Artificial Intelligence

Revamping prior authorization: How AI and automation could boost care and revenue

October 5, 2021 3:19 pm

Ninety-four percent of physicians say archaic prior authorization processes result in care delays for patients, an American Medical Association (AMA) survey reveals. Nearly one in three say these delays have resulted in an adverse event for a patient in their care.

Further, breakdowns in prior authorization processes are the root cause of revenue leakage for healthcare organizations, resulting in 30% of preventable write-offs. As a result, two out of five physicians have staff who focus exclusively on prior authorizations, according to the AMA study.

These are signs that the current approach to prior authorization — a highly manual process that costs healthcare providers and health plans $767 million a year, up from $627 million in 2019 — needs an electronic refresh. Today, more healthcare organizations are exploring the use of automation and artificial intelligence (AI) to transform prior authorization to reduce the administrative burden while strengthening patient throughput, revenue and employee retention.

The need for a more modern approach

Prior authorization holds the lowest electronic adoption rate among eight medical transactions studied by the Council for Affordable Quality Healthcare (CAQH). Manual prior authorizations cost providers $10.26 per transaction in 2019, according to the 2020 CAQH Index Report, compared with $3.64 per electronic transaction and $7.07 per partially electronic transaction. For health plans, the cost equates to $3.14 per manual transaction, compared with just 12 cents per electronic or partially electronic transaction (those taking place through a portal).

Converting partially electronic prior authorizations to fully electronic transactions presents the largest per-transaction savings opportunity for the industry identified by CAQH.

For years, challenges associated with prior authorization — from administrative inefficiency to delays in care to the impact on revenue — prompted calls for reform. “The reality is that it’s an inefficient, paper-based process for many providers, and the rules are confusing. In fact, the people who work for the insurance companies don’t always know the rules around prior authorization,” said Jonathan Shaker, executive director, surgical business operations and patient access, New England Baptist Hospital, Boston. “It’s not uncommon for a staff person to talk with a health plan representative and be told that prior authorization is not required, while another employee could call the same health plan and be told by another representative that it is needed.”

In 2018, several national organizations, including the AMA, cited five areas that could result in meaningful improvement, such as automation to improve transparency and efficiency. Yet the AMA states that little progress has been made toward improvement in these areas.

It’s one reason why some healthcare organizations are exploring the use of AI-fueled technologies to remove barriers throughout the prior authorization process. It’s an approach that holds strong potential to reduce administrative demands on physicians and patient access staff while helping to ensure timely access to care.

“I definitely think there are ways that AI or data science types of approaches could help ease the prior authorization burden,” said April Todd, senior vice president, CAQH. “For example, AI could proactively find information in a patient’s chart or medical history that could be shared with a health plan to support authorization. There are some opportunities there.”

While Todd views adoption of AI and data science in prior authorization as “a bit of a ways off” for most providers and plans, research released by Black Book in late August 2021 reveals use of AI-based tools in healthcare revenue cycle is rising: 58% of healthcare organizations are using artificial intelligence in one basic form or another, and 92% expect widespread industry implementation in the next five years, despite operational constraints and financial priorities resulting from the COVID-19 pandemic.

Exploring AI’s potential for prior authorization

As healthcare organizations explore ways to automate time-consuming and labor-intensive processes related to prior authorization, KLAS Research interviewed seven early adoptersa and found that investments in AI to enhance their approach achieved:

  • A streamlined, organized prior authorization process
  • Increased prior authorizations without the use of additional staff
  • Reductions in staff time 
  • Decreased denials

For New England Baptist Hospital, a large orthopedic specialty hospital, making the move toward an automated, AI-based approach to prior authorization wasn’t just a matter of patient throughput, satisfaction and revenue. The frustrations related to prior authorization processes also made it tough to keep patient access staff in these roles. “There was a lot of turnover due to the frustrations of dealing with the inefficiencies in prior authorization created by insurance companies,” Shaker said.

When coupled with the denials rate related to prior authorization, “We knew we needed to figure out another way to do this,” Shaker said.

“We’re a small specialty hospital, with just under 150 beds. It used to take 11 days, on average, from the time an order was submitted to the time the patient was scheduled because of inefficiencies related to prior authorization. At one point, things got so bad, a number of private practices started to divert imaging orders from our hospital to a standalone imaging center. Worse, challenges with prior authorization led to last-minute cancellations for procedures, threatening utilization of our most expensive resource: the operating room.”  

In 2018, New England Baptist Hospital invested in an AI-based prior authorization platform that automatically determines whether a prior authorization is required based on a review of medical necessity policies and requirements from more than 40,000 health plans. The platform then submits clinical documentation from the EHR to payers and tracks the status of prior authorization submissions, all while producing reports that increase visibility into and control over the authorization process. “Now, there’s a detailed audit trail that enables us to see who interacted with the authorization and when information was sent to the payer,” Shaker said. “We’ve gained much more insight into this process.”

Within a few months, New England Baptist Hospital’s patient access department reduced scheduling time for procedures from 11 days to just 2.3 days, on average, while telephone abandonment rates dropped from 16% to 2.9%. The use of AI also has dramatically boosted revenue and staff satisfaction. “We’ve reduced write-offs by 30% and cut turnaround times by 78%, all while improving staff productivity by nearly 25%,” Shaker said. “The increased efficiency we gained enabled us to promote one patient access representative to the role of quality analyst over our information access department. It’s a much better utilization of her skill set, and it reflects a new avenue for employee growth that became possible by automating administrative processes.”

Making the move toward AI and automation

While electronic adoption for prior authorization increased from 13% to 21% in 2019 — the highest increase in electronic adoption among medical transactions assessed by the 2020 CAQH Index — it remains the most time-consuming transaction for medical providers. Prior authorizations take an average of 20 minutes per manual transaction, 13 minutes for partially electronic authorizations and eight minutes for fully electronic transactions — the HIPAA standard. “Some providers reported spending as much as an hour to complete a manual prior authorization,” according to the 2020 CAQH index.

That’s why solutions that automate prior authorization with the help of AI demonstrate strong potential to make a difference in healthcare revenue cycle. “This is about helping patient access staff do their work better,” Shaker said. “Prior authorization is only going to become more complicated over time. To the extent that you can find a tool that helps automate this process and make it more efficient, the better able your organization will be to ease prior authorization demands for physicians and staff while enhancing access to timely care and reducing costs.” 

The business case for automating prior authorization

  • 85% of physicians say the prior authorization burden is “high” or “extremely high”
  • Physicians complete 40 prior authorizations per week, on average
  • It takes providers 20 minutes to complete a manual prior authorization
  • Manual prior authorizations cost $10.26 per transaction for healthcare providers
  • Providers and health plans could save $417 million annually by automating prior authorizations

Sources: 2020 CAQH Index Report and the 2020 AMA Prior Authorization Physician Survey

About Olive

Olive is the automation company creating the Internet of Healthcare. The company is addressing healthcare’s most burdensome issues through automation —delivering hospitals, health systems and payers increased revenue, reduced costs, and improved efficiency. People feel lost in the system today and healthcare employees are essentially working in the dark due to outdated technology that creates a lack of shared knowledge and siloed data. Olive is driving connections to shine new light on healthcare processes, improving operations today so everyone can benefit from a healthier industry tomorrow. To learn more about Olive, visit

This published piece is provided solely for informational purposes. HFMA does not endorse the published material or warrant or guarantee its accuracy. The statements and opinions by participants are those of the participants and not those of HFMA. References to commercial manufacturers, vendors, products, or services that may appear do not constitute endorsements by HFMA.


a. KLAS Spotlight, “Verata Health (Now part of Olive): End-to-end prior authorization,” KLAS Research, 2021.


googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text1' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text2' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text3' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text4' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text5' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text6' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text7' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-leaderboard' ); } );