For consumers, the credit card fraud prevention experience is seamless, which leaves many health payers wondering why they can’t access the same technology. The short answer? It’s complicated.
It’s easy to see how health payers, who have become disenchanted with promises by technology suppliers to “fix fraud,” find credit card fraud models alluring. As a consumer, it’s simple: Your credit card company keeps a record that identifies a specific purchasing behavior. Whenever there’s a deviation from that, the issue escalates and you receive a text along the lines of “[first name], did you just spend $XXXX at [store name] in [location]?” Your simple “yes” or “no” response can either resolve or further escalate the issue and prevent further damage if your card has indeed been compromised.
“Credit card companies use a combination of technology and humanity to fight fraud, employing automated fraud detection algorithms across massive amounts of data collected from millions of customers and hundreds of millions of cards. Once a transaction is flagged as a possible problem, humans can follow up, contacting the customer. Even this is being increasingly automated, with some card holders receiving texts asking them to verify a suspicious transaction on their account.”
We know as a health payer you hear it all the time: “Here’s the latest and greatest solution to hit the market! It’s going to drastically reduce fraud!” You’re a savvy consumer yourself, so any proposed solution is likely met with a healthy dose of skepticism. Still, you may see the benchmarks set by other industries and wonder, “Why can’t our health plan’s fraud solution be as effective as those offered by my credit card company?”
It’s a great question and one worth asking, particularly as we see a bit of fallout from overused and overhyped tech buzzwords like “AI” and “machine learning.” But as an industry expert we work closely with explains, the reason healthcare fraud tactics differ from other industries, particularly finance, is complicated. That’s because healthcare is vastly more complex than the payment solutions industry.
How Credit Card Fraud Analytics Work
As we described above, the consumer experience around capturing and preventing credit card fraud is perceived as easy and effective. But let’s look into the business aspect of how credit card companies manage fraud, which is generally a model that healthcare companies cannot directly follow.
High false positive rates in your SIU would likely have you unnerved, but did you know the credit card industry averages an eye-popping 90% false positive rate? While your department’s high false positive rate may send costs soaring, the cost of such a high false positive rate for credit card companies is… hardly anything. Credit card companies are able to automate this activity and stay pretty hands-off because the analytics are simpler.
Health payers have vastly different variables to consider when analyzing healthcare fraud making it incredibly complex. So complex, in fact, that health plans will have a hard time ever moving away entirely from utilizing human intelligence to work claims. With so many factors at play, health technology’s true focus has to be putting forth better leads so that fraud analysts don’t waste their time pursuing red herrings. As many SIU leaders will tell you, technology solutions they see today are flagging claims that can easily be dismissed by human eyes.
This blindspot is in large part due to a siloed approach and outdated technologies, according to ClarisHealth VP of Program Integrity Mark Isbitts. “It’s understandable that health payers are frustrated with current fraud solutions on the market today — they’re falling short and lack the robustness seen in other industries.”
Mastercard is entering the Health Fraud Detection space… is that a good thing?
The credit card giant has financial expertise, but can they embed themselves in the complex world of healthcare so easily? It looks like this development is, realistically, a technology that can plug into a larger more healthcare-specific fraud solution to power data security and more robust analytics, says Isbitts.
Mastercard itself says that, in addition to security that is aimed at preventing data breaches, they “will also offer products aimed at using AI and machine learning to help payers curb fraud, waste and abuse, as well as predictive analytics to enable providers to use more effective billing strategies and improve their revenue cycle management. It’s an extension of similar products already offered to other industries, including major fast-food chains and large financial institutions.”
As we’ve shown in this article, healthcare fraud is substantially more complex than the type of fraud that plagues financial institutions, and fraud as a whole is a smaller problem than waste and abuse. Still, improved analytical capabilities and a tightening of security is “a very good thing for the payer space,” says Isbitts. Read the article here.
Why Fraud Detection is Harder for Healthcare
Health payers play a high-risk game if they allow too many false positives to trickle down into cases. While in the finance industry consumers aren’t typically bothered by credible inquiries into their spending patterns, providers on the other hand feel threatened by too many inquiries into claim legitimacy. As consumers, being checked up on by your credit card company makes you feel secure. Not the case in healthcare, where the already delicate balance payers have with providers can easily be thrown askew by a false allegation.
On the surface, using credit card fraud detection as a baseline works, say industry experts. However, though the model is correct, the challenge is looking at the underlying analytics, which are different. Unsupervised methods to track fraud in healthcare come at a high cost. Fraud scoring, as it’s termed, has to be determined differently in healthcare and currently, solutions are falling short both in terms of technology offered and the promises made.
The current hit rate for advanced technologies is somewhere between 32-35%. Advanced technology, like machine learning and predictive analytics, can only work as well as the data that feeds it. Credit card fraud technology can work within simpler frameworks and run effectively with a high false positive rate all while boosting customer satisfaction. That’s not the case for healthcare.
Consider this: if a health plan’s fraud detection solution looks at medical history and flags a probable FWA case, there are variables that AI logic may struggle with. Let’s say that a male visits an OB. Your fraud system may flag that claim as fraudulent, based on an isolated incident, without the ability to intelligently link a potential gender change from the medical record. Because of HIPAA and other factors, health plans cannot simply text a patient to ask them if this was a legitimate episode of care.
Yet, there are ways that advanced technology solutions can overcome these perceived stop points. ClarisHealth is currently collaborating with a team of fraud analytics experts to develop a modern fraud detection solution that can be robust and effective for health plans.
The Future of Fraud Detection for Payers
Advanced analytics is the future of healthcare fraud detection and prevention, but healthcare is complex enough that human expertise, intervention and detection is the gold standard. For a fraud solution to be effective, it has to take into account myriad data points and overcome data inconsistencies to persist in providing an investigator with a higher rate of accurate leads. We term this approach Total Payment Integrity and it’s essentially the ability to connect data points across platforms for single-source of truth.
Payers’ short-term goal is to eliminate data silos and decrease case time for the SIU, both of which can be accomplished by reducing dependency on outdated processes. Also crucial is the ability to have clearer insights into details surrounding providers, who account for the majority of fraud cases. These details may include a better, more comprehensive understanding of a provider’s relationship with other practices and improved geographical metrics to quantify certain claims.
An ideal — and achievable — fraud solution likely consists of a series of integrated data points with the application of advanced analytical concepts (AI) that look more comprehensively at information than current systems allow. To deliver this, health payers will need to be able to rely on a technology provider that is agile — someone who can develop dynamic solutions that stay abreast of modern schemes while also expanding basic functionality and improving “simple” shortcomings.
NOW'S THE TIME FOR TOTAL PAYMENT INTEGRITY
Talk to ClarisHealth about how Pareo® can improve your health plan’s payment integrity processes while preparing for broader goals of interoperability, real-time data sharing, and more.