Another District Court Expresses Interest in Expanded Use of Statistical Sampling

Posted by Scott Stein and Brenna Jenny

We recently reported on several decisions supporting the expanded use of statistical sampling in FCA cases to support liability determinations. On November 5, 2014, a federal district court in Nevada joined the recent opinions of other courts in expressing receptiveness to the expanded use of statistical sampling beyond the context of establishing damages. The relator, who formerly worked for the defendant hospital system (“Renown”) as its director of clinical documentation, brought a FCA suit alleging that Renown improperly submitted claims for short-stay inpatient admissions that should have been billed as outpatient claims. In June 2014, Renown produced data on claims for short-stay admissions spanning three-and-half years and several diagnosis-related group (“DRG”) codes. Relator, however, argued that Renown should be ordered to produce all Medicare claims data necessary to enable her to develop a statistical sampling plan capable of estimating the false claims submitted by Renown during the full course of the alleged upcoding scheme.

In an unusual step, the district court held an evidentiary hearing on the motion to compel during which statistical experts from both sides testified. Following supplemental briefing, the court ruled that Renown must produce claims data for all 28 DRGs implicated by relator’s upcoding allegations for the six years prior to relator’s filing of her complaint, i.e., the maximum time period allowed under the FCA statute of limitations. The court concluded that the “discovery is permissible pursuant to Fed. R. Civ. P. 26(b), it will insure thoroughness and transparency of the data universe, and it will result in the development of a valid and reliable statistical sampling plan.” The court expressly declined to rule on the ultimate admissibility of evidence, deferring those issues until the pre-trial stage.

A copy of the court’s decision can be found here.