Innovations with ambient artificial intelligence, to date, have focused on aiding clinicians and doctors in hospitals and exam rooms. But care transitions and coordination could be the next frontier for clinical AI.
Startup Suvi Health developed an ambient AI platform for patients to close gaps at critical transition points and support the handoff between hospital care teams and patients and their families.
The hospital bedside is often where the information gap opens up. Patients misunderstand instructions, families miss key details and information gets lost between shift changes. Suvi Health says it closes that gap by listening to bedside conversations and turning that information into a shared plan that patients, families and care teams can actually follow, according to the company.
The platform is designed to be used inpatient as well as post discharge in the home. At home, the patients can use the recorded conversations from their hospital stay to remember complex medical instructions, share accurate details with family and caregivers, review medication schedules and prepare for follow-up appointments.
"I think we've all been in the shoes of someone who has a family member in the hospital, and sometimes you're an hour away or sometimes you're many hours away, and you're wondering what the doctor said or what the nurse said, and you're thinking that as this person's loved one and possibly even a caretaker that it would be nice to be in the room. It would be nice to know what was being said and how the patient was reacting to that, and what the steps were towards recovery," said Kelly Benning, CEO and co-founder of Suvi Health.
"The idea of Suvi Health was based on personal experiences with wanting to know and understand those conversations and really being able to put that in the hands of the patient for them to better understand their care while they're in the hospital, which I think is exceptionally powerful," she told Fierce Healthcare.
“Healthcare doesn’t need more technology competing for clinicians’ attention, it needs systems that quietly remove friction and help people stay connected during some of the most complex moments of care,” said Shawn Albert, Suvi Health chief technology officer and co-founder. “We built Suvi Health’s ambient AI platform to work quietly in the background so care teams can focus less on coordination and more on patients and families.”
Suvi Health was incubated within the 1842 Studio, a venture studio supported by the 1842 Fund, a VC fund anchored by the University of Notre Dame and managed by venture builder Alloy Partners. In addition to investing in startups emerging from Notre Dame campus, the 1842 Fund funds new startups launched from its venture studio.
“Suvi is a great example of the opportunities we seek to address through our unique ‘research co-creation’ approach,” said Mike Joslin, partner at the 1842 Fund, in a statement. “Efforts like these are built around combining academic research, clinical expertise from the industry and our venture studio playbook to drive real-world impact.”
"We worked heavily with Mayo to understand what some of their biggest challenges were, and one of those came up as length of stay," Benning said.
Heart failure patients often face complex care plans involving multiple diagnoses, medications, side effects, comorbidities, rehabilitation and follow-up appointments. By helping patients better understand and manage their care, Mayo Clinic Florida and Suvi Health aim to improve outcomes and potentially reduce hospital stays.
The work with Mayo Clinic helps to "clinically ground the app" while ensuring it is built for privacy, security and the responsible use of protected health information, she noted.
The startup partnered with researchers at the University of Notre Dame and clinicians at Mayo Clinic Florida, whose clinical expertise and foundational research helped inform its approach to addressing complex healthcare delivery and AI challenges, executives said.
"With Notre Dame, we are utilizing their expertise in how to build the application itself to drive engagement. They have behavioral experts there and app-building experts that have been assisting us with the project as well. The idea was to ensure that we were grounded in a problem that was recognized by most hospitals across the country, and then working with Mayo has led to a great level of discipline that has allowed us to make sure that we are clinically effective, especially without interrupting the clinician workflow or the EHR workflow," she noted.
Suvi Health uses AI to augment and support clinical decision-making, not replace the expertise, judgment, and accountability of healthcare professionals, company executives said.
"When healthcare teams are stretched thin and families are overwhelmed, we believe technology should step in, quietly, to listen, support and help everyone move forward together toward healing and home," Benning said.
Suvi Health's platform serves as an AI companion for patients and their families to capture guidance, decisions and encouragement without interrupting care or adding work for clinicians. Patients can use the app chatbot to ask routine questions and the AI's answers are grounded only in that specific care conversation or the Mayo Health Information Library, Benning said.
"We built a bounded knowledge engine to ensure any answers in the chatbot can only come from the conversations that they had or from the Mayo Health Information Library. If it cannot pull an answer from anywhere else; it will not hallucinate or give any wrong answers," she said.
When a patient's question needs clinical judgment, the app will loop in the care team and physician.
"We designed Suvi to be different than other ambient AI technologies in the room and we're using a device connectivity, so nobody has to press a button and nobody has to interrupt the workflow, including the clinicians, nurses and care teams," Benning said. "The devices connect to each other when that care team member enters the room and automatically starts a microphone on the patient's phone that records that conversation with that care team member, and when the care team member leaves, the conversation is turned off. What is unique about this is that in almost real-time, the patient can then log into the Suvi app and see that conversation in real time."
The AI also can provide personalized nudges tied to each patient's health and recovery goals.
"From those care conversations, we will be sending out a task list or things that need to be completed in order for that patient to be discharged. They might see things on the list like do your bed exercises or your spirometer today or your chair exercises, and this really helps the patient and the care teams and the patients' families understand what needs to be accomplished in order for them to be discharged," she said.
Hospitals already use tools for care coordinators, discharge planners and patient engagement, but Benning asserts that Suvi Health pulls together all this information with the patient and caregivers in mind.
"Right now, when patients have conversations with these different care team members, whether it be discharge, social services, a nutritionist or the doctor, those care conversations could be captured on the back end and they go into the EHR record, but the patient never has access to them. In this way, the patient can put together the full story of their continuum of care while they're in the hospital—everything from the doctor visit to the surgeon to the nurse to the discharge coordinator, and that also goes to the friends and family as well that they select to be able to view the app," Benning noted. "We're allowing that patient to build an entire care continuum conversation within the app, within the chatbot, to understand how each unique conversation plays into the next conversation so that the patient has a very clear view of what every person has shared with them during that visit."
Patients can have up to 17 or 18 conversations a day, and less than 80% of that is remembered, Benning noted. "By capturing all of these conversations together, the patient can get a very good, clear, confident understanding of their care plan and that goes for discharge tools as well. Everything in that care gap that could occur during discharge will be recorded, and they can review it at home because they will then be able to take the Suvi app home as well," she said.
For nurses and clinicians, Suvi Health acts as a beside companion that repeat care instructions to save nurses from answering repetitive questions and can help keep patients on track with tasks necessary for discharge, such as specific exercises.
"On the physician side, we're working with a cardiologist at Mayo, and he's particularly excited to see what kinds of questions are coming into the chatbot because that will give his teams an idea of where they're leaving conversations without giving the right information to the patients. They're quite excited to see how they can improve their level of communication with patients based on the data from the app and the ability for patients to share those conversations with friends and family," Benning said.
Through the Mayo Clinic Florida pilot, Suvi Health will assess the provider and patient satisfaction, the ambient AI's accuracy levels and how well it supports patient care without interfering with clinician workflows. "We'll be looking at certain endpoints to ensure that the pilot is successful," Benning said.
