Ambient AI Sets the Stage for Clinical Decision Support

Healthcare executives who took part in this week’s Winning Edge panel say ambient AI is helping clinicians reduce administrative stresses, but it will soon be used to add value to the patient encounter.

Health systems like Providence and Cedars-Sinai are embracing ambient AI to improve clinician workflows and reduce administrative tasks, but the real value will come when AI adds clinical value to the doctor-patient encounter.

That’s the long view taken by executives from the two health systems during Tuesday’s The Winning Edge panel, sponsored by Microsoft. And it points to the future of Ai in clinical care as a decision support tool.

Sara Vaezy, Chief Transformation Officer at Providence, and Yaron Elad, MD, FACC, Cedars-Sinai’s Chief Medical Information Officer, said their organizations have both gradually rolled out ambient AI tools to physicians to help them spend less time in front of a computer and more time in front of their patients. The technology is designed to accurately capture the conversation and put that information into the medical record, reducing the amount of time clinicians spend documenting the encounter.

Elad said it’s reasonable to expect that the tool reduces time spent on the computer by 10% to 20%, especially time spent after hours, even at home, tidying up the notes for the patient record. That reduces the clinician’s administrative workload and gives the doctor more time to face the patient—something that not only benefits the doctor, but makes the patient feel more valued as well.

That might not appeal to a CFO looking for financial ROI, but it does set the foundation for improved provider well-being, better patient engagement and satisfaction, and eventually improved clinical outcomes.

That said, the two executives and Jared Pelo, MD, CMIO for Microsoft’s Health and Life Sciences unit, pointed out that clinicians have to “own” the technology and have the time to get used to it on their own terms. Healthcare leaders should not tell them that AI will improve their productivity, such as giving them more time to see new patients; instead, as clinicians settle into their new workflows, they may find the time to address new productivity goals.

Vaezy said Providence is tracking several different metrics on ambient AI, including efficiency and appointment times, to get a baseline on how the tool could have an impact on productivity in the future. Just as important, they’re charting provider and patient satisfaction, and asking clinicians if they’d be disappointed if the tool were taken away, whether they’d recommend AI, whether this tool improves documentation and would this capability compel them to stay with the organization or stay in medicine.

Both Providence and Cedars-Sinai are developing AI tools in other areas as well, including in-basket messaging and chart summarization. And while Vaezy and Elad said the next big advances should come in how AI can be used to improve nurse workflows, both are particularly looking forward to how AI evolves as a clinical decision support tool.

Vaezy pointed out that AI, for the most part, is now being used to reduce complexity and remove administrative burdens, but within two to three years the technology’s value will be in adding to the provider’s toolbox and giving more value to the provider-patient encounter. That might mean coding the encounter and identifying and pushing tests and other appointments.

Elad, meanwhile, said he envisions an AI overlay that gives clinicians almost real-time clinical decision support. This would support the true definition of AI as augmented, rather than artificial, intelligence.

Stay tuned to HealthLeaders this Friday for the YouTube recording of this Winning Edge panel.