Medical reviewers, nursing staff, and clinicians are required to examine medical records for claims and verify patient data related to prior authorizations and claims, also respond to specific questionnaires associated with each claim. Currently, this process is conducted manually, where they meticulously scrutinize each document to verify its accuracy and provide responses to the questionnaire. To assist in their assessments, they refer to both the provided claim document and the CMS document. Subsequently, based on their findings and responses to the questionnaire, they make recommendations regarding the claim as approval, denial or requesting further information. The duration of this entire process varies from a few hours to several days, depending on the number of pages in each document. 

Our medical review solution, developed with the assistance of GenAI, streamlines this process by employing an LLM to automatically respond to questionnaires. This solution helps the nursing staff to improve their productivity by 30-50% by automating the review of medical records. 

Anonymous nationwide payer client

Government payer entity

Industries: Healthcare, RCM

Provider Systems: Epic

Payer Portals: eviCore, United Healthcare, Magellan

Challenge

A large healthcare provider needed to ramp up their automation strategy to offset labor instability and attrition. Currently the staff manually have to go out to payer portals, submit requests and check statuses of Prior Authorizations, sometimes multiple times. Through Task Mining AI (Artificial Intelligence) analysis, it was determined that staff were spending about 30 – 40% of their time submitting requests and checking the status for 19,000 prior authorizations per month.

Prior Authorization Requests: 3 bots processing prior authorization requests in 3 separate portals (EviCore, Magellan and UHC).

• EviCore: 923 transactions per month
• Magellan: 595 transactions per month
• UHC: 7,063 transactions per month

Prior Authorization Statusing: 3 bots processing prior authorization statuses in 3 separate portals (EviCore, Magellan and UHC).

• EviCore: 5,536 transactions per month
• Magellan: 3,572 transactions per month
• UHC: 986 transactions per month

Solution:

RPA bots are a much more accurate and effective solution to verify the status of patient health authorizations, RPA bots can repeatedly check requests under review for status changes and updates.

CampTek Software was engaged because of its healthcare and Epic experience. Through task mining, AI models were created, and results were reported back to the customer with enormous success. The next phase of design and build all occurred in a brief time. The bot now in production includes business rules, exception handling and daily reporting functionality. It is also supported by CampTek as an ongoing managed service solution for the provider.

The Bot is now saving 2,696 FTE (full-time employees) hours per month or 32,352 hours (about 3 and a half years) per year. The FTE capacity gain is 17 FTEs (full-time employees), and the ROI (Return on Investment) is $1,085,000 per year.

Any healthcare provider