Hospitals are facing unprecedented pressure to adopt AI while keeping costs, risk and complexity under control. Three digital executives told Becker’s that governance, not gadgets, will determine whether AI improves care or becomes another layer of operational burden.
At the University of Mississippi Medical Center in Jackson, Chief Data Analytics Officer Sandeep Rustagi said one of his primary responsibilities is helping operational teams make sense of a fast-moving technology landscape.
“I would say my role is more of a chief liaison officer between data and technology and business leaders,” he said during an interview with the “Becker’s Healthcare Podcast.” “I really want to help the operational leaders navigate all the challenges of data and technology that we are running into today, especially with AI being everywhere.”
That pressure makes thoughtful product selection essential. As a state-funded organization, his team avoids jumping on every pilot.
“We want to be very careful with how we select the products. We don’t want to do wasted trials,” he said. “We don’t want to have wasted spend. Sometimes delayed value is okay, rather than having wasted trials or wasted spend.”
Regulatory scrutiny, ethics and equity must be part of governance.
“AI is moving just way too fast, and state level and federal level regulations are still catching up in terms of where we want to be,” he said. “As a state funded organization, we want to make sure that we are flexible and nimble, but at the same time, we are very cautious.”
“We want to make sure that the products that we select have a diverse representation of our population demographics,” he said. “Also, we want to make sure ethics and equity are taken care of in the product selection that we have.”
Mr. Rustagi said his team often prioritizes solutions already inside the EHR to avoid unnecessary clutter. Governance also means ensuring technology truly fits the organization’s strategy and delivers the right value for return on investment.
At Madison, Wis.-based UW Health, CMIO Joel Gordon, MD, has a similar strategy. Rather than launching standalone AI initiatives, he places AI under existing operational structures — a framework he calls “AI enablement.”
“The documentation team is going to stay owner of the documentation problems and solutions and all the other areas and and all the other clinical decision support.”
That mindset keeps AI grounded in real operational needs and builds initiatives into UW Health’s ecosystem. Dr. Gordon, who frequently partners with UW Health’s security and compliance leaders, noted that governance must start before organizations engage regulators.
“First clean up your own yard and your organization,” he said. “What are your data structures? What are your security protocols, your overall manufacturing of your product stack, whatever it is you’re producing and developing.”
Philadelphia-based Jefferson Health CIO Luis Taveras, PhD, said the explosion in digital health and remote monitoring vendors increases the need for disciplined decision-making.
“One piece of advice that I would have is to be careful not to have too many solutions. There’s so many different vendors out there trying to sell us something because they believe we have a problem, not because we really think we have a problem,” he said.
Focus tightly on validated needs and identifying the best solutions for current platforms, he advised. Choosing too many point solutions creates operational strain.
“I don’t want to proliferate hundreds and hundreds of different applications out there because it really, really makes it much more cumbersome from an operational perspective,” he said.
Ultimately, CIOs must think like enterprise leaders, not technologists.
“When you think of healthcare leaders and think about the CIO, the role of CIO, I’m just one of the general managers that runs the organization. And my focus is IT, but I have to really, really focus on the strategy of the entire organization,” said Dr. Taveras.
