News|Articles|April 29, 2026

Automating prior authorization without AI

Author(s)Denise Myshko
Listen
0:00 / 0:00

Key Takeaways

  • EHR-triggered data extraction at prescription signing enables instant PBM-criteria matching, delivering near-real-time medication prior authorization decisions without human review or AI.
  • Pilot metrics showed 18-second approvals, 11% fewer denials, 17% fewer appeals, and prescriber abandonment falling from 22% to 4%.
SHOW MORE

Andrew Mellin, M.D., discusses Surescripts’ pilot program for prior authorization system, which pulls clinical data from EHRs and matches it against payer criteria.

Streamlining the prior authorization process doesn’t need AI to automate the handling of requests. In fact, some of the biggest gains are already coming from the adoption of tools such as real-time prescription benefits, prior authorization automation, and electronic benefit verification. These technologies that connect with the electronic health system (EHR) to automate prior authorization for medications can result in faster processing and a lower rate of denials, finds a pilot program that began in 2025 with Surescripts, Fairview Health Services, and Cleveland Clinic, as well as the pharmacy benefit manager Optum Rx.

In 2025, this automated system approved prior authorizations in 18 seconds, with an 11% lower rate of denials and a 17% lower rate of appeals. Importantly, the percentage of requests abandoned by prescribers went from 22% before automation to 4% last year. There has been a 19.9% increase in prior authorizations processed electronically and a 26.1% increase in authorizations for specialty medications being processed electronically.

Surescripts has launched the automation program for prior authorization across 20 health systems, including Cleveland Clinic, Fairview Health Services, Garnet Health, The Ohio State University Medical Center, UNC Health and Watson Clinic.

In an interview, Andrew Mellin, M.D., VP and chief medical information officer at Surescripts, talked about his company’s system, which works by pulling clinical data directly from the electronic health record at the moment a prescription is signed. If that data aligns with the pharmacy benefit manager’s coverage criteria, the authorization is granted instantly.

More than 45,000 prescribers have received approvals through the system. And Optum, in a previously published analysis, reported 88% fewer appeals for medications covered under the program.

Mellin described an example at UNC Chapel Hill, one of the first health systems to go live, where a patient was prescribed a GLP-1 medication just before a snowstorm was expected to shut down much of North Carolina. The prior authorization was approved in 20 seconds. The patient picked up the medication and made it home before the snow hit.

Surescripts automation does not use AI. “There are a lot of ways people are looking at prior authorization and AI,” Mellin said. We believe it is going to play and will continue to play a big role. But one of the factors that we thought a lot about when we built this was that we didn’t want a human review process. AI is not perfect yet, and that's why you need humans to review things.”

Mellin said, however, that as AI continues to evolve, it will become a big part of Surescripts' future strategy as well. “AI will continue to get better at some point, and the industry will fully trust it without human review. That’s the future. We wanted a process that doesn’t have human review and used a different approach to make sure we establish total trust in the way it works.”

Within Surescripts’ prior authorization request process, approvals can happen quickly based on whether the evidence aligns with the policy. Mellin said that if the system is unable to verify the evidence, the prior authorization request is not denied but marked as having insufficient evidence. That request then moves to the process that the provider was using previously.

Mellin said the system has a 34% approval rate, but that can vary by type of medication and diagnosis. For example, a GLP-1 prescription written for diabetes is likely to move quickly through the system, but for a GLP-1 for weight loss, there is variability in payer policy, or the evidence may not be sufficient in the EHR.

Broader automation efforts

Surescripts’ prior authorization automation effort is integrated with a real-time prescription benefit tool used by approximately 970,000 prescribers, which provides out-of-pocket cost information for patients and lower-cost alternatives. The company processed more than 1 billion such transactions in 2025. When prescribers switch to a lower-cost option, patients save an average of $77 per prescription; for specialty medications, that figure jumps to $817, Mellin said.

The system will provide doctors information about whether the plan or PBM has a lower-cost mail-order pharmacy option. "Doctors may not know that at the point of care, especially for these specialty medications,” Mellin said. “The plans can return that pricing differential and make it very clear to the provider that there could be a large savings, especially like a mail-order pharmacy.”

Surescripts has also extended its electronic benefit verification service to life sciences companies and patient assistance programs, which historically had to make phone calls to determine a patient’s insurance eligibility before enrolling them in copay card programs. The new service allows those organizations to process eligibility checks at scale, replacing a manual process that had created delays in getting patients started on specialty medications.


Latest CME