OpenEvidence brings ACP guidance into point-of-care decision support
OpenEvidence has reached a licensing agreement with the American College of Physicians to embed clinical guidance from the organization and published content from Annals of Internal Medicine into its AI-powered clinical decision support platform. The deal ties a large evidence base to a workflow that physicians can use during live patient encounters, when decisions about diagnosis, medication selection, and follow-up often need to be made in seconds rather than after a literature search.
The arrangement matters because it places peer-reviewed internal medicine content directly inside a tool designed for rapid clinical querying. For physicians who already rely on the platform, the practical effect is not abstract. It means guidance statements, practice points, and journal content can appear at the same moment a clinician is weighing options for a patient with multiple comorbidities, polypharmacy, or an uncertain presentation.
Internal medicine cases fit the platform’s highest-demand use case
Internal medicine often creates the kind of diagnostic and therapeutic complexity that software vendors like to describe but clinicians experience in very concrete terms: one visit can require sorting out heart failure symptoms, interpreting thyroid nodules, and checking drug interactions across a long medication list. OpenEvidence’s chief medical officer framed that reality directly, describing a clinic session that can move from one organ system to another without warning.
The new licensing agreement is built for exactly that setting. ACP’s content brings a structured evidence base into a platform that physicians can query at the bedside or between encounters, and the integration is designed to anchor answers in recommendations already familiar to internists. A citation trail is central here. In a setting where AI outputs can be fast but opaque, immediate access to the underlying source material gives clinicians a way to verify the basis for the answer before acting on it.
A single internist’s workday often hinges on whether a recommendation is current, specific, and defensible. Access to ACP guidance inside the same interface may reduce the friction between information retrieval and clinical action.
ACP’s content adds institutional weight to the platform
ACP is the largest medical specialty organization in the United States, with 163,000 members across internal medicine, subspecialties, and medical students in more than 172 countries. The college develops clinical guidelines, guidance statements, practice points, and best practice advice, and those formats matter in day-to-day care because they are meant to translate evidence into decisions physicians can actually use.
Jan K. Carney, MD, president of ACP, said real-time access to the college’s guidance and Annals of Internal Medicine content can help physicians make better-informed clinical decisions and spend more time on patient care. The logic is straightforward: if a physician can pull an authoritative recommendation without leaving the workflow, less time is spent chasing references and more time is available for patient counseling, medication reconciliation, and follow-up planning.
For ACP, the partnership extends its reach into a digital environment already used heavily by clinicians. For OpenEvidence, it adds a recognized layer of specialty authority in internal medicine, the largest specialty on the platform. The match is strategic, but it is also clinical. The best decision support systems do not just answer questions quickly; they answer them with content that reflects how internists practice medicine in real settings, where uncertainty is common and the consequences of error can be immediate.
AI decision support depends on provenance, not speed alone
The value of integrating ACP material into an AI tool is not simply that more text becomes available. Clinical reliability depends on provenance, and provenance depends on being able to trace a recommendation back to a trusted source. When a system serves up an answer about anticoagulation, thyroid evaluation, drug interactions, or chronic disease management, the citation is the safeguard that allows a physician to confirm the recommendation rather than accept it on faith.
That feature is especially important in internal medicine, where clinicians frequently manage overlapping conditions and competing risks. A therapy that is appropriate in one disease state may be hazardous in another, and a recommendation that looks simple on the surface may need to be adjusted for age, kidney function, bleeding risk, or concurrent medications. In that environment, the promise of AI is not replacement of judgment. It is faster access to the right evidence at the moment judgment is being exercised.
OpenEvidence said the majority of U.S. physicians use the platform daily, on average. If that pattern holds, the ACP integration has the potential to place specialty guidance into a very large number of routine encounters. Even modest improvements in evidence retrieval can matter when multiplied across repeated decisions about preventive care, chronic disease management, and acute symptom evaluation.
One citation at the bedside can change the tone of an entire visit.
What the partnership signals for clinicians at the point of care
This agreement reflects a broader shift in how physicians interact with medical literature. Traditional guideline documents and journal articles remain essential, but they are often accessed outside the pressure of a full clinic schedule. A decision support platform changes the timing of access, bringing recommendations into the same sequence as the clinical question itself.
The model is appealing for internists because their work spans many domains at once. A clinician may be managing heart failure in one room, reviewing a thyroid abnormality in the next, and reconciling a half-dozen medications in the third. In that context, the ability to surface ACP guidance and journal evidence immediately can tighten the loop between question, reference, and action.
Still, the core issue remains clinical trust. The utility of the partnership will depend on whether the integrated content stays current, whether the answers remain transparent, and whether physicians find the citations sufficient for real-world decision-making. Those are not cosmetic features. They determine whether AI functions as a reliable clinical assistant or just another layer of digital noise.
For now, the deal gives OpenEvidence a stronger evidence backbone and gives ACP another route to deliver its guidance at the point of care. The meaningful question is not whether clinicians need faster access to trusted evidence. They already do. The harder part is building a system that delivers that evidence quickly, clearly, and with enough source fidelity to support the next decision in a crowded clinic schedule.