Automated prior authorization using ABByy and UiPath

Case study

Automated prior authorization using ABByy and UiPath

The prior authorization process is meant to reduce the complexities involved in the patient care journey and keep a check of healthcare costs. With the right prior authorization process in place, patients should receive the required care at the right time and in the best possible environment. However, payers have to go through a number of complicated medical policy guidelines and regulations just to make sure patients get the right medication and treatment, which should also be economical. Providers also scrutinize these requests to avoid duplication of services. For example if the same type of scan is ordered for a second time within a span of one or two weeks by a different specialist, the provider will verify if the specialist has seen the scan that was performed by the earlier specialist. This type of verification may delay the prior authorization process.

Healthcare providers also require prior authorization for a range of treatment options, such as tests, prescriptions, lab tests, surgeries, and so on. A good amount of their time is consumed managing the prior authorization process, as it involves dealing with various payers, erratic communication channels, different payer processes, and other payer related challenges.

According to a recent American Medical Association survey, over 91% of prior authorization requests have resulted in delayed care delivery and more than 26% of claims had to wait for more than 3 business days to get a decision from the payers. The repercussion of such delays in treatment can be long-term and place a huge burden on patients, physically and psychologically.

There is immense opportunity for the healthcare industry to reduce costs through automated prior authorization automation. As per the latest report from CAQH (Council for Affordable Quality Healthcare), industry initiatives are already making substantial improvements in the prior authorization process and recommends that it is imperative for all the stakeholders to move towards a more automated prior authorization workflow.

Our client is a US-based not-for-profit health insurance organization that provides insurance coverage to over 6 million members. They wanted to automate the processing of prior authorization requests, which are received from a large network of hospitals present in the states they service.


The client needed an automated system that could gather PA requests from various sources, process the requests, and respond to them quickly. The solution had to:

  • Obtain PA requests through e-mail and ePA (electronic Prior Authorization).
  • Execute e-mail classification and document identification.
  • Run custom rules to verify the eligibility of the requests based on plans and state laws.


  • PA requests were received through multiple communication channels, requiring a lot of time and manual effort to process.
  • PA requests and related data were submitted in different formats, as prescribed by individual state laws.
  • The processing rules vary, based on the members’ health plan.
  • The processing rules had to be frequently updated or reconfigured to include new medications, tests, and therapies.


Imaginea proposed the implementation of a UiPath RPA bot to process each PA request as an entity. Requests received by email and fax are to be converted to PDF. Then the PDF PA requests are processed using OCR and ABBYY FineReader to extract key information from the forms. The bots will:

  • Extract and categorize data
  • Complete eligibility verification of the members
  • Verify the prior authorization requests
  • Manage the response verification workflow

Tech stack

How our solution helped

With the automated prior authorization system, our client can now scale their processing and respond to millions of requests every year, with an average TAT of less than two days.

Overall approach

First, the prior authorization requests received through scanned emails/fax are extracted and categorized into standardized formats by the RPA bots. Note that the requests received via the customer’s portal are already in a standard format.
Next, the bots query the member database to retrieve the plan information and verify the eligibility of the PA requests. Post the eligibility check, the requests are verified against the custom clinical/business rules.

Finally, before the bots create a workflow queue for the manual review process, adjudication responses or additional clinical input requests are framed. In the manual process, a reviewer approves the response and sends it to the provider.

The following image illustrates the complete process in detail:


  • Up to 60% reduction of operational costs in processing of prior authorization requests.
  • 40% reduction in the number of prior authorization appeals.
  • Specialists are now able to focus their time and effort on complex requests.

Talk to us