PALM VALLEY HEALTH CARE, INCORPORATED, Plaintiff-Appellant v. ALEX M. AZAR II, SECRETARY, U.S. DEPARTMENT OF HEALTH AND HUMAN SERVICES; PALMETTO GBA, L.L.C., Defendants-Appellees
No. 18-41067
United States Court of Appeals for the Fifth Circuit
January 15, 2020
Appeal from the United States District Court for the Southern District of Texas
Before OWEN, Chief Judge, and HAYNES and COSTA, Circuit Judges.
With annual spending topping half a trillion dollars, Medicare is the largest recipient of federal funds after Social Security and defense.1 With so many tax dollars at stake, Congress created an administrative process through which Medicare can recover overpayments. Because of the massivе number of claims, an audit of each one is not feasible. So federal law allows Medicare to investigate a select number of claims from a provider. If the audit of that sample reveals “a sustained or high level of payment error,” Medicare can extrapolate that overpayment rate to a larger number of similar claims.
An audit of Palm Valley Health Care, a home health care provider, revealed that a significant percentage of the sampled claims did not meet Medicare coverage requirements. Extrapolating that overpayment rate to all claims paid over the relevant time period resulted in a repayment demand of more than $12 million. Palm Valley brings constitutional, statutory, and evidentiary challenges to that decision. Finding no error, we AFFIRM.
I.
The Department of Health and Human Services (HHS), acting through a Medicare contractor, audited claims Pаlm Valley submitted between July 1, 2006, and January 31, 2009. Palm Valley was selected for review because it had submitted an unusually large number of claims involving five or more consecutive home health care episodes. An episode is sixty days long
Out of those thousands, the contractor sampled 54 and concluded that 29 of them provided services to bеneficiaries who were not eligible for home health care. Medicare will cover home health services if the beneficiary is homebound, under the care of a physician, in need of skilled services, and under a plan of care.
Palm Valley sought review of the overpayment finding. It argued that the beneficiaries qualified as homebound and thus were eligible for home health services. It also challenged the sample the auditor used and the еxtrapolation methodology used to reach the $12 million repayment figure. Notably, however, Palm Valley did not press a defense Medicare allows for a provider that “did not know, and could not reasonably have been expected to know”
For the arguments Palm Valley was asserting, it had many opportunities to make them. Administrative review of overpayment decisions has several stages. The first allows a provider to seek a redetermination from the auditor.
Although each stage of administrative review has a statutory deadline, HHS routinely fails to meet those dates. From start to finish, the average appeal takes about five years, far in excess of the statute‘s approximately one year. Maxmed Healthcare, Inc. v. Price, 860 F.3d 335, 344-45 (5th Cir. 2017). The statute recognizes that the agency may not meet the deadlines. If HHS fails to meet the review deadline at any stage, a provider may escalate its appeal and immediately jump to the next stage of review.
Palm Valley appealed through the entire administrative process. At the redetermination stage, the contractor determined that one partial denial among the 29 was in error. At the reconsideration stage, the independent contractor found that Medicare did not cover the claims for 29 beneficiaries. The ALJ subsequently reviewed the overpayment determinations and concluded that 27 claims in the sample of 54 should not have been paid. The Medicare Appeals Council mostly affirmed the ALJ‘s decision, but concluded that the claims for two beneficiaries previously found to be uncovered were eligible claims. Full administrative review thus reduced the number of ineligible claims from 29 to 25, shaving a meaningful amount off the $12 million that Palm Valley originally owed.
Consistent with recent practice, the ALJ and Medicare Appeals Council issued their decisions roughly one-and-a-half years and three years after Palm Valley sought their review. Although HHS did not come close to meeting either 90-day deadline, Palm Valley did not escalate its appeal.
Palm Valley finally sought review in district court. The court affirmed the decision of the Medicare Appeals Cоuncil.
II.
Palm Valley argues that HHS violated due process by failing to meet the statutory deadlines at each stage of the administrative process. We have difficulty
Due process typically requires “some form of hearing . . . before an individual is finally deprived of a property interest.” Matthews v. Eldridge, 424 U.S. 319, 333 (1976). Palm Valley received a hearing before it had to give any money back to Medicare. Its argument is essentially that it had too many hearings—really too many appeаls of a hearing—and that they took too long. It does not cite any Supreme Court or circuit level authority finding a due process violation for delays occurring during an administrative appeals process.4 And violation of a statutory deadline does not automatically mean a lack of due process; the Constitution, not statutes, determine the minimum procedures that due process requires. See Cleveland Bd. of Educ. v. Loudermill, 470 U.S. 532, 538, 538-39, 542 (1985) (recognizing that state statute created a рroperty interest but applying Matthews’ balancing test to determine what level of process the Due Process Clause required).
Even assuming the possibility of a due process violation arising from prolonged appeals of a hearing, Palm Valley failed to take advantage of the statutory escalation procedure that would have allowed it to expedite the process. Nor did it seek a federal court injunction to try and prevent recoupment. See Family Rehab., Inc. v. Azar, 886 F.3d 496 (5th Cir. 2018). One problem with allowing this after-the-fact complaint about delay is that it lets the provider have it both ways. Palm Valley took advantage of every opportunity it had to undo the overpayment finding, and it took some bites out of the apple with partial success on some appeals. Only now, after going through every appeal, does it challenge the delay. Contrast id. (involving provider that brought action in district court oncе recoupment began because it faced a three-year delay before an ALJ).
The timing of Palm Valley‘s due process challenge also means it cannot show the substantial prejudice that is necessary for a due process claim to succeed. United States v. Lober, 630 F.2d 335, 337-38 (5th Cir. 1980) (per curiam). Palm Valley does not brief any theory of substantial prejudice, and none is apparent from the record. The roughly four-year delay did not affect HHS‘s ability to evaluate Palm Valley‘s claims, as the evidence that the agency drew on at each stage existed when Palm Valley first requested redetermination. Nor did the delay cause financial harm to Palm Valley. During the slow appeals process, HHS was not engaging in recoupment, the process by which Medicare holds back payments on new claims to cover overpayment findings that are still being appealed and thus may be overturned. Contrast Family Rehab., 886 F.3d at 503 (holding that a district cоurt had subject matter jurisdiction to consider a due process claim that a plaintiff subject to recoupment brought). In other words, HHS did not seek to recover any overpayment until the end of the appeals process. And as we have noted, Palm Valley preferred repeated review—with each level providing a new opportunity to succeed—to expeditious resolution of its claims. That choice to pursue each level оf review shows that Palm Valley saw some benefit from pursuing multiple appeals despite the known delay each phase caused. Palm Valley has
The district court correctly rejected the due process claim.
III.
Turning from the process of the agency review to its substance, we consider Palm Valley‘s argument that the ALJ and Medicare Appeals Council applied the wrong definition of “homebound.” An individual is homebound if he or she “has a condition . . . thаt restricts the ability of the individual to leave his or her home except with the assistance of another individual or the aid of a supportive device . . . or if the individual has a condition such that leaving his or her home is medically contraindicated.”
Palm Valley argues this was too demanding a standard. It argues that for an individual to qualify as homebound, the condition of the person “should“—but not must—“be such that there exists a normal inability to leave home and that leaving home requires a considerable and taxing effort by the individual.”
But Palm Valley raised this argument for the first time in the district court. A federal court reviewing an agency determination will not ordinarily consider arguments that a litigant could have raised before the agency but chose not to. Gulf Restoration Network v. Salazar, 683 F.3d 158, 174–75 (5th Cir. 2012). HHS limits the Medicare Appeals Council‘s review to objections a represented party asserts challenging the ALJ‘s ruling.
IV.
We next consider Palm Valley‘s argument that HHS lacked substantial evidence for its finding that 25 beneficiaries were not homebound. Substantial evidence7 supports a finding that a patient is not homebound if “more than a mere scintilla” of evidence support the determination. Salmond v. Berryhill, 892 F.3d 812, 817 (5th Cir. 2018) (quotation omitted).
That low evidentiary bar wаs met. The main thrust of Palm Valley‘s evidentiary challenge is that the Medicare Appeals Council and ALJ relied too heavily on interviews of individuals (including the beneficiaries themselves), who indicated the beneficiaries were not homebound. That testimony was unreliable, Palm Valley argues, because a significant amount of time, sometimes several years, had passed between the claims and the interviews.
As a general matter, Palm Valley is right that memories often fade over time. Basic principles of our legal system, like statutes of limitations and the right to a speedy trial, reflect that view. But the fact that passage of time may be grounds for impeaching testimony does not render that testimony irrelevant. This is the difference evidence law recognizes between relevancy and probative value. Compare
The Appeals Council considered these factors when evaluating the interviews as part of its claim-by-claim review of the homebound question. Take, for instance, beneficiary R.L. The Council‘s determination that R.L. was not homebound reliеd in part on an interview with R.L.‘s daughter two years after the dates of service. R.L.‘s daughter described how two years earlier R.L. was able to drive to the barbershop and to visit his daughters. Also consider beneficiary F.D. Interviews with F.D. and a staff member in the building where F.D. lived revealed that F.D. was alert and frequently traveled outside the home, without assistance, for activities like shopping and visiting friends. For both these claimants, contemporary records corroborated the testimony. Clinical records demonstrated that R.L.‘s medication changes could be managed without skilled care at home and that his diagnoses were unlikely to leave him confined to home. And Palm Valley‘s records showed multiple days when it had sent someone to F.D.‘s residence, but the patient was not home.
Palm Valley‘s critique of the interviews is unavailing given the deferential standard of review. The Appeals Council carefully evaluated the evidence on eaсh claim in an 86-page opinion. Substantial evidence supports HHS‘s determination that many beneficiaries were not homebound.
V.
The agency finding we have just upheld—that 25 claims were not eligible for payment because the patients were not homebound—does not on its own cause Palm Valley much financial harm. But the extrapolation of that overpayment rate to Palm Valley‘s more than 10,000 claims does. It thus is not surprising that Palm Valley also challengеs that statistical analysis.
As we noted at the outset, HHS may use statistical extrapolation to determine overpayment amounts when the Secretary determines that “there is a sustained or high level of payment error.”
Palm Valley thus challenges not whether it was appropriate to use sampling and extrapolation, but the statistical methods the agency uses when performing those tasks. It contends that the methodology does not pass muster under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993), because it has not been peer reviewed or generally accepted in the relevant scientific community. Daubert, however, does not apply in agency proceedings. See Nat‘l Taxpayers Union v. U.S. Soc. Sec. Admin., 302 F. App‘x. 115, 121 (3d Cir. 2008); Bayliss v. Barnhart, 427 F.3d 1211, 1218 n.4 (9th Cir. 2005); Niam v. Ashcroft, 354 F.3d 652, 660 (7th Cir. 2004). It interprets
In looking at the extrapolation from that substantive standpoint, we see no reversible error. The methodology that the agency employed resulted in a random sample of 54 of the 10,699 claims, the audit of which provided an unbiased estimate of the actual average overpayment for all 10,699 claims. See
Palm Valley argues that the sample was too imprecise—or more simply, that the Medicare contractor did not use a large enough sample. But as the Medicare Appeals Council recognized, demanding a larger sample to marginally increase the precision of an estimate “ignore[s] real world constraints imposed by conflicting demands on limited public funds, constraints which CMS chose to incorporate into the statistical sampling guidelines.” The extrapolation methodology may be imperfect, but it is the product of a complex balance of interests. At a minimum, it constitutes substantial evidence in support of the agency‘s decision. Cf. Maxmed, 860 F.3d at 343 (“Congress clearly envisioned extrapolation in overpayment determinations involving home health agencies like [the plаintiff], and the Secretary‘s reliance on extrapolation as a tool was justified.“). If anything, the extrapolation methodology is provider friendly. The extrapolation does not assume that the average overpayment in the random sample occurred for the universe of claims. Rather, the agency assumes that the average overpayment for all claims is equal to a number that there is a 90% chance is smaller than the actual overрayment. See
We see no error in the extrapolation.
*
*
*
The judgment is AFFIRMED.