Electronic health records have long represented both potential and pain. But with the rise of AI, their role could shift from documentation burden to clinical intelligence engine.
AI-powered capabilities — such as ambient listening tools, predictive analytics and clinical decision support — are gaining traction among leading health systems and vendors seeking to improve efficiency and reduce clinician burnout.
Somerville, Mass.-based Mass General Brigham is already scaling up its use of ambient documentation tools that capture patient-provider conversations and convert them into structured clinical notes. Rebecca Mishuris, MD, chief medical information officer, told Becker’s that early results from the pilot program point to improved provider satisfaction and workflow optimization — two key benchmarks for AI’s potential in reimagining the EHR experience.
“We don’t need to prove that it [the ambient AI tool] helps with burnout anymore — we’ve proven that,” Dr. Mishuris said.
While some systems are leaning on third-party tools, others are building from within. Stanford Health Care in Palo Alto, Calif., is piloting ChatEHR, an internally developed, AI-powered assistant embedded directly into its EHR platform. The tool allows clinicians to ask questions, generate summaries and retrieve patient data using natural language prompts — all in an effort to streamline EHR use and reduce cognitive load. The pilot is currently limited to a group of 33 physicians, nurses and physician assistants.
Major EHR vendors are also evolving. In April 2023, Epic expanded its integration with Microsoft’s Azure OpenAI Service, bringing generative AI capabilities to its platform. Health systems like UC San Diego Health and Madison, Wis.-based UW Health are testing tools that can summarize lengthy patient portal threads and draft message responses for clinicians to review — incremental features that promise real-time relief without overhauling workflows.
Additionally, Oracle is developing a new EHR system built on Oracle Cloud Infrastructure. The platform incorporates advanced AI capabilities and is designed to be intuitive, eliminating traditional menus in favor of voice commands, according to the company. Oracle also has its Clinical AI Agent, formerly known as the Clinical Digital Assistant, a voice-activated tool integrated into the Oracle Health EHR that generates clinical notes and suggests follow-up actions, aiming to reduce administrative burdens for clinicians.
Still, progress comes with constraints. One of the biggest hurdles remains interoperability. Many EHR vendors continue to operate in closed ecosystems, limiting the seamless integration of AI tools and hindering the exchange of models across institutions.
That’s why many hospital leaders believe the future of AI in EHRs will be defined by more than individual tools — it will hinge on open, intelligent platforms. These platforms must support plug-and-play AI solutions, facilitate data sharing and evolve alongside clinical needs.
