July | August 2017



 

States Use Data Analytics to Reduce Medicaid Fraud

by Debra Miller
The June 16, 2015, Boston Globe headline that the state Medicaid program—called MassHealth— needlessly spent half a billion dollars was alarming to both supporters and critics of the $10.8 billion program.
State Auditor Suzanne Bump, first elected to office in 2010, concluded in the audit report that $233 million was spent unnecessarily when MassHealth paid providers directly for services that should have been paid by managed care organizations that receive a set amount to cover health care costs for their members. Another $288 million could have been saved if the MassHealth contracts with the managed care organizations were clearer about what services should have been covered.
“MassHealth has failed to fully realize the cost savings potential of managed care organizations,” said Bump in a press release at the time of the audit. “To avoid paying for health services twice, MassHealth must know exactly what it should and should not pay for. … I am pleased that MassHealth has pledged to use this audit to strengthen its operations.”
Bump is proud of the savings the auditor’s office has realized through substantial recoupments and criminal prosecutions, but more important to her are systemic improvements made in state program administration.
“Certainly, audits of providers punish wrongdoing, serve as deterrents to others and lead to recoupments, but the greatest fiscal impact comes about when the MassHealth program is able to improve its operations, for then the benefits are realized over the long term,” she said.
“Data is king in today’s world,” Bump said. “The more data that is accessible, the more effective auditors can be in making government work better.”
Major public benefits programs such as Medicaid are jointly funded and administered by the states and the federal government, but the states have the responsibility to identify instances of fraud and recover misspent funds. Recovered funds are returned to the federal government in proportion to their share of program funding. Administrative changes at the state level to prevent fraud and abuse in the first place save money for both the states and the federal government.
Bump described how her office uses data analytics to identify billing trends and anomalies that indicate potential billing fraud. Her staff has access to all operational and claims data in real time. Audit staffs don’t rely solely on claims data; they perform site visits to ask follow-up questions, review member files and request supporting documentation.
California has built one of the most comprehensive anti-fraud efforts for Medicaid of any state in the nation. According to Karen Johnson, chief deputy director of the Department for Health Care Services, the administrative agency for Medicaid, her agency will continue to work aggressively to develop new techniques to prevent and identify provider and beneficiary fraud.
“(California Medicaid) has adopted a ‘zero tolerance for fraud’ philosophy,” Johnson said.
DHCS is expanding its use of data analytics. Johnson pointed out that under the old model of chase-and-pay when you tried to recover the money paid incorrectly, there was a chance that the criminal had already spent it before the state could get it back.
Front-end, anti-fraud measures are consistent with new program integrity requirements of the Affordable Care Act. These efforts in California include pre-enrollment investigations as well as provider outreach and education. One of the key elements of the pre-enrollment and re-enrollment efforts is a background check and onsite review of providers prior to enrollment to ensure the legitimacy of providers before they are admitted to the California Medicaid program.
California regularly conducts error studies to protect against billing mistakes by fee-for-service providers and minimize the risk of overpayment. The studies are used to monitor emerging fraud trends, make informed decisions on the allocation of fraud control resources and identify program areas most at risk of provider payment errors.
Data analytics “can help us concentrate our efforts so that we may readily identify issues and suspend payments, which in turn means the system can make sure more of its resources pay for the care that people need,” Johnson said.
Johnson also said that good work in this area isn’t cheap. She said DHCS is constantly striving to maintain an adequate staff with sufficient training and skills to keep abreast of the latest fraud trends and utilize the latest technologies and techniques to detect and eliminate fraud, waste and abuse.
John Dougall, state auditor in Utah, stressed his role as a steward of taxpayer funds, whether they are state or federal funds.
“We have the responsibility to ensure that welfare program benefits are going to the right people for the right purposes,” he said. “It is important that only those eligible receive benefits.”
Dougall said his office is currently working on a performance audit of Medicaid. Like Massachusetts, the state of Utah is moving away from a fee-for-service model.
“The old incentive was to treat more people and bill for more services. Utah has gone with ACOs (accountable care organizations) to establish a new incentive for more accountable care,” Dougall said. “But people will figure out how to abuse that system as well. We hope to balance that with our audit efforts.”
Accountable care organizations, or ACOs, are voluntary networks of doctors and hospitals that provide coordinated care to patients. The ACO is paid a set fee and the ACO members share in the financial risk of providing unnecessary care and reap rewards from preventive and coordinated care that saves money.
Techniques used to abuse programs are getting more sophisticated, Dougall said, but data analytics can help states stay ahead of abuse patterns and trends.
His office recently completed a performance audit of the food stamp program—known as SNAP—that used data analytics to proactively identify and investigate potential misuse of SNAP funds. The audit concluded that the Utah Department of Workforce Services, the administrator of the SNAP program, could —and should—use data analytics to identify SNAP recipients who make the majority of their purchases out of state, request excessive replacement electronic benefit transfer—or EBT—cards, and make frequent or large even-dollar transactions.
“Data analytics are what credit card companies use,” Dougall said. “If you purchase several big ticket items on the same day with your credit card, or make a purchase out of state, you are likely to get a call, or maybe even have your card frozen.”
“It shouldn’t be any different with state government services.”