Health systems lack AI governance: 'Generative AI is not a one-time fix'

Health systems executives are cautiously optimistic about artificial intelligence in healthcare. The technology has great promise to improve clinical and administrative functions, but most organizations are lacking strong governance for AI and generative AI.

Pittsburgh-based UPMC's Center for Connected Medicine and KLAS Research partnered to survey almost three dozen health system executives about AI risks and potential benefits. Just 16% of the survey respondents said they had systemwide governance policies, although many had formed governance committees that include senior executives to oversee AI.

The three ways respondents thought AI would most benefit healthcare include:

1. Improving efficiency
2. Adding visibility to clinical decisions
3. Repetitive task automation

Seventy-three percent of the executives surveyed have plans to adopt AI solutions via EHR vendors because the integration is easy. But Jeffrey Jones, UPMC Enterprises' senior vice president of product development, cautions leaders against moving too quickly.

"Before adopting generative AI technologies in healthcare, it's crucial for executives to clearly define their objectives and establish measurable benchmarks," Mr. Jones said. "Regular evaluations are essential to adjust strategies as necessary. Generative AI is not a one-time fix, but a dynamic tool that requires attention and calibration."

Many of the health systems surveyed reported having a cross-functional team with multiple health system executives accountable for deploying AI. CIOs and chief data and analytics officers were often the leaders responsible for AI oversight, while CEOs, chief technology officers and COOs were also involved in AI strategy.

One chief technology officer who participated in the survey said their system created a general analytics and AI governance steering group to advise on AI.

"The landscape is changing so quickly that I don't think we will ever get to the point that we create a policy; as soon as we create a policy, it is out of date," the CTO said.

Security is another big focus for healthcare organizations as they embrace AI. Health systems trust the large companies, including Epic and Microsoft, to incorporate AI into existing platforms with HIPAA compliance in mind. However, systems are less trusting of startups and new vendors because the security standards for dealing with healthcare information are so strict. Nine of the 30 respondents had a security vetting process for third-party AI solutions and seven included business associate agreements.

After a health system decides to deploy a new AI-driven technology, leaders focus on clinician and team member adoption, according to the report. Some systems have rolled out learning sessions and pilot committees to accelerate collaboration and alleviate skepticism.

Eleven of 23 respondents said showing proven benefits and return on investment for AI helps generate buy-in, while five said their system provides training and education. Three leaders selected champion users to lead the way, and another three elevated correct use cases.

Copyright © 2024 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.


Featured Whitepapers

Featured Webinars