Cleveland Clinic’s AI strategy: Partner if you can, build if you must

Rohit Chandra, PhD, chief digital officer at Cleveland Clinic, came to the organization from more than two decades in big tech rather than medicine or health administration.

Four years into his tenure, he said the health system is working to understand how to apply a technology-first mindset to healthcare delivery.

“The Clinic chose to say that technology is going to be a bigger, more strategic, disruptive and transformational force in healthcare,” Dr. Chandra said during Becker’s 16th Annual Meeting. “How can you bring a technology-first mindset, keeping in mind of course that the objective function is healthcare transformation?”

When it comes to AI, Dr. Chandra said Cleveland Clinic’s progress with the technology is best measured not by the number of tools deployed, but by whether they are implemented at scale.

“We have taken a few big problems, developed solutions for those big problems, and then deployed them at scale,” he said. “And deploying them at scale is never easy in healthcare and in large organizations like us.”

That focus is critical in a market with what he described as “about 6 to 8 billion dollars a year” in venture investment.

“The opportunities are endless and there’s a huge risk of distraction,” Dr. Chandra said. “We need to say no to many more things than we say yes to, just because of attention units.”

To avoid that, he said every investment must align with enterprise priorities.

“It’s not about a technology goal. I don’t have a technology objective. I have a healthcare objective,” he said. “If you can draw a line of sight to those macro-level objectives, then at least you’re working on the right thing.”

On the question of building versus buying, Dr. Chandra said Cleveland Clinic prefers to partner.

“The glib answer to that question would be partner if you can, build if you must,” he said.

He added that in AI, traditional vendor evaluation approaches no longer apply.

“In the AI space, if anybody claims that they have a fully finished product, that’s generally not true just because AI is in the early stages,” Dr. Chandra said. “So you have to be prepared that you’re actually not buying a product, you’re actually buying a company.”

That shifts how his team evaluates partners.

“Trying to assess the company in terms of their trajectory, their ambition, their culture and fit and their alignment with your objectives is actually more important than the specifics of the product today,” he said.

He also questioned the value of traditional procurement processes.

“Technology is moving fast and if we follow a playbook of an RFP and an assessment of five vendors and some PowerPoint decision-making fabric, I think it’s impractical in the technology landscape today,” Dr. Chandra said.

Instead, Cleveland Clinic emphasizes pilots.

“If you design a good pilot with a clear outcome at the end, it’s good for us, it’s good for the company, it’s good for the industry,” he said.

Looking ahead, Dr. Chandra said revenue cycle management remains a major opportunity for AI.

“Revenue cycle management is a gift that keeps on giving,” he said, pointing to prior authorization, denials and appeals as areas for continued automation.

“That’s an area where it’s safe to do,” he added. “Safe meaning you don’t have clinical risk in doing that.”

Nursing workflows present a more complex challenge.

“It’s not that you can find a single big hammer that can solve a lot of problems,” he said. “Nursing work is a lot more a collection of lots and lots of discrete things that nurses do throughout the day.”

Cleveland Clinic is also exploring AI for patient engagement.

“We’ve been piloting [AI tools] for about a year,” Dr. Chandra said, describing a future where organizations can “engage patients at scale with high-quality AI-powered solutions.”

Across use cases, he said one lesson has been consistent.

“Change management is the hardest problem and technology is usually the easier part of the equation,” Dr. Chandra said. “Trying to modify or adjust or realign the workflows of 30,000 people is not something that just happens because you think you have a better mousetrap,” he said.

His advice to peers is to focus on meaningful problems and work closely with end users.

“Whatever technology you bring in, you do it in partnership with the folks that are going to use it,” he said. “If not, then sooner or later you’re going to fail.”

The post Cleveland Clinic’s AI strategy: Partner if you can, build if you must appeared first on Becker's Hospital Review | Healthcare News & Analysis.