Statistical Sampling Growing in Acceptance – and Importance – in False Claims Act Cases (Part 1 of 3)
Attorneys representing whistleblowers and the government are increasingly turning to statistical sampling for evidence of both liability and damages in False Claims Act cases. The decision to sample rather than directly review every claim is often driven by necessity when the allegedly false claims are too voluminous to permit individual analysis.
Using a statistical sample as evidence involves having an expert take a large body of evidence (like 50,000 allegedly false claims), and draw a much smaller, random, representative sample of several hundred claims that is then practical to review individually for falsity. The result of this review is then extrapolated to the unreviewed claims: for example if 150 of 300 sample claims are determined to be false, then it can be determined with a reasonable degree of certainty that 25,000 of 50,000 total claims are false, with a margin of error around 5%.
Whether statistical sampling can be used to prove both liability and damages is a particularly important question in False Claims Act cases involving healthcare fraud. As the United States Supreme Court recently recognized in Tyson Foods, Inc. v. Bouaphakea, 136 S. Ct. 1036, 194 L. Ed. 2d 124 (2016): “In many cases, a representative sample is ‘the only practicable means to collect and present relevant data’ establishing a defendant’s liability.” 1 Healthcare False Claims Act cases often involve tens of thousands of individual claims, and the time and expense of reviewing each claim individually can rapidly outstrip the damages at issue.
The Supreme Court’s decision in Tyson, a Fair Labor Standards Act case, is therefore very important to False Claims Act whistleblowers because it affirmed that a statistical sample, like any other evidence, can be used to address liability in any action: “A representative or statistical sample, like all evidence, is a means to establish or defend against liability.” 2 However, the Supreme Court observed that “whether or not a statistical sample is permissible evidence of liability turns not on the form a proceeding takes—be it a class or individual action—but on the degree to which the evidence is reliable in proving or disproving the elements of the relevant cause of action.” 3
Tyson resolved any remaining doubt about whether statistical samples may be used to address voluminous records that cannot realistically be presented claim-by-claim. Lower courts weighing the reliability of statistical samples offered in False Claims Act cases to prove liability and/or damages, however, come out on both sides of this fact-specific question. Cases on both sides of this question are collected in U.S. ex rel. Michaels v. Agape Senior Cmty., Inc., No. CA 0:12-3466-JFA, 2015 WL 3903675, at *7 (D.S.C. June 25, 2015), order corrected, No. CA 0:12-3466-JFA, 2015 WL 4128919 (D.S.C. July 6, 2015), and aff’d in part, appeal dismissed in part sub nom. United States ex rel. Michaels v. Agape Senior Cmty., Inc., 848 F.3d 330 (4th Cir. 2017).
Cases rejecting sampling include:
United States ex rel. Crews v. NCS Healthcare of Illinois, Inc., 460 F.3d 853, 857 (7th Cir.2006) (rejecting relator’s attempt to establish FCA liability based upon percentages rather than proof of actual false claims); United States ex rel. El–Amin v. George Washington Univ., 533 F.Supp.2d 12, 31 n. 9 (D.D.C.2008) (requiring the relators to set forth specific evidence for each individual claim); United States ex rel. Hockett v. Columbia/HCA Healthcare Corp., 498 F.Supp.2d 25, 66 (D.D.C.2007) (court held that it is “imperative for [R]elator[s] to produce real evidence to support [their] contention [s]….”); United States v. Medco Physicians Unlimited, No. 98–C–1622, 2000 U.S. Dist. LEXIS 5843, at *23 (N.D.Ill. Mar. 15, 2000) (declining to allow statistical sampling and explaining that the parties provided “no case law or other authority to support such a request”); United States ex rel. Trim v. McKean, 31 F.Supp.2d 1308, 1314 (W.D.Okla.1998) (declining to allow a statistical sample “to find a percentage of false claims from all claims submitted by [defendant] ….”).
Id. And cases allowing sampling include:
United States v. Rogan, 517 F.3d 449, 453 (7th Cir.2008) (appellate court rejected the argument that the district court had to address each of the 1,812 claims at issue and held that “[s]tatistical analysis should suffice;” however, all 1,812 claims were objectively false); United States ex rel. Ruckh v. Genoa Healthcare, LLC, No. 8:11–CV–1303–T–23TBM, 2015 WL 1926417, at *3 (M.D.Fla. Apr.28, 2015) (district court expressed an inclination to allowing statistical sampling and extrapolation by rejecting Friedmanand explaining that Friedman does not stand for the proposition that statistical sampling cannot be used in large-scale qui tam cases); United States ex rel. Martin v. Life Care Ctrs. of Am., Inc., Nos. 1:08–cv–251, 1:12–cv–64, 2014 U.S. Dist. LEXIS 142660, 2014 WL 4816006 (E.D.Tenn. Sept. 29, 2014) (same); United States v. Fadul, No. CIV.A.DKA 11–0385, 2013 WL 781614, at *2 (D.Md., Feb., 28, 2013) (finding the extrapolation method acceptable for Medicare and Medicaid claims under common law theory of payment by mistake); United States v. Chen, No. 2:04–cv–00859, 2009 WL 1683142 (D.Nev.2009) (jury found physician liable under the FCA for submitting 3,544 false claims, but parties only analyzed 37 claims at trial after the physician conceded that the referral request and services provided were the same for each of these claims); United States ex rel. Loughren v. UnumProvident Corp., 604 F.Supp.2d 259, 263 (D.Mass.2009) (extrapolation is a reasonable method for determining the number of false claims so long as the statistical methodology is appropriate); United States ex rel. Barron v. Deloitte & Touche, LLP, Civil No. SA–99–CA–1093–FB, 2008 WL 7136869, *2 (W.D.Tex. Sept.26, 2008) (in a Daubert hearing, the court excluded the statistician’s testimony, but recognized the relevance of statistical evidence); United States ex rel. Doe v. DeGregorio, 510 F.Supp.2d 877, 890 (M.D.Fla.2007) (court relied on an extrapolated overpayment figure derived from a prior Government audit when calculating the pre-judgment remedy figure in the subsequent FCA action); United States v. Cabrera–Diaz, 106 F.Supp.2d 234, 242 (D.P.R.2000) (court entered default judgment against defendants for treble the damages sustained based off the estimated overpayments that a prior Government audit revealed); United States v. Krizek, 859 F.Supp. 5, 7 (D.D.C.1994) supplemented, 909 F.Supp. 32 (D.D.C.1995) aff’d in part and remanded, 111 F.3d 934 (D.C.Cir.1997) (physician agreed upon sampling process to prove liability of the 8,002 potentially false claims, appellate court agreed that physician consented to sampling).
Michaels, 2015 WL 3903675 at *7.
Parts two and three of this post discuss two important recent District Court opinions that came out on either side of permitting the use of a statistical sample to prove liability and damages in False Claims Act cases.
The second Part of this three-part blog post discusses a leading False Claims Act case that accepted a statistical sample as evidence of both liability and damages. For more information, or to discuss a potential case with us, please contact us.
- Tyson Foods, Inc. v. Bouaphakeo, 136 S. Ct. 1036, 1046, 194 L. Ed. 2d 124 (2016) (citing Manual of Complex Litigation § 11.493, p. 102 (4th ed. 2004)). See also United States ex rel. Martin v. Life Care Centers of Am., Inc., 114 F. Supp. 3d 549, 571 (E.D. Tenn. 2014) (“Given the large number of claims that can be submitted by a single entity to be reimbursed by Medicare, it is often not practicable to do a claim-by-claim review of each allegedly false claim in a complex FCA action.”). ↩
- Tyson Foods, 136 S. Ct. at 1046. ↩
- Id. (citing Fed. R. Evid. 401, 403, and 702). ↩