The $111,705 the state Medicaid system gave to Crystal Edgington bought a lot of things. It went to a home on a golf course in Sequim, Wash. It went to Seahawks season tickets. All that money was supposed to provide in-home care for her mother. Instead, from May 2008 to January 2012, she illegally transferred it to her father.
They were caught, condemned in press releases from the Attorney General's Office and sentenced to 30 days in jail for Medicaid fraud. And her case is just a drop in the bucket.
"The units of measure for losses due to health-care fraud and abuse in this country are hundreds of billions of dollars per year," Harvard health-care fraud expert Malcolm Sparrow testified before Congress in 2009. "We just don't know what the first digit is."
He predicted that by taking health-care fraud seriously, the nation could save as much as 10 to 20 percent of what it spends on Medicare and Medicaid. Today, with the number of those on Medicaid dramatically expanding, the risk of fraud — on the part of everyone from doctors to petty thieves — may be even greater.
Traditionally, Washington state has been on the forefront of anti-fraud innovation; it was one of the first states to track down Medicaid fraudsters using data analysis. In spring 2012, the state Legislature nearly unanimously passed a bill, sponsored by Rep. Kevin Parker (R-Spokane), to explore using similar technology to prevent fraudulent payments. Parker believes it could immediately save the state millions, but the Washington State Health Care Authority — with the power to do something about it — is skeptical, and so far hasn't been willing to take the next step.
Medicaid fraud cases can take years and years to wind through the court system, so it's better to not pay the crooks to begin with. Parker's bill tasked the Health Care Authority to seek out contractors who could provide a "predictive analytic" system. Using algorithms similar to those banks and credit card companies use to detect credit card fraud and identify theft, they'd analyze data to automatically detect and prevent suspicious transactions.
"We would be the first state in the country leveraging technology in the private sector and applying it to our public sector approach," Parker said in a February 2012 video message. Relatively conservative estimates, he says, indicate that preventing more fraud could save Washington as much as $300 million per biennium. Similar programs have saved many times more than they've cost.
"It would buy a heck of a lot of teachers," he says. His bill passed both houses with only a single vote — a Republican — against it. In a 2012 fundraising email, Parker cited "aggressive steps to stop Medicaid fraud" as one of his major accomplishments.
But after more than a year passed with little apparent progress, it seemed to Parker that the HCA was dragging its feet. At the beginning of this month, he finally heard: The plan to use predictive analytics has been dead since fall 2012.
"If I would have known, I would have been calling them every single week, gently nudging them along," Parker says. "I trusted them to get the job done."
While the HCA sent 19 vendor requests for information and analyzed their responses, it never went any further.
"Really, it was pretty clear to us once we got all of that together — the technology really shows some promise, but it's not really a mature process at this point," says Cathie Ott, HCA division director. "We didn't consider it to be cost-effective."
The law encouraged the state to issue "requests for proposal" after requesting information, but only if vendors could save the state money without adding cost or delays to the state Medicaid system.
Ott says the HCA vigorously examined vendors' responses, but says many were vague about the anticipated costs and benefits. Some of the vendors focused on detecting fraud after it had occurred, instead of preventing it in real time. Many relied on vastly different definitions of "predictive analytics." (Sparrow, the health-care fraud expert, dismisses "predictive analytics" as mostly a marketing term for slightly more sophisticated analysis of data.)
Parker's disappointed. He says it's not unusual for legislation to run into trouble when it hits the state bureaucracy, but he trusted the HCA, giving them the flexibility they requested in his bill.
"The Health Care Authority did not follow the intent of the legislation," Parker says. "They did the bare minimum."
He doesn't buy the notion that the technology isn't ready. Parker says he recently contacted two of the vendors — LexisNexis and Emdeon — and they told him that despite HCA's skepticism, the technology exists and could be swiftly implemented.
"LexisNexis assures me that they could have their version adapted to [Washington] state's existing system and have it up and running in no more than a month," Parker writes in an email.
LexisNexis gave scant specifics to Washington state on how much its program could save, according to HCA's assessment document. Emdeon provided significantly more detail, but HCA still wasn't satisfied.
"They did not go into any specifics about their cost model," Ott says.
Beyond looking over the initial answers, however, the agency never sought out those specifics. Asked whether HCA contacted Emdeon or any other vendor for more information, Ott says it hadn't.
Louis Saccoccio, executive director of the National Health Care Anti-Fraud Association, says the federal government and state agencies nationwide have been moving toward using data models to prevent fraud, instead of just detecting it. Starting in May, he says, states have been allowed to use federal funds to data-mine for fraud.
After contracting with LexisNexis to create an identity fraud pre-verification system for public assistance programs like Medicaid, Florida has already saved far more money than it anticipated. In September, Massachusetts launched a Medicaid fraud-prevention program to automatically identify suspicious claims before reimbursing them.
But Ott says that Washington, unlike Massachusetts or the federal Medicare system, relies increasingly on separate "managed care" organizations to handle the ground-level details of Medicaid, making statewide analysis of fraud difficult.
Parker isn't deterred. He says he's willing to keep working with HCA, but there's a lot of money the state has missed out on by not using the technology.
"I am prepared to run further legislation if the HCA does not follow through on the true intent of the legislation to force their hand for the sake of taxpayers, the poor and the most vulnerable," Parker says.♦