As we noted in two of our prior posts in the Insider blog, the government has long touted its ability to rely upon data mining as a means of detecting fraud in the federal health care system, and has initiated a host of investigations and prosecutions based on its analysis of claims data from theMedicare and Medicaid programs. Yet any approach that relies on data mining rests on a fragile foundation, because the quality of the information upon which the government relies has often been in doubt. As we explained in the first of our two prior posts on this topic, an HHS Regional Inspector General testified in June 2012 that much of the data used to identify overpayments and fraud is not “current, available, complete, [or] accurate.” Subsequently, in a post from November 2012, we described the concerns that two United States Senators raised regarding the effectiveness of the “Fraud Prevention System Program” (“FPS”), which is intended to use “predictive analysis” to reduce fraud, waste, and abuse in the Medicare program.
Given the increased attention that commentators and government officials have paid to the need for accurate claims data in federal health care programs, one might have hoped that now, nearly two years later, the problems would be addressed and the government’s data rendered more reliable. Unfortunately, while improvements may have been made, cracks in the system continue to emerge. Just two days ago, the Department of Health and Human Service’s Office of the Inspector General (“HHS-OIG”) issued a report regarding flaws in the “Medicaid Interstate Match” program, which is intended to reduce improper Medicaid payments by identifying beneficiaries who are enrolled in the Medicaid programs of more than one state. Although the report does not directly suggest that problems in the Medicaid Interstate Match program will necessarily hinder fraud investigations or result in the targeting of innocent Medicaid participants, it yet again underscores the fundamental problems that exist in relying on data mining in the federal health care system.
As the report from the HHS-OIG indicates, the Medicaid Interstate Match is part of a broader data collection program (the so-called “Public Assistant Reporting Information System,” or “PARIS”) that utilizes Medicaid enrollment data in order to determine if the same person is receiving government healthcare benefits from more than one source or from more than one state. The Medicaid Interstate Match in particular can identify situations in which beneficiaries remain enrolled in a given state’s Medicaid program despite the fact that they have moved to another state and are receiving benefits from that second state’s program. Although a beneficiary’s receipt of funds from more than one state’s Medicaid program often results from a failure to timely report a change of address, rather than from an intent to defraud, the program can nonetheless save the Medicaid program significant amounts of funds that beneficiaries are not entitled to receive. Accordingly, as of October 1, 2009, the Social Security Act required each and every state to participate in the Medicaid Interstate Match, and the Centers for Medicare and Medicaid Services (“CMS”) is responsible for issuing guidance regarding such participation.
Nonetheless, as the HHS-OIG revealed in its report from earlier this week, the participation of the various states in the Medicaid Interstate Match program is significantly limited. Indeed, of the four steps that HHS-OIG has identified as comprising “participation” in the Medicaid Interstate Match (notably, CMS has not itself ever defined the term “participation,” even though such participation is a pre-requisite for obtaining federal funds), many of those steps simply are not taken. For example, whereas state participation in the Medicaid Interstate Match requires that a state submit its enrollment information so that data can be matched with that from other states, HHS-OIG determined that, for a sample three month period (the Medicaid Interstate Match is to conducted on a quarterly basis), 14 states did not submit Medicaid enrollment data for all of their beneficiaries, and for those 14 states, on average only 46 percent of the relevant data was provided. As another example, in order for the Medicaid Interstate Match program to play a meaningful role in detecting improper benefits payments, data that appears to signify a match between beneficiaries in more than one state must be verified, to confirm that there is not a “false positive.” However, the HHS-OIG report determined that the states did not verify nearly 70 percent of the matches that were identified, in part because the enrollment data submitted by the states was incomplete. As a result of these and other flaws, the HHS-OIG report indicates that for the three-month period under examination, not a single improper Medicaid payment was recovered through the use of the Medicaid Interstate Match.
What is one to make of HHS-OIG’s report regarding the failings of the Medicaid Interstate Match? Initially, there is the somewhat anti-climactic and even obvious conclusion reached by HHS-OIG, which is that “CMS should issue guidance to states on the requirement for participating in the Medicaid Interstate Match.” Wisely, perhaps, CMS “concurred” with this recommendation.
More important, though, are two observations that may be of particular relevance to those who practice in the health care fraud arena. First, HHS-OIG notes in its report that, according to CMS, “5.8 percent of all Medicaid payments made in fiscal year 2013 were improper, representing $14.4 billion in Federal expenditures.” The government frequently cites to such massive figures as proof of rampant fraud, waste, and abuse that allegedly exists in government health care programs. But in fact, the HHS-OIG report provides some much needed perspective, indicating that 57% of the “improper” Medicaid payments result from more prosaic, mundane issues, such as the “eligibility errors” that arise when a beneficiary moves from one state to another and does provide Medicaid with a change of address. Fraud in the Medicaid program may still be a dramatic problem, but when “improper payments” are the result of such “eligibility errors” rather than fraud, the true scope of the problem can better be understood.
Second, even though the Medicaid Interstate Match is intended to detect eligibility mistakes rather than fraud, the flaws in the collection and use of Medicaid data still raise significant concerns. As this blog has noted, reliance upon inaccurate data and flawed methodology can result in innocent participants in the health care system being exposed to expensive, lengthy, and potentially ruinous audits and investigations. Further, while efforts to correct data inaccuracies and address problems in program implementation may sometimes have a useful effect, systemic issues relating to the size of the Medicare and Medicaid programs, the nature of the bureaucracies that surround them, and the difficulty of coordinating federal-state interactions in such a complex area may make it impossible for health-care data mining to ever be a fully reliable source of investigative decisions. For the regulators, auditors, investigators, and prosecutors who rely upon government data when determining whether to bring their power to bear on those who are the subjects of costly and burdensome health-care investigations, all of these issues must carefully be weighed in the balance.