California-based Stanford Health Care is piloting an internally developed, AI-backed software designed to revolutionize clinician interaction with the EHR.
Nigam Shah, MD, PhD, chief data science officer at Stanford Health Care, is leading the development team for ChatEHR, which allows clinicians to ask questions, request summaries and pull specific information from a patient’s medical record. ChatEHR is built directly into Stanford’s EHR to maximize clinical workflow.
The pilot is available to a small cohort of 33 physicians, nurses and physician assistants. The technology is secure and designed for information gathering; not medical advice.
ChatEHR, which has been in development since 2023, facilitates a more streamlined and efficient way for clinicians to interact with patient records.
“This is a unique instance of integrating [large language model] capabilities directly into clinicians’ practice and workflow,” said Michael Pfeffer, MD, chief information and digital officer at Stanford Health Care and School of Medicine, in a news release. “We’re thrilled to bring this to the workforce at Stanford Health Care.”
Stanford is still working on automation to evaluate tasks, such as determining whether to transfer patients between hospitals or units. Dr. Shah and his team are using an open-source framework for real-world large language model evaluation, MedHELM, to evaluate ChatEHR. His goal is to scale ChatEHR to all clinicians and the team is working on more features to ensure accuracy.
“We’re rolling this out in accordance with our responsible AI guidelines, not only ensuring accuracy and performance, but making sure we have the educational resources and technical support available to make ChatEHR usable to our workforce,” said Dr. Shah in the release.
