Advanced technology helps with revenue cycle management by solving mundane tasks such as managing payer masters.
This article is part 1 of a two-part series. Stay tuned for part 2.
Artificial intelligence promises to bring many benefits to healthcare organizations, but perhaps one of the most underappreciated is the benefits this technology, combined with machine learning, can bring to the entire insurance claims resolution process.
It offers a wide range of possibilities, but which one offers the greatest value for money? Over time, it has become clear that the greatest returns to innovation are achieved when users work independently of software designed to accelerate and optimize processes.
Here and in the next article, we review the cash acceleration component of revenue cycle management (RCM) after claims submission. It is estimated that one in three claims requires some type of review and additional work to resolve, so this is a critical time to get support.
As these new technologies are applied to revenue cycle management processes, it is important for those overseeing these processes to understand how AI and ML can improve business processes.
Payer arrangement
One aspect of RCM that AI and ML can address is managing payer masters and assigning payer IDs to payer names and addresses.
Sounds easy, right? This is a critical component because the payer master is provided to RCMs and billing companies in a separate file or access point from patient information. Payer master is critical to the revenue process because you need to use electronic transactions to drive revenue, and you need accurate payer IDs.
Because electronic transactions require a correct match between the payer’s name and address and the payer ID, inaccurate data can result in significant disconnections. This is risky, as current approaches require intensive manual processes, are prone to errors, and can compromise a seamless claims resolution process.
A payer master may contain hundreds of repetitions of the same payer name for the same state. For example, a Blue Cross plan in New Jersey might be labeled Horizon, BCBS, Blue New Jersey, Horizon Blue Cross, or incorrectly labeled Anthem.
Other payers’ data may be inaccurate or outdated, which can further confuse the process. For example, payer addresses can change frequently, and some addresses on file may be more than 10 years old or no longer exist due to mergers or obsolescence of business addresses.
This is where advanced technology can have an impact. For example, a revenue cycle management company with over 93,000 payer name and address records used automation to assign payer IDs to payer names and addresses. This was completed in 3-4 hours and 99% accuracy was confirmed in the final validation step of the process.
The company said it simulated the same process as a manual process and determined that it would take more than a week manually, with an estimated time of about 5 hours per day. The accuracy with the manual process was estimated to be only about 75%.
Accelerate your efforts with AI
This all works with the contribution of machine learning, a subset of artificial intelligence.
To understand the synergy, think of AI as a brain. AI creates systems with human-like intelligence. Machine learning is a real learning process, a mechanism by which the brain becomes smarter through experience and data, rather than being taught all the rules.
Email is a simple example of machine learning. The application uses spam filters. Rather than having developers create rules for every message that might be spam, this feature uses machine learning to learn what spam looks like from the thousands of emails it receives.
When machine learning is applied to payer IDs, the technology can match addresses and organization names, even when they are operating in the same state in duplicate. This is done by first parsing the address into its components and then calculating a similarity score using a fuzzy match algorithm that incorporates the address and name data. Think of fuzzy matching as a technique for identifying and linking records that aren’t exact matches, but are similar enough to be considered the same entity.
Although beyond the scope of this article, a more advanced approach uses a hybrid technique that combines phonetic and semantic similarity, a multipass system, to first narrow down a large number of candidates and then use a more accurate method for the final match.
Using AI and machine learning to organize payer details and create accurate payer masters is just one way these advanced technologies can be used to cost-effectively and significantly impact the revenue cycle and a healthcare provider’s bottom line.
In the next article, we will explore other ways in which advanced computing technology can bring additional efficiencies to the claims adjudication process. It is clear that the savings that can be achieved are of great benefit to healthcare organizations.
Ken Poray is the CEO of Integrex Health and has 20 years of experience in processing claims status, including EDI and web portal transactions.
This article is part 1 of a two-part series. Stay tuned for part 2.
