U.S. District Court of D.C. Invalidates Decision Denying Hospital Reimbursement for Indigent Bad Debt Based on Third-Party Information

Alert
April 5, 2022
6 minutes

On March 29, 2022, the U.S. District Court for the District of Columbia issued a favorable decision for the plaintiff hospitals in Sentara Hospitals v. Azar in their challenge to an adverse decision by the Centers for Medicare & Medicaid Services (“CMS”) denying reimbursement for their bad debt claims under the Medicare Act.1 This Alert summarizes the key portions of the Court’s analysis and discusses the implications of the decision for similarly situated hospitals. The decision, if it stands, could strengthen hospitals’ appeals of adverse bad debt adjustments based on hospitals’ use of third-party data to assist them in making indigence determinations.

Background

Under the Medicare regulations, Medicare patients’ unpaid deductibles and coinsurance, or bad debts, are reimbursable by CMS if certain requirements are met, including, as relevant here, that the debt is “actually uncollectible.”2 The Provider Reimbursement Manual (the “Manual”) provides that for bad debt to be “actually uncollectible,” hospitals must either undertake “reasonable collection efforts” or determine that the patient is indigent using the hospital’s own “customary methods.”3 Pertinent here, the Manual further provides “guidelines” for hospitals to use in applying their “customary methods,” including that (1) a “patient’s indigence must be determined by the provider”; (2) the “provider should take into account a patient’s total resources,” including “an analysis of assets . . . , liabilities, income and expenses”; and (3) the provider should document “the method by which indigence was determined.”4

For the 2010 through 2013 hospital cost years at issue in this case, Sentara, a Virginia-based health care system, determined that patients were indigent using two approaches. First, Sentara considered available financial assistance applications and other documentation of resources provided by patients, and used information obtained from Equifax to verify the accuracy of this information.5 That patient-provided information included, for example, “tax returns, bank statements, Social Security statements, W-2s, mortgage statements, [and] disability statements.”6 Second, when Sentara was unable to obtain a financial assistance application or other documentation regarding resources from the patient, Sentara considered only the information from Equifax, including information about patient mortgages and unused lines of credit, as well as an income predictor score, a payment predictor score, and a bankruptcy navigator index score.7

Sentara challenged the Medicare contractor’s denial of reimbursement for Medicare bad debt for fiscal years 2010 through 2013. The Provider Reimbursement Review Board issued a largely favorable decision for Sentara in August 2020, finding that “Sentara’s methods for determining indigency met [Medicare] requirements.”8 Then, in October 2020, the CMS Administrator issued a decision overturning the Board and denying reimbursement, reasoning that Sentara failed to determine indigence in accordance with the Manual and claiming incorrectly that a purported 2020 retroactive amendment to the bad debt regulations supported the reversal.9 Ropes & Gray filed a complaint on behalf of Sentara challenging this decision in Federal District Court, arguing that the Administrator’s decision was arbitrary and capricious, contrary to law, and not supported by substantial evidence.10

Court’s Ruling and Reasoning

On March 29, 2022, the District Court found that the Administrator’s decision was not supported by substantial evidence because (1) Sentara adequately analyzed patient resources, (2) Sentara itself, not Equifax, made indigence determinations relating to each patient, and (3) Sentara adequately documented its indigence determinations. The Court set aside the Administrator’s decision and ordered CMS to reimburse Sentara for its cost years 2010 through 2013 indigent bad debt plus interest.

Sentara Adequately Analyzed Patient Resources

First, the Court addressed whether Sentara complied with section 312(B) of the Manual, which states that hospitals “should take into account a patient’s total resources,” including “assets, income, liabilities and expenses.”11 As a threshold matter, the Court considered whether this provision contains a mandatory requirement that hospitals conduct an analysis of patients’ resources in determining indigence.12 In a previous case, Baptist Healthcare Systems v. Sebelius,13 another judge in the same Court determined that “the word ‘should’ in [this section] is a suggestion, not a mandate, and that § 312 ‘does not create a mandatory asset test.’”14 Ultimately, here, the Court found that it did not need to consider this question because, even if “should” means “must” and section 312(B) imposes mandatory requirements, Sentara fully satisfied them.15

The Court found that the record established that Sentara properly considered patient resources. When Sentara was unable to obtain a financial assistance application or other documentation regarding patient resources directly from the patient, Sentara considered Equifax-furnished information, including data elements relating to patient resources as well as the Equifax scores.16 First, Equifax reports information on patient assets, including cars, homes, and monetary judgments.17 Second, Equifax reports reflect income through the income predictor score.18 The Court noted that Equifax validates income predictor scores “against a national database of employer-reported income,” and that Sentara confirms the scores against patient-reported income whenever possible.19 Third, Equifax reports information documenting or analyzing patient liabilities and expenses, including mortgages, auto loans, judgments, liens, and expected monthly payments.20 And when available, Sentara considered patient-reported assets, income, liabilities and expenses through “tax returns, banks statements, Social Security statements, W-2s, mortgage statements, [and] disability statements.”21 Ultimately, the Court found that “[i]t is hard to imagine what else a provider could do to reasonably assess the patient’s ability to pay.”22

Sentara Itself Made Indigence Determinations

Second, the Court addressed whether Sentara satisfied section 312(A) of the Manual, which requires the hospital, rather than the patient, to make indigence determinations.23 The government argued that Equifax, rather than Sentara itself, made Sentara’s indigence determinations, but the Court found this argument unpersuasive. The Court held instead that Sentara made its own indigence determinations because it analyzed, rather than “rubber-stamped,” low Equifax scores. In particular, the Court explained that Sentara used Equifax scores to segment patient accounts into categories based on likelihood to pay, and then considered additional data elements beyond Equifax scores, such as mortgages and unused lines of credit, before itself making final indigence determinations.24

Sentara Adequately Documented Patient Indigence

Third, the Court addressed whether Sentara satisfied section 312(D) of the Manual, which states that hospitals should provide documentation of the method by which they determined indigence to their Medicare contractor.25 In its decision, the Administrator found that Sentara failed to provide appropriate documentation because Equifax’s model was “not auditable.”26 The Court rejected this argument, holding that this section requires only that Sentara document the method it relied on—in this case, that it analyzed Equifax scores to determine indigence.27 The Court concluded that Sentara provided such documentation to its Medicare contractor.28

For these reasons, the Court found the Administrator’s decision unsupported by substantial evidence in violation of the Administrative Procedure Act, invalidated that decision, and ordered that CMS reimburse Sentara for its fiscal years 2010 through 2013 indigent bad debt and pay litigation interest on the amounts in controversy.29

Looking Ahead

The government has until May 28 to appeal the District Court’s decision. If it stands, the decision could strengthen hospitals’ appeals of adverse indigent bad debt adjustments based in part or in whole on use of third-party data. In particular, if a hospital used patient resource information or scores from a third party to assist it in making indigence determinations for cost reporting periods beginning prior to October 1, 2020, the hospital could rely on the Sentara decision to argue that this practice satisfied the Manual’s requirements. In doing so, the hospital should, to the extent possible, develop evidence establishing that the third-party data being utilized included information on patients’ assets, income, liabilities, and expenses as well as evidence establishing that the hospital conducted its own analysis of this information in making its indigence determinations.

If you have any questions related to this decision, please feel free to contact your usual Ropes & Gray advisor or one of the authors listed below.

  1. See Sentara Hosps. v. Azar, No. 20-cv-3771 (CRC), 2022 WL 910514 (D.D.C. Mar. 29, 2022).
  2. Id. at *1.
  3. Id.
  4. Id. at *2; see also Provider Reimbursement Manual (Part I), § 312.
  5. Sentara Hosps., 2022 WL 910514, at *3.
  6. Id. at *3, *7 (alteration in original).
  7. Id. at *2.
  8. Id. at *3.
  9. See Sentara Healthcare Bad Debt CIRP Groups v. Palmetto GBA c/o Nat’l Gov’t Servs., CMS Adm’r Dec. at 15-16 (Oct. 22, 2020). The provisions of the amended regulation that the Administrator invoked were actually effective for cost reporting periods beginning on or after October 1, 2020, see 85 Fed. Reg. 58,432, 58,996-99 (Sep. 18, 2020), and those provisions could affect the government’s position on indigent bad debt going forward.
  10. Sentara Hosps., 2020 WL 910514, at *4.
  11. Id.at *4-5.
  12. Id.
  13. 646 F. Supp. 2d 28 (D.D.C. 2009).
  14. Sentara Hosps., 2020 WL 910514, at *5 (citing Baptist, 646 F. Supp. 2d at 30).
  15. Id. at *6
  16. Id. at *6-8.
  17. Id. at *6.
  18. Id.
  19. Id.
  20. Id.
  21. Id. at *7.
  22. Id.
  23. Id.
  24. Id. at *8.
  25. See id.
  26. Id.
  27. Id.
  28. Id. at *9.
  29. Id.