Writing for the New York Times, Teddy Rosenbluth explains how some doctors are turning to generative artificial intelligence (AI) to streamline the prior authorization process and "operate at the same level as companies that have essentially infinite resources." Advisory Board's Mallory Kirby analyzes the impact on plan-provider dynamics and outlines the importance of strong plan-provider partnerships.

Doctors turn to AI to reduce prior authorization burden

Prior authorization was designed by insurance companies to rein in the use of unnecessary and expensive treatments to keep healthcare costs down. However, doctors say they cannot keep up with the amount of documentation and paperwork required by insurers for prior authorization.

According to a survey from the American Medical Association (AMA), doctors and their staff spend an average of 12 hours a week submitting prior authorization requests. Many doctors also say that the process is burdensome and negatively impacts patient health.

"If you want to see a physician go apoplectic at a cocktail party, mention prior authorizations," said Robert Wachter, the chair of the medicine department at the University of California, San Francisco.

To reduce the burden of prior authorization, some doctors are now turning to generative AI to write their request letters or appeal denials.

"We haven't had legislative tools or policymaking tools or anything to fight back," said Azlan Tariq, who has a rehabilitation medicine practice in Illinois. "This is finally a tool I can use to fight back."

According to Tariq, the time he spends on prior authorization has been halved since he started using Doximity GPT, a HIPAA-compliant version of the chatbot. The tool, which uses information from his patients' medical records and insurers' coverage requirements, has also made his letters more successful.

Before using AI, Tariq said that only around 10% of his prior authorization requests were approved by insurers. Now, around 90% of his requests are approved.

Generative AI has also helped doctors at small practices appeal prior authorization denials. In the AMA survey, almost half of doctors said that when they didn't appeal a claim denial, it was partly because they didn't have the time or resources to go through a lengthy appeals process.

Michael Albert, an obesity medicine specialist in Oklahoma, said that AI has allowed his small telehealth practice to go from almost never appealing denials to sending 10 to 20 appeals per week. AI has helped his practice "operate at the same level as companies that have essentially infinite resources," Albert said.

Could prior authorization become an AI 'arms race'?

According to Rosenbluth, some experts are concerned that the prior authorization process could become an AI "arms race" as both insurance companies and physicians use the technology to draft letters about coverage.

Although AI is still primarily used by individual doctors, companies are also trying to make the technology more mainstream. For example, Epic, one of the largest EHR companies in the United States, is testing an AI prior authorization tool among a small group of physicians. Several major health systems are also currently piloting Doximity GPT.

Insurance companies are also looking for ways to use AI in their processes. According to Jeff Levin-Scherz, a health policy expert at the Harvard T.H. Chan School of Public Health, most health plans are likely at least evaluating how to use AI in their claims review process, if they haven't been using them already.

Separately, Chris Bond, a spokesperson for AHIP, said that insurers welcome ways to streamline the prior authorization process, including those involving the "appropriate use of A.I."

As doctors use AI to write prior authorization letters more quickly, Wachter said he had "tremendous confidence" that insurance companies would also use AI to get better at denying these requests.

"You have automatic conflict," Wachter said. "Their A.I. will deny our A.I., and we'll go back and forth."

According to Wachter, he hopes that, in the future, more advanced AI technology will help insurers and providers create a new system that relies more on automation than just sending letters back and forth.

For example, insurance coverage could be automatically approved based on AI analyzing the most up-to-date scientific literature. An AI tool could also notify doctors that an expensive drug they're about to prescribe could be substituted with a cheaper, but similarly effective medication.

For now, doctors are using AI to have a new, albeit potentially temporary, foothold over insurance companies when it comes to prior authorization, Rosenbluth writes.

"It is definitely an A.I. race," Tariq said, "but I think it wouldn't be fair if the physicians didn't have A.I." (Rosenbluth, New York Times, 7/10)