8 Key Insights on Advancing AI in Clinical Care

Clinical AI is top of mind for CMOs and clinical leaders. Here are eight insights they shared that you need to know.

The HealthLeaders AI in Clinical Care Mastermind program reached a milestone today, with healthcare provider executives discussing key opportunities and challenges in Atlanta.

In the months leading up to today's in-person meeting in Atlanta, executives from 10 health systems and hospitals held several virtual calls and collaborated with HealthLeaders editors on stories about how their organizations are adopting and managing AI tools. The program will conclude by the end of the year, with publication of a final report.

Determining return on investment from AI tools was a pressing issue discussed during today's event.

"There are a couple of areas where there is 100% clear ROI from AI, said James Blum, MD, chief health information officer at University of Iowa Health Care. "In coding, there is clearly an ROI. With AI-based coding, I can decrease my claims denial rate. There is also an ROI with clinical documentation improvement."

Thinking about AI as infrastructure is the right play for health systems to determine ROI, according to William Sheahan, senior vice president and chief innovation officer at MedStar Health.

"For things like coding, you can find a third party and plug their tool into your data infrastructure and generate ROI," Sheahan said.

Generating an ROI from AI tools involves preparing staff members and leaders to benefit from the technology, Sheahan explained.

"There is a whole separate body of work for us that is about people and process to train clinicians, back-office staff, administrators, and leaders on how to use AI to make themselves more efficient," Sheahan said. "Ultimately, that is what is going to deliver ROI over time. We can make our business more efficient."

The impact of AI on clinicians and the role of clinicians in hospitals and health systems was a hot topic at today's event. Participants said it is highly unlikely that AI will replace clinicians.

"You are always going to need clinicians," said Hoda Asmar, MD, MBA, executive vice president and chief clinical officer at Providence. "AI will never replace clinicians, but it will affect what clinicians do and not do. AI can provide simplification and ease the way for the people who are delivering care."

The AI in Clinical Care Mastermind program was a golden opportunity for participating health systems and hospitals, executives said.

"We talked about how fast AI is changing and there is a lot of information coming out," said Roopa Foulger, vice president of digital innovation development at OSF HealthCare, "When you are collaborating with practitioners who are dealing with the same issues, you can share ideas and avoid potential pitfalls."

In addition to the difficulty of determining the ROI of AI tools, the Mastermind program focused on several other daunting challenges, including applying AI in the right clinical areas and addressing the cost, Asmar said.

"The other part is making AI benefit the people it is supposed to benefit the most," Asmar said. "We need to understand that the clinicians and the care teams on the frontline need to see some benefit from AI. They need to feel that things are changing for the better—not like the example of the struggles with the EHR."

Insights from virtual calls and HealthLeaders stories

As part of the AI in Clinical Care Mastermind program, HealthLeaders held several virtual calls with participating executives and featured each executive in a HealthLeaders online story. The following are the primary themes and findings from these virtual calls and stories:

• ROI is elusive: Many small projects are showing early success, but that doesn’t translate into scalability or sustainability. There is a challenge in balancing financial ROI with clinical ROI, which can take longer to develop.
   
• Early wins in ambient listening: Many health systems and hospitals are launching ambient listening tools to capture the clinician-patient encounter. Often these tools also capture coding opportunities. These tools are reducing clinician stress and "pajama time" conducting documentation after work hours.
   
• Bots are becoming popular: Many health systems and hospitals are experimenting with AI agents in population health and public health programs, garnering success in engagement and increased appointments. They are exploring where else bots can be used to help doctors with clinical care.
   
• Governance is an issue: Many health systems and hospitals are handling AI governance on their own, with dedicated committees, but they worry about stunting innovation. Challenges include determining who sits on these committees and making sure all bases are covered.
   
• Outsourcing versus in-house: Many health systems and hospitals are outsourcing AI development because they do not have the knowledge or analysis capabilities to do the work on their own. Some are waiting on their EHR provider to develop tools for them, while others are looking at startups for innovative ideas. Those with in-house capabilities are doing what they can and partnering with vendors for the rest.
   
• Cost concerns: AI is expensive, and it requires a lot of data and data storage, which can also be expensive. Costs are limiting innovation and AI development, particularly for small and rural health systems and hospitals.
   
• Pressure on vendors: Many health systems and hospitals are taking extra precautions in negotiating with AI vendors. They are requiring proof of concept and ROI up front, shortening contracts to three years or less, and asking for details about how data is collected and used.
   
• Generative and predictive AI: As AI evolves, healthcare executives will ask the technology to do more, such as predicting treatment plans and clinical outcomes. A key factor will be building clinician trust in AI tools. A pressing question is how AI will be measured against a doctor's observations and experience?

The HealthLeaders Mastermind series is an exclusive series of calls and events with healthcare executives. This Clinical AI Mastermind series features ideas, solutions, and insights on advancing clinical technology. This program is sponsored by Microsoft-Nuance, Rapid AI, and Ambience. Please join the community at our LinkedIn page.

To inquire about participating in an upcoming Mastermind series or attending a HealthLeaders Exchange event, email us at exchange@healthleadersmedia.com.