cms fraud


Big Data And Beating Medicaid Fraud

, , , ,

cms fraud

Helen Sieger used to own a nursing and rehabilitation home in the Bronx, not far from the Hudson River. She ran the home from the mid-90s until 2009 when she was arrested for medical misdoings. She was engaged in a criminal partnership with a hospital social worker named Frank Rivera.

Sieger would pay Rivera $300 for each patient he referred to her facility with a bonus of $1,000 for every tenth one he would throw her way. To add insult to injury, she was fraudulently claiming money from the Medicaid system.

After she was arrested, she skipped bail and fled to Miami where the authorities eventually tracked her down in a hotel, but she died in custody in 2011 before any repayments could be made.

But in June, New York State managed to claw back $2.5 million from Sieger’s estate, heralded as proof that New York State won’t ignore medical fraud.

“When we have a case involving a criminal scheme that robs Medicaid, our prosecutors will do whatever it takes to restore those stolen funds–whether that criminal is alive or we’re forced to settle with their estate,” said New York’s state attorney general Eric Schneiderman in a statement.

The government can hardly be blamed for reveling a while in this landmark victory because “it’s very hard to get money back and when you do it’s usually cents in dollar,” Markus Fromherz, Xerox’s chief innovation officer in healthcare, told HealthBiz Decoded.

“There’s a lot of pharmacy fraud,” says Fromherz.

It’s estimated that medical fraud in the U.S. totals 75 billion, that’s roughly the GDP of Ecuador. Add in waste and unnecessary or inefficient services and that figure jumps to somewhere between $130 billion and $210 billion. Crucially, Fromherz says, “fraud is criminal and means there’s an intent. Waste just means they’re careless.”

State governments by and large outsource the complex process of Medicaid fraud detection to companies with information management expertise, like Xerox, which has successfully implemented Medicaid Management Information Systems approved by CMS in 31 states and counting.These information systems use sophisticated search algorithms to detect suspicious claims submitted by providers.

Repeat prescriptions for drugs are often written up by a physician in batches of three but the patient may end up only needing two of those refills – and Fromherz says one of the most common fraudulent activities is for pharmacies to bill Medicaid for the third prescription even though the patient didn’t ask them to do so.

Xerox uses a rules-based system to detect a skeptical looking claim. “There are millions of records that have to be combed through,” says Fromherz.

If a pharmacy or a doctor’s office breaks one of those pre-determined rules consistently, then the system will flag it.

“You’re looking for duplicated billing or for pharmacies that use repeat prescriptions and always use all of the refill options,” explains Fromherz.

The trouble with this technique is that you’re relying on past patterns to detect fraud, and felons have a talent for being intelligently inventive and finding new ways to cheat the claims procedure.

“Criminals literally test the system. They put claims through to see if it works,” says Fromherz.

That’s why there’s been a move away from the sole reliance on rules-based systems towards a detection method that uses predictive behavior.

“Descriptive looks back and predictive looks forward,” says Fromherz. “Xerox uses both descriptive (rules-based) and predictive detection.”

Xerox is working to implement their fraud detection before the government pays a claim, rather than retrospectively inspecting records.

Fromherz advises Medicaid and private insurers not to pay out if they suspect a claim is fraud, but to send the bill back and say “prove this isn’t fraud.”

“It’s better to just not pay for fraud in the first place,” he concluded, “rather than get involved with lengthy and expensive court cases with criminals like Sieger.”