Prior Authorization - To Automate Or Not To Automate?

Background

There is often talk about million dollar problems, but what about billion dollar problems?

It's no secret that processing prior authorization requests are expensive, inefficient and most importantly, the cause of delays in patient care.

Given the amount of manual work involved with processing authorizations, 麦肯锡 predicts that 50 to 75% of manual tasks within PA can be automated via a mix of AI & machine learning. This is an estimate back in 2022 right before the boom of generative AI. (McKinsey Next-Gen Prior Authorization 2022 ).

What is prior authorization (PA)?

Prior authorization is a process where healthcare providers must obtain approval from a patient's insurance company before delivering specific treatments, medications, or services to ensure coverage and reimbursement.

The goal for providers is to ensure treatments are medically necessary and that they are covered by the patient's insurance in order to be reimbursed.

Why does it matter to automate?

Bottom-line: If prior authorization stalls it can delay patient care.

This is especially important for terminally-ill patients who require care on a timely basis. Therefore, it's not simply about improving efficiency. It's that inefficiency negatively impacts the lives of humans. Providers have a duty to standardize and automate as much as possible to streamline the process for their staff, patients and the industry as a whole.

Patient care has to be at the center of every technology strategy.

Prior Authorization by the Numbers

A 2023 survey by CAQH found that healthcare providers spend $494 million annually on PA processes as a whole. This is 30% more than 2022 due to the increase in volumes. Providers indicated that they are spending 16 minutes to process each PA via a portal, making it one of the most time consuming administrative tasks throughout healthcare operations (2023 CAQH Healthcare Survey ).

The burden that PA bestows on providers is not light. According to a response from a survey from the American Hospital Association, 95% of health systems reported an increase in the amount of time spent by staff in seeking PA approval. This is basically ALL healthcare providers that were involved in the survey.

To put this in context, a single 355-bed psychiatric facility reported needing 24 full-time employees dedicated to processing authorizations. This is coming at a time where providers are dealing with workforce shortages & a hunt for quality talent (American Hospital Association Prior Authorization Reform).

Prior Authorization Steps - Bird's Eye View

PA involves a number of steps that require frequent back and forth communication with providers and insurers. This inefficiency is the result of extensive policy manuals on rules that are ever-changing, adding to the amount of manual work here. These delays significantly impact providers and payers. The process of PA helps avoid fraud yet causes serious bottlenecks in processing. This is because preventative treatments and low cost services like podiatry often do not require PA, while 79% of the highest cost services do require PAs.

How can automation & AI play a part?

Ultimately, providers have been leveraging automation & artificial intelligence to take over all of the low-IQ work from the PA process - including all data reconciliation, data entry and rules-based actions.

In many cases, providers still need to manually contact call centers via the phone with wait times averaging ~25 minutes. However, the industry is slowly shifting towards electronic forms of PA submissions. This still requires a significant amount of time spent logging into a system and performing a myriad of data entry for these requests.

The latter is the portion that providers are investing most of their efforts to automate, given the rules-based, administrative nature of these tasks.

Submission Automation - Entering Prior Authorization Requests onto Payer Portal Website

Here's an example of how an automation works to assist providers in submitting PA requests:

• Bot selects patient & payer: Bot logs into EHR system to search for particular patient (based on name, DOB, etc.) & select specific payer (e.g. United Healthcare, Availity, BCBS, etc.)
• Bot pulls list of PAs: Bot pulls list of PAs to be processed for each specific payer
• Bot finds provider on payer portal: Bot logs into payer portal's website to select specific provider (based on their location)
• Bot enters patient demo data into payer portal: Bot enters patient demographic information from patient chart to payer portal (e.g. name, DOB, gender, address, member ID, group #, clinic address, etc.)
• Bot enters provider data into payer portal: Bot enters provider details into payer portal (e.g. date of submission, type of patient, nature of condition, date of surgery, etc.)
• Bot enters additional patient data into payer portal: Bot enters additional patient details into payer portal (e.g. frequency of symptoms, average pain intensity, overall health condition)
• Bot updates patient chart: Bot copies confirmation number, adds it to patient chart & adds next activity date to two days out
• Bot adds authorization to patient chart: Once results are received, bot then goes back to payer portal to find completed authorizations, prints the authorization as a PDF & adds it to the patient chart

This automation assists in the drudgery that providers deal with when submitting PA requests, freeing up staff to higher-IQ portions of the process.

What's the takeaway of all this?

The takeaway here is that maintaining PA as a fully manual process has a pervasive negative impact on patient care. If the financial impact and burden on administrative staff are not enough, patient care always will be.

So, the question remains: To automate or not to automate?

With the benefits of automation and AI clearly outweighing the drawbacks, the answer for many is clear: respect your staff and your patients - automate.

References

1. McKinsey Next-Gen Prior Authorization 2022
2. 2023 CAQH Healthcare Survey
3. American Hospital Association Prior Authorization Reform
4. Healthcare Administration Problem McKinsey 2021
5. CAQH 2021 Survey & Data - Prior Authorization