By Jeffrey H. Brochin, J.D.
Use of statistical sampling by contractor raises a presumption of validity as to the amount of an overpayment.
A federal district court in Virginia has upheld the determination of an Administrative Law Judge (ALJ) who ruled that a Medicare contractor’s use of statistical sampling methodology from which an overpayment of Medicare benefits was extrapolated, followed the law and the six-step procedure as set forth in the Medicare Program Integrity Manual (MPIM). Therefore, a presumption of validity as to the amount of overpayment existed, and the provider failed to overcome that presumption (Livinrite, Inc. v. Azar, June 17, 2019, Ellis, T.).
CMS contractor’s investigation. In March 2010, AdvancedMed, a CMS contractor, opened an investigation of the provider, a Medicare-certified home health services company, based on a complaint that the provider was admitting patients who did not qualify for home health care services, and, had continued to provide physical therapy to patients even after they reached their maximum level of potential. After determining that a full audit of the provider would not be feasible, the contractor conducted a statistical sampling of the Medicare claims paid, by reviewing services provided to 30 randomly selected beneficiaries, and then extrapolating the overpayment determinations to estimate the total amount the provider was overpaid by Medicare.
Basis of provider challenge. The provider filed an action against HHS seeking a reversal of a decision by the Medicare Appeals Council (MAC) that the provider had been overpaid by approximately $1 million for Medicare claims submitted between January 1, 2008, and June 30, 2019. The provider challenged the MAC determination that ten claims were not covered, that a valid statistical sampling methodology was used to derive—through extrapolation—the total overpayment amount, and that the provider was not entitled to a waiver of liability for the overpayment. For the reasons stated below, the MAC determination was affirmed.
Extrapolation permitted by MPIM. In light of the substantial volume of Medicare claims submitted by providers, the MPIM authorizes contractors to use statistical sampling and extrapolation as the method for calculating overpayments. However, certain steps must be followed in order to use statistical sampling for the calculation, including: (1) the contractor must select the period to be reviewed; (2) the universe, sampling unit, and sampling frame must be defined; (3) the contractor must choose a sampling method and implement the method; and (4) the sampling method used must be classified as a "probability sampling."
The use of statistical sampling by the contractor creates a presumption that the agency action was valid. The provider must overcome this presumption by demonstrating either (i) the sample was not statistically valid, or, (ii) the contractor’s determinations of overpayment with respect to specific units in the selected sample were incorrect.
Statistical validity challenge. A challenge to the statistical validity of the sample must be predicated on the actual statistical validity of the sample as drawn and conducted. If a particular probability sample design was properly executed, then the provider’s assertions as to the resulting estimates not being statistically valid will fail. The court noted that "a probability sample and its results are always valid." The provider also argued that the contractor’s decision should be set aside because the MAC’s determination that the contractor’s sampling methodology could be accurately replicated, was arbitrary and capricious, and that the extrapolation overpayment calculation could not be replicated based on the materials in the record.
Sufficient documentation provided. The court noted that failure to supply the provider with sufficient documentation to recreate the sampling frame and sample would effectively deprive the provider of its right to challenge the statistical validity, in the instant case, the record showed that the provider was in fact supplied with appropriate documentation to enable it to replicate the sampling frame and the sample.
For the foregoing reasons, the decision of the ALJ as to use of extrapolation and the statistical validity of the calculation was affirmed.
The case is No. 1:18-cv-00603.
Attorneys: Ashley Hudson Morgan (Liles Parker PLLC) for LivinRite, Inc. Catherine M. Yang US Attorney's Office, for the Alex M. Azar, II.
Companies: LivinRite, Inc.
MainStory: TopStory CaseDecisions CMSNews HomeNews MedicareContractorNews PaymentNews ProgramIntegrityNews ProviderNews VirginiaNews
Interested in submitting an article?
Submit your information to us today!Learn More
Health Law Daily: Breaking legal news at your fingertips
Sign up today for your free trial to this daily reporting service created by attorneys, for attorneys. Stay up to date on health legal matters with same-day coverage of breaking news, court decisions, legislation, and regulatory activity with easy access through email or mobile app.