The House Ways and Means Subcommittee on Oversight recently held a hearing on the federal government’s use of data analysis to confront Medicare Fraud. The hearing featured testimony from Dr. Shantanu Agrawal, Deputy Administrator and Director of the Center for Program Integrity at the Centers for Medicare and Medicaid Services (CMS) and Gary Cantrell, Deputy Inspector General for Investigations at the Office of Inspector General (OIG).
Medicare fraud “remains a serious, evolving threat,” Ways and Means Oversight Subcommittee Chairman Peter Roskam (R-IL) began the hearing. “I want to emphasize just how big of a problem this is. Last year, the federal government lost $124.7 billion dollars in improper payments across 124 programs. Of that $124 billion, one program accounted for $60 billion—or nearly half of the losses: Medicare.”
Pay and Chase
“Historically, CMS has used a method called ‘pay and chase’ in processing Medicare payments, first paying for a charge, and then later looking back to check on the validity of the transaction and potentially trying to claw back the money if the payment was made improperly,” Roskam stated. “As you can imagine, that strategy isn’t very effective. Time and again we have seen fraudsters bilk the system for a few million dollars, shut down, and pop up under a new name to run their scams all over again. The Medicare program is getting outsmarted by these methods and the proof is in the unacceptably high rate of improper payments each year.”
Kirk Ogrosky, a former prosecutor and now partner at Arnold & Porter LLP, provided a concise overview of the issues at hand in his testimony. “Fraud will not be reduced or eradicated with a ‘pay-and-chase’ enforcement system that relies on criminal prosecution and civil litigation,” Ogrosky stated. “With advances in the ability to analyze claims data, the goal of the system should be to detect fraudulent claims when they are submitted, identify the perpetrators, and to use prosecution sparingly to punish and deter.”
Rep. Roskam stated: “In 2010, I proposed a new approach to help CMS work smarter. Instead of ‘pay and chase,’ CMS should use the same kind of cutting-edge predictive analytics technology that private companies use successfully to look at transaction data in real time and identify potentially fraudulent charges—stopping the payment before the money goes out the door.” This is similar to what private credit card companies use to identify potentially fraudulent charges and stop payments while they further investigate claims. Indeed, the panel heard testimony from Visa, a private company whose “global rate of fraud is 6 basis points—meaning 99.4 percent of the $10 trillion dollars in payments it processes globally are fraud-free.” Roskam added: “That’s quite an impressive track record, and one we hope to learn a thing or two from.”
Big Data: The “Fraud Prevention System”
The system created by CMS to incorporate data analytics to protect Medicare is called the Fraud Prevention System, or FPS. “In its first year, FPS got off to a rocky start—the Health and Human Services Inspector General could not even certify any of the system’s results,” stated Rep. Roskam. In its second year, ending in July 2013, the Inspector General certified that the system had returned one dollar and thirty-four cents for each dollar invested that year, totaling around $54.2 million in savings. “Now $54.2 million dollars is a lot of money, but it is quite literally a drop in the bucket when compared to the $60 billion that Medicare programs lost last year,” Roskam noted.
CMS Testimony
“Historically, CMS and our law enforcement partners have been dependent upon ‘pay and chase’ activities, by working to identify and recoup fraudulent payments after claims were paid,” Shantanu Agrawal of CMS acknowledged. “Now, CMS is using a variety of tools, including innovative data analytics, to keep fraudsters out of our programs and to uncover fraudulent schemes and trends quickly.”
He described CMS’s data tools in detail:
Since 2011, CMS has been using [FPS] to apply advanced analytics on all Medicare fee-for-service claims on a streaming, national basis by using predictive algorithms and other sophisticated analytics to analyze every Medicare fee-for-service claim against billing patterns. The system also incorporates other data sources, including information on compromised Medicare cards and complaints made through 1-800-MEDICARE. When FPS models identify egregious, suspect, or aberrant activity, the system automatically generates and prioritizes leads for review and investigation by CMS’ Zone Program Integrity Contractors (ZPICs). The ZPICs then identify administrative actions that can be implemented swiftly, such as revocation, payment suspension, or prepayment review, as appropriate. The FPS is also an important management tool, as it prioritizes leads for ZPICs in their designated region, making our program integrity strategy more data-driven.
He also described a variety of other methods for detecting a preventing fraud, including CMS’s improved coordination with law enforcement. View Agrawal’s testimony here
OIG Testimony
Gary Cantrell, Deputy Inspector General for Investigations for OIG called combatting Medicare fraud a “top priority,” and stated: “We use data analytics to detect and investigate Medicare fraud and to target our resources for maximum results.” He stated that these results have included “almost $15 billion in investigative receivables and more than 2,700 criminal actions in the past 3 years.”
OIG uses data in order to detect and investigate fraud, and to target the use of their limited resources. “OIG is a front-runner in the use of data analytics to detect and investigate health care fraud,” states Cantrell. “We use innovative analytic methods to analyze billions of records and data points to identify trends that may indicate fraud, geographical hot spots, emerging schemes, and individual providers of concern.”
He notes that at the “macro-level,” OIG analyzes “data patterns to assess fraud risks across Medicare services and provider types and geographically to prioritize and deploy our resources.” Then, “[a]t the micro-level, we use data analytics, including near-realtime data, to identify fraud suspects and conduct our investigations as efficiently and effectively as possible.”
Cantrell also walked through a particular example of how OIG integrates various strategies to fight fraud:
We combine our field intelligence with data analytics to assess vulnerabilities across the program and to deploy our special agents to investigate the most egregious cases of suspected fraud. For example, we worked with OIG’s evaluators to develop indicators of questionable billing for Part D drugs that may be associated with fraud and abuse based on our experience with prescription drug investigations. OIG evaluators designed studies using sophisticated data analytics to identify questionable billing by retail pharmacies, prescribers with aberrant patterns, individuals writing prescriptions without authority to prescribe, and Schedule II drugs billed as refills. These studies generated numerous law enforcement leads, resulting in multiple ongoing investigations. They also identified systemic vulnerabilities in the Part D program and made recommendations to CMS to better prevent fraud.
“The need to protect the Medicare program and the beneficiaries it serves from fraud and harm has never been more important,” Cantrell concluded. “OIG, working with our internal and external partners, will continue using data analytics to target our resources for maximum results.” View Cantrell’s testimony here.