Stanford Health Care, headquartered in Palo Alto, California, is often ahead of the curve when it comes to technology innovation. As a result, health informatics leaders at provider organizations across the country can look to Stanford for lessons.
This is the goal educational session case study health system will lead the HIMSS AI in Healthcare Forum, scheduled for September 5-6 in Boston. The session will focus on how Stanford is approaching the implementation and evaluation of generative AI applications.
Troy Foster, director of digital health at Stanford Health Care, is scheduled to speak at the session. We spoke with him to get a taste of what he’ll be discussing in Boston.
Q. What will you focus on during the HIMSS AI in Healthcare Forum and why is it essential in healthcare today?
AND. Given growing concerns about burnout and workforce constraints, many technology companies are focusing on artificial intelligence and other solutions to address these issues.
One of the more promising possibilities that is quickly gaining popularity is the area of ambient voice. The technology’s ability to offload significant amounts of clinical documentation time will allow physicians to spend more time with patients.
Q. What is an example of ambient voice and genAI in action in your organization?
AND. Stanford Health Care recently completed a pilot with some of our providers who are using an integrated ambient voice solution between Epic Rover and Microsoft DAX CoPilot, leveraging the proprietary ChatGPT engine to populate Epic astute fields with suggested summaries from the patient consultation.
Once patient consent is obtained, providers simply begin recording as the patient comes in, and within minutes of the consultation being completed, providers can review, edit as needed, and then approve the documentation in Epic. This potentially saves our practices a significant amount of time instead of having to create all of the notes from scratch.
Q. What are some of the takeaways you hope session participants can take away and apply to their service organizations?
AND. While this is potentially groundbreaking technology for practitioners, it is still very up-to-date and needs to be refined. Early feedback is that some love it and some hate it—with a wide range of opinions in between. Some providers exploit the note-writing process to review the consultation and make appropriate clinical decisions.
ChatGPT summaries are still often too verbose or imprecise and require extensive editing, which for some people negates any potential time savings. Academic medical centers with many residents, fellows, and other practitioners create documentation problems.
In miniature, this technology has huge potential, but is still in the development phase.
Healthcare facilities will need to ensure appropriate socialization and monitoring of the implementation and exploit of voice recognition technology, but it is a technology that has enormous potential to have a significant and positive impact on alleviating burnout and increasing the capacity of healthcare workers.