Fighting LTC insurance fraud with last century’s tools?

Fraud Sentry will change that.

If I had been talking about fraud last summer, I’d be citing US insurance fraud in the US, running at $80 billion. As an industry, we’ve been using this number for almost three decades.

We all knew it was wrong – probably wildly wrong. It’s no surprise to any of us that somehow, the insurance industry has gotten used to putting up with old, incomplete, unreliable fraud data – and was also struggling to find new ways to tackle the increasingly complex issues in this ballooning problem.

Then, in August 2022, the Coalition Against Insurance Fraud updated it to ‘at least’ $308.6 billion every year, as the amount that is stolen from American consumers.

The previous number was wrong for many reasons – it only included fraud in the property and casualty lines and a few other ancillary lines. Even more bafflingly, it had not even been updated to reflect inflation which would have swelled it to $152 billion in 2022.

Certainly, at illumifin, we’ve been thinking for a while about the ways in which we could revolutionize the fight against fraud in long-term care (LTC) – and it’s definitely a fight that someone needed to take on.

The nation’s LTC insurers paid out $12.3 Billion in claims during 2021, according to the American Association for Long-Term Care Insurance and there were $1.2 billion of fraudulent long-term care claims.

It’s a specifically difficult nut to crack in the LTC market, where we face unique, and very human challenges. Insurance fraud is a huge financial problem in all insurance lines, but specialty lines like LTC are currently underserved when it comes to fraud solutions.

Fraud detection is still highly manual and ineffective; human referrals are often too late; an vulnerable population poses additional unique challenges, and (back to my point about failings in data across the industry), there is a lack of industry fraud databases (from claims or providers) or other benchmarking.

Finally, LTC data is difficult to find, clean, visualize, and analyze in meaningful ways.

Let’s be really clear on this: claimants are victims. Providers take advantage of claimants, by drawing down their policy limits. Every dollar that isn’t applied to real care, is a dollar that could have been spent on care. And many of these policies concern vulnerable people, often senior citizens, so they are being put at risk of actual harm, not ‘just’ financial risk.

That’s something we knew we had to take on, and it became obvious that we should bring automated fraud detection to our suite of services, which is why we came up with Fraud Sentry, a product where carriers, fraud experts, and analytics intersect to fight fraud.

Carriers no longer need to worry about the cumbersome processes involved in identifying fraud. With Fraud Sentry and our expert staff, cases are identified proactively, and risk management stakeholders can feel confident in leaving these critical activities in the hands of LTC industry experts.

For carriers who don’t have the expertise or resources for development, it provides them with fraud analytics as well as augmenting carriers’ existing analytics with a slick user interface, cross-carrier insights, customized searches, and investigator-based data insights/visualizations.

We developed this to pair advanced technology with LTC fraud experts to surface fraud, waste and abuse (FWA) risks throughout the LTC claims cycle. The program automates fraud detection by combining a proprietary, machine learning- based risk scoring system, red flag alerts, and network analysis.

Through Fraud Sentry, we can now create a fraud database of known FWA cases, providers, agents, and claimants for the industry to leverage, while at the same time, solving the ‘false positives’ dilemma by providing advanced tools to help investigators find risky claims more quickly and more accurately.

When you think about fighting insurance fraud (and unfortunately, not enough people do) you might have the image of some guy hiding behind the bushes at the golf course parking lot, taking photos of a claimant carrying 30 pounds of clubs out of his trunk with one hand (despite attesting he can’t get out of bed unaided in the morning).

Maybe your fraud prevention team still includes the guy with the fedora and the turned-up collar in the parking lot. But shouldn’t it also include the tools developed by the data scientists and engineers?

— Jeffrey Ferrand · Vice President of Fraud Services at illumifin