AHIP 2024: Not just a buzzword—payers envision AI’s future

You can’t spell AHIP without AI.

That was certainly the case last week in Las Vegas as thought leaders from the country’s biggest health plans discussed, perhaps more than any other topic, how artificial intelligence is shaping their businesses and could rapidly change the industry forever.

You’d be hard-pressed to find a company or industry not exploring the potential of AI today, and health systems and plans are no exception. But the rate to which technology is improving is causing heads to spin among even long-time observers of AI in healthcare.

“I’ve been in this field for 30 years,” said Charles DeShazer, M.D., chief quality and health advocacy officer for Cigna. “I’ve never had trouble keeping up until now. I’ve never seen anything like this in my career.”

“It is really difficult,” Stacy Byers, senior vice president for design and digital at Highmark Health, told Fierce Healthcare. “You almost have to use AI to keep up with AI.”

Health plans in attendance outlined countless use cases already making daily operations more efficient and could, in the not-so-distant future, lead to much more personalized care for members.

Many of the advances speed up data submission and prior authorization delays on the backend. Johns Hopkins Health Plan Chief Information Officer Joydeep Mukherjee said to Fierce Healthcare there are fewer barriers to data exchange between the health plan and Johns Hopkins Medicine, now that the health plan implemented the Epic Payer platform less than a year ago. Clinical data exchange is now enabled between the two entities. In July, the system will attempt to implement real-time prior authorization.

“That will be a huge step forward,” he said. “Up until then, there hasn’t really been a strong value proposition from the perspective of the provider entity.”

The health plan will start making that process more seamless, but noted instanteous prior auth decisions won’t happen immediately. Over time, automation will handle more of the challenges currently causing administrative burden. Johns Hopkins is hoping to increase productivity, as defined by the amount of time a nurse or medical director spends approving a prior auth decision, by 40%.

No longer, said Mukherjee, will physicians have to submit claims by electronic fax, a workflow that is not yet obsolete. He said he hopes these efforts will convince other plans and providers to seamlessly transfer data between Epic EMR and the Epic Payer platform.

Mukherjee also led a panel with DeShazer and Elevance Health Chief Clinical Digital Officer Vardini Vishwanathan on AI in delivering whole health.

Elevance Health is currently using AI to better understand the needs of care members, but could one day “infuse intelligence into an EMR” and build out more personalized care, said Vishwanathan.

DeShazer said Cigna uses AI in “low-risk” areas like data summarization of Medicare Advantage benefit information. He says AI will be used to better comb through all data available to health plans.

“I think in the near future we are going to have digital twins on each one of our members,” he explained. “We’re going to model ‘what would happen if we did X, Y or Z’? What are key interventions for that scenario?”

He said, through cognitive computing capabilities, health plans will soon be able to model individual patients on a granular level. For example, a plan could test how an intervention affects a patient with diabetes, rheumatoid arthritis and is living in a food desert, showing the potential of personalized modeling.

Byers believes “synthetic personas” will soon help plans like Highmark Health to validate concepts, gather feedback and scale operations more efficiently.

Doug Henry, vice president of psychiatry and behavioral health at Allegheny Health Network, teased a not-yet-released study showing AI can produce a mental health questionnaire score for an interaction between a care manager and patient, in less than a minute, without a mental health questionnaire being administered.

One of the paper’s findings is that the AI tool was more accurate in recording a correct score than a doctor, as validated by five independent psychologists. Highmark is working with an AI company that looks at the quality of the vocalization as well a semantic analysis.

Henry predicts AI will soon transform behavioral health as it has in chemo care.

“The science of chemotherapies evolves so quickly there's no way that any single oncologist could stay on top of it,” he said, noting that AI scans recent literature to determine the best drug therapy for a patient. Henry believes a similar AI-determined treatment plan is possible, and necessary, in the mental health space.