On July 1, CMS will begin using predictive modeling technology to help fight Medicare fraud, Health and Human Services Secretary Kathleen Sebelius and agency officials announced.

PHILADELPHIA — On July 1, CMS will begin using predictive modeling technology to help fight Medicare fraud, Health and Human Services Secretary Kathleen Sebelius and agency officials announced Friday.

By analyzing the 4.5 million claims that pass through its systems daily, Medicare's new state-of-the-art analytics will be able to identify potential fraud on a nationwide basis instead of evaluating one claim at a time, Sebelius said. Similar to that used by credit card companies to spot fraud, the technology should move CMS beyond a "pay and chase" approach, she said, by helping to stop fraudulent claims before they are paid.

"Medical data is in lots of different pots," Sebelius told reporters at a summit on health care fraud in Philadelphia. "We wanted to be able to spot the doctor claiming to be in six cities billing for the same procedure on the 16th of June."

Funds for the project came from the Small Business Jobs Act of 2010, which provided $100 million for the implementation of a predictive model approach, according to Peter Budetti, director of the CMS Center for Program Integrity.

CMS awarded the contract for developing the system to Northrop Grumman, which partnered with National Government Services (NGS) and Federal Network Systems, owned by Verizon. The technology will allow CMS to check Medicare's data mass across claims and a wealth of other information such as enrollment records and stolen provider and beneficiary identification numbers.

The system algorithms will check original claims when they are submitted — by beneficiary, provider, service origin or other patterns — to identify billing patterns and evaluate the validity of each claim, according to the agency. Potential problems will be flagged, and an "alert" and "risk scores" for those claims will be assigned. The alerts will allow CMS to prioritize claims for additional review and investigative or other enforcement actions.

The risk-scoring process will build over time as additional factors are added in, Budetti said.

Sebelius said that in a pilot for the new program, CMS linked public information available from court records, addresses, medical licenses and lists of providers and suppliers excluded from federal health care programs.

"Today's contract represents the greatest scrutiny ever applied to Medicare's payments," she said. "Suddenly, it's a lot harder for the rotten apple to blend in with the bunch."

Read a press release from CMS.