65% of U.S. Doctors Quietly Using OpenEvidence AI for Clinical Decisions
Key Takeaways
- ▸OpenEvidence has achieved 65% adoption among U.S. doctors, processing 27 million clinical encounters in April alone
- ▸Primary use case (60% of queries) is clinical decision-making—physicians asking how to treat specific patient conditions and comorbidities
- ▸Adoption has been rapid and organic, driven by peer recommendation rather than institutional mandates or formal rollouts
Summary
OpenEvidence, an AI-powered chatbot designed for medical professionals, has achieved remarkable adoption across the United States, with approximately 65% of American doctors—roughly 650,000 physicians—actively using the platform. The tool has processed nearly 27 million clinical encounters in April alone, with an additional 1.2 million doctors using it internationally, according to company representatives. Doctors leverage OpenEvidence primarily for clinical decision-making (60% of searches), medical knowledge verification, discharge note writing, and exam preparation, with the platform delivering quick, tailored responses that serve as a digital peer consultation.
The adoption has been notably organic and informal, with doctors discovering and adopting the tool through peer recommendation rather than institutional mandates. Physicians across different specialties, clinic sizes, and geographic regions—from rural South Dakota to urban Boston—report using OpenEvidence to quickly verify medication side effects, confirm diagnostic procedures, and answer urgent clinical questions during patient encounters. The tool's accessibility, free availability, and mobile functionality have made adoption remarkably easy, enabling physicians to answer questions faster than traditional methods.
However, the rapid and largely unmonitored adoption has raised concerns among healthcare experts and researchers. Critics point to the lack of rigorous scientific studies on OpenEvidence's impact on patient outcomes, potential AI hallucinations and incomplete answers, and the risk that excessive reliance could erode physicians' critical thinking and evaluation skills. Despite these concerns, many in the medical community view OpenEvidence as a valuable time-saving tool that can improve patient care when used appropriately. Harvard researchers are currently examining 90 million queries submitted since 2024 to assess the tool's real-world clinical impact.
- Concerns persist about AI accuracy, lack of rigorous patient outcome studies, and potential erosion of physician critical thinking skills
Editorial Opinion
OpenEvidence's explosive adoption represents a fundamental shift in medical practice that regulatory bodies and medical institutions appear unprepared to address. While the tool's efficiency gains are undeniable, the lack of rigorous oversight and validation of patient safety outcomes is troubling—healthcare is essentially running a large-scale, uncontrolled experiment on clinical practice. Medical establishments must establish evidence-based guidelines for AI tool use in clinical settings before efficiency gains compromise physician competency and long-term patient safety.
