AI technology takes notes for doctors, freeing up time for patients

Stanford Health Care clinicians will soon have access to an AI-powered app that can securely listen to patient interactions and automatically draft clinical notes. The app, which was recently tested in a Stanford Health Care pilot program, uses ambient speech recognition technology to create summaries that outline key clinical details.

It’s standard practice for doctors and other medical professionals to take notes during consultations, and they often summarize patients’ medical histories and record details of treatment plans afterward. DAX Copilot, an ambient listening technology developed by Nuance Communications, a Microsoft unit, promises to help ease much of that workload.

“This can be a meaningful way to enable clinicians to spend more time with patients and reduce the burden of administrative and non-clinical tasks that are a major cause of burnout,” said Niraj Sehgal, MD, chief medical officer at Stanford Health Care.

Through this new effort and others like its RAISE Health initiative, Stanford Medicine aims to proactively integrate new AI-powered technologies effectively and responsibly to foster better provider-patient connections. “This is about empowering clinicians to do more of what they love, which is interacting with patients and focusing on that engagement and therapeutic relationship,” said Michael Pfeffer, MD, chief information officer at Stanford Health Care.

listen

To use the app, doctors first get consent from their patients, then securely record the conversation through an app on the patient’s smartphone before continuing with the consultation. Once the recording stops, an algorithm processes the data and creates a draft clinical note after a few seconds.

The technology can distinguish between friendly small talk and discussions of relevant health information, essentially becoming an invisible assistant that selectively prioritises the relevant parts of a patient’s medical history and consultation details.

Clinicians can then review their notes, add or edit them, and approve the final version of the patient’s electronic health record. Throughout the process, all conversations and data are secure and HIPAA compliant, says Associate Chief Medical Informatics Officer Patricia Garcia, MD, who helped run the pilot program with Clinical Informatics Fellow Steven Marr, MD. Clinicians can add notes to the patient’s chart using voice prompts during the encounter. These notes are then incorporated into the documentation.

Chief Medical Information Officer Christopher Sharp, MD, who helped lead and participated in the pilot program, is excited about how the automated generative AI app can help strengthen doctor-patient relationships: “One of the most striking things for me was being able to take my eyes off the keyboard and face the patient and really listen. I didn’t have to expend a lot of mental or physical energy documenting the moment because I knew everything shared in the conversation was being recorded.”

Reduce stress

Starting in fall 2023, 48 physicians from a range of specialties, including primary care, cardiology, orthopedics, rheumatology and neurology, tested the technology.

In a pilot study, approximately 96% of physicians said the technology was easy to use, 78% said it sped up clinical documentation, and about two-thirds said it saved them time.

“That’s a win here,” says Gary Fritz, director of applications at Stanford Health Care. “Saving, say, an hour will help rebalance workforce overload and cognitive load for healthcare workers.”

Stanford Medicine clinical and technology leaders plan to roll out the app to all Stanford Health Care care providers, including physicians, nurses, physician assistants, residents and medical students.

Technology advancements are on the way too, including the ability to customize note styles, suggested orders, and even edit drafts using natural language.

“This has the potential to be transformative in how we deliver clinical care. AI tools will never replace clinicians, but they can replace parts of the workflow,” Segal said. “As people become more comfortable using AI-powered technology, it creates fertile ground for the continued adoption of other workplace tools that support healthcare workers, allowing them to focus on providing better care to patients.”