Posted by Jaime L.M. Jones and Brenna Jenny
Yet another district court has joined recent opinions (as reported here and here) permitting relators and the government to proceed to trial on the basis of extrapolations of FCA liability from a sample of submitted claims. See United States v. AseraCare Inc., No. 12-cv-00245 (N.D. Ala. Dec. 4, 2014). Following the government’s intervention in a suit against hospice care provider AseraCare, it submitted an expert witness opinion based on a review of a sample of 233 claims. The expert concluded that the patients in 124 of the claims failed to meet the criteria for medically indicated hospice care. The government then sought to extrapolate this finding to the full universe of 2,181 claims submitted by AseraCare.
AseraCare filed a motion for partial summary judgment, arguing that the government had failed to introduce any evidence regarding any of the claims other than the 124 that its expert testified were unsubstantiated. These remaining claims broadly fell into three categories: those the expert agreed were substantiated; those the Medicare Administrative Contractor or ALJ had found to be substantiated (regardless of the expert’s opinion); and those claims which were signed by a physician attesting to the medical necessity of the hospice care. The court disagreed that the government had failed to produce any evidence regarding these claims. Noting that “statistical evidence is evidence,” the court ruled that the 124 claims would serve as statistical evidence of the falsity of the remaining claims in the pool, and left it to the jury to decide the relative weight properly applied to the government’s statistical evidence and AseraCare’s evidence related to the circumstances of individual claims.
The court’s decision to allow the jury to weigh the government’s statistical evidence was premised only on a Supreme Court case relating to the type of statistical regression analyses sufficient to allow a court to infer a pattern or practice of racial discrimination under Title VII. In doing so, the court did not distinguish that case on the basis that the means appropriate for proving classwide liability for a pattern or practice of discrimination under Title VII do not necessarily transfer to the still novel use of statistical sampling to establish liability under the False Claims Act, which attaches only to the submission of individual false claims.
A copy of the court’s opinion can be found here.