Grant Medical Center v. Eric D. Hargan

United States Court of Appeals

FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued September 28, 2017 Decided November 17, 2017

No. 16-5314

GRANT MEDICAL CENTER, ET AL.,

APPELLANTS

v.

ERIC D. HARGAN, ACTING SECRETARY, UNITED STATES

DEPARTMENT OF HEALTH AND HUMAN SERVICES,

APPELLEE

Appeal from the United States District Court

for the District of Columbia

(No. 1:15-cv-00480)

Daniel C. Gibson argued the cause for appellants. With

him on the briefs was James F. Flynn.

Weili J. Shaw, Attorney, U.S. Department of Justice,

argued the cause for appellee. With him on the brief was

Michael S. Raab, Attorney.

Before: GARLAND, Chief Judge, TATEL, Circuit Judge, and

GINSBURG, Senior Circuit Judge.

Opinion for the Court filed by Circuit Judge TATEL.

2

TATEL, Circuit Judge: One of our sister circuits, the Sixth,

ruled that the Centers for Medicare and Medicaid Services’

(CMS) method for counting hospital beds conflicted with the

plain language of the applicable regulation. CMS amended the

regulation to permit its preferred counting method but—central

to this case—applied the Sixth Circuit’s interpretation to

hospitals located within that circuit until the revised regulation

took effect. Appellants, hospitals in the Sixth Circuit, challenge

CMS’ decision to acquiesce to the Sixth Circuit’s ruling. Given

that obeying judicial decisions is usually what courts expect

agencies to do, the hospitals face an uphill battle. The district

court found that the agency acted reasonably, and we agree.

I.

Medicare reimburses hospitals for providing inpatient care

through the Inpatient Prospective Payment System. 42 U.S.C.

§ 1395ww(a), (d). Under that system, Medicare pays hospitals

a fixed amount for each patient regardless of the actual costs

incurred. 42 C.F.R. § 412.2(a). In order to account for certain

differences among hospitals, the reimbursement formula

includes several supplemental adjustments. See 42 U.S.C.

§ 1395ww(d). Two such adjustments are at issue here: the

Indirect Medical Education (IME) adjustment, which

supplements payments to hospitals that train medical residents,

id. § 1395ww(d)(5)(B), and the Disproportionate Share

Hospital (DSH) adjustment, which supplements payments to

hospitals that serve a disproportionate share of low-income

patients, id. § 1395ww(d)(5)(F). Both adjustments turn, in part,

on the number of inpatient beds at the hospital. Also, due to the

particularities of the formulas, hospitals claiming the IME

adjustment generally benefit when the bed count is lower, see

42 C.F.R. § 412.105, while hospitals claiming the DSH

adjustment benefit when the bed count is higher, see id.

§ 412.106.

3

Under this “complex and highly technical regulatory

program,” Thomas Jefferson University v. Shalala, 512 U.S.

504, 512 (1994) (quoting Pauley v. BethEnergy Mines, Inc.,

501 U.S. 680, 697 (1991)), counting beds is no simple matter.

A hospital’s bed count is calculated according to an intricate

formula set forth at 42 C.F.R. § 412.105(b). Prior to October 1,

2003, that regulation provided:

[T]he number of beds available in a hospital is

determined by counting the number of available bed

days during the cost reporting period, not including

beds or bassinets in the healthy newborn nursery,

custodial care beds, or beds in excluded distinct part

hospital units, and dividing that number by the

number of days in the cost reporting period.

42 C.F.R. § 412.105(b) (2002).

At issue in this case are two types of beds occasionally

used for inpatient care but unmentioned in section 412.105(b)’s

express exclusions: “swing beds” and “observation beds.”

Swing beds, found primarily in small rural hospitals, change in

reimbursement status depending on whether the facility is

using the bed for acute care or skilled nursing care. See

Medicare Program; Changes to the Hospital Inpatient

Prospective Payment Systems and Fiscal Year 2004 Rates, 68

Fed. Reg. 45,346, 45,418–19 (2003). Observation beds are

short-term beds used for outpatient care when a patient has not

been formally admitted to the hospital. See id. Even though

section 412.105(b) did not expressly exclude swing or

observation beds, the “longstanding policy” of CMS, which

administers Medicare on behalf of the United States

Department of Health and Human Services (HHS), was to

4

exclude these beds when calculating bed counts. See id.; Joint

Stipulations ¶ 2.

In 2001, two Kentucky hospitals that fell short of the bed

count needed to qualify for the DSH adjustment challenged

CMS’ interpretation of section 412.105(b). See Clark Regional

Medical Center v. HHS, 314 F.3d 241, 242 (6th Cir. 2002). The

hospitals argued that excluding swing and observation beds

conflicted with section 412.105(b)’s plain text. In Clark

Regional Medical Center v. HHS, the Sixth Circuit agreed,

explaining that “[b]ecause the regulation specifically lists

certain types of beds that are excluded from the bed count, but

does not list swing or observation beds, the plain meaning of

the regulation suggests that it is permissible to count swing and

observation beds.” Id. at 247.

In response to the Sixth Circuit’s decision in Clark, CMS

amended section 412.105(b) through notice-and-comment

rulemaking to expressly exclude swing and observation beds.

See Medicare Program; Proposed Changes to the Hospital

Inpatient Prospective Payment Systems and Fiscal Year 2004

Rates, 68 Fed. Reg. 27,154, 27,205–06, 27,229 (May 19, 2003)

(notice of proposed rulemaking); 68 Fed. Reg. at 45,470 (final

rule). CMS explained that, despite its longstanding policy of

excluding swing and observation beds, “courts have applied

our current rules in a manner that is inconsistent with our

current policy and that would result in inconsistent treatment.”

68 Fed. Reg. at 45,416 (discussing Clark). The effective date

of the revised regulation was October 1, 2003. Id. at 45,346.

CMS has taken two additional actions relevant to the issue

before us. First, to address reimbursement claims for patients

discharged prior to the effective date of the revised regulation,

the agency issued Joint Signature Memorandum 109 (JSM-

5

109). For hospitals located within the Sixth Circuit, CMS stated

that it would comply with Clark and include swing and

observation beds in the total bed count. But for hospitals

located outside the Sixth Circuit, CMS maintained its policy of

excluding swing and observation beds from the total bed count.

Second, in St. Vincent Mercy Medical Center v. Blue Cross

Blue Shield Association, CMS Adm’r Dec., 2008 WL 6468508

(Nov. 17, 2008), CMS affirmed its commitment to follow

Clark and JSM-109 for pre-October 2003 reimbursement

claims at hospitals within the Sixth Circuit. In that case, an

Ohio hospital challenged CMS’ decision to comply with Clark

and include observation beds when calculating total beds for

purposes of the DSH adjustment. The Administrator rejected

the claim, reasoning that “[g]enerally, when a court determines

that an agency’s interpretation is inconsistent with the language

of the regulation, an agency may recognize that court’s

interpretation and apply the court’s interpretation uniformly,

thereafter, within the jurisdictional bounds of the interpreting

court.” Id. at *9. This approach, the Administrator explained,

ensured that “all similarly situated providers are treated the

same for the applicable cost reporting periods” and facilitated

“the orderly administration of a complex and time sensitive

program.” Id. at *10. In the years following St. Vincent, CMS

has continued to apply Clark to reimbursement claims for pre-

October 2003 discharges filed by hospitals within the Sixth

Circuit. See, e.g., Clinton Memorial Hospital v. Blue Cross

Blue Shield Association, CMS Adm’r Dec., 2010 WL 5570983,

at *9 (July 26, 2010).

Appellants in this case are ten Ohio hospitals (“Hospitals”)

operating within the Sixth Circuit. For reasons we need not

probe, the Hospitals would like swing and observation beds

excluded from their total bed count when calculating

6

reimbursements for discharges prior to October 1, 2003. The

Provider Reimbursement Review Board (PRRB)—the body

responsible for initially hearing Medicare reimbursement

disputes—rejected the Hospitals’ claim. See OhioHealth 2004

Clark Bed Days Group v. BlueCross BlueShield Association,

PRRB Dec. No. 2015-D1, 2015 WL 10739301, at *9 (Jan. 29,

2015). The PRRB explained that the “inclusion of observation

bed days and swing bed days” for discharges prior to October

2003 “was correct . . . as all the providers are located within the

Sixth Circuit and the Clark decision is controlling legal

precedent.” Id. The PRRB also noted that it “concurs with the

Administrator in St. Vincent that the ‘separation of powers

doctrine requires administrative agencies to follow the law of

the circuit whose courts have jurisdiction over the cause of

action.[’]” Id. at *8 (quoting St. Vincent, 2008 WL 6468508, at

*9).

The Administrator declined to review the PRRB’s

decision, and it became final. See 42 U.S.C. § 1395oo(f)(1).

Exercising their prerogative under a venue-choice provision,

the Hospitals challenged the decision in the United States

District Court for the District of Columbia. Id. (providing that

an appeal “shall be brought in the district court of the United

States for the judicial district in which the provider is located

. . . or in the District Court for the District of Columbia”). The

district court granted summary judgment in favor of the

Secretary, see Grant Medical Center v. Burwell, 204 F. Supp.

3d 68, 71 (D.D.C. 2016), and this appeal followed.

“Because we apply the same standard of review as the

district court, we proceed de novo, as if [the plaintiff] had

brought the case here on direct appeal.” Tenet HealthSystems

HealthCorp. v. Thompson, 254 F.3d 238, 244 (D.C. Cir. 2001).

Thus, we review CMS’ decision under the Administrative

7

Procedure Act to determine whether it is “arbitrary, capricious,

an abuse of discretion, or otherwise not in accordance with

law.” 5 U.S.C. § 706(2)(A).

II.

The Hospitals believe that we can disregard the

acquiescence issue because CMS erred when it calculated bed

counts under the pre-2003 version of section 412.105(b) rather

than the revised version promulgated after Clark. Because the

parties agree that the revised regulation would require

excluding swing and observation beds—the Hospitals’

preferred outcome—the Hospitals see this as the simplest path

to relief. Unfortunately for the Hospitals, this simple solution

runs into a simple problem: their case concerns reimbursement

claims for discharges made prior to October 1, 2003, while the

revised regulation applies only to “discharges occurring on or

after October 1, 2003.” 68 Fed. Reg. at 45,346 (emphasis

added).

Undaunted, the Hospitals advance two arguments to

circumvent this clear effective-date provision. Neither is

convincing.

First, invoking the general rule that a “court is to apply the

law in effect at the time it renders its decision,” Bradley v.

School Board of City of Richmond, 416 U.S. 696, 711 (1974),

the Hospitals argue that the PRRB should have applied the

revised regulation because it was the law “in effect” when the

Board rendered its 2015 decision. See Appellants’ Br. 29. But

under a contrary presumption, “congressional enactments and

administrative rules will not be construed to have retroactive

effect unless their language requires this result.” Bowen v.

Georgetown University Hospital, 488 U.S. 204, 208 (1988).

Although our court has grappled with this “apparent

8

inconsistency in presumptions,” Gersman v. Group Health

Association, Inc., 975 F.2d 886, 889 (D.C. Cir. 1992), this

ambiguity exists “only in the absence of statutory terms clearly

directing the choice between retroactive and prospective

application,” id. at 888. Given that we apply a statute

prospectively when “congressional intent is clear,” Kaiser

Aluminum & Chemical Corp. v. Bonjorno, 494 U.S. 827, 837

(1990), we see no reason why the principle should be any

different when the agency’s intent is clear in a regulation.

Next, the Hospitals contend that the PRRB should have

applied the revised regulation retroactively because it

“clarified” rather than “changed” the law and because the

clarification inures to their benefit. This misses the mark in two

respects. First, even if the revised regulation merely reiterates

the law outside the Sixth Circuit, it still marks a “change” from

the interpretation of section 412.105(b) that CMS had

acquiesced to after Clark within that circuit until the revised

regulation took effect. Second, the Hospitals have this circuit’s

retroactivity law backwards: while we have prohibited

retroactive application of a rule that disadvantages a party by

“effect[ing] a substantive change from the agency’s prior

regulation,” National Mining Association v. Department of

Labor, 292 F.3d 849, 860 (D.C. Cir. 2002), we never require

agencies to apply rules retroactively even where it would be

permissible for them to do so. See Mountain Solutions, Ltd. v.

F.C.C., 197 F.3d 512, 520 (D.C. Cir. 1999) (“Because

rulemakings are generally prospective, there would appear to

be no basis for the court to fault the Commission for failing to

give [appellants] the benefit of its new rule.” (citations

omitted)). Had CMS decided to apply the revised regulation

retroactively, it might have sought to justify that decision by

presenting arguments similar to those advanced here by the

Hospitals. But CMS chose to apply the revised regulation only

9

prospectively, and the Hospitals have given no reason to

overturn that determination.

Focusing on the original bed-counting regulation, the

Hospitals argue that even if the old rule applies, the Sixth

Circuit’s Clark decision can have no bearing on our analysis of

whether CMS’ interpretation of the regulation was arbitrary or

capricious. Instead, they argue, having brought their case in this

circuit, they are entitled to a ruling based on this court’s

independent review of the issue.

This mischaracterizes the question before us. To

understand our task, we need look no further than the statute

that gives us jurisdiction, 42 U.S.C. § 1395oo(f)(1), which

states that “[p]roviders shall have the right to obtain judicial

review of any final decision of the Board.” In this case, then,

we must determine whether the PRRB’s final decision to

follow Clark when calculating reimbursement claims for

discharges prior to the revised regulation’s effective date was

“arbitrary, capricious, an abuse of discretion, unsupported by

substantial evidence, or otherwise not in accordance with the

law.” Sentara-Hampton General Hospital v. Sullivan, 980 F.2d

749, 755 (D.C. Cir. 1992) (citing 5 U.S.C. § 706(2)(A) and

(E)). In doing so, of course, we evaluate the reasons CMS gave

for complying with Clark. But ignoring Clark altogether would

require us to disregard the context and basis of the very

decision we are reviewing. See Independent Petroleum

Association of America v. Babbitt, 92 F.3d 1248, 1258 (D.C.

Cir. 1996) (holding that Department of the Interior’s

interpretation of a rule was unreasonable because it

“constitute[d] an unexplained departure” from the agency’s

prior adoption of a Fifth Circuit decision).

10

We turn, then, to the issue at the heart of this case: whether

CMS acted arbitrarily or capriciously when it decided to follow

Clark by excluding swing and observation beds when

calculating bed counts at the Hospitals’ facilities.

The Hospitals argue that the PRRB’s decision to acquiesce

in this case was arbitrary and capricious because it relied on the

faulty premise that the agency was required to follow Clark.

According to the Hospitals, the only rationale the PRRB gave

for following Clark was that “‘the separation of powers

doctrine requires administrative agencies to follow the law of

the circuit whose courts have jurisdiction over the cause of

action.[’]” OhioHealth, 2015 WL 10739301, at *8 (quoting St.

Vincent, 2008 WL 6468508, at *9).

Were this characterization of the PRRB’s decision correct,

we would have to reverse because, in this circuit, an agency

need not always acquiesce to an adverse ruling. See Johnson v.

U.S. Railroad Retirement Board, 969 F.2d 1082, 1093 (D.C.

Cir 1992) (“Although the decision of one circuit deserves

respect . . . ‘it need not be taken by the [agency] as the law of

the land.’” (quoting Givens v. United States Railroad

Retirement Board, 720 F.2d 196, 200 (D.C. Cir. 1983))).

Nonacquiescence may be particularly justified where, as here,

it occurs in the context of a “broad venue statute [that] often

forces the agency to act without knowing which circuit court

ultimately will review its actions.” Id. at 1091.

The Hospitals, however, take too narrow a view of the

PRRB decision. Where “an agency merely implements prior

policy,” as CMS did here by following St. Vincent, “an

explanation that allows this court to discern ‘the agency’s path’

will suffice.” WLOS TV, Inc. v. F.C.C., 932 F.2d 993, 995

(D.C. Cir. 1991) (quoting Hall v. McLaughlin, 864 F.2d 868,

11

872–73 (D.C. Cir. 1989)). And here, the PRRB’s repeated

references to St. Vincent make the agency’s path abundantly

clear. See OhioHealth, 2015 WL 10739301, at *8 (noting the

argument that “Providers here are situated similarly to . . . St.

Vincent Mercy Medical Center and that the Board should rule

consistently with . . . the Administrator’s decisions in St.

Vincent”). Our task, then, is to determine whether CMS’

incorporation of St. Vincent demonstrates that the agency

properly understood it had no obligation to acquiesce to Clark.

Reading the PRRB’s decision alongside the CMS

Administrator’s ruling in St. Vincent, we think it evident that

CMS’ view is not simply that it was required to acquiesce. See

Tourus Records, Inc. v. DEA, 259 F.3d 731, 738 (D.C. Cir.

2001) (holding that, although a court may not accept appellate

counsel’s post hoc rationalizations for agency actions, it may

consider contemporaneous documents outlining the agency’s

reasoning). To be sure, the Administrator did state that “the

separation of powers doctrine requires administrative agencies

to follow the law of the circuit whose courts have jurisdiction

over the cause of action,” but the context reveals that the

Administrator was merely describing the general disapproval

of intracircuit nonacquiescence. See St. Vincent, 2008 WL

6468508, at *9 n.21 (describing Johnson’s discussion of “true

intra-circuit refusal of an agency to recognize adverse

controlling case law” (citing Johnson, 969 F.2d at 1092)).

Immediately after making this observation, the Administrator

recognized that the principle of complying with adverse

judicial rulings “is more problematic when an agency is faced

with venue uncertainty.” Id. at *9. Nonetheless, the

Administrator explained, CMS “decided to apply the JSM to

ensure the orderly administration of a complex and time

sensitive program and despite the venue uncertainty.” Id. at *10

(emphasis added). Considering both the PRRB’s decision here

and the Administrator’s decision in St. Vincent that it relies

12

upon, we think it clear that CMS recognized its discretion not

to follow Clark but made a reasoned decision to do so.

The Hospitals also argue that the PRRB acted arbitrarily

and capriciously because it failed to give a rational explanation

for treating hospitals in the Sixth Circuit differently from

providers located elsewhere. Even if Clark constrained CMS’

options for counting beds in the Sixth Circuit, the Hospitals

believe that the better approach would have “treat[ed]

[providers] similarly based on their common interests vis a vis

the bed counting regulation (DSH versus IME).” Appellants’

Br. 64. The Hospitals style this argument as both an arbitrary-

and-capricious challenge and an equal-protection challenge.

Appellants’ Br. 15, 17. As they acknowledged at oral

argument, however, these two versions are “fundamentally

indistinguishable.” Oral Arg. Rec. 14:17–:27. Accordingly, we

consider these arguments together, reversing only if “‘the

agency offers insufficient reasons for treating similar situations

differently.’” Muwekma Ohlone Tribe v. Salazar, 708 F.3d

209, 215 (D.C. Cir. 2013) (quoting County of Los Angeles v.

Shalala, 192 F.3d 1005, 1022 (D.C. Cir. 1999)); cf. Ursack,

Inc. v. Sierra Interagency Black Bear Group, 639 F.3d 949,

955 (9th Cir. 2011) (noting that where “no suspect class is

involved” the “equal protection argument can be folded into the

APA argument”).

The mere fact that the Hospitals suggest an alternate

approach hardly means that CMS acted arbitrarily or

capriciously in choosing the approach it did. To the contrary,

the existence of different law in different circuits provides a

perfectly adequate reason for distinguishing between providers

based on location. As this circuit’s nonacquiescence case law

establishes, once a court rejects an agency’s interpretation of a

regulation, the agency is not free to simply ignore the ruling

13

within that court’s jurisdiction. Heartland Plymouth Court MI,

LLC v. NLRB, 838 F.3d 16, 22 (D.C. Cir. 2016) (noting that

intracircuit nonacquiescence may constitute bad faith if not

“clearly asserted and accompanied by a preservation of

arguments for Supreme Court or en banc review”). Rather, the

agency must choose between compliance or nonacquiescence

and, if opting for the latter, it must “specify[] its arguments

against adverse precedent” and seek further judicial review. Id.

Given that seeking en banc review or certiorari after Clark was

unrealistic—after all, CMS could correct the problem simply

by revising the regulation—it is hardly surprising that the

agency chose the former: acquiescing to the Sixth Circuit and

promptly revising the regulation. In explaining its decision, the

agency stated that complying with Clark would ensure that “all

similarly situated providers are treated the same for the

applicable cost reporting periods” and maintain “the orderly

administration of a complex and time sensitive program . . .

despite the venue uncertainty in PRRB cases.” St. Vincent,

2008 WL 6468508, at *10. Perhaps there were more elegant

solutions to CMS’ predicament, but our task is not to test

whether the agency chose “the best solution, only a reasonable

one.” Petal Gas Storage, L.L.C. v. FERC, 496 F.3d 695, 703

(D.C. Cir. 2007).

As a final matter, although the Hospitals never squarely

challenge the Sixth Circuit’s reading of section 412.105(b), we

emphasize that CMS’ acquiescence would have been

unacceptable had it rested on an unreasonable interpretation of

the regulation. See Holland v. National Mining Association,

309 F.3d 808, 812, 819 (D.C. Cir. 2002) (rejecting agency’s

contention that acquiescing nationwide to an interpretation

compelled by court order “cannot violate that ‘not in

accordance with law standard of the APA’”). We would be

unable to sustain an administrative action that conflicted with

the applicable statutory or regulatory text just because a sister

14

circuit has approved the interpretation or because the agency

acts out of “administrative concerns, such as the desirability of

uniformity.” Id. at 818; see also Atchison, Topeka & Santa Fe

Railway v. Peña, 44 F.3d 437, 446 (7th Cir. 1994) (cautioning

against agency action seeking to ensure that the “applicable

rule of law be settled [rather] than that it be settled right”

(quoting Burnet v. Coronado Oil & Gas Co., 285 U.S. 393, 406

(1932) (Brandeis, J., dissenting))), aff’d sub nom. Brotherhood

of Locomotive Engineers v. Atchison, Topeka & Santa Fe

Railway, 516 U.S. 152 (1996). That, however, is not the case

here. As CMS explained when it first articulated its

acquiescence decision in JSM-109, “[t]he regulations’ text was

silent” on the issue of counting swing and observation beds at

the time of the Clark decision. Given this ambiguity, CMS

believed that the Sixth Circuit’s interpretation was permissible,

even if not required, and we do as well.

III.

For the foregoing reasons, we affirm the district court’s

grant of summary judgment to CMS.

So ordered.