Predictive analytics still has a long way to go in preventing healthcare fraud
Several weeks ago, I was talking with Jeff Baird, a healthcare attorney with Brown & Fortunato P.C. in Amarillo, Texas, about the recent spike in compounded pharmaceutical fraud when he uttered a familiar phrase.
"I don't know if you've ever gone to a pizza arcade where you have the Whac-A-Mole--do you know what I'm talking about?"
I could already see where he was going.
"What'll happen is the government is whacking this mole," Baird said. "Then there will be another code that will pop up and people will chase after that. It's your natural cat and mouse game between the private sector and government regulators."
Whac-A-Mole? Cat and mouse? Does anyone have a tail we can chase?
The compounding pharmaceutical industry is the latest sector to absorb blows from that infamous padded hammer. What's discouraging is that it still seems to be the government's primary fraud-fighting tool. In fact, the latest rash of fraudulent claims from the compounding pharmaceutical industry underscores an approach that is still decidedly more reactive than proactive.
Compounding pharmacy claims are making headlines now because officials have linked bogus compounded creams to half a billion dollars in fraudulent payments. But the signs were there several years ago. In 2013, the second-largest insurer in Massachusetts, Harvard Pilgrim Health Care, dropped coverage of compounded pharmaceuticals, in part because of a 171 percent increase in claims, and a three-fold jump in costs, according to The Boston Globe. In 2014, Express Scripts announced it would stop paying for more than 1,000 ingredients after seeing similar spending spikes.
In other words, the fraud risks associated with compounded drugs were no secret.
With fewer options, "code-killers" (Baird's term for providers that overbill one vulnerable reimbursement code), put the screws to Tricare. But the signs of abuse were already there. By the end of fiscal year 2014, Tricare's annual spending on compounded pharmaceuticals had increased five-fold in three years, from $95 million in 2012 to $514.4 million in 2014, according to statistics provided by Tricare to FierceHealthPayer: Antifraud.
By November 2014, Tricare was scrambling to respond to these spikes, but, as with most government programs, the changes were not exactly swift. By May, Tricare implemented a new screening process that eliminated medically unnecessary prescriptions, effectively bludgeoning the mole. But during the five months in between, Tricare spent nearly $1.5 billion on compounded pharmaceuticals. Now Department of Justice investigators are tasked with the onerous, virtually impossible task of clawing back that money.
There is a lot of talk about how predictive analytics can save the healthcare industry from itself, but this recent, incredibly expensive scheme is a blunt reminder that the transition toward a preventive approach to fraud is still in its infancy. Stopping improper payments from going out the door is still a pipe dream, particularly when it comes to government-run programs.
That's not to say there aren't pockets in which predictive analytics is making its mark. The chief data officer at the Office of Inspector General says analytics played a significant role in last year's national fraud bust, although she admits the government is still "trying to develop new approaches to identify unknown, undetected and emerging patterns." Certain private payers, like Aetna, have implemented data models that identify sophisticated schemes and cut down on the amount of time fraudsters have to accumulate payments.
But that's exactly what happened with compounding pharmacies that were overbilling insurers over the last several years: They were given time to accumulate payments while investigators fumbled for a hammer.