Using big data and machine learning to reduce Medicare fraud

March 19, 2019

Fraud is a problem across industries, with fraudsters becoming more and more savvy as access to technology increases. It is an especially large problem in transactional environments, like banking, insurance, telecom and the like. It’s also a growing concern in the healthcare space, as Medicare fraud is thought to be responsible for losses pushing the $65 billion mark each year. The National Healthcare Anti-Fraud Association has an even higher estimate, believing losses due to fraud are closer to $80 billion annually, and others still put the estimate at closer to $200 billion.

Traditional methods to combat fraud have been largely unsuccessful in rooting out problems and identifying fraudulent behaviors. Aside from unintentional errors, such as coding mistakes, an article in Datameer says that the most common intentional fraud types include:

  • – Billing for services not rendered
  • – Billing for more expensive services/procedures, or “upcoding”
  • – Charging extra fees for things that should be included, or “service unbundling”
  • – Performing medically unnecessary services

Big data holds the promise of beginning to curb these losses. And in the healthcare arena this data is really, really big. The Datameer article says “Given the sheer volume of information out there and the disparate sources from where they came from, combing through the data for any signs of [fraud] is akin to looking for a needle in a haystack. The data is deep and the data is wide. We’re talking about millions, even billions of healthcare and claims records, with each record having up to 300 attributes.” Beyond the buzzword of “big data” is the very real opportunity to train machine learning models to discriminate between “normal” claims and “anomalous,” or potentially-fraudulent ones.

The only way to get a handle on data like this is through smart utilization of technologies like cloud computing and machine learning (ML). ML is an excellent tool for fraud detection, as it uses continually evolving algorithms to process massive amounts of data, identifying anomalies or red flags, and alerting its human counterparts for follow-up review or audit. Because the machine is continually learning, the more data it processes, the “smarter” it becomes at finding patterns and trends. Rather than manually poring over dozens of spreadsheets, a modern fraud AI system involves: 1) big data; 2) processed on cloud computers; 3) running machine learning algorithms.

According to a study by the Florida Atlantic University’s College of Engineering and Computer Science, machine learning represents a “breakthrough” that could lead to a huge – billions of dollars – impact on fraud. In the study, as written about in this Inquisitr article, researchers matched “fraud labels to the data, checking against provider details, payment and charges, procedure codes, total procedures performed, and medical specialty.” They turned the data over to the computer and let it compare data, and start to sort out unusual behaviors and flag them.  

The study was promising about the future of this technology in saving taxpayers significant costs each year that are related to fraud. This makes sense, and it’s easy to identify three sources of savings including: automating manual workflows, making existing human fraud experts 10-100x more efficient; flagging fraud before the claim is paid, avoiding costly and time-consuming litigation; and clearing a claim, avoiding the cost of an unnecessary claim audit. In all of these cases, the human expert still makes the final determination, but the AI does most of the “legwork,” serving up reliable and data-driven recommendations.

The field of artificial intelligence and machine learning is opening up access to advanced fraud fighting techniques across the board. We have more data available to us than ever before, yet many businesses don’t know how to use that data to their advantage. We’ve seen that big data is exactly what fraud-fighting AI systems need to do their job well, and such systems more than pay for themselves. This article in The Innovation Enterprise also illustrates AI’s impact on everything from retail to email spam, and from Medicare to tax evasion.

Delta Brain’s solution is grounded in AI, backed by experts with experience deploying this technology into multiple real-world industries from aerospace to finance. The robust pattern detection and predictive analytics systems that Delta Brain offers not only proactively match care with individuals and provide personalized risk-reduction when it comes to patient care, but are also excellent tools for fighting fraud. Human auditors equipped with a Delta Brain AI system are better able to pinpoint fraud within millions of claims, and ultimately this empowers them to save their organizations massive amounts of money.

Have Questions About Our Products?

Drop us a line and we’d be happy to answer any questions you have.

Contact Us
Read Next
As Value Based Care Models Gain Traction, Pressure to Adopt New Technology Grows

By making faster, more data-driven decisions to drive better patient outcomes, healthcare organizations are increasingly able to do the right

Read More
What does telemedicine have to do with logistics? New MedCity News article

Telemedicine has a lot of moving pieces. In many ways its challenges can be likened to the trucking or package

Read More
Modernizing our approach to managing the shrinking healthcare workforce

With healthcare facing a workforce shortage—ranging from physicians to non-clinical staff and everyone in between—it’s more important than ever to

Read More
The Role of AI in Value-Based Healthcare: Part One

We address this issue in a series of three posts on our blog, this is part one. Parts two and

Read More
Delta Brain featured in Becker’s Hospital Review

Becker’s Healthcare is the leading source of cutting-edge business and legal information for healthcare industry leaders. One of their key

Read More
Can #AI ease healthcare’s administrative burdens? Becker’s Hospital Review and Harvard Business Review think so (and so do we)

Becker’s Hospital Review (check out our the Delta Brain feature on their blog here), recently posted an article with this

Read More