Charlotte, N.C.-based Advocate Health called over 15,000 hypertension patients with conversational AI and cleared nearly 10% as having normal blood pressure, highlighting the technology’s potential to close care gaps, a leader told Becker’s.
“We’re hoping that as we pursue this at scale, we actually can make that an antiquated term, this idea of ‘gap closure,'” said Andy McWilliams, MD, vice president of AI clinical product development and co-chair of AI governance at Advocate Health.
In the December test run, the 69-hospital system used a solution from startup Hippocratic AI to reach out to more than 15,000 hypertension patients who, over the past year, hadn’t gotten their blood pressure rechecked as recommended (data science teams had identified the patients in the EHR).
The AI prompted patients to take their blood pressure at home if they had a cuff. If their levels were elevated, or they raised other health concerns, they were connected with a human nurse.
Over 12 days, about 1,200 patients reported normal blood pressure, while another 274 had abnormal readings and were connected to care. The AI also flagged nonhypertension-related health issues in roughly 400 more individuals, creating “a net-new access point,” as Dr. McWilliams put it.
These rates were better than Advocate’s usual cold-call outreaches and will serve as a baseline on which to improve moving forward, Dr. McWilliams said. “It affirmed what our hypothesis was, which is that when you reduce the friction for patients to engage, the engagement gets much better.”
The agents introduce themselves as AI, and patients can opt out and follow up with a human — few chose this option. If patients are busy, they can tell the AI a better time and it’ll call back then. “In addition to making care gaps obsolete, we want to make wait times obsolete,” Dr. McWilliams said.
With AI, Advocate is looking for improvements in three areas: experience, outcomes and efficiency, Dr. McWilliams said. So the health system is measuring this tool’s effects on patient and staff satisfaction, hypertension rates, and new patients reached vs. the number of staffers that would’ve been required in the past.
Advocate nurses helped design and develop the application with Hippocratic AI, as well as the escalation protocol. Because untreated high blood pressure is dangerous, human nurses needed to be available on the back end if the patient’s condition warranted it.
Going forward, Advocate plans to follow up on hypertension patients with the virtual assistant within days, not weeks or months, and “plug them into more of a continuous experience, as opposed to a one-off phone call,” Dr. McWilliams said. “This allows us to think more about resource abundance and not have to think of resource scarcity.”
Advocate is building and testing that AI now, with hopes to go live this summer and scale through the end of the year.
“That’s part of why we wanted to move slowly at first, so we could understand what our staffing needs to look like on the back end, so that when you make X number of calls, you’re going to receive Y number of real-time escalations,” he said. “We think December’s pilot was probably the worst-case scenario in terms of escalation rates and the staffing that would be needed.”
In another small trial, Advocate found the AI to be even more effective when primary care providers told patients to expect its call.
Advocate also has conversational AI agents to schedule referrals, book Medicare wellness visits, and reach out to patients at risk for abdominal aortic aneurysms for screenings.
In addition, the health system is exploring the technology to resolve inbound calls without human involvement. And the hypertension use case could be applied to a variety of chronic diseases, as well as medication adherence.
“This, for us, feels like a very tangible example of separating through the hype of where AI can really change healthcare experiences for our patients and improve outcomes,” Dr. McWilliams said.
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
