In the health policy arena, information technology is seen as a panacea to improve efficiency, reduce repeat testing, and save billions. While those are some expected benefits of electronic medical record-keeping, its greater money-saving potential may lie in helping to uncover fraud.

The National Health Care Anti-Fraud Association (NHCAA.org) conservatively estimates that 3 percent of all health care spending--about $68 billion--is lost to fraud. A more jaundiced, or perhaps realistic estimate by the FBI and CDC puts the rate at 10 percent, a crippling $226 billion loss yearly.

Suddenly the urgency for electronic medical record adoption is in sharp focus because the best way to spot fraud is by using high-tech data mining tools.

One company at the forefront of this effort is HealthCare Insight, South Jordan, UT. HCI explained to DOTmed News that medical identity theft is a multi-faceted problem. To begin with, in a nation of nearly 50 million uninsured the temptation is great for families to misrepresent who is eligible for insurance, a well meaning but nonetheless illegal activity.

"The intent might be just to get somebody coverage. The fact is that [the health insurance policy] wasn't underwritten that way. So there are perpetrators who add others falsely or give their card to a family member or someone else to get care, and it is still fraud," said Joel Portice, COO of HCI.

While health care consumers are sometimes the perpetrators or victims of medical ID theft, all too often the bad actors are not the patients but health care professionals.

"The whole idea around stealing identity is not only on the patient side but it's also on the provider side where billing schemes use a legitimate provider's identification number and name but set up a phony address," Portice said. Once the scheme is in place, reimbursements can be diverted to the bad guys.

HCI combats the problem for public and private insurers by using sophisticated software tools that sift through claims looking for something amiss. Rule-based technology and predictive analytics identify patterns and schemes that indicate fraud -- a kind of virtual magnifying glass.

"Our approach is to look at fraud from the perspective of it being perpetrated by a provider, or using their provider identification information; for example somebody stealing their ID," Portice said. "We look at profiles and patterns of provider data. We process millions of claims each month to establish those patterns and scores that give us some kind of probability of fraud risk."

Red flags include the obvious such as claims on behalf of deceased patients, to more subtle address information or mismatching medical information -- telltale signs that something may be very wrong. "It could be a spike of claims, the type of claims, the types of patients, the frequency and date on which these things occur. There are a number of different schemes that indicate some kind of risk."

While Portice will not give details on the company's proprietary technology, the company's software and services define and find these patterns. In some cases, provider information is stolen and misused unknown to the doctor, such as for gaining prescription drugs under false pretenses.

"In that case the provider is innocent. But if the provider is falsifying data in records and submitting information then he is the perpetrator," Portice said, acknowledging that this is the most common scenario.

The Bush administration and now the Obama administration have both placed a priority on shifting to electronic health records. Considering the potential to use software tools for fraud monitoring by public insurers, the urgency is clear.

"Electronic records will help tremendously not only in the administration and reduction of medical errors but in the identification of potential ID theft," Portice said.

To prevent medical ID fraud, he had these suggestions for health care providers: Be careful about how you release your identification information. And review remittance statements from payors to ensure the claims and amounts are consistent and reconciled in your own practice.