We previously reported on the Supreme Court’s opinion earlier this year in Tyson Foods v. Bouaphakeo, a non-FCA case that upheld the use of statistical sampling to establish liability in a Fair Labor Standards Act suit, but which offered important narrowing limitations that we argued were applicable to FCA cases (see here). Relying in part on Tyson Foods, a district court recently refused to allow a relator to use extrapolation to establish FCA liability, finding that extrapolation was inappropriate in a case based on claims of medical necessity. See United States ex rel. Wall v. Vista Hospice Care, Inc., No. 07-cv-00604 (N.D. Tex. June 20, 2016).

The relator alleged that defendant VistaCare, a hospice provider, knowingly submitted hospice claims for patients ineligible for hospice care. Relator’s expert statistician identified a universe of 12,000 patients who received hospice care from VistaCare for at least one year during the relevant time period. From this universe, the statistician selected a stratified sample of 291 patients. Relator’s clinical expert then reviewed the medical records of these 291 patients and opined on their hospice eligibility. The expert reviewed the records with a so-called “lenient” standard of review, under which he concluded that a patient was ineligible for hospice if “no reasonable physician could have concluded the patient was eligible.” All hospice claims for such patients were, according to the relator, false claims. To yield the total number of false claims at issue, the statistical expert extrapolated the ineligible rate within the sample to the population of 12,000 patients.

The district court concluded that extrapolation was not appropriate to establish liability in this case. Quoting Tyson Foods, the district court explained that “[t]he permissibility of statistical sampling turns on ‘the degree to which the evidence is reliable in proving or disproving the elements of the relevant cause of action.’” Questions relating to hospice eligibility are, as even the relator’s expert acknowledged, highly subjective. Furthermore, each claim implicated “different patients, different medical conditions, different caregivers, different facilities, different time periods, and different physicians.” The district court held that such diversity among claims rendered inappropriate proof of liability through representative evidence.

The district court did observe in a footnote that extrapolation could “be appropriate where the evidence establishes that a defendant’s objective approach was similar in all cases.” In reaching this conclusion, the court was noticeably hesitant to permit the relator’s own over-reaching to serve as a justification for extrapolation. The Supreme Court acknowledged in Tyson Foods that representative evidence is sometimes “the only practicable means to collect and present relevant data establishing a defendant’s liability.” However, the district court was unmoved by the burden the relator voluntarily took on of seeking “to pursue all potential false claims submitted in fourteen states over nearly a decade, of which she did not have personal knowledge.” In the alternative, the court ruled that even if extrapolation were appropriate in this case, flaws in the statistical expert’s sampling methodology did not permit valid extrapolation.

Having concluded that extrapolation was inappropriate, the court moved on to consider the circumstances in which the highly subjective claims within the sample could be “false” for purposes of FCA liability. As the court explained, “an FCA claim about the exercise of [clinical] judgment must be predicated on the presence of an objectively verifiable fact at odds with the exercise of that judgment, not a matter of questioning subjective clinical analysis.” For example, the court suggested a relator might be successful if she could present evidence that a physician never actually assessed a patient’s medical condition. In contrast, the court determined that the relator had only marshaled the “subjective clinical analysis” of her physician expert. Such dueling expert opinions fell short of creating a factual issue as to the falsity of the hospice certifications. Neither could the relator’s evidence as to corporate culture overcome this flaw. Although relator had presented some evidence of practices “that could jeopardize the proper exercise of physician judgment,” the relator was missing “a causal link between Defendants’ policies, a few instances where medical information was allegedly falsified, and actual false or fraudulent certifications and claims.”

Finally, the court considered the relator’s argument that VistaCare violated the FCA by falsely certifying compliance with the AKS. According to the relator, VistaCare’s various bonus programs for its employees—generally rewarding them for facilitating new admissions—violated the AKS. The district court disagreed, finding that the bona fide employee exception to the AKS applied. This exception shields compensation to individuals “for employment in the furnishing of” services covered by the federal healthcare programs. The relator advanced a narrow view, limited to employees directly providing such services. However, the district court adopted the broader scope advanced by VistaCare, interpreting the exception to cover payments for services including marketing and sales functions. Again ruling in the alternative, the district court held that even if the payments did not satisfy this AKS exception, summary judgment was nonetheless appropriate because the relator had not adduced evidence of any actual false certifications. Furthermore, the Fifth Circuit requires “some connection between kickbacks, referrals, and claims,” and the relator failed to establish one.

This district court’s opinion promises to significantly shape the debate over the appropriate use of statistical sampling, both in the hospice context (including the sampling case currently on appeal in the Fourth Circuit, as discussed here) and as the question arises in other contexts.

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