Law enforcement, federal agencies, and private whistleblowers are increasingly working alongside data scientists to identify fraudulent practices by analyzing vast data sets. The application of data analytics has notably amplified the investigation of health care claims. It has furnished examiners with the ability to scrutinize and probe irregular billing practices, which typically involves the review of vast sets of data.
The Emerging Role of "Professional Relators” and Data Whistleblowers
As technology progresses, so does the emergence of a new breed of "professional relators." These relators deviate from the traditional whistleblower archetype. Historically, whistleblowers were insiders armed with direct knowledge garnered from their employment. By contrast, these modern professional relators are principally dependent on their mastery of data science.
Integra Med Analytics LLC (Integra) illustrates this transformative change. Integra’s proficiency in data analytics enables it to spot potential fraud indicators within public data repositories, a feat that has earned it the title "data whistleblowers." The arrival of these players heralds a significant paradigm shift in whistleblowing, moving away from direct observation and toward a more analytical, inference-based approach.
Integra achieved a major coup in 2021 by securing its first government intervention in a False Claims Act (FCA) lawsuit against a collection of skilled nursing facilities based in New York. The intricate statistical analyses of publicly accessible data sets allowed Integra to expose unusual patterns suggestive of fraudulent activities. This landmark achievement underscores the potential impact non-employee whistleblowers can make on FCA enforcement. As of now, the case remains ongoing as all involved parties continue to navigate the complexities of the legal proceedings.
This innovative whistleblowing model has not escaped controversy. Courts have reacted to these pioneering techniques with discernable skepticism. Notably, the Fifth and Ninth Circuits have underscored the necessity of concrete evidence in fraud cases, dismissing Integra’s reliance on statistical anomalies as proof of fraud. The courts emphasized that these anomalies could equally signify legitimate business strategies or lawful adaptations to intricate billing regulations. They further cautioned that Integra’s statistical methods offered a potential explanation, not conclusive proof of fraud.1
These cases underline the critical importance of entertaining innocent explanations before deducing fraud from statistical irregularities. They also cast uncertainty over the sustainability of leveraging FCA's qui tam provisions purely for profit, absent genuine knowledge of fraud. With grave costs for defendants and the court system associated with government fraud allegations, it is essential for relators to provide substantive facts demonstrating fraud rather than relying solely on statistical inferences. This necessitates a balanced approach that acknowledges the potential of data analysis in qui tam actions while also emphasizing the importance of concrete evidence of fraudulent activities.
The Intricacies of Statistical Sampling and Extrapolation in FCA Enforcement
The incorporation of data analytics into the mechanisms of FCA enforcement has stimulated a profound dialogue concerning the intersection of technology, jurisprudence, and due process rights. Two pivotal lawsuits — United States ex rel. Michaels v. Agape Senior Community, Inc., and United States ex rel. Martin v. LifeCare Centers of America, Inc. — manifest these dynamics.
In Agape Senior Community, the plaintiff-relators and the government alleged that a network of South Carolina nursing homes utilized a widespread fraudulent scheme to submit fraudulent claims for medically unnecessary services. The plaintiffs argued for the use of statistical sampling to define the scope of liability, asserting that a comprehensive, claim-by-claim review would be financially prohibitive and unduly time-consuming. The U.S. District Court for the District of South Carolina rejected this proposal, emphasizing the necessity for an intricate, individual claim assessment to ascertain the medical necessity of each claim.2 On appeal, the Fourth Circuit abstained from issuing a definitive statement on the deployment of statistical sampling in FCA lawsuits, instead leaving it as a question of fact subject to the district court's discretion.3
Conversely, in LifeCare Centers, the U.S. District Court for the Eastern District of Tennessee affirmed the practicality of statistical sampling and extrapolation in establishing FCA liability while also noting its limitations.4 Plaintiff-relators — former employees —– accused Life Care Centers of America of inflating Medicare claims for unnecessary services. Despite objections to statistical extrapolation, the court accepted its usage for proving liability when claim-by-claim review is impracticable. In doing so, the court noted that courts had discretion regarding the weight given to the use of statistical sampling and extrapolation and referenced its endorsement in other contexts to determine damages. This verdict represents a careful equilibrium between technological advancements and due process, confirming that statistical evidence, although valuable, cannot solely satisfy the FCA's scienter requirement. Stated differently, it cannot, without more, demonstrate that a defendant knowingly committed fraud.
The rulings in these cases spotlight the need for an equilibrium — one that acknowledges the efficiency and potential of statistical sampling but also ensures the protection of due process rights. The use of data analytics in FCA enforcement must adhere to stringent audit procedures, ensure transparency, and offer defendants sufficient opportunity to challenge statistical extrapolations. Only such a balanced approach can truly harness the power of data analytics while preserving fairness and efficiency in FCA prosecutions.
1 See generally, U.S. ex rel. Integra Med Analytics, L.L.C. v. Baylor Scott & White Health, et al., No. 19-50818 (5th Cir. May 28, 2020); Integra Med Analytics, L.L.C. v. Providence Health Servs., No. 19-56367 (9th Cir. Mar. 31, 2021).
2 See United States ex rel. Michaels v. Agape Senior Community, Inc. No. 0:12–3466–JFA, 2015 WL 3903675 *7–*8 (D.S.C. Jul. 6, 2015)
3 See United States ex rel. Michaels v. Agape Senior Community, Inc., No. 15-2145 25–26 (4th Cir. Feb. 14, 2017).
4 See United States v. Life Care Ctrs. of Am., 114 F. Supp. 3d 549, 560 (E.D. Tenn. 2014).