During a congressional hearing last week, an exchange occurred between a congressman and government officials 
that’s happened dozens of times before. The congressman asked a relatively simple question: How much fraud is taking place within government programs?
And the government officials repeated the answer that’s been offered just as many times in responses: We don’t really know.
In this instance it was Rep. Gerry Connolly (D-Va.) pressing the issue. Of the $136.7 million in improper payments reported across the federal government in fiscal year 2015, he asked, how much was tied to fraud?
“We do not have a commonly accepted methodology measuring the extent of fraud,” said Sheila Conley, deputy chief financial officer for the U.S. Department of Health and Human Services.
“I don’t know and I don’t want to guess,” said David Mader, controller of the Office of Federal Financial Management at the Office of Management and Budget.
A representative for the U.S. Department of Treasury merely shook his head.
“I’m frustrated by this because I don’t know how we devise strategies that try and get at this if we’re not willing to put some percentage or number [on it],” Connolly replied.
Connolly is echoing an ongoing frustration that has plagued these types of hearings for several years. Shantanu Agrawal, M.D., the deputy administrator and director for the Center for Program Integrity at the Centers for Medicare & Medicaid Services (CMS), fields this question routinely, usually with the same basic response: Improper payments are a poor measure of fraud.
“The improper payment rate is not a true measure of fraud; it is really more a measure of perhaps waste and abuse,” Agrawal said during a congressional hearing in 2014, noting that the primary driver behind the improper payment rate is insufficient documentation.
He made the same point earlier this year as lawmakers grilled him on CMS’ rising improper payment rate.
“I can’t give you an actual estimate of fraud either as it relates to the improper payment rate or on its own,” he said. “There just isn’t a validated statistical methodology for generating a fraud rate.”
In years past, Agrawal’s predecessor, John Spiegel, offered a similar response. During a congressional hearing in 2011, Spiegel refused to pin down a number, even as former Florida Rep. Cliff Stearns pressed him, referencing a 2009 "60 Minutes" report that estimated Medicare fraud totaled $60 billion a year.
The problem with calculating fraud, as numerous fraud experts and just about every government official faced with this question have pointed out, is that it’s difficult to define, and even harder to parse out from what might be classified as innocent errors--a forgotten form or insufficient documentation. Officials say these make up the majority of improper payments, and lack intent, a defining characteristic of fraud.
True fraud, on the other hand, is stuck in some kind of measurement purgatory. No one is willing to commit to one number or percentage, and the only measure we do have--improper payments--isn’t all that good, which leads to a whole lot of shoulder shrugging.
The problem is that a lack of any measurable fraud statistics grinds these discussions around improper payments to a screeching halt. In the absence of a definitive starting point, antifraud efforts are, in some respects, rendered rudderless.
CMS is equipped with vast troves of data that it has integrated into its Fraud Prevention System, a program that the agency claims saved $42 billion over a two-year period. That’s an estimate that carries its own methodological baggage, but that didn’t stop CMS from proudly announcing it.
As difficult as healthcare fraud is to track, it’s hard to believe that CMS is utterly helpless to come up with even a vague estimate. Existing fraud data the agency is using to prevent payments, coupled with the Department of Justice’s newfound interest in data analytics, could serve as a starting point.
For example, how many improper payments lead to fraud referrals or investigations? How many providers are classified as billing outliers, and how many of those lead to fraud convictions? These measurements are certainly imperfect, but that doesn’t mean they are worthless.
On the bright side, that elusive figure might not be that far out of reach. During the hearing in May, Agrawal hinted that the agency is currently working on a “pilot approach” to measuring fraud. If that figure is finally unveiled, perhaps we’ll all be able finally put this perpetually nagging question to rest, and move on to more precise fraud-prevention efforts. -- Evan @HealthPayer