Back to the stories

Large-scale JAMA study characterizes how clinicians use an AI clinical information platform

Score 9.7,

Healthcare AI

A JAMA paper now gives us the first large-scale, peer-reviewed look at how clinicians actually use an AI clinical information platform.

Harvard Medical School and Brigham and Women's Hospital led the study, which examined how roughly 950,000 physicians, physician assistants and nurse practitioners used OpenEvidence over the platform's early rollout. Monthly activity climbed from only a few hundred users to more than 700,000.

Why this matters: peer-reviewed data about real clinicians using an AI tool at scale is rare. Most prior evidence has been small pilots or vendor reports. This paper finally shows patterns of everyday use rather than lab results.

How clinicians used it is telling. The most common reason was to support clinical decision-making, especially treatment and management questions. Clinicians also asked about topics outside their own specialties, treating the tool like a quick consultant they can check during care.

What changes now is practical. Health systems and regulators gain credible usage data to shape pilot programs and policies. Vendors can no longer rely only on demos; they will be judged against this real-world footprint. At the same time, usage does not prove improved patient outcomes or safe deployment. Hospitals still need governance, validation, and integration work before these tools change care quality.

What's next: watch for broader rollouts, independent benchmarks and any regulatory reactions. The central question remains open: will heavy real-world use translate into safer, better care?