Ropes & Gray Issues Public Comment on Centers for Medicare & Medicaid Services Proposed Rule

WASHINGTON, July 18 -- Stephanie A. Webster, partner at Ropes and Gray LLP, has issued a public comment on the Centers for Medicare and Medicaid Services' proposed rule entitled "Medicare Program; Hospital Inpatient Prospective Payment Systems for Acute Care Hospitals and the Long-Term Care Hospital Prospective Payment System and Proposed Policy Changes and Fiscal Year 2021 Rates; Quality Reporting and Medicare and Medicaid Promoting Interoperability Programs Requirements for Eligible Hospitals and Critical Access Hospitals". The comment was written on July 10, 2020, and posted on July 14, 2020:

* * *

We write concerning the proposed Inpatient Prospective Payment System ("IPPS") rule for federal fiscal year ("FFY") 2021 that was published by the Centers for Medicare & Medicaid Services ("CMS") in the Federal Register on May 29, 2020. 85 Fed. Reg. 32,460. In particular, we submit for your consideration the following comments on the proposed revisions of Medicare bad debt policy, the Disproportionate Share Hospital ("DSH") uncompensated care payment and Provider Reimbursement Review Board ("PRRB") electronic filing systems provisions of this proposed rule.

PROPOSED REVISIONS OF MEDICARE BAD DEBT POLICY

For the reasons discussed below, we urge the agency to abandon its unlawful proposal to change retroactively through new regulations various Medicare bad debt policies, many of which have been the subject of litigation. 85 Fed. Reg. at 32,867. The agency's proposal to give the bad debt provisions some, as yet undefined, retroactive effect violates the Medicare statute's general prohibition on retroactive rulemaking. Moreover, the retroactive application of the bad debt provisions to periods prior to 2012 would violate the bad debt moratorium, which prevents the agency from making changes to bad debt policies that were in effect in 1987.

In light of the economic hardship brought on by the COVID-19 pandemic, it is more critical than ever for the agency to preserve Medicare bad debt reimbursement, a critical source of funding for hospitals, rather than attempting to place further restrictions upon it. In the end, CMS's proposals seem geared toward bolstering the agency's litigation position in bad debt appeals that have been filed by hospitals across the country rather than toward any public interest.

I. If Effective for Prior Cost Years, the Agency's Proposal Would Violate Medicare's Restrictions on Retroactivity

Retroactive rulemaking is strongly disfavored, and there can be no retroactive rulemaking absent "an express statutory grant." Bowen v. Georgetown University Hospital, 488 U.S. 208, 208-09 (1988). See Northeast Hospital Corp. v. Sebelius, 657 F.2d 1, 13 (D.C. Cir. 2011) (Defendant's agency "may not promulgate a retroactive rule absent express congressional authorization."). The notice and comment provision of the Medicare statute reinforces that general prohibition and permits retroactive rulemaking in Medicare only in very limited circumstances. 42 U.S.C. Sec. 1395hh(e). Under the statute, retroactive rulemaking is only permitted when "necessary to comply with statutory requirements" or when the "failure to apply the change retroactively would be contrary to the public interest." Id./1

When Congress adopted Medicare's prohibition on retroactive rulemaking, it did so in an effort "to ensure that Medicare's rules are not generally applied retroactively." H.R. Rep. No. 108-74, Div. II, A at 42 (2003) (Ways and Means Rep.). The measure was intended to "prohibit[] the government from imposing regulations retroactively" and from "changing the rules of the game and then punishing providers for noncompliance." Medicare Regulatory and Contracting Reform Act of 2001, Hearing on H.R. 3391 before the House of Representatives, 107th Cong. 8793 (2001) (statement of Rep. Nancy Lee Johnson). The legislative history further indicates that the public interest exception was intended to be applied only where beneficial to providers. See H.R. Rep. No. 107-288, Div. II at 26 (2001) (Ways and Means Rep.) (providing that a substantive change in Medicare rules may not be applied retroactively except where "necessary to comply with statutory requirements" or where it "would have a positive impact on beneficiaries or providers and suppliers").

CMS has proposed to codify retroactively what it incorrectly characterizes as "longstanding bad debt policies" that CMS claims have existed in Medicare guidance for many years. But that it simply not the case. For example, the Provider Reimbursement Manual ("PRM") currently provides that hospitals "should" consider a patient's income, assets, liabilities and expenditures in making a determination regarding the patient's indigence. See PRM, Part I, Sec. 312. But, multiple courts have held that there is no mandatory requirement that providers conduct a test of expenses and liabilities in order to claim reimbursement for their indigent bad debt. See Baptist Healthcare Sys. v. Sebelius, 646 F.Supp.2d 28, 34 (D.D.C. 2009) (noting that CMS has stated "unequivocal[ly]" that "a hospital may determine its own individual indigency criteria"); see also Harris Cnty. Hosp. Dist. v. Shalala, 863 F. Supp. 404 (S.D. Tex. 1994), aff'd on other grounds, 64 F.3d 220 (5th Cir. 1995) (finding that the word "should" as used in the manual is precatory and does not require the use of an asset test for bad debt reimbursement). Despite CMS's characterization of these changes as a codification of "longstanding bad debt polic[y]," the proposals would necessarily change the legal consequences of furnishing care to low-income patients and are, therefore, retroactive. Northeast Hosp., 657 F.3d at 14, 17 (finding that "[a] rule that 'alter[s] the past legal consequences of past actions' is retroactive" and that the 2004 DSH rule was impermissibly retroactive because it attached "new legal consequences to hospitals' treatment of low-income patients during the relevant time period").

Moreover, CMS's rationale is untenable when considered in light of Congress's intent to help providers - not hurt them - in adopting the general prohibition on retroactive rulemaking. CMS has invoked the public interest exception to the general prohibition on retroactive rulemaking, claiming, among other things, that the codification of these alleged "longstanding" policies will serve the "important public interest" of "avoiding confusion as to which longstanding policy should be applied for which cost reporting period." 85 Fed. Reg. at 32867. But in the process of supposedly "avoiding confusion" as to the Secretary's bad debt policy for a given point in time, the proposed rule, if adopted retroactively, would seemingly upset hospitals' settled expectations, and could ultimately lead to CMS unlawfully clawing back Medicare bad debt reimbursement that was properly paid to hospitals under the actual policies and guidance applicable to prior time periods. Although the agency claims that the retroactive effective date of the rule "does not affect prior transactions or impose additional duties or adverse consequences upon providers or beneficiaries," the proposed rule says that the bad debt provisions would be "effective for cost reporting periods beginning before, on, and after the effective date of this rule." Id. at 32867. Any proposed rule should at least state with particularity the cost years to which it would apply.

Moreover, the agency routinely adopts regulations that result in different policies applying to different points in time, without resorting to retroactive rulemaking. The agency's treatment of outpatient observation days in the calculation of a hospital's count of available beds is a prime example. Prior to 2003, CMS's regulation included outpatient observation days in the number of available beds and days used in calculating the Medicare DSH payment. See Health Alliance Hosps. Inc. v. Burwell, 130 F. Supp. 3d 277, 289-99 (D.D.C. 2015). In 2003, CMS amended the regulation to exclude outpatient observation bed days from the count of available beds and days. See 68 Fed. Reg. 45346, 45418-19, 45470 (Aug. 1, 2003). In 2004, CMS amended the DSH regulation again to include observation bed days in the count of available beds and days when the patient was admitted as an inpatient after observation. 69 Fed. Reg. 48916, 49096-98, 49245-46 (Aug. 11, 2004). And, in 2009, CMS once again amended the available beds regulation to revert back to the 2003 rule excluding observation bed days from the count of available beds and days. See 74 Fed. Reg. 43754, 43908 (Aug. 27, 2009). Despite the potential for "confusion as to which longstanding policy should be applied for which cost reporting period," see 85 Fed. Reg. at 32867, CMS never invoked the retroactive rulemaking provisions to apply changes in the bed count rules retroactively.

And for good reason. Just as there was no compelling public interest in that instance to apply the rule retroactively, there is no compelling public interest here either. As noted above, Congress intended the public interest exception to be invoked only in situations when it would help providers. Here, the multitude of bad debt provisions that CMS has proposed to adopt with unlimited retroactive effect would almost universally work to the detriment of hospitals, permitting the government to claw back or withhold bad debt reimbursement to which hospitals were entitled under existing policy. It is not, and cannot be, in the public interest to deprive safety net hospitals of an important source of revenue when they are most in need of it. In the wake of the COVID-19 pandemic and resulting declines in patient revenue, it is more important now than ever for hospitals to be able to rely on the agency's existing policies and guidance without the agency attempting to adopt new rules after-the-fact that may ultimately result in the agency clawing back funds properly paid to hospitals under existing law at the time.

Hospitals throughout the country have challenged many of the policies that CMS is now attempting to codify retroactively on a variety of grounds, and there have been numerous court decisions addressing those challenged policies. For example, hospitals have challenged the Secretary's alleged policy barring hospitals from claiming bad debt for accounts that are still at a collection agency after 120 days of collection activity in no less than four court actions in the District of Columbia. See Foothill Hosp. v. Leavitt, 558 F. Supp. 2d 1 (D.D.C. 2008); District Hosp. Partners v. Sebelius, 932 F. Supp. 2d 194 (D.D.C. 2013); Lakeland Reg'l Health Sys. v. Sebelius, 958 F. Supp. 2d 1 (D.D.C. 2013); Community Health Systems, Inc. v. Burwell, 113 F. Supp. 3d 197 (D.D.C. 2015). Other hospitals have challenged the agency's alleged requirement to consider expenses and liabilities in making indigence determinations. See, e.g., Baptist Healthcare Sys., 646 F.Supp.2d at 34; Harris Cnty. Hosp. Dist., 863 F. Supp. at 404. Numerous other cases on these issues are pending before the PRRB.

Although the agency has given the pretext that the bad debt proposals will avoid confusion over which policies apply to which cost years, the agency appears to be merely attempting to strengthen its litigation position and avoid further court action over its improperly-adopted bad debt policies. In particular, the proposal appears to be in response to the Supreme Court's decision in Azar v. Allina Health Services, where the Court held that even interpretive rules that establish a substantive legal standard must be adopted through notice and comment rulemaking. Permitting the agency to cure its failure to have engaged in the required notice-and-comment rulemaking when it initially adopted its existing bad debt policies through the promulgation of a retroactive rule years or decades after the fact is exactly the type of situation that Congress intended to prevent in adopting the general prohibition on retroactive rulemaking in the first place. See Medicare Regulatory and Contracting Reform Act of 2001, Hearing on H.R. 3391 before the House of Representatives, 107th Cong. 8793 (2001) (statement of Rep. Nancy Lee Johnson) (stating that "there will be no more changing the rules of the game and then punishing providers for noncompliance"). That would turn on its head the limited public interest exception for retroactive rulemaking.

In sum, the proposed rule's explanation here is hardly the type of compelling reason needed to justify retroactive rulemaking. To the contrary, the proposed rule seems geared toward bolstering the government's defense of its improperly adopted policies, not benefitting providers. CMS's limited retroactive rulemaking authority does not give it the power to retroactively change the rules that applied, years or even decades after the fact, to the detriment of providers' bad debt reimbursement claims. Imposing additional restrictions on Medicare bad debt reimbursement and withholding this critical source of funding from hospitals, particularly in this time of economic hardship, can hardly be justified as being in the "public interest."

II. The Agency's Proposal Violates the Bad Debt Moratorium

The agency's proposal to retroactively codify various bad debt provisions also violates the statutory bad debt moratorium that was in effect until 2012. In 1986, in response to proposed changes to the agency's bad debt policy, Congress enacted legislation that has become known as the "bad debt moratorium." In the Omnibus Budget Reconciliation Act of 1987, Congress prohibited the Secretary from making any changes in the bad debt reimbursement policies that were in effect as of August 1, 1987. See Pub. L. No. 100-203, Sec. 4008(c), 101 Stat. 1330, 133055 (codified, as amended, at 42 U.S.C. Sec. 1395f note); see also Hennepin Cty. Med. Ctr. v. Shalala, 81 F.3d 743, 751 (8th Cir. 1996) ("The Secretary may not retroactively apply a more stringent interpretation of [the bad debt reimbursement rules existing on August 1, 1987]."). Even though the bad debt moratorium was repealed prospectively, it remains in effect with respect to cost reporting periods beginning before October 1, 2012. See Middle Class Tax Relief and Job Creation Act of 2012, Pub. L. 112-96, Sec. 3201(d), 126 Stat. 156, 192-3 (2012).

Courts have found that several of the bad debt policies that CMS has proposed to codify retroactively are in violation of the bad debt moratorium. For example, the U.S. District Court for the District of Columbia found in Foothill Hosp., 558 F. Supp. at 9-11, and District Hosp. Partners, 932 F. Supp. 2d at 205-06, that disallowing Medicare bad debt on the ground that it had not been returned from a collection agency violated the bad debt moratorium. In addition, a decision issued just last week found that the requirement that hospitals receive remittance advice from the State Medicaid agency before claiming Medicare bad debt for dual eligible beneficiaries violated the Medicare bad debt moratorium. See Kindred Healthcare, Inc. v. Azar, No. 18-650 (D.D.C. July 1, 2020), slip op. at 17-18.

While we urge the agency to rethink its retroactive bad debt proposals altogether, at a minimum, those proposals should be made effective only for cost reporting periods ending on or after October 1, 2012 in order to avoid conflicting with the statutory bad debt moratorium. CMS's retroactive proposal appears to be premised on the fact that "the bad debt moratorium is no longer in existence," 85 Fed. Reg. at 32867, but the rulemaking does not explain how the unlimited retroactive effect of the proposed rule squares with the bad debt moratorium's prohibition on changing CMS's bad debt policies that were in effect in 1987 and applied for any cost reporting periods beginning before October 1, 2012.

Because the proposed rule includes no limits on its retroactive effect, as currently written, providers are concerned that CMS would attempt to apply the new regulations to any cost reporting period beginning prior to the effective date of the rule, including those cost years that are still governed by the bad debt moratorium. The proposed rule seems intended to allow the government to make an end run around the moratorium and to strengthen the government's litigation position in appeals of this nature challenging the agency's attempts to modify longstanding bad debt policies.

DSH UNCOMPENSATED CARE PROVISIONS

First, we ask that CMS meet the requirements of the Administrative Procedure Act ("APA") and the Medicare Act by providing meaningful explanations for the calculations and figures contained in the proposed rule relating to the DSH payment for uncompensated care costs. Second, we submit that CMS should satisfy its legal obligation to furnish interested parties with advance opportunity to comment on new calculations based on the more recent data that CMS intends ultimately to use for its final rule. Third, we request that CMS reconsider its decision not to reconcile final DSH payments for uncompensated care with actual data for cost reporting periods covering FFY 2021, especially in light of the drastic changes in estimates used by CMS in determining the first factor (Factor 1) in the DSH uncompensated care payment for FY 2021, as compared with the estimates used in calculating that Factor for previous FFYs. Fourth, we note that CMS should not calculate Factor 3 by comparing audited data for some hospitals with unaudited amounts or amounts reflecting the contractor's preliminary, non-final adjustments for other hospitals.

CMS proposes to reduce the total amount of DSH uncompensated care payments by more than half a billion dollars from the figure paid to hospitals in FY 2020 (a decrease from $8.350 billion to $7.816 billion). Given the unprecedented and precarious financial situation hospitals find themselves due to the COVID-19 health crisis, this unexplained and ahistorical decrease in total DSH funding could not have come at a worse time for our nation's safety net hospitals. Furthermore, CMS's failure to even address the impact of COVID-19 on the calculations for FY 2021 is a further deficiency in the calculations.

I. Inadequate Explanation of Calculations in Proposed Rule

The proposed rule explains that the process for determining the uncompensated care pool (Factor 1) for FFY 2021 will be the same as it was for FFY 2020, which continued the process established in the FFY 2014 final rule. See 85 Fed. Reg. at 32,747. Namely, according to the proposed rule, Factor 1 will be based on a recent estimate by the Office of the Actuary that Medicare DSH payments would be $15.359 billion in 2021 if it were not for the changes to DSH methodology made by the Affordable Care Act ("ACA"). Id. at 32,748. As in prior years, the 2021 proposed rule provides an inadequate explanation for how the Actuary's Office came to this number. According to the supplemental data file/2 and the proposed rule, 85 Fed. Reg. at 32,748, the agency extrapolated from the actual 2017 DSH data by applying certain unexplained update factors to derive an estimated payment for 2021. Some of these update factors are much different than they were in previous years, a change which also goes unexplained in the proposed rule.

The agency has unlawfully provided too little explanation for the key assumptions, data and methods used to calculate Factor 1 for hospitals to be able to meaningfully comment on this calculation in the proposed rule. This is a clear violation of the APA, as evidenced by the court's decision on the inadequately explained assumptions underlying the 0.2% rate reduction that CMS imposed in connection with its adoption of the two-midnight policy. See Shands Jacksonville Med. Ctr. et al. v. Burwell, 139 F. Supp. 3d 240, 260-65 (D.D.C. 2015). Other cases hold likewise. See e.g., Allina Health Servs. v. Sebelius, 746 F.3d 1102, 1110 (D.C. Cir. 2014) ("[A]n agency's failure to disclose critical material, on which it relies, deprives commenters of a right under Sec. 553 to participate in rulemaking.") (internal quotation marks omitted); American Radio Relay League, Inc. v. FCC, 524 F.3d 227, 237 (D.C. Cir. 2008) (agencies must make available the "technical studies and data" relied upon in a rulemaking); Appalachian Power Co. v. EPA, 251 F.3d 1026, 1035 (D.C. Cir. 200 l) (agencies must explain the assumptions and methods used in predictive computer models); Connecticut Light & Power, Co. v. Nuclear Regulatory Comm'n, 673 F.2d. 525, 530 (D.C. Cir. 1982) ("To allow an agency to play hunt the peanut with technical information, hiding or disguising the information that it employs, is to condone a practice in which the agency treats what should be a genuine interchange as mere bureaucratic sport.").

In addition to not providing enough explanation in the proposed rule, this year CMS also failed to include any discussion or even reference impact of the COVID-19 crisis in calculating the estimates for FY 2021. Despite the fact the Secretary declared a public health emergency on January 31, 2020, the proposed rule, published initially on May 11, 2020, makes no effort to detail how the agency factored the COVID-19 crisis into its estimates. Given the monumental impact COVID-19 has had on hospitals specifically, and the national economy generally, the failure to address the impact of the crisis is an invalid decision-making process under the APA. See Motor Vehicle Mfrs. Ass'n of the U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983) (agency action is arbitrary and capricious where agency has "entirely failed to consider an important aspect of the problem"); Dept. of Homeland Security v. Regents of the Univ. of California, 2020 WL 3271746 (U.S. June 18, 2020) (finding agency decision to terminate Deferred Action for Childhood Arrivals program was arbitrary and capricious because it failed to address key aspects of program).

Without further details about the calculations and the data underlying them, hospitals do not have the legally required opportunity to comment on whether the starting point for extrapolation is correct, whether all adjustments made to that starting point are proper or whether the adjustments have been made consistently throughout the proposed rule. Based on the information furnished, it is unclear what Medicaid expansion level(s) are presumed by CMS in calculating Factors 1 and 2, including whether the agency has consistently accounted for the level of Medicaid coverage across both Factors 1 and 2, what adjustment CMS made to Factor 1 to account for the presumed Medicaid expansion level for FFY 2021 and how CMS calculated the DSH payment impact that would result from the estimated expansion in FFY 2021 in the absence of the ACA changes to the DSH payment calculation.

We also have concerns regarding the explanations in the proposed rule on the calculation of Factor 2. Specifically, while the agency is no longer required by the statute to use CBO ("Congressional Budget Office") data, the agency has not adequately explained how the Actuary derived the estimates it decided to use and the evidence supporting them in calculating Factor 2. Furthermore, in projecting coverage levels for FY 2021, the proposed rule assumes, without adequate explanation, an under-reporting of Medicaid coverage "due to a perceived stigma associated with being enrolled in the Medicaid program or confusion about the source of their health insurance." 85 Fed. Reg. at 32,750. Further, there is nothing in the proposed rule to indicate that the agency has applied this same presumption of under reporting in calculating Factor 1, where increased Medicaid coverage would serve to increase expected DSH payments. It appears that the agency has applied internally inconsistent assumptions on Medicaid expansion between Factors 1 and 2 with no explanation.

As an example of these unsupported assumptions, the proposed rule for FFY 2015 provided no notice and no explanation whatsoever as to how the agency accounted for the effect of Medicaid expansion in its determination of Factor 1. In response to further comments on this point, the Secretary identified the percentages by which the agency assumed that traditional DSH payments would increase for FFY 2014 (4.9%) and FFY 2015 (3.4%) due to Medicaid expansion. 79 Fed. Reg. 49,854, 50,011 (Aug. 22, 2014). The preamble to the final rule for FFY 2015 said that these factors were "lower than commenters may have expected due to the assumption [by the agency's actuaries] that the expansion population is healthier than the rest of the Medicaid population and will utilize fewer hospital services." Id. The agency also makes this same claim in the FFY 2021 proposed rule. 85 Fed. Reg. at 32,749 ("[O]ur actuaries have assumed that the new Medicaid enrollees are healthier than the average Medicaid recipient and, therefore, use fewer hospital services."). But, here again, the rulemakings neither discussed nor described the basis for, or the data relied upon by CMS, in making that "assumption." See 79 Fed. Reg. at 50,011; 85 Fed. Reg. at 32,749. The final rule for FFY 2015 also did not discuss or describe how CMS determined, or the data it used to determine, that Medicaid expansion would increase the traditional DSH payment by the estimated increase percentages.

Further, the effect of Medicaid expansion on the agency's projection as to the traditional DSH payment that would have been made for FFY 2018 has varied erratically, without explanation, in the agency's successive rulemakings for FFYs 2018 through 2021. The effect of Medicaid expansion for FFY 2018 is reflected in an "other" category of adjustments included in the CMS table data, reflecting the factors used to calculate Factor 1 for FFY 2017 in the final rule for FFY 2017. In the FFY 2018 final rule, this estimate was 2.36%. This FFY 2018 estimate inexplicably rose to a 2.77% adjustment in the FFY 2019 final rule and to a 3.18% adjustment in the FFY 2020 final rule. In the FFY 2021 proposed rule, this figure representing changes in Medicaid coverage for FFY 2018 has again inexplicably increased to 5.08%. Had this 5.08% figure been applied consistently across the impacted fiscal years, Factor l would have been greater for FFYs 2018 through2020. CMS has not ever explained anywhere why this single factor, supposedly reflecting the Factor 1 effect of Medicaid expansion for FFY 2018, has changed from one year's rulemaking to the next year's rulemaking. Given the amount of money at stake, even the smallest changes in these factors can have large impacts on the payments to DSH hospitals.

CMS's opaque, unexplained and entirely inconsistent approach to this critical aspect of the DSH uncompensated care payment calculation is not acceptable. First, the agency should by now have within its possession or control the enrollment and utilization information from the States that have expanded Medicaid, which would either support or refute the assumption that the Medicaid expansion population is "healthier than the average Medicaid recipient and, therefore, uses fewer hospital services." 85 Fed. Reg. at 32,749. The agency cannot rely solely on an unsupported "assumption" when it has the ability to readily determine if its "assumption" is supported by the facts. That approach produces a rule that is arbitrary, capricious and unsupported by substantial evidence.

Second, by failing to provide any meaningful explanation as to the basis for that assumption, CMS has still failed to provide adequate detail for the hospitals to comment in a meaningful way on the proposed rule for FFY 2021. The above-described history of the "Other" adjustment factor that CMS uses to estimate Factor 1 highlights the problem with CMS's lack of adequate explanation. The unexplained changes in this particular factor is just one example of issues with the update factors overall. CMS needs to explain how it has calculated each of the various adjustment factors that are applied in arriving at the proposed Factor 1 estimate for FFY 2021 and the reasons why a given factor, like the "Other" adjustment factor for 2018, has changed from publication of the final rule for FFY 2018 through the publication of this proposed rule for FFY 2021. In addition, CMS must explain why the agency believes it is proper to continually change these factors from one rulemaking to the next when the agency has otherwise taken the position that all of these estimates should be fixed when made and should not be reconciled with data that becomes available later.

Third, the update factors that CMS apparently applied to extrapolate an estimate of DSH payments that otherwise would be made for 2021 appear to depart from the Actuary's prior, historic estimates of annual increases in the aggregate DSH payment and the estimates used to calculate Factor 2. As discussed by agency personnel at the March 2013 American Health Law Association conference, the Actuary's prior estimates as to aggregate DSH payments made to hospitals for FFYs 2000 through 2012 indicated that aggregate DSH payments increased from $5.18 billion for fiscal year 2000 to $11.93 billion for fiscal year 2012. Those estimates reflect an average increase of more than 7%per year from 2000 through 2012. Yet, in projecting total DSH payments for periods after 2012 (particularly in their estimates for changes from 2013 to 2014, and from 2014 to 2015), the agency has assumed much smaller increases, and in some cases has actually projected decreases in the total DSH payments, despite unprecedented and ongoing Medicaid expansion by the States under the ACA during these periods. In fact, the agency has estimated a significant decrease in total estimated DSH payments from the final FFY 2020 rule ($16.583 billion) to this FFY 2021 proposed rule ($15.359 billion), a decrease of approximately 7.4% in total DSH payments, which marks a significant, and unexplained, departure from previous trends. These internally inconsistent estimates also are irreconcilable with agency estimates that project a decrease in the uninsured rate during the same period and have been used to reduce the uncompensated care DSH payment pool through Factor 2. Using these figures for the purposes of Factor 2 calculation, CMS has estimated that the uninsurance rate has decreased from 14% in 2013 to 9.5% in 2021 due to Medicaid expansion, but has not fully carried over this increase in Medicaid participation to the Factor 1 calculation, which would lead to a greater increase in the overall DSH calculation.

Furthermore, other government reports have contradicted many of the most important assumptions the agency relied on in calculating the DSH uncompensated care payments. In September 2017, the CBO issued a report on nationwide health insurance levels, which concluded that the ACA had insured fewer individuals than previously estimated. CBO, Federal Subsidies for Health Insurance Coverage for People Under Age 65: 2017 to 2027 (Sept. 2017)./3

As discussed in the previous rulemakings, increased coverage under the ACA was a big driver in the agency's presumption of expanded coverage through Factor 2. Coverage levels less than what was estimated by the agency therefore suppressed payments to hospitals. Additionally, in the President's 2018 Economic Report, the Administration noted that not only was coverage expansion less than initially expected, but it was also due more to Medicaid expansion than was initially projected as well. Economic Report of the President (Feb. 2018)./4

Since Medicaid coverage has a large impact on the traditional DSH calculation, these recent findings suggest that Factor 1 was understated by the agency as well.

These findings further support what has been presented in comments during all previous rulemakings: namely, that estimates of expanded insurance coverage and estimated DSH payments were unsupported by the facts and highly uncertain. This is all the more the case given the COVID-19 crisis that has had, and will continue to have, a drastic impact on hospitals and the national insurance rate. Reports are now confirming that previous estimates were off, which is all the more reason the agency should have implemented a system that reconciled payments once later data was received. See 42 C.F.R. Sec. 412.106(g)(l)(iv) (2014).

II. Opportunity for Comment on Calculations in Final Rule

Aside from failing to explain the numbers and the evidentiary support for the assumptions actually contained in the proposed rule, CMS's rulemaking is also faulty because it will use different data and calculations for the final rule without any opportunity for the hospitals to comment. According to the proposed rule, CMS bases its Factor 1 calculations on the September 2019 update of the Medicare Hospital Cost Report Information System ("HCRIS"). 85 Fed. Reg. at 32,748. With respect to Factor 3, CMS used a February 19, 2020 HCRIS update of the 2017 Medicare cost reports for the Worksheet S-10 uncompensated care data. Id. at 32,756. CMS makes clear that it will use different data to determine DSH payments for uncompensated care in the final rule; CMS states it "intends to use more recent" HCRIS data for the calculation of Factor 1 and that it "intends to use the March 2020 update of HCRIS" Factor 3 Worksheet S-10 uncompensated care data. Id. at 32,748, 32,756. And, once again, the Secretary's proposed rule fixes payment amount determinations at the time when the agency will promulgate the final IPPS rule for 2021. Id. at 32,747.

The proposal to determine the amount of hospitals' new DSH payment based on data first released with the final rule, and on which hospitals will have no meaningful opportunity to comment, violates the notice and comment rulemaking requirements prescribed by the APA, 5 U.S.C. Sec. 553, and the Medicare statute, 42 U.S.C. Sec. 1395hh. See Allina Health Servs., 746 F.3d at 1110 (holding "an agency's failure to disclose critical material" deprives commenters the ability to participate in rulemaking). While CMS should address comments related to the impact of COVID-19 on their estimates in the final rule, by failing to include any discussion of this important topic in the proposed rule, hospitals will further be denied their opportunity to meaningfully comment on any impact of this important factor on their DSH uncompensated care payments.

III. Use of Best Data Available and Reconciliation Process

The above-described problems would be avoided if the amount of the DSH uncompensated care payments were finally determined, as traditional DSH payments are determined, when final payment determinations are issued in Notice of Program Reimbursements ("NPRs") for hospitals' cost reporting periods covering FFY 2021 and prior FFYs. See Dist. Hosp. Partners, L.P v. Burwell, 786 F.3d. 46, 56 (D.C. Cir. 2015) (holding "agencies do not have free rein to use inaccurate data"); Cape Cod Hosp. v. Sebelius, 630 F.3d 203, 214-15 (D.C. Cir. 2011) (prospective payment system does not permit agency to "continue using" inaccurate data "after the error in the data on which the Secretary has relied was brought to her attention"); Baystate Med. Ctr. v. Leavitt, 545 F. Supp. 2d 20, 41 (D.D.C. 2008) (explaining the repeated recognition in case law that in order for CMS to produce sufficiently accurate figures, it must use the most reliable data available).

In fact, CMS has taken an arbitrarily inconsistent approach as to finality of the DSH payments for uncompensated cares costs that reflects CMS's own recognition of the need to use the best available data. As noted above, CMS indicated in the proposed rule that the payment amounts published in the Supplemental Data File along with the final rule will not change. 85 Fed. Reg. at 32,747. But continuing the approach CMS took for FFYs 2014 through 2020, eligibility for the DSH payment for FFY 2021 is not determined at that time. Id. at 32,746. Rather, the DSH payment amount reflected in the supplemental file will not be paid to a hospital that does not ultimately qualify for a traditional DSH payment based on its actual number of low-income patient days for 2021. See id.; 42 C.F.R. Sec. 412.l06(g)(l),(h)(2). Thus, if a hospital does not end up with a sufficient number of low-income patient days to produce a disproportionate patient percentage that is great enough to qualify for the traditional (or "empirically justified") DSH payment, then the hospital will not be paid any DSH amount for uncompensated care costs, and any DSH uncompensated care payments made to the hospital on an interim basis during the course of the year will have to be refunded to the Secretary when the NPR is issued, years after the end of the hospital cost reporting periods. Final should not sometimes mean final and sometimes mean non-final.

To achieve the most accurate payments and avoid this inconsistency, CMS should use the traditional payment reconciliation process to calculate final payments for uncompensated care costs pursuant to Section l886(r)(2) of the Social Security Act. We do not object to CMS using prospective estimates derived from the best data available to calculate interim payments for uncompensated care costs in a FFY after 2013. Interim payments should be subject to later reconciliation, however, based on estimates derived from actual data for the FFY. This is the traditional process that the program has used to calculate Medicare payments for inpatient hospital services, including Medicare DSH payments made to hospitals over the past three decades. This process would not be any more time consuming either, as only a small fraction of all cost reports for FFY 2018 (322 out of 6,121 cost reports, or 5% of the total) have been finally settled as of the March 2020 HCRIS update.

Reconciliation is all the more important for the FY 2021 rulemaking, as the country is facing unprecedented upheaval due to COVID-19, and it would be impossible for anyone, including CMS, to accurately estimate the factors involved in calculating DSH uncompensated care payments. For example, even assuming the estimated 9.5% national uninsured rate in FY 2021 used to calculate Factor 2 was reasonable, that figure was created projecting from historical figures, and was calculated well before the economic downturn associated with the COVID-19 health crisis. See 85 Fed. Reg. at 32,750-51. Given that the unemployment rate has ballooned from 3.6% in January 2020 to 11.1% in June 2020,/5 it logically follows that the rate of uninsurance would have risen as well since employer-provided insurance would have decreased. While the estimates used to calculate Factor 2 in the final rule could and should be revised now that CMS can take into account the COVID-19 crisis, the most accurate approach would be to reconcile this data years later at the time hospital cost reports are settled and the relevant data is available.

Section 1886(r)(2) of the Act contemplates just such a reconciliation approach. The statute provides for the use of several estimated "factors" to calculate the uncompensated care cost payment under Section 1886(r)(2) of the Act. The statute requires the Secretary to determine the aggregate amount of DSH payments that would be made absent the new uncompensated care payment (Factor 1), the change in the uninsured rate (Factor 2) and the level of uncompensated care furnished by each hospital and all hospitals in the aggregate (Factor 3). But, the statute only requires that part of Factor 2 be finally determined in advance and not based on actual data for a given FY, and only for FYs 2014 through 2017. Specifically, Section 1886(r)(2)(B)(i)(I) of the Act directs the Secretary to calculate the change in the uninsured percentage for a FY from 2014 to 2017 compared to a baseline uninsured percentage in FY 2013 that is determined "based on the most recent estimates available from the Director of the Congressional Budget Office" before a House vote in 2010. By not imposing other similar restrictions on the data used for the factors, Section 1886(r)(2) of the Act calls for the Secretary to determine the other factors and the other parts of Factor 2 used in the calculation of the payment for uncompensated care costs based on actual data for the specific FY in which those costs are incurred. Using this approach, the factors would still be "estimates," but the estimates would be more accurate to the extent they would be derived from actual data for the FY and not mere extrapolations from priorperiod figures.

IV. Use of Worksheet S-10 Data in Calculation of Factor 3

CMS should not calculate Factor 3 in the DSH uncompensated care formula by comparing audited FY 2017 uncompensated care amounts for some hospitals with unaudited amounts or amounts reflecting the contractor's preliminary, non-final adjustments for other hospitals. The statute requires a determination of each hospital's "uncompensated care" as a percentage of the total "uncompensated care" for all hospitals. The comparison of audited uncompensated care amounts for some hospitals with unaudited amounts for the vast majority of all hospitals creates a distortion guaranteed not to produce an appropriate determination of hospitals' respective shares of uncompensated care. That skewed methodological approach is both ultra vires and otherwise arbitrary and capricious. See UnitedHealthcare Insurance Co. v. Azar, 330 F. Supp. 3d 173 (D.D.C. 2018) (overturning CMS's comparison of unaudited fee-for-service data with audited Medicare Advantage patient data in computing alleged "overpayments" to Medicare Advantage plans); H. Lee Moffitt Cancer Ctr. & Research Institute Hosp. v. Azar, 324 F. Supp. 3d 1 (D.D.C. 2018) (ruling that CMS's denial of cancer center payment adjustment for 2011 is invalid and not immune from administrative and judicial review because the denial violated the plain language of the statute and was ultra vires).

V. Agency Accountability and Preclusion of Review

The need for meaningful engagement on concerns raised in the rulemaking process is all the more important given the statutory provision precluding administrative and judicial review of estimates of the Secretary in Section 1886(r)(3) of the Act. Given the agency's broad interpretation of this provision to foreclose all judicial and administrative review, including any administrative appeals process, it is critical that the agency correct errors in the agency's calculations and estimates that are identified by commenters during the rulemaking process. The preclusion of review provision leaves intact the agency's responsibilities under the APA, 5 U.S.C. Sec. 553, and the rulemaking provisions of the Medicare Act, 42 U.S.C. Sec. 1395hh. See Yale New Haven Hosp. v. Azar, 2020 WL 2204197 (D. Conn. May 6, 2020) (finding that the FFY 2014 rulemaking was defective and that the hospital's challenge to these rulemaking infirmities was not prohibited from judicial review by Section 1886(r)(3)). And the preclusion of review provision in no way affects the agency's mandate to use "appropriate data" in its estimates. 42 U.S.C. Sec. 1395ww(r)(2)(C).

REVISED REGULATIONS TO ACCOUNT FOR AND MANDATE PRRB ELECTRONIC FILING

We are concerned with the Secretary's proposal to require electronic filing for all appeals before the PRRB. Now is not the time to impose new requirements on hospitals, and the Office of Hearings Case and Document Management System ("OH CDMS") does not allow users the flexibility and security that is required with appeal filings.

We recommend that CMS withdraw its proposal to mandate electronic filing of PRRB appeals. As stated in the proposed rule, the Board has heard provider appeals since 1972. 85 Fed. Reg. at 32,865. Paper documents were the only acceptable form of appeal filing for the first 46 years of the Board's history, and hospitals have developed their internal appeal processes accordingly. Id. Under normal circumstances it would be unreasonable to require hospitals and their representatives to abandon their time-honed appeal processes completely, on sixty days' notice, as early as this November. 85 Fed. Reg. at 32,866. However, hospitals are not operating under normal circumstances. The COVID-19 pandemic continues to harm communities and strain hospital resources, with no end in sight. CMS has rightfully recognized the need for flexibility in hospital operations by loosening requirements and granting waivers as needed. See, e.g., COVID-19 Emergency Declaration Blanker Waivers for Healthcare Providers (June 25, 2020)./6

The proposal to require electronic filing, which does not address the impact of COVID19 on hospitals, takes the opposite approach to provider appeals and would impose a new appeal process, with short notice, during an unprecedented global pandemic. We believe that forcing such requirements on hospitals at this time would be a mistake, and we urge the Secretary to table this proposal.

If the Secretary persists in imposing these requirements on hospitals, we recommend that the OH CDMS system be modified to better accommodate provider appeals. As currently constructed, the electronic filing system takes a "one-size-fits-all" approach to appeal filings and does not allow hospitals any flexibility to address exigent circumstances, either with the filing process or with the content of an appeal. For example, the current proposal does not account for situations where a provider or representative is unable to complete the electronic filing process, either due to problems on the filer's end, or problems with the electronic filing system. If the Board insists on paperless appeals in all cases,/7 then it should follow the lead of the federal district court's CM/ECF system by allowing appellants to e-mail filings to a dedicated e-mail inbox when filing through the electronic system is not possible. See Civil ECF Filing Pointers, United States District Court for the District of Columbia (Oct. 2019)./8

("If technical difficulties arise (on our end or yours), documents may be emailed to the Clerk's Office[.]").

In addition, the current proposal is too rigid in its treatment of provider appeals, which often require flexibility that is not built into the OH CDMS System. For example, PRRB instructions require all filers to make certain certifications regarding their appeals and possible appeals by related hospitals. See PRRB Rules 6.5, 12.10. However, the OH CDMS system allows hospitals and representatives no flexibility in how to accurately frame these certifications to reflect the hospital's unique circumstances. The electronic system presents users with a series of blanket statements, and if the user does not check the corresponding boxes in the interface, the appeal cannot proceed. These online certifications require users to conclusively state that the appeal issues are not pending in any other appeal, which imputes knowledge of facts that cannot be established (i.e., a user can never know with absolute certainty whether another party has mistakenly filed an appeal on the same issue). Under the current system, in a paper-based appeal, a provider representative may honor the Board's certification requirements, but may couch those statements to be more accurate. We recommend that CMS change this element of the filing system to allow providers and their representatives more flexibility in their certifications.

The rigidity of the online filing system, and its failure to account for different types of appeals also results in a time-consuming filing process. For each case, a provider or representative must separately enter multiple data elements, including the provider name, FY end, Medicare Administrative Contractor, audit adjustment numbers and amounts in controversy; and must separately upload each document that comprises the appeal, such as the NPR, issue statement, authorization of representative letter, audit adjustment report excerpts and cost report excerpts. If the filing requires exhibits, then each exhibit must be uploaded separately, and it is not uncommon for appeals to rely on dozens, if not hundreds, of exhibits. This process will be especially time-consuming in appeals that challenge rulemaking notices in the Federal Register, as those appeals are typically brought by large hospital chains, face the same filing deadlines for each hospital, and rely on the same source documents. As the electronic filing system is currently designed, an appeal of a final rule would require the representative to upload essentially the same documentation for each hospital in the appeal group, whether that group includes two hospitals or fifty hospitals. To make this process more efficient, the Board should take steps to facilitate the mass upload of appeal documents, including allowing group appeals the ability to use a single uploaded document for every hospital in a group appeal.

Finally, we have concerns about quality control and information security on the OH CDMS system. As opposed to a paper filing where the Board only receives the final product, the electronic filing system requires a piecemeal approach to appeal filings, in which discreet data elements must be manually entered by the user. This increases the chance of user error and could have unintended consequences for providers. To minimize these risks, we recommend that the OH CDMS interface be updated to allow hospitals to submit appeal filings in their complete, final form, with minimal data entry. Under the district court's CM/ECF document management system, for example, the only data entry required with respect to the parties is identification of the names and addresses of the plaintiffs. Further, as noted above, the time consuming process of filing an appeal and entering the required data elements is time-consuming, which could leave the data vulnerable.

Outside of appeal filings, we also have data security concerns with the OH CDMS user enrollment process, which requires applicants to provide the agency with their Social Security Numbers and purports to run a limited credit check for each individual that wants to use the system. The federal government is well aware that data breaches can affect even the best-designed and most well-intentioned records systems. See Cybersecurity Incidents, Office of Personnel Management (describing theft of personal information of over 25 million federal employees and job applicants in 2015, including theft of more than 21.5 million individual Social Security Numbers)./9

In this age, against that backdrop, it seems reckless and inappropriate for CMS to require this sort of personal information for individuals that are seeking to file institutional appeals, and we urge the agency to reconsider this element of the application process.

We appreciate the opportunity to comment on these proposals. Thank you for your consideration of these comments.

Sincerely,

Stephanie A. Webster

* * *

Footnotes:

1/ Congress has reinforced the prohibition against retroactivity by prohibiting the Secretary from taking action "against a provider of services . . . with respect to noncompliance with . . . a substantive change for items and services furnished before the effective date of such a change." 42 U.S.C. Sec. 1395hh(e)(1)(C).

2/ A link to the supplemental data file for the proposed rule is available at: https://www.cms.gov/files/zip/medicaredshsupplemental-data-file-fy-2021-proposed-rule.zip

3/ Available at https://www.cbo.gov/system/files/l I 5th-congress-2017-20 I 8/reports/53091-fshic.pdf

4/ Available at https://www.whitehouse.gov/wp-content/uploads/2018/02/ERP _2018_Final-FINAL.pdf

5/ U.S. Bureau of Labor Statistics, available at https://data.bls.gov/timeseries/LNS14000000?years_option=all_years.

6/ Available at https://www.cms.gov/files/document/summary-covid-19-emergency-declaration-waivers.pdf.

7/ The Board seemingly realizes that some hard copy documents are necessary for provider appeals, as it still requires appellants to submit paper copies of the voluminous Schedules of Providers that are filed in group appeals. See PRRB Rule 20.1, Commentary (explaining that, despite its move toward electronic filing, "until further notice, the Board is still requiring a hard copy of the Schedule of Providers and its accompanying supporting documentation.")(emphasis in original). If the Board insists on receiving paper copies of appeal filings for its own convenience, then providers and their representatives should be extended the same courtesy: the option to file paper-based appeals.

8/ Available at https://www.dcd.uscourts.gov/sites/dcd/files/CivilFilingPointers.pdf.

9/ Available at https://www.opm.gov/cybersecurity/cybersecurity-incidents/.

* * *

The proposed rule can be viewed at: https://www.regulations.gov/document?D=CMS-2020-0052-0002

TARGETED NEWS SERVICE (founded 2004) features non-partisan 'edited journalism' news briefs and information for news organizations, public policy groups and individuals; as well as 'gathered' public policy information, including news releases, reports, speeches. For more information contact MYRON STRUCK, editor, [email protected], Springfield, Virginia; 703/304-1897; https://targetednews.com