Medicare Fraud: Has Anti-Fraud Predictive Modeling been Successful Thus Far?

April 6, 2012
By Brady & Associates on April 6, 2012 4:41 PM |

Last summer, the Brady & Associates Whistleblower Blog published an article that described the new Medicare fraud detection system implemented by the government to detect potential health care fraud before it occurs. The statistical modeling system evaluates large amounts of data and recognizes patterns that lead to fraud. By applying this model to individual claims, the system can evaluate the risk associated with each claim. The claim may then be flagged as a potential fraud item prior to payment. Other agencies have improved on their pursuit of healthcare fraud, which has resulted, according to the proponents of the system, in about $4.1 billion in taxpayer money recovered in fiscal year 2011.

Ted Doolittle, Deputy Director of the Center for Program Integrity (CPI), a new department under the Center for Medicare & Medicaid Services (CMS), answered questions about the modeling system in a recent interview with Beckers's Hospital Review. During that interview, the questioner asked Doolittle about the $4.1 billion recovery and his agency's focus. Mr. Doolittle pointed out that a large percentage of this figure represents money recovered from qui tam lawsuits. Qui tam suits are commenced by individuals, private attorneys and the Justice Department on behalf of the U.S. Government. Thus, not all of the recovery can be associated with the Center's efforts.

Procedures have been expanded to detect fraud by providers and suppliers. The predictive model is applied to determine risk. One of the goals of the modeling system is to screen providers likely to commit fraud. Getting the high risk providers out first is essential. Automated public records allow the agency to evaluate providers in the system once they have been enrolled. Any changes, such as changes in business affiliations, are automatically detected. Additional information can then be gathered. Early fraud detection makes it more difficult for questionable providers to get in and to stay in. This is different from how the Medicare system worked in earlier years. Previously, once a provider was enrolled, they had good standing until revalidation, which rarely, if ever, resulted in disenrollment. In addition, new procedures have been implemented to make it easier for good providers to enroll.

Medicare claims processing is another area with enhanced investigation. During this phase, fraudulent claims and providers making them identified. When a fraudulent claim is identified, it is essential that the provider of the claim be eliminated from the Medicare program. The modeling capability allows the agency to detect details about the claims and compare those claims against the algorithms in the model. Mr. Doolittle commented that no one in the health care arena is doing such a large-scale analysis and that CMS continues to add complexity to the system to generate more alerts. This should result in continuous improvement in the probability of identifying undesirable providers and eliminating them from the program for good.

Not everyone, however, is convinced that the new modeling program is a success. According to a recent article in the Huffington Post, the million dollar computer program has not given taxpayers a good return on their investment. Since its inception, the program has saved taxpayers just under $8,000 in claims. Proponents of the program, however, say that it is unfair to judge the soundness of the new modeling software off the small amount it has recovered. Doolittle said, "Suspending payments is only one way of stopping the money. . . There's lots of ways of stopping the money, and we are using them all. Looking at payment suspensions only - that's an unsophisticated view that doesn't give you a full picture of our activities." Other features of the new predictive modeling system must be considered to fully understand the benefit it is providing to the health care industry. Proponents claim that when those other benefits are factored into the analysis, the potential savings in the first six months could reach $20 million.

Others who find the program troubling highlight that the purpose of the predictive modeling program was to stop Medicare's previous "pay-and-chase" method of addressing fraud. Under pay-and-chase, claims were paid, and afterwards the government would attempt to track down and recoup the claims that were paid out erroneously. This was never an effective strategy. The new system allows Medicare to flag suspicious claims and investigate before the claim is paid to the provider. Some, however, do not believe that the new system has been successful in ending pay-and-chase. "The whole idea for creating this technology was they were going to be able to end pay-and-chase," said Hank Walther, formerly the head of the Justice Department's health care fraud division. "But we haven't yet seen evidence of its success."

Though the predictive program has only prevented $7,591 in fraudulent payments so far, it is important to remember that the system is new and contractors are still training those responsible for enforcement, proponents say. Increased familiarity with the system should improve the numbers. However, the ultimate goal of the system is to get rid of providers that are likely to commit fraud.

The proponents of the new system hope to expand it to impact hospital Part A and Part B claims processing. As new mathematical algorithms are developed, operators will be able to target specific types of claims or provider types, such as orthopaedic surgery or other health care sub-groups.

Sources:
Where Does Predictive Modeling Stand? Q&A With CMS Center for Program Integrity Deputy Director Ted Doolittle, by Bob Herman, published at BeckersHospitalReview.com, March 21, 2012.

Anti-fraud effort disappoints, by Kelli Kennedy and Ricardo Alonso-Zaldivar, published at HuffingtonPost.com, February 23, 2012.

See Our Related Blog Posts:
Federal Crack Down on Medicare Fraud Reaps Record Recovery of $4.1 Billion
$25 Billion Agreement Reached on Mortgage Fraud