Federal Court Decision in Nebraska Receives National Attention: US District Court Upholds ZPIC Sampling Methodology for Extrapolation to Determine Medicare Overpayment
On January 22, 2014, the U.S. District Court for the District of Nebraska granted a motion for summary judgment by the U.S. Department of Health and Human Services (“HHS”), upholding an extrapolation from a statistical sampling of claims for chiropractic services, to recover $37,580 in Medicare overpayments.
Wisconsin Physician Service (“WPS”) is a Medicare Zone Program Integrity Contractor (“ZPIC”) for HHS, and is charged with conducting expanded post-payment medical review of claims submitted by Medicare providers to Medicare beneficiaries. In this case, the ZPIC reviewed claims submitted by Schuldt Chiropractic Wellness Center for services provided to 75 beneficiaries from January 2008 through March 2010. The statistical sampling consisted of 214 claims, representing 445 services billed, out of a “universe” of 5,098 services billed for 154 beneficiaries. As a result of the review, the ZPIC concluded that the claims had a 99.55% error rate, equating to an actual overpayment of $11,376.13 for the 445 services for which claims were submitted. Based on that data, the ZPIC extrapolated a total overpayment of $126,041 to Schuldt for the universe of claims.
Schuldt appealed. WPS affirmed its earlier decision on redetermination, following review by a different examiner, using the same statistical method. Schuldt requested reconsideration by a qualified independent contractor (“QIC”), which affirmed the decision by WPS. Schuldt then requested a hearing before an Administrative Law Judge (“ALJ”).
The ALJ hearing was telephonic, with no representative from HHS making an appearance. The ALJ reviewed the 445 services billed, considering issues of coverage, liability, overpayment waiver and the extrapolation. On the first three of the four issues, the ALJ found that 344 of the 445 services were properly billed to Medicare and that Schuldt should be paid for them. Schuldt presented testimony of a statistical expert and the ALJ consulted his own statistical expert. The two experts opined that the methodology used by WPS was not reliable and should not be used for purposes of extrapolating the findings to a larger universe beyond the samples. The theory was that “correlation” could have skewed the WPS sample, because individual beneficiaries in the sample may have had multiple claims for similar services. The experts explained that a larger sample of claims should have been used, or the samples should have been selected from a larger number of beneficiaries. On the extrapolation issue, the ALJ held that WPS’s methodology was “insufficiently reliable to be used for the purpose of estimating an overpayment to a larger universe than the sample itself.”
The Medicare Appeals Counsel (“MAC”), on its own motion, conducted a de novo review of the ALJ’s decision on the single issue of whether Schuldt met its burden of proving that the statistical sampling methodology used by WPS was invalid and insufficiently reliable to be used for extrapolation. The MAC found that the ALJ erred in finding the sampling methodology and extrapolation to be invalid. The MAC reversed the ALJ’s decision on that issue alone and did not disturb the ALJ’s findings on coverage, liability or overpayment waiver. WPS issued a recalculation of projected overpayment applying the ALJ’s findings on coverage, liability and overpayment waiver, reducing the overpayment amount to $37,580. Schuldt paid that amount to WPS.
The District Court’s standard of review was limited to determining whether HHS’s action, reflected in the MAC decision, was supported by substantial evidence in light of the record as a whole. The Memorandum and Order prepared by Chief US District Judge Laurie Smith Camp explains that sampling “only creates a presumption of validity as to the amount of an overpayment which may be used as the basis for recoupment. The burden then shifts to the provider to take the next step.”
Judge Camp’s analysis outlines the strategies that may be successful for a provider bearing the burden of disproving a claim for recoupment of Medicare overpayment:
1. The provider may dispute denials of claims in the sample. Schuldt successfully did this before the ALJ, who reduced the number of denied claims from 445 to 101.
2. The provider may challenge the statistical validity of the extrapolation from the sample. Schuldt did this to the satisfaction of the ALJ, but not to the satisfaction of the MAC.
Weighting the process against the provider is the HHS Medicare Program Integrity Manual (“MPIM”) which allows for smaller statistical samples with less precise results, offset by the direction that overpayments be assessed at the lower level of confidence intervals, giving the benefit of the doubt regarding the range of overpayment to the Medicare provider. Further, Medicare regulations require ALJs and the MAC to give “substantial deference” to manual instructions.
In the end, it appears that it was the standard of review that carried the day for HHS’ successful motion for summary judgment. The Court noted that the ALJ’s decision was thorough and well-reasoned, and that the Court might have reached a different conclusion if the case had come before it de novo. Given the standard of review, the Court found that there was substantial evidence in the record as a whole to support the MAC’s conclusion that Schuldt failed to meet its burden of demonstrating that WPS’s sampling methodology and overpayment extrapolation were invalid. The Court, referring to the D.C. Circuit Court of Appeals’ decision in Chaves County Home Health Service, Inc. v. Sullivan, M.D., 931 F. 2d 914 (1991), illustrated further strategies that could have helped Schuldt meet its burden to discredit the extrapolation:
a. An extrapolation may be challenged by presenting evidence of a different sample from the same universe of claims, showing a lower rate of denials, or otherwise show that the projection is not a true estimate of the rate of denials in the non-sample universe.
b. A provider may also establish the validity of all or a sufficient number of its actual claims to demonstrate that the HHS projection is factually impossible of correctness. The Court noted that this approach would have been effective in this case, to the extent that the error rate had been determined at 99.55%.
Schuldt has appealed the District Court’s decision to the Eight Circuit Court of Appeals. So there may be more to be learned from this case in terms of challenging Medicare recoupments. In the meantime, Medicare providers are well-advised to prepare promptly and with qualified experts to debunk a questionable statistical sampling and extrapolation for purposes of Medicare recoupment.