has said more than half of U.S. physicians use the platform regularly.
At the Becker’s 32nd Annual Meeting: The Business and Operations of ASCs, taking place October 29-31 in Chicago, ASC leaders, surgeons and healthcare executives will explore strategies to drive growth, enhance operational performance, navigate reimbursement challenges and prepare for the future of ambulatory surgery. Apply for complimentary registration now.
