ViVE 2024: Health systems are making big bets on AI. Here's how

LOS ANGELES — AI took center stage at the ViVE 2024 conference this week and health systems across the country are deploying the tech to tackle issues from easing the drudgery of medical documentation to making it easier for patients to get information about their medical care.

"The buzz is all about, 'What is everybody doing with gen AI?' I think everybody is placing bets from a technology standpoint and the question becomes, how do you focus it?," said Rebecca Kaul, Ph.D., senior vice president, chief of innovation and transformation at Northwell Health, while on stage at ViVE.

New York-based Northwell is making sizable bets on AI. The health system launched an "AI catalyzer" that aims to drive innovation and improvement in quality, safety and efficiency through the use of artificial intelligence. 

"It's about making a lot of industry language models accessible for innovators across the system to solve their own problems, and then identifying core problems across the system. We've identified our top five priorities to just go after," Kaul said.

"The question becomes, how do you start to get your feet wet? So, a lot of capabilities around reducing administrative burden, which ultimately brings the human back to the clinical experience, because if you take the burden out of the system, then caretakers can focus on the patients," she noted. "That's a good place to start with some of these emerging technologies that are low risk and you can create a lot of value and you're able to show some early wins."

Kaul joined executives from Providence, Sutter Health and Houston Methodist during on-stage discussion to share how health systems are investing in innovation.

As clinician burnout levels continue to rise, health systems are eager to try out tech tools like voice-enabled ambient listening to help draft doctors' notes during patient exams.

"As we start to think about this world of increased automation, it's about bringing people back to people, not replacing people with technology," Kaul said.

Northwell also formed a joint venture with startup studio Aegis Ventures, armed with $100 million, to create novel AI tools for patient care. 

In Texas, Houston Methodist is using technology and AI to redesign the care experience for patients and boost efficiency in areas like supply chain management. As one example, the health system is developing what it calls "smart hospital rooms" with tech like AI cameras and smart sensors for safety monitoring and bio-buttons for remote monitoring. The rooms will have voice-controlled room controls for temperature, lighting and window shades and will enable virtual medicine by connecting clinicians with their patients remotely.

"It's about how are you utilizing tech to create the experiences that you want for your patients who are getting services from you?," said Michelle Stansbury, vice president of innovation and IT applications at Houston Methodist, while speaking on the same panel Tuesday afternoon.

"As health systems, we're all experiencing financial pressures, so how are you utilizing tech to create those operational efficiencies that are helping with those financial pressures?" Stansbury said.

Historically, hospital IT departments spent years testing out new technologies before scaling them across the organization.

In the age of AI, things are moving fast. Houston Methodist aims to be agile by piloting new tech with select service lines to prove the impact, which enables the health system to "succeed fast, fail fast” with innovation projects. 

"We're trying to make sure that what we are focused on, we can scale quickly, and get it done in a very agile and fast manner within our organization. I come from a traditional IT background. It was like, "Let's try this. It's not working? Let's keep trying. Two years later, we're still trying to make it work.' Not anymore. You do it, if it doesn't work, then you fail it, move on and learn from it. If it does work, then scale it across as quickly as you can," Stansbury noted.

Out in Washington, Providence is tapping into AI and generative AI to revamp its patient navigation capabilities and chatbots to help patients communicate with the health system and quickly get information.

"We re-platformed everything when it comes to our navigation and chat capabilities," Sara Vaezy, chief strategy and digital officer at Providence said during the Tuesday afternoon session.

"We can now understand when someone comes to us with a text-based sort of chatbot request to complete some sort of simple task. We used to not be able to parse out the intent if it was complex, and now we can. That is a very real difference. It's material," she said.

She added, "Generative AI does have a lot of hype associated with it, but I think it's worth the hype."

"The other area within generative AI that we're making big bets on is actually on tooling and infrastructure," Vaezy said. "Classic AI is more deterministic and generative AI more probabilistic. You have to have transparency around the inputs, you've got to be able to do testing and validation, you've got to be able to do ongoing monitoring and telemetry and you need to do it at every phase of the journey. That tooling and infrastructure, how you do ML (machine learning) ops and LLM (large language model) ops within that context is super, super critical so that you can start to prepare for what are you going to do to scale it out beyond one use case," she said.

The health system also is doubling down on using AI to develop what Vaezy called "identity-driven personalization."

"Personalized content, next-best actions, task partners that can create a personalized experience, that was not possible before. It was customized, but it wasn't personalized. We are making big bets in how we're pulling all of the different pieces of data and data stores together to drive something like that, and do it across, not just our own assets, but across an ecosystem," she said.

She added, "We think that there's also a lot of potential in clinical decisioning and a clinical decisioning engine that can actually be provider-facing and bring context from the clinical record as well as best practices and things from the clinical domain combined with a semantic search to drive clinical decisions."

In Northern California, Sutter Health is tapping into tech and AI to make healthcare simpler and to engage patients outside of the traditional healthcare ecosystem, said Chris Waugh, vice president and chief design and innovation officer at Sutter Health on Tuesday afternoon.

"One huge bet for us at Sutter and our innovation team is we're going to all the places where care doesn't currently exist, because those decisions every day, wherever the human body is, through a bunch of micro adjustments throughout their everyday life, and those decisions are not made in an eight-minute appointment with your primary care doc every six months," Waugh said. "We're going to the 99.9% of the time when some type of health delivery ecosystem doesn't exist, where the actual health is happening."

Sutter Health is focused on using apps, remote monitoring, data analytics and AI for digital navigation and support.

The health system launched Scout, a non-clinical digital program that supports teens and adolescents to help them build more resiliency to better manage their everyday mental health. 

"We launched that in all 50 states. Within eight weeks, we had adoption all over the place. So, we're breaking rules that the traditional system might think, 'We only think about our geography or we only think about existing care delivery and making it better.' We think we can reinvent care models like primary care and come up with some pretty disruptive solutions that the current system might not be thinking about," Waugh said.

Like many other health systems, Sutter Health also is focused on saving clinicians' time by reducing administrative and paperwork burdens.

"How are we using AI to save time? It's simple things like in-basket charting. How do we save a ton of hours in the organization to create efficiency? And how do we present ourselves much more often in the everyday lives of people where health is happening?" he noted.

Back on the East Coast, Boston Children's Hospital has been at the forefront of integrating AI into healthcare. The hospital has been working with OpenAI on a HIPAA-compliant ChatGPT program that aims to develop prompts for AI and support clinical decision-making. The ChatGPT program also can guide medical residents through clinical cases.

John Brownstein, Ph.D., SVP and chief innovation officer at Boston Children's Hospital, said the organization's approach to AI is both "top-down" and "bottom-up."

"We're thinking about the lower-risk use cases, places where we have vast amounts of data that we want to put into a chat interface. At the same time, we've actually tried to democratize access to these use cases within the institution," he told the ViVE audience during another panel discussion focused on generative AI.

"It's nice to see the use cases that people are building. In fact, it's not one particular domain. We're seeing an equal split, a third in clinical decision-making, a third in administration and a third in research," he noted.

In fact, last year, Boston Children's was one of the first healthcare providers to hire an AI prompt engineer.