To increase efficiency in administering Medicare, private insurance companies (“Carriers”) contract with the Secretary of Health and Human Services to administer certain aspects of Medicare Part B. It is Carriers who make the determination of whether the Medicare provider’s claimed services are medically necessary. Carriers calculate the amount of Part B payments due to the provider and Carriers pay the claims. Subsequent to payment to the provider for claimed services, Medicare providers, including, physicians, home health care, physical therapists, and ambulance companies, are subject to post-payment audits of whether the claimed services were “ . . . reasonable and necessary for the diagnosis or treatment of illness or injury. ” Anghel v. Sebelius, 2012 WL 6212843 (E.D. NY 2012).
The Social Security Act itself authorizes Carriers to conduct post-payment audits. (42 U.S.C. §1395u(a).) Guided by the Medicare Program Integrity Manual (“MPIM”), Carriers use a statistical sampling to estimate the total amount of overpayments to a provider.
The first step in conducting a post-payment audit is to get a “probe sample” of the provider’s billings. The Carrier looks to determine whether there is a likelihood of overpayment to the provider. The Carrier may then request a statistically valid random sample (”SVRS”) from the provider. The Carrier uses the SVRS to extrapolate to the provider’s total billings for the purpose of making a reasonable approximation of the total overpayment. If the Carrier determines that there was an overpayment, the Carrier may recoup Medicare overpayments from future Medicare payments to the provider. The methodology for the statistical extrapolation has issues which are coming forth as the process is used more and more.
Some troubling issues with statistical extrapolation:
- A sample of as few as 30 Medicare claims may be valid Balko v. Sebelius __ F. Supp.2d __ (W.D. PA 2012) (2012 WL 6738246);
- During the appeals process, if the initial determination is reversed, the extrapolation shall be adjusted, however a new sampling process is not needed (Id.);
- There may be an unreliability of the design used to create the sampling group Miniet v. Sebelius, 2012 WL 2930746 (S.D. Fl. 2012); and
- The statistical methodology need be only valid, not the most precise methodology (Id.).
Attorneys experienced in handling issues related to Medicare Part B may assist providers in understanding the process of and defending against demands from Medicare for overpayments, in particular when demands are based on statistical extrapolation.