Faster Submissions or Faster Decisions? Prior Authorization Needs Both

There are two ways to make prior authorization faster and help patients start therapy sooner. AI speeds up the work—finding information and drafting responses for a clinician to review. Automation removes the work altogether.

With Surescripts Prior Authorization Automation, everything starts moving the moment a prescription is signed. The solution gathers the clinical information from the electronic health record (EHR), sends it to the pharmacy benefit manager (PBM) for review and returns a determination—in most cases without the clinician seeing a prior authorization at all. The request is simply approved if the clinical evidence in the chart is aligned with the patient’s benefits. 

What happens when that process isn't automated? A patient waits. Consider two patients, one who is able to pick their medication up from the pharmacy moments after it’s prescribed and one who must wait days, weeks or more. They have the same diagnosis and same prescription, but a delay in access puts them on two different clinical trajectories. 

Prior authorization delay is what separates those two patients. 

Health systems are moving quickly to shorten that wait with artificial intelligence (AI), and there’s no doubt that prior authorization AI is genuinely useful. AI tools help the provider and staff find answers to prior authorization questions, but they still require a human to review the information, refine answers and submit the prior authorization request.

What Automation Does That AI Does Not 

Automation solves prior authorization differently. It doesn't just help a clinician or their staff work through a prior authorization, it takes the prior authorization completely off their desk. When the evidence in the EHR meets the payer's criteria, no one initiates, completes, submits or follows up on anything, and there's nothing for the care team to review. 

This is what it means to accelerate decisions. Payers get the information they need up front, automatically, so the determination itself moves faster. 

Prior Authorization Automation drives the following results:  

• Approvals delivered in a median time of 18 seconds when all criteria are met 
• An 11% lower rate of denials due to lack of information and a 17% lower rate of appeals 
• Requests abandoned by prescribers falling from 22% before automation to 4% after 

The automated path also asks remarkably little of care teams. It uses data clinicians are already documenting inside the workflow they already use, so there's no new interface, login or additional steps to complete at the end of a visit. 

Where Prior Authorization AI Works Best

AI is very useful for complex prior authorizations. When applied in these use cases, AI prior authorization tools can: 

• Search the patient chart for the specific documentation a payer is asking for, instead of leaving a clinician to hunt for it 
• Draft responses to prior authorization questions for a human to review, edit and approve 
• Reduce the time burden of requests that cannot be handled any other way

The Case for Automation and AI  

AI does not replace automation, and automation does not make AI unnecessary. Health systems and patients benefit from both, and a health system that deploys only one is only solving part of the problem. The role of each tool is straightforward: 

• Automate as many decisions as possible: Where clinical data in the EHR can be matched against published criteria, automation handles the request from end to end, with no clinician effort or review. This covers a substantial share of medication prior authorizations today, and that share grows as more medications come into scope. Every request routed through Prior Authorization Automation is one your team never has to touch, so the goal should be to run as many as possible through it. 
• Apply AI when automation is not available: For more complex prior authorizations, AI tools that search for relevant documentation or draft responses keep the work moving with a clinician in the loop. 

The key takeaway is to start with automation. Automation should be the default path, and AI-assisted workflows should handle the exceptions.

Think Beyond the Submission 

Most health systems are asking how to complete requests faster. The better question is which prior authorizations should require no human effort at all. 

Answering that question means connecting clinical information, decision criteria and payer systems so those requests resolve themselves. That's the difference between a request that happens automatically in the background and one that sits with the care team—and the difference between a patient who starts therapy today and one who waits.