Tennessee approves use of statistical sampling in False Claims Act cases
Statistical sampling is regularly used in employment discrimination, antitrust and toxic torts cases to minimize cost and effort in cases with large data sets, but has only recently begun to be used to determine damages in False Claims Act (FCA) cases.
Andrew Silver, of Tycko & Zavareei, in an article in The National Law Review, explores Tennessee’s federal court recent ruling approving statistical sampling in FCA cases to determine liability. This will be useful in lawsuits involving a large number of claims, for which performing claim-by-claim reviews would be time consuming and burdensome.
US ex rel. Martin v. Life Care Centers of America
In this case, the defendant is the owner of more than 200 nursing facilities across the country. The plaintiff alleged that Life Care submitted fraudulent claims to Medicare by pressuring its staff to target patients for whom they could provide medically unnecessary services.
Reviewing every relevant claim in this case, which spans a seven year period, would have been time consuming and costly. Instead, the government sought to utilize data from a random sample of 400 patient admissions from 82 of the Life Care facilities to draw conclusions about the total 54,396 patient admissions and total 154,621 claims.
Life Care moved for summary judgment, claiming that statistical sampling is not appropriate in this context, but the court denied the motion. The court determined that statistical sampling could in fact be used to prove that a false claim was made, adding that although specific factors and variables could affect each claim, it did not negate the accuracy and usefulness of statistical sampling in a case such as this. The court also deemed statistical sampling appropriate for proving materiality of the statements in the total ‘universe of claims’.
The defendant challenged the methodology of the government’s statistical expert witness, but the court found that the expert need not be an expert in healthcare, just in statistics, to provide valid analysis in this case.
This ruling provides litigants with an important statistical tool in cases that involve a particularly large number of claims. The court said if it “were to reach the conclusion urged by the Defendant – that a claim-by-claim review is required in every FCA action and that statistical sampling is never permissible – potential perpetrators of fraud would be emboldened by the fact that a claim-by-claim review is often impractical,” adding that “large-scale perpetrators of fraud would reap the benefits of such a system.”
To read a more detailed discussion of statistical sampling, peruse Chris Haney’s article on Law360, in which he explores a number of aspects of statistical sampling in this case, including: the evidence of specific claims, falsity, knowledge, due process, representativeness, size, examination and randomness of the sample, precision and confidence. To learn more about representative sampling, visit Dwight Steward, Ph.D.‘s page on sampling.