At Health Evolution’s annual Summit, CEOs from leading health systems, health plans, and life sciences companies, as well as select policymakers, investors, retailers, and technology leaders, convene to examine the industry’s most pressing issues and emerging opportunities. This unmatched gathering of prominent leaders allows executives to have candid and cross-industry dialogue that sets the industry’s agenda for the year ahead.
Our 2024 Summit featured in-depth assessments of how unprecedented advancement in artificial intelligence is driving transformation in health care—and how leaders and their organizations must evolve in response. In a Big Discussion entitled “AI at the Forefront: Confronting the New Leadership Imperatives for Care Delivery Innovation” and a Capstone Conversation on AI in health care, Molly Coye, Executive in Residence, AVIA; David Entwistle, President & CEO, Stanford Health Care; Gianrico Farrugia, MD, President & CEO, Mayo Clinic; Vinod Khosla, Founder, Khosla Ventures; and Suchi Saria, PhD, Founder & CEO, Bayesian Health, and Director of the Machine Learning, AI and Healthcare Lab, Johns Hopkins University, shared how CEOs of the nation’s leading health care institutions are guiding their organizations to leverage AI-powered tools, what’s needed to ensure the industry can maximize benefits from AI, and what an AI-powered health care system could achieve in the future.
Using AI to support—not distract from—strategic priorities
With many of the world’s governments, companies, and innovators giving much attention to AI’s potential benefits and drawbacks, some health care leaders may be unsure whether AI tools are overhyped or a worthwhile investment. But executives at the nation’s leading health care institutions recognize that embracing AI now is imperative to achieving their organizations’—and the industry’s—goals.
At Stanford Health Care, the organization’s strategic priorities focus on what’s best for Stanford’s patients and how the organization can provide “the best care delivery models possible right now,” Entwistle said. Stanford’s digital health program is integral to advancing those priorities, and AI has great potential to be a powerful addition, he noted.
Entwistle said the integrated strategic plan aligning the School of Medicine and adult and pediatric health care delivery systems includes a guiding standard to be “digitally driven.” That priority encourages leaders to think about leveraging AI, as well as other technologies, to help the organization and staff achieve KPIs and other goals. To support that effort, Stanford Medicine has held “a series of webinars for the entire organization devoted to AI,” focusing on how staff can “educate, bring perspectives, and share best practices, learnings—and sometimes failures” they experienced testing and refining AI tools, Entwistle said. “I think it’s that integration that allows this work to come forward. Staff must see the leadership team speaking it, talking about it, and doing it, for them to feel that this is something that’s part of our organization,” he added.
Overall, leaders at Stanford Medicine emphasize that AI can provide solutions for both the organization’s and patients’ needs. “This is not chasing technology or going after the latest, greatest shiny thing,” Entwistle said. “There are real-world applications of AI tools that benefit our patients, and that’s what gets us excited. We can really make a difference in the work we’re doing.”
Leaders at Mayo Clinic take a similar approach, with the organization already using hundreds of AI applications to support medical practice and its strategic goals. “Mayo Clinic has a strategy called Bold. Forward., which has three interconnected objectives: The first is ‘more cures.’ We firmly believe that we need to create more cures for the world, both for acute and chronic diseases. The second is ‘connect,’ bringing people and data together to create new knowledge, and then convert that knowledge into scalable end-to-end solutions. And the third is ‘transform,’ which is creating and then moving health care to a collaborative platform model,” Farrugia shared. He added, “AI plays a very significant role there. So much so that we currently have over 230 AI applications that run in our practice every day.”
Farrugia said facilitating digital transformation in health care is key to Mayo Clinic’s strategy, and that includes leveraging AI to drive change. “For us, the digital transformation of health care is not just a part of Mayo Clinic, it is Mayo Clinic. So that means every department and committee at Mayo Clinic has the authority, and the responsibility, to incorporate digital tools—which can include AI—into everything it does,” he said.
Beyond driving strategic priorities at the organizational-level, leaders noted that leveraging AI across the health care ecosystem could help the industry achieve long-held goals.
For instance, Saria said organizations can use AI in efforts to expand high-quality care in underserved communities. “With data learning and AI technology, we can bring some of the best expertise in the country, which right now is localized to some institutions, and democratize it. We can make it more cheaply accessible. We can make people in small towns have access to the best specialists, which we previously couldn’t do. So in some sense, what is truly exciting is that, for the first time, we have the opportunity to bring forward these public health aspirations we’ve had and really make them happen through AI,” she explained.
Farrugia added that AI tools already are helping organizations across the ecosystem improve clinical care. “The opportunities in the clinical space are so significant that it’s an area where you can actually quickly improve health care and do so in a way that is sustainable,” he said.
Further, organizations could use AI solutions to help address the health care industry’s growing workforce strain and shortages, leaders said. “We have to figure out a way to do and serve more with the workforce that we have, and AI tools could help us do that,” Saria said.
With all the potential uses in health care, Saria said AI advancements present an opportunity to drive notable change. “I unequivocally believe this is a truly transformative opportunity, because there are win-wins here. There can be better clinical outcomes, better provider experience, and financial improvements that can come from being more effective and efficient” with the help of AI, Saria asserted.
Laying the groundwork to maximize benefits from AI
But realizing benefits from AI advancements is not a guarantee for the industry, leaders said. “There’s a difference between capability to do something and doing something,” Khosla noted.
So how do industry leaders ensure that AI tools “will actually generate results we can count on and measure?” Saria asked.
“Part” of the solution “is having leadership that can really talk both talks: not just the digital talk, but the everyday challenges of running the core business combined with digital, so it doesn’t feel like there are AI people or tech visionaries out there just saying things when they don’t understand health care organizations day to day,” Saria offered. “The most important ingredient is having the right people in the room feeling like they’re ready to shape this and make it happen,” she continued.
Entwistle agreed and said leaders should start by focusing on AI tools that can help solve their day-to-day operational challenges to begin seeing benefits.
Likewise, Farrugia said organizations must “go back and look at the way they’re running their systems and what they need to change in order to get the maximum benefit from AI.” He noted, “There are tangible, real, diffusible effects from AI that are making patients’ lives better and are scalable.”
However, Farrugia asserted that health care organizations must invest in both their physical and digital foundations to fully realize AI’s promise. “Health care infrastructure is getting old. It needs to be rejuvenated and looked after to integrate digital solutions into physical care settings. If we don’t look after both the physical and the digital infrastructures, we’re going to end up with less of a good result than we could when it comes to AI,” he said.
And as organizations make those investments, leaders must encourage staff to incorporate and leverage AI tools in ways that support their work. “If we truly think AI is an opportunity to reset the bar, cure more disease, and create better patient outcomes, then we have to firmly decentralize how we deal with AI in our institutions,” Farrugia said, because “decentralizing AI is essential to getting maximum impact and trust.”
In practice, Farrugia explained that leaders at Mayo Clinic “made the decision upfront that our departments need to look at AI solutions the same way they would look at any other clinical application they need to deploy within their departments.” That means, he said, “they need to be responsible for having the expertise to create and implement an AI tool to bring maximum benefit to the patient. Then, Mayo’s role is to provide the compute power, create and work within partnerships to support the tools, and make sure that nothing is done that would hurt either a patient or another department.”
Similarly, leaders at Stanford focus on “developing a platform within the organization where AI is something that’s part of what we do,” Entwistle said. “We’ve said to our teams, ‘what are the opportunities to apply enablers and match AI to what we’re actually doing?’” he noted.
But to truly maximize AI’s potential, health care executives must go beyond strategies to reap benefits at their own organizations and ensure they’re sharing learnings and best practices across the ecosystem, leaders said.
“We need to share information as we’re all testing, trialing, and modeling these activities, because if we were to keep that internal, we’re not going to have the learnings that I think we could,” Entwistle said. For example, he said organizations should share results from evaluating whether certain AI tools would add clinical value for patients and providers. In addition, organizations should share any learnings that could inform best practices and guardrails for using AI in health care, Entwistle said.
Farrugia said health care organizations—particularly leading, innovative organizations—”have an obligation to step up and work with others to create a way forward that is scalable and diffusible.” Leading organizations also should bring their “unique perspective” and expertise to create guidelines for safe use of AI in health care. “It is up to us to use our advantage to allow others to use innovations with us and create a better system for all,” Farrugia urged.
Envisioning what an AI-powered health care system could achieve
With strategic, integrated, and open approaches to fostering AI use in health care, the industry could achieve goals that now feel out of reach, leaders said.
“I truly believe we should think of AI as a tool that—together with automation, workforce education, and all the other things that we have available—can allow us to seize an opportunity to create the health care system that we expect and deserve,” Farrugia said, noting that AI offers an “opportunity to scale that is different from other opportunities.”
For example, Khosla noted the value of “a learning health care system” in which “somebody in Boston does something, and a person in Louisiana learns from it.” That hasn’t been easy to achieve thus far, because “with a million practicing physicians, you’re not going to get that aggregation of knowledge except in an AI system,” Khosla said. “We will get to a learning health system only because of AI,” and “I think that’s a big upside,” he contended.
For instance, Khosla noted that “some of the most important tools in medicine”—which include people’s genetics and other traits that can indicate treatment efficacy or predict future conditions—currently aren’t used to their fullest potential. Pairing AI tools with genetic and other screenings could make it “much easier” for health care organizations to implement predictive and genomic medicine that wouldn’t be possible without AI, because it would require “too much data for humans to use” on their own, Khosla said.
In addition, AI tools could drive significant cost savings and efficiency gains the industry has long struggled to realize, leaders said.
“The operational side is probably one of the richest opportunities and areas of cost savings and efficiency gains for AI, because some of the back-office administrative tasks, where you’re thinking about billing and callbacks and looking at records and all the things that go into that, is where these tools really enable us,” Entwistle explained.
Khosla said AI tools could eliminate staff’s need to physically input data into software systems and handle tasks such as “claims processing, preauthorizations,” and others. “I think AI will take a lot of the administrative burden” out of health care, which could substantially reduce costs, he portended.
But the most “exciting” future Khosla envisioned “is on the clinical side,” with the possibility to use AI tools to provide “free or near-free primary care,” he said. “It’s very easy to imagine” low-cost, high-quality “full primary and mental health care that is-AI driven and includes ongoing, high-touch services like nutrition care and disease management,” Khosla said. “I believe we can significantly reduce chronic care risk with more precision and daily interaction” through AI tools, “and that will become the definition of primary care: including disease management, nutrition, personal wellness advice, and full mental health services,” he said.
“None of this is affordable if a human being has to do it all,” Khosla noted. Instead, he said the industry “should think of physicians as overseeing the care that an AI tool might provide.” And in doing so, “patient panels and being in at-risk payment structures will become easier and more important,” because “AI will be much more objective in recommending the best course of action with the best budget to treat a condition,” Khosla contended.
So, Khosla raised, if leaders take holistic and strategic approaches to adopting and advancing AI across the health care ecosystem, “could we dramatically drop the cost of health care” while also increasing accessibility and improving quality of care?
“Yes, we can,” he answered, adding, “We have a historic opportunity over the next 10 years to radically change.”
