Diamonds from Barbados: The possibility of real-time fraud prevention

My credit card company called me awhile back asking if I had just bought diamond jewelry in Barbados. With powerful analytic subtlety, the bank's anti-fraud unit flagged this purchase as a tad aberrant for someone who normally buys, say, pot roast at Stop & Shop. I told them it wasn't me, they stopped the transaction and reissued my card. It was a sweet piece of real-time fraud prevention.

Have you wondered why we don't do likewise in health insurance, why we're not finding and stopping fraudulent and abusive claims for the whole customer base before payment?

Financial professionals are quick to ask this. They're also quick to point out that payers have been slow to use biometrics to fight identity theft. Insurers are seen as trailing other industries in leveraging technology to solve business problems. As Aetna CEO Mark Bartolini said, "it's difficult moving the model."

Thirty years ago I learned that retrospective review is slow and expensive, and that we need more prepayment monitoring to cut financial loss. Experts still say that: "If fraud goes on for a period of time--say it took a year to identify by traditional means--there's lots of money that's been paid off and the potential to recover that money is frequently very slim," Michigan Blues' Doug Cedras told FierceHealthPayer in an exclusive interview.

But special investigations unit professionals tell me wistfully that they don't see how anti-fraud technology similar to what banks use can apply to health insurance.

Certainly there are obstacles. How would a solution like this work for out-of-network provider claims, for instance?

And some fraud schemes aren't apparent at the individual claims level, which is a giant fur ball in the cat's throat. A subscriber receiving high doses of OxyContin for long periods, for example, may or may not be a problem; but a medical practice where most patients fit this pattern is worth a closer look. Is there an anti-fraud solution advanced enough to see how individual claims fit into a larger picture?

Another issue is that claim deviations can be appropriate. Think of a subscriber who's been healthy all his life but then develops a disease, causing all kinds of specialty care claims to show up for the first time in his history. How would a solution similar to what's used in banking adjust for that?

Finally, what if there's an illness epidemic? Could a system to stop aberrant claims before payment recognize and adapt to that?

So I shake my head wistfully, too, at the thought of a front-end solution that stops fraud in its tracks the way banks do. Besides all my questions, maybe I've heard too many exaggerated claims from vendors.

But I keep returning to the idea that fraud requires a "yes" from someone or something on our side of the table. That puts us--at least initially--in a position of power. Before anyone defrauds an insurer, we--meaning payers, the government, employer groups, or claims and enrollment processing systems--must first allow payment, approve a participating provider application, allow enrollment of individual or group health plan members. If only payers had the ability to say yes to the right claims and people and no to the wrong ones in real time.

All of which brings me to something my best boss taught me. When faced with a difficult task, many at my former company would throw their hands up and say "it's impossible;" but he'd scoff in their faces. "I don't believe that," he'd reply reflexively.

He's a leader, and his words moved us off facile answers. We dug deeper for solutions. We became better problem solvers and realized that while my boss wasn't always right, more was possible than we supposed.

It's been said that while technology isn't a silver bullet, it is a mechanism for positive change. So I hold out hope for real-time fraud prevention in health insurance, though I know it won't come easy. But the rewards for fraud fighters could be as fabulous as diamonds from Barbados. - Jane (@HealthPayer)