Hays Medical Center v. Azar ( 2020 )

FILED
United States Court of Appeals
Tenth Circuit
PUBLISH April 21, 2020
Christopher M. Wolpert
UNITED STATES COURT OF APPEALS Clerk of Court
TENTH CIRCUIT
HAYS MEDICAL CENTER; MERCY
HOSPITAL LEBANON; MERCY
HOSPITAL ARDMORE, INC.; NORTH
PLATTE NEBRASKA HOSPITAL
CORPORATION, d/b/a Great Plains
Medical Center; HANOVER HOSPITAL,
INC.; KNOX COMMUNITY HOSPITAL;
LABETTE COUNTY MEDICAL
CENTER; MEMORIAL HOSPITAL OF
SWEETWATER COUNTY; NEWMAN
MEMORIAL COUNTY HOSPITAL, d/b/a
Newman Regional Health;
NORTHWESTERN MEDICAL CENTER,
INC.; POCATELLO HOSPITAL, LLC,
d/b/a Portneuf Medical Center,
Plaintiffs - Appellants, No. 17-3232
and
RICHLAND MEMORIAL HOSPITAL,
Plaintiff,
v.
ALEX M. AZAR, II, in his official
capacity as Secretary of Health and Human
Services, *
Defendant - Appellee.
*
Pursuant to Fed. R. App. P. 43(c)(2), Eric Hargan is replaced by Alex M.
Azar II as the Secretary of Health and Human Services.
Appeal from the United States District Court
for the District of Kansas
(D.C. No. 2:15-CV-09893-JTM-GEB)
Daniel J. Hettich (Elizabeth N. Swayne with him on the briefs), King & Spalding
LLP, Washington D.C., for Plaintiffs - Appellants.
Carleen M. Zubrzycki, Attorney, Appellate Staff (Chad A. Readler, Acting
Assistant Attorney General, Stephen McAllister, United States Attorney, and
Michael S. Raab and Abby Wright, Attorneys, with her on the brief), United
States Department of Justice, Washington, D.C., for Defendant - Appellee.
Before TYMKOVICH, Chief Judge, HOLMES and PHILLIPS, Circuit Judges.
HOLMES, Circuit Judge.
Plaintiff-Appellants are eleven rural hospitals (the “Hospitals”). They
challenge the methodology that the U.S. Secretary of Health and Human Services
(the “Secretary”) uses to calculate their Medicare reimbursements. The question
before us is whether that methodology is arbitrary and capricious. We hold that it
is not. Accordingly, exercising jurisdiction under 28 U.S.C. § 1291, we affirm
the district court’s judgment.
I
This case plunges us into the “labyrinthine world” of Medicare
reimbursement. Adirondack Med. Ctr. v. Sebelius, 740 F.3d 692, 694 (D.C. Cir.
2
2014). 2 In order to navigate this world, we first lay out the necessary background,
which we do in three steps. First, we outline the basics of the Medicare
reimbursement system. Second, we explain the Secretary’s methodology at issue
and trace its application over the years. Third, we recount how this appeal
unfolded.
A
The Medicare program provides federally funded healthcare to persons over
sixty-five years old, as well as to disabled persons. See Social Security
Amendments of 1965, Pub. L. No. 89-97, § 102, 79 Stat. 291 (1965) (codified as
amended at 42 U.S.C. § 1395 et seq.). Medicare enrollees obtain their healthcare
through different “parts.” This case involves what is known as Medicare Part A.
See 42 U.S.C. §§ 1395c–1395i-5. 3 That part provides eligible persons with
2
Throughout this opinion, we cite to four decisions bearing the case
caption (or some substantially similar version thereof) Adirondack Medical
Center v. Sebelius. Three out of four of these decisions were different
dispositions within the same case. As such, the citations for each of these three
decisions contain a Roman numeral, corresponding to the chronological order in
which the respective disposition was rendered in that case (e.g., “Adirondack I,”
“Adirondack II,” and “Adirondack III”). The fourth decision is Adirondack
Medical Center v. Sebelius, 740 F.3d 692 (D.C. Cir. 2014), which, despite bearing
the same case caption as the aforementioned three decisions, is not connected to
that same case. Accordingly, that decision is not denoted by a Roman numeral
identifier as are the other three, but rather is cited to hereinafter in traditional
shortcite form (e.g., Adirondack Med. Ctr., 740 F.3d at 692).
3
Notwithstanding certain adjustments to dates and other minor
changes, the statutory and regulatory provisions involved here have remained
substantively unchanged over the relevant years. Thus, for convenience, we cite
3
insurance covering, among other things, certain inpatient hospital services. See
id. § 1395d(a). The Secretary administers this program through the Centers for
Medicare & Medicaid Services (“CMS”), 4 a division of the Department of Health
and Human Services.
The Secretary reimburses participating hospitals for certain inpatient care
they provide to Medicare Part A patients through what is called the “inpatient
prospective payment system,” a regime established by Congress in the 1980s
aimed at encouraging more cost-efficient management of medical care by
hospitals. See Social Security Amendments of 1983, Pub. L. No. 98-21, § 601, 97
Stat. 65, 149 (1983) (codified as amended at 42 U.S.C. § 1395ww). Through this
system, the Secretary reimburses hospitals for inpatient services on a prospective
basis, doing so “at a fixed amount per patient, regardless of the actual operating
costs they incur in rendering these services.” Sebelius v. Auburn Reg’l Med. Ctr.,
568 U.S. 145, 149 (2013). “By establishing predetermined reimbursement rates
that remain static regardless of the costs incurred by a hospital, Congress sought
‘to reform the financial incentives hospitals face, promoting efficiency in the
the most recent iteration of these provisions. When a provision has been changed
and that change is relevant for purposes of this appeal, we note that change and
cite the controlling language.
4
For simplicity’s sake, we use the term “the Secretary” as a catchall to
cover any governmental actor—including CMS—involved in administering the
hospital-reimbursement scheme at issue here.
4
provision of services by rewarding cost/effective hospital practices.’” County of
Los Angeles v. Shalala, 192 F.3d 1005, 1008 (D.C. Cir. 1999) (quoting H.R. R EP .
N O . 98–25, at 132 (1983), as reprinted in 1983 U.S.C.C.A.N. 219, 351).
1
At its most basic, a hospital’s reimbursement for a given Medicare Part A
patient is the mathematical product of two figures—the payment rate and the
patient’s diagnosis-related group weight. We turn first to providing an overview
of the latter figure, the diagnosis-related group weight (the payment rate is
explored in the immediately succeeding subsection infra). Ascertaining this
figure begins with the Secretary sorting the Medicare Part A patient into a
“diagnosis-related group” (periodically referred to hereinafter as “DRG”). These
groups are essentially “categor[ies] of inpatient treatment” that reflect the
differing costs of treating various types of diagnoses, with each Part A patient
being sorted into a group at the time of discharge, based on that patient’s
diagnosis. Adirondack Med. Ctr., 740 F.3d at 694 n.1; see also 42 U.S.C.
§ 1395ww(d)(4)(A). The Secretary assigns each diagnosis-related group a
“weighting factor” that “reflects the relative hospital resources used” to treat
patients in that group “compared to [patients] classified within other groups.”
§ 1395ww(d)(4)(B). There are several hundred different diagnosis-related
groups, with individual weights “ranging from less than 1.000 to more than
5
7.000.” Adirondack Med. Ctr. v. Sebelius (Adirondack II), 29 F. Supp. 3d 25, 30
(D.D.C. 2014). The more complicated and expensive the diagnoses, “the greater
the weight assigned to that particular [diagnosis-related group] will be.” County
of Los Angeles, 192 F.3d at 1008. The Secretary then multiplies the Medicare
Part A patient’s diagnosis-related group weight by the payment rate (again, we
discuss this latter figure infra). The product of that calculation is the hospital’s
reimbursement for treating a particular patient. 5 See id.; see also 42 U.S.C.
§ 1395ww(d)(3)(D).
In order to account for changes in resource consumption, Congress
recognized that it would be necessary to periodically recalculate the diagnosis-
related group weighting factors. Changes to the Inpatient Hospital Prospective
Payment System and Fiscal Year 1991 Rates, 55 Fed. Reg. 35,990, 36,008 (“FY
1991 Final Rule”). Accordingly, at Congress’s direction, the Secretary annually
“adjust[s]” the diagnosis-related group “classifications and weighting factors.” Id.
(citing § 1395ww(d)(4)(C)(i)); see also Changes to the Hospital Inpatient
Prospective Payment Systems and Fiscal Year 2010 Rates, 74 Fed. Reg. 43,754,
43,895 (Aug. 27, 2009) (“FY 2010 Final Rule”) (noting that the Secretary makes
5
We offer a simplified example. A hospital treats two patients.
Patient 1 had a heart attack, placing her in a diagnosis-related group with a weight
of 3.0. Patient 2 had the flu, which has a diagnosis-related group weight of 1.0.
Assume that the Secretary pays the hospital using a payment rate of $1,000. For
Patient 1, the hospital would get $3,000. For Patient 2, it would get $1,000.
6
“changes to the [diagnosis-related-group] classifications and weighting factors”).
This adjustment in the classifications and weighting factors is designed “to reflect
changes in treatment patterns, technology . . ., and other factors which may change
the relative use of hospital resources.” § 1395ww(d)(4)(C)(i). 6
We focus here on the Secretary’s annual adjustment of the diagnosis-related
group weighting factors, or, in the common parlance of regulators, his
“recalibration” of the weights. The Secretary starts this process by assembling “a
dataset of recent patients.” J.A. at 159 (Mem. in Opp’n to Pls.’ Mot. for Summ. J.
& in Supp. of Def.’s Cross-Mot. for Summ. J., filed Oct. 20, 2016); see also 55
Fed. Reg. at 36,033 (FY 1991 Final Rule) (“One of the basic issues in
recalibration is the choice of a data base that allows us to construct relative DRG
weights that most accurately reflect current relative resource use.”). Based on this
dataset, the Secretary then calculates the average cost of treating patients across all
diagnosis-related groups and “the average cost of treating a patient in each
diagnosis-related group.” J.A. at 159. The Secretary then divides the average cost
for each diagnosis-related group by the overall average cost. See id. Using this
ratio, he then assigns a new weighting factor to each diagnosis-related group. See
6
For context, we note that the process of adjusting the diagnosis-
related group classifications entails, inter alia, moving some diagnoses from one
diagnosis-related group to another and sometimes creating new diagnosis-related
groups altogether. See, e.g., 55 Fed. Reg. at 36,010 (FY 1991 Final Rule) (adding
thirteen new DRGs).
7
id.; see also 42 C.F.R. § 412.60(a) (noting that the Secretary “assigns, for each
DRG, an appropriate weighting factor that reflects the estimated relative cost of
hospital resources used with respect to discharges classified within that group
compared to discharges classified within other groups”). 7
The recalibration process has the potential of leading to higher payments to
hospitals systemwide. However, Congress requires the Secretary to recalibrate “in
a manner that assures that the aggregate payments . . . are no greater or less than
those that would have been made [pre-recalibration].” 42 U.S.C.
§ 1395ww(d)(4)(C)(iii). In other words, Congress mandates that aggregate
payments remain budget neutral notwithstanding recalibration. See, e.g.,
Adirondack II, 29 F. Supp. 3d at 32 (“[T]he annual DRG recalibration must ‘be
made in a manner that assures that the aggregate payments . . . for discharges in
the fiscal year are not greater or less than those that would have been made for
discharges in the year without such adjustment.’” (omission in original) (quoting
§ 1395ww(d)(4)(C)(iii)), aff’d sub nom. Adirondack Med. Ctr. v. Burwell, 782
F.3d 707 (D.C. Cir. 2015)).
7
Here is a grossly simplified example. Say, after recalibration, the
overall average cost of treating Medicare patients is $1,000. And assume that the
average cost of treating patients in diagnosis-related group A is $1,500. Dividing
$1,500 by $1,000 (i.e., $1,500 ÷ $1,000) yields a new (post-recalibration)
weighting factor of 1.5 for that diagnosis-related group.
8
To achieve budget neutrality in aggregate payments, the Secretary takes two
steps. The Secretary’s first step is to “normalize” the recalibrated diagnosis-
related group weights to offset any change in the average weight caused by
recalibration. 74 Fed. Reg. at 4395 (FY 2010 Final Rule). The Secretary has
explained that normalization negates any increase in average diagnosis-related
group weights by “equating the average case weight after recalibration to the
average case weight before recalibration.” Changes to the Hospital Inpatient
Prospective Payment Systems and Fiscal Year 1994 Rates, 58 Fed. Reg. 46,270,
46,291 (Sept. 1, 1993) (“FY 1994 Final Rule”). 8 More specifically, the Secretary
applies a numerical adjustment “in order to ensure that the average [diagnosis-
related group] weight after recalibration is equal to the average [diagnosis-related
group] weight prior to recalibration.” Proposed Changes to the Hospital Inpatient
8
When the Secretary refers to normalized average diagnosis-related
group weights, he frequently uses the phrase “average case weight.” See, e.g., 74
Fed. Reg. at 43,896 (FY 2010 Final Rule); Changes to the Hospital Inpatient
Prospective Payment Systems and Fiscal Year 2006 Rates, 70 Fed. Reg. 47,278,
47,322 (Aug. 12, 2005) (“FY 2006 Final Rule”). Throughout this opinion, to
avoid introducing yet another term for readers to keep straight, we do not use the
phrase “average case weight.” We instead use “average diagnosis-related group
weight” or “average DRG,” which is the terminology the parties most often use.
See, e.g., Aplts.’ Opening Br. at 11 (“[W]hen the Secretary calculates the
hospital-specific rate by dividing by the hospital’s average DRG weight.”);
Aplee.’s Resp. Br. at 7 (referring to “the average diagnosis-related group weight”
(emphasis omitted)).
9
Prospective Payment Systems and Fiscal Year 2010 Rates, 74 Fed. Reg. 24,080,
24,184 (May 22, 2009) (“FY 2010 Proposed Rule”). 9
Normalization, however, “does not necessarily achieve budget neutrality
with respect to aggregate payments to hospitals because payment to hospitals is
affected by factors other than average [diagnosis-related group] weight.” 58 Fed.
Reg. at 46,291 (FY 1994 Final Rule); see also 74 Fed. Reg. at 43,895 (FY 2010
Final Rule) (noting that “our analysis has indicated that the normalization
adjustment does not usually achieve budget neutrality with respect to aggregate
payments to hospitals” (emphasis added)). For example, Congress requires the
Secretary to apply wage-index adjustments to hospital reimbursement rates to
account for varying labor costs of hospitals located in different geographic areas
of the country, as well as in urban and rural settings, and to do this “in a manner
that assures that aggregate payments to hospitals are not affected by the change in
the wage index.” 58 Fed. Reg. at 46,346 (FY 1994 Final Rule); see 42 U.S.C.
9
A simplified example might help. Assume that the average
diagnosis-related-group weight before recalibration was 2.0. And say that after
recalibration the average weight is 2.5. To offset this increase, the Secretary
would apply a normalization factor of 0.8 (i.e., 2.0 ÷ 2.5) to each patient’s post-
recalibration diagnosis-related group weight. See, e.g., Aplee.’s Resp. Br. at 25
(“To normalize the average weights across the two years (i.e., to ensure that the
average [diagnosis-related group] weight does not change), the Secretary would
multiply each patient’s diagnosis-related group weighting factor by 3/6 (the
pre-recalibration average weight, divided by the post-recalibration average
weight).”).
10
1395ww(d)(3)(E) (“[T]he Secretary shall adjust the proportion . . . of hospitals’
costs which are attributable to wages and wage-related costs, of the DRG
prospective payment rates . . . for area differences in hospital wage levels by a
factor (established by the Secretary) reflecting the relative hospital wage level in
the geographic area of the hospital compared to the national average hospital wage
level.”); see also 58 Fed. Reg. at 46,291 (FY 1994 Final Rule) (“Under the
Medicare prospective payment system, different payment rates are calculated for
hospitals located in rural, urban, and large urban areas.”).
Payment to hospitals is also affected by Congress’s requirement that the
Secretary “provide for an additional payment amount for” teaching hospitals “with
indirect costs of medical education.” § 1395ww(d)(5)(B); see Rush Univ. Med.
Ctr. v. Burwell, 763 F.3d 754, 755 (7th Cir. 2014) (“Teaching hospitals provide a
valuable service to the public by training the next generation of doctors and
medical professionals, but that benefit comes at a price: such hospitals experience
significantly higher per-patient care costs than their non-teaching counterparts. To
compensate them for taking on this extra financial burden, the federal Medicare
program provides additional reimbursement for expenses beyond the immediate
costs of patient care. One such adjustment is for ‘indirect medical education’
(IME) costs.”); Henry Ford Health Sys. v. Dep’t of Health & Human Servs., 654
F.3d 660, 663 (6th Cir. 2011) (“Under the Medicare program, teaching hospitals
11
receive additional payments, above and beyond the reimbursement rate for treating
Medicare patients, to cover the ‘direct’ and ‘indirect costs of medical education.’”
(quoting § 1395ww(d)(5)(B), (h))).
In light of such factors, normalization of the average diagnosis-related
group weights will not necessarily result in budget neutrality of aggregate
payments systemwide. That is because those factors affect—and frequently
increase—the hospitals’ payment rate and, in computing a hospital’s
reimbursement, “a patient’s [normalized] diagnosis-related group weight is
multiplied by the hospital’s applicable [payment] rate.” Aplee.’s Resp. Br. at 7. 10
10
The Secretary argues that “[g]iven the interaction of other Medicare
payment adjustments . . . normalization usually does not lead to budget
neutrality.” Aplee.’s Resp. Br. at 24. He contends that this is especially so when
the high-weight diagnosis-related group weights are disproportionately found in
hospitals that receive higher payments because of adjustments, like the wage
index. See id. (“[I]f urban hospitals with high wage-index adjustments have a
disproportionate number of patients with a high diagnosis-related group weighting
factor, then normalization will not result in budget neutrality.” (emphasis added));
see also 55 Fed. Reg. at 36,074 (FY 1991 Final Rule) (noting that “urban
hospitals generally have a greater proportion of cases concentrated in the
high-weighted DRGs . . . weights will tend to increase the average DRG weight
for these hospitals and, therefore, raise their payments under the prospective
payment system”). The Secretary offers a mathematical hypothetical to support
this view in which normalization of average diagnosis-related group weights does
not result in budget neutrality in part because a patient with a comparatively high
diagnosis-related group weight is admitted to a hospital with a very high payment
rate “due, for instance, to a high wage index or an adjustment for teaching
hospitals.” Aplee.’s Resp. Br. at 24.
12
Accordingly, to ensure budget neutrality in aggregate payments, the
Secretary takes a second step: he calculates “a budget neutrality adjustment” to
account for those other factors, such that the adjustment is in fact “budget
neutral.” 58 Fed. Reg. at 46,291 (FY 1994 Final Rule); see 55 Fed Reg. at
36,073–74 (FY 1991 Final Rule) (noting the “interactive effect of the wage index
and DRG weights on aggregate payments” and that the Secretary applied a “budget
neutrality adjustment factor” to the hospital-specific rates because unless he did
so, he would not be able to “meet the statutory requirement that aggregate
payments neither increase nor decrease”). “This budget-neutrality adjustment
takes the form of a multiplier that the Secretary applies in calculating the
hospital’s base payment rate[.]” Aplee.’s Resp. Br. at 8.
Since Fiscal Year (“FY”) 1994, the Secretary has applied the budget-
neutrality adjustment “in a cumulative manner,” commencing with the FY 1993
adjustment factors for that inaugural cumulative computation. Aplts.’ Opening Br.
at 10; see 70 Fed. Reg. at 47,429 (FY 2006 Final Rule). That is to say, he “does
not remove the prior year’s budget-neutrality adjustment before calculating and
applying the current year’s adjustment.” Aplee.’s Resp. Br. at 8; see id. (“[A]ll
diagnosis-related group budget-neutrality adjustments that were applied each fiscal
year from FY 1993 onward are permanently incorporated into a later year’s
rates.”); see also 58 Fed. Reg. at 46,346 (FY 1994 Final Rule) (explaining that the
13
“budget neutrality adjustment factor is applied . . . without removing the effects of
the FY 1993 budget neutrality adjustment”). The Secretary then applies this
cumulative budget-neutrality adjustment factor to the payment rate used in the
reimbursement formula. See 74 Fed. Reg. at 43,895 (FY 2010 Final Rule).
2
Having defined and described the term “diagnosis-related group weight,” we
turn now to explicating the other major component undergirding a hospital’s
reimbursement: the payment rate. The Secretary reimburses most hospitals using
what is called the federal rate. The starting point in calculating the federal rate in
a given year is the “average standardized amount,” which is essentially the
estimated average per-patient operating costs of inpatient hospital services for all
hospitals within a geographic region. See 42 U.S.C. § 1395ww(d)(1)(A)(iii),
(d)(3)(A); 42 C.F.R. § 412.64. Each year, the Secretary updates the standardized
amount to account for variables such as inflation and wage levels. See 42 U.S.C.
§ 1395ww(b)(3)(B)(i), (d)(3)(E). Once an updated standardized amount is
calculated, the Secretary multiplies it by the cumulative budget-neutrality
adjustment. 58 Fed. Reg. at 46,358 (FY 1994 Final Rule) (in context of discussion
about calculating the federal rate for FY 1994, explaining that the budget-
neutrality adjustment “is applied to the standard Federal payment rate”). To
calculate the actual reimbursement for a specific patient’s inpatient care, the
14
mathematical product of the standardized amount and the budget neutrality
adjustment is multiplied by the patient’s diagnosis-related group weight. 42
U.S.C. § 1395ww(d)(3)(D).
Although, as discussed, the above-noted federal rate is the mode of
reimbursement that is employed for most hospitals, because certain rural hospitals
“provide critical services to the underserved and uninsured, Congress has adopted
special payment provisions for them.” J.A. at 286 (Dist. Ct. Mem. & Order, dated
Aug. 31, 2017) (quoting Adirondack II, 29 F. Supp. 3d at 32). As relevant here,
those rural hospitals are “sole community” and “Medicare-dependent” hospitals. 11
See, e.g., Cmty. Hosp. v. Sullivan, 986 F.2d 357, 358 (10th Cir. 1993) (noting
Congress has evinced special concern for sole community hospitals). It is
undisputed that each of the Hospitals here is either a sole community or Medicare-
dependent hospital.
11
A sole community hospital is one that is (1) “located more than 35
road miles from another hospital,” (2) “the sole source of inpatient hospital
services reasonably available” to Medicare Part A patients in the area, and (3)
“located in a rural area and designated by the Secretary as an essential access
community hospital.” 42 U.S.C. § 1395ww(d)(5)(D)(iii). A Medicare-dependent
hospital is one (1) that is “located in a rural area,” (2) that has fewer “than 100
beds,” (3) that “is not classified as a sole community hospital,” and (4) “for which
not less than 60 percent of its inpatient days or discharges . . . were attributable to
inpatients entitled to benefits under [Medicare Part A].” Id.
§ 1395ww(d)(5)(G)(iv).
15
Chief among the special payment provisions Congress has adopted for sole
community and Medicare-dependent hospitals is the option of selecting
reimbursement for covered healthcare based on a “hospital-specific rate,” instead
of the federal rate. See 42 U.S.C. § 1395ww(d)(5)(D)(i), (d)(5)(G)(i)–(ii);
Adirondack Med Ctr., 740 F.3d at 695 (noting that these rural hospitals “have the
option” of receiving the higher of the hospital-specific and the federal rate
“[b]ecause these facilities typically serve underserved communities”). As its name
suggests, the hospital-specific rate is different for each eligible hospital.
The Secretary’s starting point for calculating a hospital-specific rate is the
hospital’s “target amount” in a given “base year.” See 42 U.S.C.
§ 1395ww(d)(5)(D), (d)(5)(G). A target amount is a hospital’s “historic operating
costs” in a specific year. Adirondack Med. Ctr., 740 F.3d at 695; see also 42
U.S.C. § 1395ww(b)(3)(A). Base years are certain statutorily designated years that
Congress has authorized for use in determining a hospital’s corresponding historic
operating costs—viz., its target amount—with hospitals being free to select the
base year that will yield them the highest payment. 12 See 42 U.S.C.
12
Sole community hospitals, for example, are authorized to select
either FY 1982, 1987, 1996, or 2006 as a base year. Medicare-dependent
hospitals, for their part, may select either FY 1982, 1987, or 2002 as a base year.
See 74 Fed. Reg. at 43,760–61 (FY 2010 Final Rule). The Hospitals are sole
community hospitals that the Secretary paid based on their FY 2006 base-year
rates and a Medicare-dependent hospital that the Secretary paid based on its FY
2002 base-year rate. See Aplts.’ Opening Br. at 7.
16
§ 1395ww(b)(3)(C), (b)(3)(D). Of particular relevance here, and as further
discussed infra, Congress periodically adds new base years. See, e.g., id.
§ 1395ww(b)(3)(G).
As with the federal rate, the Secretary annually adjusts the hospital-specific
rate to account for variables such as inflation and wage levels. See id.
§ 1395ww(b)(3)(B)(i), (d)(3)(E); 42 C.F.R. §§ 412.78(e), 412.79(d). He also
applies the cumulative budget-neutrality adjustment to the hospital-specific rate.
See 74 Fed. Reg. at 43,895 (FY 2010 Final Rule); see also 42 C.F.R. §§ 412.77(j),
412.79(i). Finally, to determine payment for a specific patient, the Secretary
multiplies the hospital-specific rate by each patient’s diagnosis-related group
weight, with the product being the hospital’s reimbursement.
Although both sole community and Medicare-dependent hospitals may select
a “payout” based on a hospital-specific rate, the nature of the payout “differs
slightly” between the two hospitals. Adirondack Med. Ctr., 740 F.3d at 695 n.2.
Reimbursement for sole community hospitals is “fairly straightforward”: they are
to be paid the higher of either the hospital-specific rate or the federal rate. Id.; see
42 U.S.C. § 1395ww(d)(5)(D)(i). On the other hand, Medicare-dependent
hospitals are to be paid the federal rate plus “75% of the difference between the
federal rate payment and the hospital-specific rate payment.” Adirondack Med.
Ctr., 740 F.3d at 695 n.2; see 42 U.S.C. § 1395ww(d)(5)(G)(ii)(II).
17
B
Having laid out the basics of the Medicare Part A reimbursement system, we
can focus on the part of that system at issue here—the Secretary’s methodology for
calculating the hospital-specific rate when Congress adds a new base year. In the
following subsections, we explain how that methodology works and how the
Secretary has applied it over the years.
1
When Congress adds a new base year, the hospital-specific rate for that new
year is calculated according to the following four-step formula:
Step One: Divide the hospital’s target amount for the new base year “by
the number of discharges in the base [year].” 42 C.F.R. §§ 412.78(c),
412.79(b).
Step Two: Divide the figure from step one by the hospital’s normalized
average diagnosis-related group weight for the new base year. See id.
§§ 412.78(d), 412.79(c).
Step Three: Apply an update factor to account for specific variables
such as inflation and wage levels. See id. §§ 412.73(c), 412.78(e),
412.79(d).
Step Four: Apply a cumulative budget-neutrality adjustment to the
resulting rate figure. See id. §§ 412.78(j), 412.79(i); 74 Fed. Reg. at
43,895–96 (FY 2010 Final Rule). In this context, “cumulative” means
that the Secretary does not remove the budget adjustments from years
predating the new base year. 13
13
The formula is stated here in highly simplified form to highlight the
area of contention between the parties: the interaction between step 2 and step 4.
The Secretary, for example, states the formula in eight steps, detailing matters not
directly bearing on this dispute. See Aplee.’s Resp. Br. at 29–30.
18
The amount this formula yields is the hospital-specific rate for a given
hospital in the new base year. That rate is then multiplied by each patient’s
diagnosis-related group weight, which yields the hospital’s reimbursement for each
patient’s inpatient care.
The Hospitals do not challenge steps one and three. And much of steps two
and four is also uncontroversial. The Hospitals agree, for instance, “that the
average DRG weight used as a divisor [in step two] has been normalized.” Aplts.’
Reply Br. at 17. And they “do not object to the fact that DRG weights are used as
a divisor” in step two. Aplts.’ Opening Br. at 9 n.4. Likewise, “[t]he Hospitals do
not object to prior-year adjustments being applied once.” Id. at 18. In fact, the
Hospitals do not even complain about the cumulative nature of this adjustment.
See Aplts.’ Reply Br. at 2 (“The Hospitals explicitly stated on several occasions . .
. that their contention is not that the prior-year budget-neutrality adjustments
should not be applied once . . . .”). Nor do they “necessarily take issue with the
Secretary’s decision to apply the [cumulative budget-neutrality adjustment] to the
rates instead of the weights.” Aplts.’ Opening Br. at 24.
What the parties do vigorously dispute, however, are the consequences of
the Secretary applying a cumulative budget-neutrality adjustment in step four.
Above all, they contest whether that adjustment is being applied twice—once in
step two and then again in step four. Going forward, the focal point of our
19
analysis will be unpacking how those steps interact. Before delving into that
disagreement, however, it helps to understand how the Secretary has applied his
methodology over the years.
2
As discussed supra, Congress has periodically added new base years.
Whenever this happens, the Secretary issues “rebasing” instructions to fiscal
intermediaries. 14 See, e.g., J.A. at 339–43 (CMS Transmittal A-01-123, dated
Sept. 27, 2001) (“2000 Instructions”). These are technical instructions, which tell
the intermediaries how to calculate the hospital-specific rates for the new base
year.
In 1999, Congress added FY 1996 as a base year for sole community
hospitals. See Medicare, Medicaid, and SCHIP Balanced Budget Refinement Act
of 1999, Pub. L. No. 106-113, § 405, 113 Stat. 1501 (codified as amended at 42
U.S.C. § 1395ww(b)(3)(G)). In response, the Secretary issued the 2000
Instructions. See J.A. at 339–43 (2000 Instructions). The instructions, as it turned
out, were not a model of clarity. Notably, they failed to expressly state whether
14
Fiscal intermediaries, also referred to as “Medicare Administrative
Contractors,” are private entities that, acting as the Secretary’s agent, process
Medicare claims. See 42 U.S.C. § 1395h; 42 C.F.R. §§ 421.3, 421.100. As
relevant here, reimbursement of participating hospitals’ Medicare Part A costs is
sought and made through these intermediaries, and as part of that process, the
intermediaries are responsible for actually performing the reimbursement
calculations described at length in this opinion. See 42 C.F.R. § 421.100.
20
the intermediaries were to apply a cumulative budget-neutrality adjustment to the
hospital-specific rate for the new base year. The closest the instructions came to
addressing the issue was in the statement that the FY 1996 hospital-specific rate
was to be multiplied “by a factor of 1.02547 to update it from 1996 to 2000.” Id.
at 342.
But that remark is ambiguous. On the one hand, the 1.02547 figure might
have included the budget-neutrality adjustments from FY 1993 (the first year for
such adjustments) to FY 1996. See Aplts.’ Opening Br. at 11–12 (acknowledging
this possibility); Aplee.’s Resp. Br. at 15 (arguing that the 2000 Instructions
directed fiscal intermediaries to “apply the cumulative diagnosis-related group
budget-neutrality adjustment from the first year the adjustment was applied”); cf.
Adirondack Med. Ctr. v. Sebelius (Adirondack I), 935 F. Supp. 2d 121, 132
(D.D.C. 2013) (explaining that the 2000 Instructions directed intermediaries to
apply “cumulative DRG budget neutrality adjustment factors from FY 1993 to FY
2000”). On the other hand, the intermediaries were instructed to update the
hospital-specific rate “from 1996 to 2000,” implying the budget-neutrality
adjustment was not cumulative. J.A. at 342 (2000 Instructions) (emphasis added).
In 2006, Congress added FY 2002 as a new base year for Medicare-
dependent hospitals. See Deficit Reduction Act of 2005, Pub. L. No. 109-171,
§ 5003(b), 120 Stat. 4, 32 (codified as amended at 42 U.S.C. § 1395ww(b)(3)(K)
21
(2006)). The Secretary again issued rebasing instructions. See J.A. at 324–38
(CMS Transmittal 1067, dated Sept. 25, 2006) (“2006 Instructions”). This time,
the instructions seemed to speak clearly—the Secretary did not instruct
intermediaries to apply prior years’ budget-neutrality adjustments to the new base
year. See id. at 328. Instead, he instructed that the budget-neutrality adjustments
were to be applied prospectively only, i.e., for years after the new FY 2002 base
year. See Aplee.’s Resp. Br. at 15 (acknowledging that “unlike the 2000
instructions, these instructions inadvertently failed to tell the [Medicare]
contractors to apply the cumulative diagnosis-related group budget-neutrality
adjustment from FY 1993 forward”).
Two years later, Congress added FY 2006 as a new base year for sole
community hospitals. See Medicare Improvements for Patients and Providers Act
of 2008, Pub. L. No. 110-275, § 122(a), 122 Stat. 2494 (codified as amended at 42
U.S.C. § 1395ww(b)(3)(L) (2008)). As before, the Secretary issued rebasing
instructions. See J.A. at 344–49 (CMS Transmittal 1610, dated Oct. 3, 2008)
(“2008 Instructions”). Like the 2006 Instructions, the 2008 Instructions did not
direct intermediaries to apply prior years’ budget-neutrality adjustments to the new
base year; again, they were directed to apply the adjustments only
prospectively—that is, for years after FY 2006. See id. at 349; see also Aplee.’s
Resp. Br. at 15 (“In 2008, when Congress added FY 2006 as a possible base year
22
for sole community hospitals, the technical instructions that CMS first sent to
contractors contained the same mistake as the 2006 instructions regarding
Medicare-dependent, small rural hospitals.”).
But about six weeks later, the Secretary revised the 2008 Instructions in a
Joint Signature Memorandum. See J.A. at 350–53 (Joint Signature Mem., dated
Nov. 17, 2008). In this memorandum, the Secretary no longer instructed
intermediaries to apply the budget-neutrality adjustments prospectively only, as he
had in his initially issued 2008 Instructions. Rather, he instructed the
intermediaries to apply to the new FY 2006 base year all budget-neutrality
adjustments extending back to FY 1993. See id. at 350–51.
Another six months passed before the Secretary issued a proposed rule
pertaining to Medicare-dependent hospitals that explained to the public his
methodology for FY 2010. 74 Fed. Reg. at 24,183–85 (FY 2010 Proposed Rule).
The FY 2010 Proposed Rule reiterated the Secretary’s policy of applying a
cumulative budget-neutrality adjustment “to both the standard Federal rate and
hospital specific rates” without “remov[ing] the prior years’ budget neutrality
adjustment.” Id. at 24,184. To do otherwise, the Secretary explained, “would not
satisfy” Congress’s mandate to achieve budget neutrality. Id.
In explaining why a cumulative budget-neutrality adjustment was needed,
the Secretary said that “[i]f we were to remove this budget neutrality adjustment
23
factor for years prior to the base year, we believe the normalized DRG weights
applied to the hospital-specific amounts would be artificially high, thus resulting
in higher aggregate payments than permitted under the statute.” Id. This notion of
“artificially high” normalized average diagnosis-related group weights was new.
For years, the Secretary had explained that normalization negated any increase in
diagnosis-related group weights by “equating the average [diagnosis-related group]
weight after recalibration to the average [diagnosis-related group] weight before
recalibration.” 58 Fed. Reg. at 46,291 (FY 1994 Final Rule). But normalization
did not always “achieve budget neutrality with respect to aggregate payments . . .
because payment . . . is affected by factors other than average [diagnosis-related
group] weight.” Id. (emphasis added). The cumulative budget-neutrality
adjustment accounted for these other factors—not “artificially high” average
diagnosis-related group weights. 74 Fed. Reg. at 24,184 (FY 2010 Proposed
Rule). These longstanding positions of the Secretary were inconsistent with the
idea of “artificially high” normalized average diagnosis-related group weights.
Although the FY 2010 Proposed Rule introduced some uncertainty to the
regulatory landscape, it also shed substantial light on the lingering question
surrounding application of the cumulative budget-neutrality adjustment. As
explained above, the 2000 Instructions had left in doubt whether the Secretary’s
instructions required application of a cumulative budget-neutrality adjustment to
24
calculate the hospital-specific rates for the FY 1996 base year. The FY 2010
Proposed Rule removed that doubt. The proposed rule clarified that “the
instructions for implementing both the FY 1996 and FY 2006” base years applied
“cumulative budget neutrality adjustment factors” including each year’s
adjustment “since FY 1993.” Id. And those instructions were in keeping with the
Secretary’s “established policy.” Id.
The FY 2010 Proposed Rule brought the 2006 Instructions in line with that
declared policy. Recall, those instructions had directed intermediaries to apply
“budget adjustment factors” prospectively for the new FY 2002 base year—that is,
“for FYs 2003 through 2007.” Id. The failure to include the budget-neutrality
adjustments from “FYs 1993 through 2002,” the Secretary opined, was
“inconsistent with [his] stated policy of applying a cumulative budget neutrality
adjustment” and led to about a “1.74 percent” overpayment to Medicare-dependent
hospitals using the FY 2002 base year. Id. So going forward, the Secretary
announced, intermediaries were to once again apply a cumulative budget-neutrality
adjustment for all base years. Id. Of course, he admitted, this change would
potentially “lower the hospital-specific rate” for some Medicare-dependent
hospitals using the FY 2002 base year, to the point that the federal rate would
result in higher payments than the hospital-specific rate. Id. at 24,185. But the
cumulative adjustment was necessary to achieve budget neutrality and maintain “a
25
meaningful comparison between payments under the Federal rate, which is
adjusted by the cumulative budget neutrality factor, and payments based on the
hospital-specific rate.” Id. at 24,184.
The FY 2010 Proposed Rule drew many public comments. Several
commenters argued that a cumulative budget-neutrality adjustment for new base
years was unnecessary. Indeed, they believed this adjustment was already baked
into step two of the Secretary’s methodology—i.e., by dividing the hospital’s
average per-patient expenses in the new base year by its normalized average
diagnosis-related group weight. See J.A. at 312–13 (Rural Referral Center/Sole
Community Hospital Coalition, Comment Letter on FY 2010 Proposed Rule (June
25, 2009)). After all, these commenters pointed out, the Secretary had admitted
that those average diagnosis-related group weights were “artificially high”—even
after normalization. Id. at 312 (quoting 74 Fed. Reg. at 24,184 (FY 2010 Proposed
Rule)). And using these inflated weights as a divisor, the commenters explained,
resulted in a hospital-specific rate that was lower than it would have otherwise
been. See id. at 313. The same was not true for the federal rate, however, because
the Secretary did not use the inflated weights as a divisor in calculating that rate.
Hence, they argued, a cumulative budget-neutrality adjustment was appropriate for
the federal-rate hospitals, but it was unnecessary—indeed, duplicative—for sole
community and Medicare-dependent hospitals. Id. at 314–15.
26
The Secretary responded to these comments in the preamble to the FY 2010
Final Rule. See 74 Fed. Reg. at 43,895–96 (FY 2010 Final Rule). Retreating from
the “artificially high” language in the proposed rule, the Secretary rejected the
claim that the normalized average diagnosis-related group weights were too high.
See id. at 43,896. Rather, he explained that “[b]ecause the weights are
normalized,” they were no higher than they were pre-recalibration. Id. For that
reason, the Secretary denied that “the cumulative budget neutrality adjustment”
was meant to “offset[] an average [DRG] weight increase due to recalibration.”
Id. Thus, he reasoned that “[t]he cumulative budget neutrality adjustment [was]
not already being accounted for” in step two. Id. Satisfied that he put down the
confusion, the Secretary published the FY 2010 Final Rule.
C
This case unfolded against that backdrop. After the publication of the FY
2010 Final Rule, the Hospitals took issue with the Secretary’s methodology for
calculating the hospital-specific rate for new base years. And dissatisfied with
their reimbursements under that methodology, the Hospitals filed administrative
appeals with the Provider Reimbursement Review Board, an independent panel
authorized to hear appeals from the Secretary’s final determinations. See, e.g.,
J.A. at 354–81 (Hannover Hosp. Appeal Request, dated May 7, 2010); see 42
U.S.C. § 1395oo(a)(1). Though the Board concluded that it “lack[ed] the authority
27
to decide the legal question” the Hospitals raised, it granted their request for
expedited judicial review. J.A. at 322, 323 (Bd. Decision, dated Oct. 8, 2015); see
42 U.S.C. § 1395oo(f)(1).
The Hospitals then sued the Secretary in the district court. See J.A. at
50–108 (First Am. Compl., filed Feb. 16, 2016). Their complaint alleged that the
Secretary’s methodology violated the Administrative Procedure Act (the “APA”).
See id. at 59–62. Specifically, the Hospitals argued that the Secretary’s
application of the cumulative budget-neutrality adjustment to the hospital-specific
rate for new base years was “arbitrary, capricious, . . . or otherwise not in
accordance with law.” 15 Id. at 61 (quoting 5 U.S.C. § 706(2)(A)).
Both parties moved for summary judgment. The Hospitals argued that the
Secretary’s methodology was arbitrary and capricious for three reasons. See id. at
133–47 (Pls.’ Mem. of P. & A. in Supp. of Mot. for Summ. J., filed Sept. 6, 2016).
First, they claimed that the Secretary applied the same cumulative budget-
15
The Hospitals’ complaint also alleged that the cumulative budget-
neutrality adjustment was “in excess of statutory jurisdiction” under 5 U.S.C.
§ 706(2)(C) and contrary to the notice-and-comment requirements in § 553(b)–(c).
See J.A. at 59–63. On appeal, the Hospitals have waived these arguments by
inadequately briefing them. See Aplts.’ Opening Br. at 22–23 (mentioning only
§ 706(2)(A)); see also Allen v. United Servs. Auto. Ass’n, 907 F.3d 1230, 1236
n.3 (10th Cir. 2018) (ruling that party’s “cursory argument” on appeal was
“waived”); Sierra Club, Inc. v. Bostick, 787 F.3d 1043, 1060 n.18 (10th Cir.
2015) (concluding that because parties “have not developed” their argument, “it is
waived”). Thus, we need not and do not consider whether the Secretary’s
methodology violates §§ 553(b)–(c) or 706(2)(C).
28
neutrality adjustment twice—once by using inflated normalized diagnosis-related
group weights as a divisor in step two and then again in step four. See id. at
133–43. Second, the Hospitals contended that the Secretary’s methodology
yielded different payments than “would have been made had [he] . . . applied the
budget-neutrality adjustments to the DRG weights themselves.” Id. at 132. Third,
they claimed that the Secretary acted arbitrarily and capriciously by not calculating
the hospital-specific rate for new base years “based on 100 percent” of a hospital’s
base-year “target amount.” 16 42 U.S.C. § 1395ww(d)(5)(D)(i)(I); see J.A. at
143–46. For those reasons, the Hospitals asked the district court to order “the
Secretary to recalculate [their] base-year hospital-specific rates and compensate
[them] accordingly.” J.A. at 148.
16
This “based on 100 percent” language applies only to sole community
hospitals. Compare 42 U.S.C. § 1395ww(d)(5)(D)(i)(I), with id.
§ 1395ww(d)(5)(G)(ii) (omitting that language for Medicare-dependent hospitals).
But the statute does command the Secretary to reimburse both sole community
and Medicare-dependent hospitals based on their “allowable operating costs” in
the base year. See id. § 1395ww(b)(3)(C)(i)(I) (sole community hospitals); id.
§ 1395ww(b)(3)(D)(i)(I) (Medicare-dependent hospitals). Both before the district
court and on appeal, the Hospitals have read these provisions as demanding the
Secretary reimburse both sole community and Medicare-dependent hospitals for
100% of their base-year operating costs. See J.A. at 143–46; Aplts.’ Opening Br.
at 34–41, 37 n.20. For simplicity, in addressing this argument, we refer to the
“based on 100 percent” language without distinguishing between sole community
and Medicare-dependent hospitals because that language best captures the
substance of the Hospitals’ argument as it pertains to both types of hospitals.
29
The Secretary cross-moved for summary judgment. See J.A. at 150–84
(Def.’s Mem. in Opp’n to Pls.’ Mot. & in Supp. of Cross-Mot. for Summ. J, filed
Oct. 20, 2016). First, the Secretary denied that his methodology applied the
budget-neutrality adjustment twice. See id. at 181. That argument, the Secretary
noted, turned on the false premise that the normalized average diagnosis-related
group weights were “too high.” Id. Instead, normalization guaranteed that the
weights were exactly the same as they were pre-recalibration. Id. at 181 & n.10;
see id. at 181. That the FY 2010 Proposed Rule had used the phrase “artificially
high” was an unfortunate and imprecise “way of stating that without a cumulative
budget-neutrality adjustment the payments under the hospital-specific rate would
be higher.” Id. at 181 n.10. Second, the Secretary pointed out that the Hospitals
had conceded that the Medicare statute did not compel him to apply the budget-
neutrality adjustment to the weights rather than to the rates. As a result, whether
his methodology yielded a different payment than if he had applied the adjustment
to the weights was irrelevant. See id. at 176–77. Lastly, the Hospitals’ “based on
100 percent” argument failed, too, the Secretary observed: that argument wrongly
assumed that a new base year was “a kind of reset” entitling the Hospitals to 100%
of their operating costs rather than a starting point from which the Secretary must
make adjustments. Id. at 178. Thus, the Secretary argued that he was entitled to
summary judgment.
30
The district court agreed with the Secretary. It viewed the Secretary’s
choice “to apply pre-base year budget neutrality adjustments” as a “consistent and
rational” one. Id. at 295, 297 n.8. And the district court thought the Secretary had
adequately explained in the FY 2010 Final Rule why his methodology did not in
fact apply the budget-neutrality adjustment twice. See id. at 297–98. Likewise,
the court agreed with the Secretary’s view that “an amount based on 100 percent
of the hospital’s target amount,” 42 U.S.C. § 1395ww(d)(5)(D)(i)(I), was “the
starting point—not the end point—for the Secretary’s calculations,” J.A. at 301.
The district court added that it would “not second-guess the Secretary’s policy”
just because there may have been “other ways of calculating payments.” Id. at
300. And so the court denied the Hospitals’ summary-judgment motion, granted
the Secretary’s cross-motion, and entered final judgment.
II
The Hospitals now appeal from the final judgment. Their appeal raises one
issue: Is the Secretary’s methodology for calculating the hospital-specific rate for
a new base year arbitrary and capricious under 5 U.S.C. § 706(2)(A)? 17 In
17
The Hospitals style their appeal as raising three issues. See Aplts.’
Opening Br. at 6. In truth, those three issues are permutations of the same
issue—whether the Secretary’s methodology is arbitrary and capricious. We
therefore address the Hospitals’ three issues as three arguments within the one
larger legal issue. Similarly, though the Hospitals cite both §§ 706(2)(A) and
706(2)(C) in their statement of the issues, see id., their standard-of-review section
discusses only § 706(2)(A), see id. at 22. By inadequately briefing their
31
answering that question, we review the district court’s grant of summary judgment
de novo. See McKeen v. U.S. Forest Serv., 615 F.3d 1244, 1253 (10th Cir. 2010).
During our review, “[w]e apply the same standard of review as the district court.”
WildEarth Guardians v. U.S. Bureau of Land Mgmt., 870 F.3d 1222, 1233 (10th
Cir. 2017).
In Medicare-reimbursement cases, the APA supplies the standard we use to
review the Secretary’s actions. See 42 U.S.C. § 1395oo(f)(1); accord Thomas
Jefferson Univ. v. Shalala, 512 U.S. 504, 512 (1994). We will invalidate an
agency action under § 706(2)(A) only if it is “arbitrary, capricious, an abuse of
discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A);
accord W. Watersheds Project v. U.S. Bureau of Land Mgmt., 721 F.3d 1264, 1273
(10th Cir. 2013). To assess the agency’s action, we look to “the agency’s
contemporaneous explanation in light of the existing administrative record.”
Dep’t of Commerce v. New York, ---- U.S. ----, 139 S. Ct. 2551, 2573 (2019)
(recognizing “a narrow exception” when there is “a strong showing of bad faith or
improper behavior”). But we do not consider the agency’s “post hoc
rationalizations” for the action. Sorenson Commc’ns, Inc. v. FCC, 567 F.3d 1215,
1221 (10th Cir. 2009). That said, if the agency relied on impermissible
§ 706(2)(C) argument, the Hospitals waive it. See, e.g., Allen, 907 F.3d at 1236
n.3.
32
considerations or “entirely failed to consider an important aspect of the problem,”
then it acted arbitrarily and capriciously. Ukeiley v. EPA, 896 F.3d 1158, 1164
(10th Cir. 2018) (quoting US Magnesium, LLC v. EPA, 690 F.3d 1157, 1164 (10th
Cir. 2012)). The same is true if the agency’s stated rationale “runs counter to the
evidence before [it] or is so implausible that it could not be ascribed to a
difference in view or the product of agency expertise.” Id.
But APA review is narrow. And the arbitrary-and-capricious standard “is
very deferential to the agency.” W. Watersheds, 721 F.3d at 1273 (quoting
Hillsdale Envtl. Loss Prevention, Inc., v. U.S. Army Corps of Eng’rs, 702 F.3d
1156, 1165 (10th Cir. 2012)). Indeed, we presume that an agency action is valid
unless the party challenging the action proves otherwise. See Dine Citizens
Against Ruining Our Env’t v. Bernhardt, 923 F.3d 831, 839 (10th Cir. 2019).
“Our deference to the agency is ‘especially strong where the challenged decisions
involve technical or scientific matters within the agency’s area of expertise.’” Id.
(quoting Morris v. U.S. Nuclear Regulatory Comm’n, 598 F.3d 677, 691 (10th Cir.
2010)). Above all, we may not “substitute our judgment for that of the agency” on
matters within its expertise. Judulang v. Holder, 565 U.S. 42, 53 (2011) (quoting
Motor Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S.
29, 43 (1983)).
33
The parties agree that § 706(2)(A)’s arbitrary-and-capricious standard
governs this appeal. The Hospitals frame their entire argument as arising under
that subsection. See Aplts.’ Opening Br. at 22. For his part, the Secretary agrees
that we should review his actions “under the familiar standards of the [APA].”
Aplee.’s Resp. Br. at 19. And, though he does not expressly argue in the language
of arbitrary-and-capricious review, the Secretary fully engages with the Hospitals’
arbitrary-and-capricious arguments. See, e.g., id. at 22–30 (responding to the
Hospitals’ divisor argument but never using the words “arbitrary” or “capricious”).
Thus, we take it as undisputed that § 706(2)(A)’s arbitrary-and-capricious standard
applies here. 18
18
The Secretary also invokes the familiar standard from Chevron,
U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984).
See Aplee.’s Resp. Br. at 19. But after a threadbare citation, the Secretary never
again mentions Chevron, never recites its well-known two steps, and never
applies those steps to the facts here. Instead, the Secretary dresses his arguments
in “reasonableness” garb. See, e.g., id. at 20 (arguing “[t]he Secretary reasonably
adopted [his] methodology”). “Reasonableness,” however, is relevant under both
the arbitrary-and-capricious standard and Chevron. See Chevron, 467 U.S. at 844;
W. Watersheds, 721 F.3d at 1273. So the Secretary’s perfunctory and fleeting
invocation of Chevron waives his argument for Chevron deference. See, e.g.,
Dutcher v. Matheson, 840 F.3d 1183, 1203 & n.12 (10th Cir. 2016) (declining to
perform a Chevron analysis when a plaintiff raised argument for first time in a
motion to reconsider and rejecting argument that courts must “sua sponte
undertake a Chevron reasonableness analysis any time an agency interpretive
regulation is implicated”); Bronson v. Swensen, 500 F.3d 1099, 1104 (10th Cir.
2007) (“[W]e routinely have declined to consider arguments that are not raised, or
are inadequately presented, in an appellant’s opening brief.”); see also Neustar,
Inc. v. FCC, 857 F.3d 886, 893–94 (D.C. Cir. 2017) (“The FCC’s brief nominally
references Chevron’s deferential standard in its standard of review but did not
34
invoke this standard with respect to rulemaking. Consequently, it has forfeited
any claims to Chevron deference.”); Commodity Futures Trading Comm’n v.
Erskine, 512 F.3d 309, 314 (6th Cir. 2008) (“[T]he CFTC waived any reliance on
Chevron by failing to raise it to the district court.”); Faris v. Williams WPC-I,
Inc., 332 F.3d 316, 319 (5th Cir. 2003) (“This [Chevron] argument was not
presented to nor passed on by the district court, and therefore may not be
considered on appeal.”); cf. Lubow v. U.S. Dep’t of State, 783 F.3d 877, 884 (D.C.
Cir. 2015) (“The applicability of the Chevron framework does not go to our
court’s jurisdiction, and a party therefore can forfeit an argument against
deference by failing to raise it.”); cf. also Gardner v. Galetka, 568 F.3d 862, 878
(10th Cir. 2009) (answering in the negative the following question: “can the
congressionally mandated deferential standard of review [of the Anti-terrorism
and Effective Death Penalty Act] be waived by counsel?” (emphasis added)). In
any event, eliding a formal Chevron analysis does not materially alter our
resolution of the issue before us or, more specifically, our consideration of the
parties’ arguments. After all, the Secretary engages fully with the Hospitals’
arbitrary-and-capricious arguments. See, e.g., Aplee.’s Resp. Br. at 22–30. And
insofar as the Secretary’s reasonableness arguments involve a tacit claim for
deference at the second step of Chevron, the Secretary is effectively arguing that
his methodology is not “arbitrary or capricious in substance.” Judulang, 565 U.S.
at 52 n.7 (quoting Mayo Found. for Med. Ed. & Research v. United States, 562
U.S. 44, 53 (2011)). Consequently, whether we apply Chevron or arbitrary-and-
capricious review does not materially alter the outcome or change our analysis.
35
III
With the standard of review laid out, we can turn to the question before us:
Is the Secretary’s methodology for calculating the hospital-specific rate for a new
base year arbitrary and capricious?
The Hospitals answer that question in the affirmative, for three reasons.
First, they argue that the “Secretary acted arbitrarily and capriciously in applying
the same budget-neutrality adjustments twice.” Aplts.’ Opening Br. at 6. Second,
the Hospitals claim that the Secretary’s methodology is arbitrary and capricious
and contrary to Congress’s statutory command because it “yields a different
payment than had he budget-neutralized the weights themselves.” Id. Third, they
contend that the Secretary’s methodology is arbitrary and capricious because it
“would not reimburse a hospital its full base-year allowable operating costs.” Id.
The Secretary disagrees. He denies that applying a cumulative budget-
neutrality adjustment to a new base year mistakenly applies that adjustment twice.
The Secretary likewise refutes the claim that he must apply the cumulative budget-
neutrality adjustment to the diagnosis-related group weights themselves. Finally,
the Secretary disagrees that his methodology must be arbitrary and capricious
because it would not reimburse the Hospitals for 100% of their base-year operating
costs.
36
We agree with the Secretary. That is to say, for the reasons below, we hold
that the Secretary’s methodology is not arbitrary and capricious. Thus, we affirm
the district court’s judgment.
A
The centerpiece of the Hospitals’ position is their argument that the
Secretary mistakenly applies the cumulative budget-neutrality adjustment twice
when calculating the hospital-specific rate for a new base year. This double
application is, the Hospitals say, “arbitrary and capricious on its face.” Id. at 27.
“[B]ut at a minimum,” they argue, this error proves that the Secretary
misunderstood his own methodology and its consequences. Id. For those reasons,
the Hospitals claim that the Secretary’s methodology is arbitrary and capricious.
We disagree. We explain our disagreement in three sections. First, we
recount the Hospitals’ double-application argument in more detail. Second, we
conclude that this argument rests on flawed premises. Third, we reject the
Hospitals’ suggestion that the Secretary misunderstood his own methodology.
1
To understand the Hospitals’ double-application argument, recall three
features of the Medicare-reimbursement system:
Recalibration: Congress requires the Secretary to “recalibrate” (i.e.,
adjust) the diagnosis-related group weights every year. See 42 U.S.C.
§ 1395ww(d)(4)(C)(i). Because recalibration can change aggregate
payments to hospitals systemwide, Congress requires the Secretary to
37
recalibrate the weights “in a manner that assures that aggregate
payments” neither increase nor decrease. Id. § 1395ww(d)(4)(C)(iii); 74
Fed. Reg. at 43,895 (FY 2010 Final Rule).
Normalization: Achieving budget neutrality is a two-step process. To
that end, the Secretary’s first step is to “normalize” the recalibrated
weights. Normalization guarantees “that the average [diagnosis-related
group] weight after recalibration is equal to the average . . . weight prior
to recalibration.” 58 Fed. Reg. at 46,291 (FY 1994 Final Rule). For
example, in 1990, recalibration increased the average diagnosis-related
group weight by 1.35%. See Changes to the Inpatient Hospital
Prospective Payment System and Fiscal Year 1990 Rates, 54 Fed. Reg.
36,452, 36,470 (Sept. 1, 1989) (“FY 1990 Final Rule”). So to offset the
1.35% increase in average weight caused by recalibration, the Secretary
normalized the post-recalibration weights. See id.
Budget-Neutrality Adjustment: Normalization, however, does not always
“achieve budget neutrality” because “factors other than average
[diagnosis-related group] weight” affect payments. 58 Fed. Reg. at
46,291 (FY 1994 Final Rule). So the Secretary’s second step to achieve
budget neutrality is to calculate and apply “a budget neutrality
adjustment” to account for those other factors. Id.
Now recall the Secretary’s four-step methodology for calculating the
hospital-specific rate for a new base year:
Step One: Divide the hospital’s target amount for the new base year by
the number of discharges that year .
Step Two: Divide the figure from step one by the hospital’s normalized
average diagnosis-related group weight for the new base year.
Step Three: Apply an update factor to account for specific variables
such as inflation and wage levels.
Step Four: Apply the cumulative budget-neutrality adjustment.
The nub of the dispute is whether the Secretary effectively applies the
budget-neutrality adjustment twice—once in step two and then again in step four.
The Hospitals argue that they can prove that step two includes the cumulative
38
budget-neutrality adjustment. This argument rests on two premises. First, that the
normalized diagnosis-related group weights are “too high to be budget neutral.”
Aplts.’ Opening Br. at 2–3. Second, that these weights are inflated by the precise
amount necessary to achieve budget neutrality. And, because these inflated
weights are used as a divisor in step two, the hospital-specific rate for a new base
year is reduced by the exact percentage that the Secretary later reduces the rates
through application of the cumulative budget-neutrality adjustment in step four.
Consider the following hypothetical that the Hospitals say proves their
double-application argument. 19 Imagine that Congress adds FY 2002 as a new
base year. And assume that the Secretary calculates a cumulative budget-
neutrality adjustment factor of 0.986. (This figure would be the product of
budget-neutrality adjustments from 1996 to 2002.) When the Secretary multiplies
this adjustment factor by the federal rate and does the same for the hospital-
specific rate in step four, he reduces those rates by 1.4%. And, from this “1.4
percent budget-neutrality adjustment,” the Hospitals reason that “the DRG weights
have increased by 1.4 percent” from 1996 to 2002. Aplts.’ Reply Br. at 19. Put
19
The Hospitals crafted this hypothetical in their reply brief,
responding to the Secretary’s own less extensive hypothetical in which the
Secretary concludes that, at bottom, the Hospitals seek to “avoid the impact of
budget neutrality adjustments from 1996 through 2002”—the new hypothetical
base fiscal year. Aplee.’s Resp. Br. at 32. We convey that hypothetical in all
material respects, see Aplts.’ Reply Br. at 18–24, and subsequently analyze it.
39
differently, the Hospitals think that—even after normalization—the average
diagnosis-related group weights for the new FY 2002 base year are 1.4% too high.
See id. at 20–21. And by using these inflated weights as a divisor in step two, the
Secretary applies the same 1.4% budget-neutrality reduction to the hospital-
specific rate that he later applies in step four. See id.
Here is how this hypothetical would play out under the Hospitals’
assumptions. Assume that in 1996 Hospital X’s per-patient target amount was
$1,000, and its average diagnosis-related group weight that year was 1.0. So
Hospital X would have “receive[d] $1,000 for treating its average patient” in 1996.
Id. at 19. And assume that nothing changed from 1996 to 2002; that is, the per-
patient target amount in 2002 remained $1,000, and the patients that Hospital X
treated in 2002 had identical ailments to those treated in 1996. Despite this lack
of change, Hospital X’s average diagnosis-related group weight for FY
2002—even after normalization—would have increased by 1.4% to 1.014. See id.
So when the Secretary in step two divides Hospital X’s per-patient target amount
for 2002 ($1,000) by its inflated average diagnosis-related group weight (1.014),
he gets $986.19. Id. Then, in step four, “the Secretary . . . would mistakenly
believe that he had to apply another 1.4 percent reduction [in the form of a
cumulative budget-neutrality adjustment of 0.986] to the hospital’s 2002
hospital-specific rate of $986.19, leading to a hospital-specific rate of $972.38.”
40
Id. at 20. This rate is about 2.8% less than the 1996 rate. Thus, the Hospitals
posit that the Secretary mistakenly applied the 1.4% cumulative budget-neutrality
adjustment twice—once in step two and then again in step four.
2
The Hospitals’ double-application argument, however, rests on flawed
premises. The Hospitals assume that the normalized average diagnosis-related
group weights used as a divisor in step two are “artificially high” or “too high to
be budget neutral.” Aplts.’ Opening Br. at 24 & n.12. This assumption turns on
the validity of the claim “that a hospital that had an average DRG weight of 1.0 in
1996 would have a [normalized average] DRG weight of 1.014 in 2002 if it treated
the same type of patients.” Aplts.’ Reply Br. at 23; see id. at 21 (noting that “the
one point of disagreement between the parties regarding this hypothetical” pertains
to “whether the hospital’s DRG weight in 2002 would necessarily be 1.014 . . . as
the Hospitals contend”). According to the Hospitals, it is “necessarily” true that
the DRG weight would be 1.014 “because the 1.4 percent budget neutrality
adjustment means that, based solely on DRG recalibrations, the [normalized] DRG
weights have increased by 1.4 percent.” Id. at 19.
These assumptions evince a misunderstanding of the role of normalization
and the budget-neutrality adjustment. Recall that after the Secretary recalibrates
the diagnosis-related group weights, he normalizes them. The parties seem to
41
agree that normalization “ensure[s] that the average DRG weight is the same after
recalibration as it was before.” Aplts.’ Opening Br. at 10. It should ineluctably
follow, then, that under the Hospitals’ hypothetical, the average diagnosis-related
group weight of Hospital X in FY 1996 should be the same as in FY 2000—that is,
1.0. See 74 Fed. Reg. at 24,184 (FY 2010 Proposed Rule) (noting that, in
normalizing the weights, the Secretary applies a numerical adjustment “in order to
ensure that the average [diagnosis-related group] weight after recalibration is
equal to the average [diagnosis-related group] weight prior to recalibration”). In
short, normalization—not the application of a budget-neutrality
adjustment—removes any year-to-year increase in the average diagnosis-related
group weights. And, relatedly, with respect to those group weights, it is
normalization—not a budget-neutrality adjustment—that serves the statutory
mandate (though not necessarily completely) of establishing budget neutrality in
aggregate payments systemwide. See 58 Fed. Reg. at 46,291 (FY 1994 Final Rule)
(noting that normalization “does not necessarily achieve budget neutrality with
respect to aggregate payments to hospitals because payment to hospitals is affected
by factors other than average [diagnosis-related group] weight”).
Correcting these misunderstandings exposes the logical hole in the
Hospitals’ double-application argument. They assume that the existence of a “1.4
percent budget neutrality adjustment means that, based solely on DRG
42
recalibrations, the [normalized average] DRG weights have increased by 1.4
percent.” Aplts.’ Reply Br. at 19. But, by its very nature, normalization makes
that impossible. And in any event, the Hospitals’ reasoning begs the question of
what the cumulative budget-neutrality adjustment accounts for. Underlying their
reasoning appears to be their answer: the Hospitals repeatedly assert that “the
entire purpose of the budget-neutrality exercise is simply to budget neutralize the
DRG weights.” Id. at 26 (emphasis added); see id. at 1 (“The budget-neutrality
adjustments at issue here are meant to keep the recalibrated DRG weights budget
neutral and ensure that Medicare payments neither increase nor decrease based
simply on the adjustments to the DRG weights.”); see also Aplts.’ Opening Br. at
19 (noting that “since the budget-neutrality adjustments were instituted to fulfill
the Congressional mandate to budget-neutralize the DRG weights”); id. at 31
(noting that “the entire point of the [budget-neutrality] adjustment is to render
DRGs weights budget neutral”). But the Hospitals are wrong.
Congress’s statutory directive to the Secretary is not to budget neutralize the
DRG weights per se but rather to ensure that the process of recalibration of those
weights is performed “in a manner that assures that the aggregate payments” are
budget neutral. See § 1395ww(d)(4)(C)(iii) (requiring the Secretary to recalibrate
“in a manner that assures that the aggregate payments . . . are no greater or less
than those that would have been made [pre-recalibration]” (emphasis added));
43
accord Adirondack II, 29 F. Supp. 3d at 32. Despite their repeated assertions to
the contrary, the Hospitals cite no on-point authority that supports their position.
The Secretary has discretion on how to accomplish this statutory budget-neutrality
directive. See Adirondack Med. Ctr. v. Burwell (Adirondack III), 782 F.3d 707,
710 (D.C. Cir. 2015) (noting “the wide discretion afforded the Secretary to
implement the Medicare reimbursement formula, including determining how to
meet Medicare’s budget neutrality requirements”). And, as noted, he does so
primarily through the normalization process. And, notwithstanding the Hospitals’
contrary view, the budget-neutrality adjustment comes into play, not to budget
neutralize the DRG weights themselves, but rather to ensure that, after
normalization, the budget-neutrality job is done by addressing the effects of other
factors in the Medicare system, such as wage-index and teaching-hospital
adjustments, that could hinder achievement of budget neutrality. See, e.g., 42
U.S.C. 1395ww(d)(3)(E) (providing for a wage-index adjustment); id.
§ 1395ww(d)(5)(B) (directing the Secretary to “provide for an additional payment
amount for” teaching hospitals with “with indirect costs of medical education”).
In fact, the Hospitals’ own hypothetical (discussed above) undercuts their
argument. In that scenario, we assumed that the average diagnosis-related group
weights for the new FY 2002 base year were 1.4% higher than the weights from
1996. See Aplts.’ Reply Br. at 19 (“[B]ased solely on DRG recalibrations, the
44
DRG weights have increased by 1.4 percent.”). But we now know that
normalization guarantees that the 2002 weights are exactly the same as the 1996
weights. See Aplts.’ Opening Br. at 10 (explaining that normalization “ensure[s]
that the average DRG weight is the same after recalibration as it was before”).
Even so, the cumulative budget-neutrality adjustment is needed to offset a 1.4%
increase in aggregate payments caused by “other Medicare payment adjustments.”
Aplee.’s Resp. Br. at 24. So, in our hypothetical, we may derive Hospital X’s
2002 hospital-specific rate based on the following formula: its per-patient target
amount for 2002 ($1,000) divided by its normalized average diagnosis-related
group weight (1.0) and then multiplied by the cumulative budget-neutrality
adjustment (0.986). The result is a hospital-specific rate for the new FY 2002 base
year of $986.00 (i.e., $1,000 ÷ 1.0 × 0.986 = $986.00). And that rate is 1.4% less
than the 1996 hospital-specific rate. Simply put, the Secretary’s methodology does
not apply the cumulative budget-neutrality adjustment twice. 20
3
That said, the confusion that seemingly underlies the Hospitals’ double-
application argument is understandable. In fact, the Secretary arguably invited
20
This fact also scuttles the Hospitals’ claim that the Secretary’s
methodology, “[r]ather than creating comparability with the Federal rate . . . ,
destroys it.” Aplts.’ Opening Br. at 43. The premise for this argument is that the
Secretary uses “inflated” average diagnosis-related group weights “as the
divisor.” Id. As explained above, that premise is false.
45
that confusion. Recall that in the FY 2010 Proposed Rule the Secretary reiterated
that he “normalize[s] DRG weights by an adjustment factor in order to ensure that
the average [DRG] weight after recalibration is equal to the average . . . weight
prior to recalibration.” 74 Fed. Reg. at 24,184 (FY 2010 Proposed Rule); see also
58 Fed. Reg. at 46,291 (FY 1994 Final Rule) (making essentially same statement).
But, unlike past rules, the FY 2010 Proposed Rule went on to say that if the
Secretary “were to remove this budget neutrality adjustment factor for years prior
to the base year, . . . the normalized DRG weights applied to the hospital-specific
amounts would be artificially high.” 74 Fed. Reg. at 24,184 (FY 2010 Proposed
Rule). In this imprecise wording, the Secretary seemingly planted the seeds for
the Hospitals’ confusion.
In the FY 2010 Final Rule, the Secretary confronted the confusion
apparently wrought by his imprecise wording. He wrote, “[s]ome commenters
[had] asserted that the application of a cumulative budget neutrality adjustment
factor . . . doubles the impact of this adjustment on the hospital-specific rates.” 74
Fed. Reg. at 43,896 (FY 2010 Final Rule). These same commenters, the Secretary
recognized, had thought that the normalized average “weight from FYs 1993
through 2002 [had] increased and that the cumulative budget neutrality adjustment
. . . offsets this average . . . weight increase.” Id. And he observed that, on this
46
assumption, the commenters had concluded “that this budget neutrality adjustment
[was] already being accounted for [in step two] when [the Secretary] divides the
. . . average cost per discharge by the hospital’s [normalized average DRG
weight].” Id.
The Secretary then disabused the commenters of this notion. The Secretary
repeated his longstanding position that “the recalibrated DRG weights are
normalized each year . . . so that the national average [DRG] weight after . . .
recalibration is equal to the . . . average [DRG] weight before recalibration.” Id.;
see 58 Fed. Reg. at 46,291 (FY 1994 Final Rule) (reiterating same position). And
“[b]ecause the weights are normalized,” the Secretary explained, “they do not
reflect . . . average [DRG] weight change due to recalibration.” 74 Fed. Reg. at
43,896 (FY 2010 Final Rule) (emphasis added). He repeated that “DRG weights
after normalization, do not reflect . . . average [DRG] weight change.” Id.
(emphasis added). To put a finer point on it, the Secretary wrote that he
“disagree[d] with commenter’s assertions that the average [DRG] weight from FYs
1993 through 2002 increased due to recalibration.” Id. The Secretary also
disagreed with the idea “that the cumulative budget neutrality adjustment . . . for
this time period offsets an average [DRG] weight increase due to recalibration.”
Id. For that reason, he explained, “[t]he cumulative budget neutrality adjustment
is not already being accounted for [in step two] when [the Secretary] divides the
47
FY 2002 average cost per discharge for a hospital by the hospital’s [normalized
average DRG weight] for FY 2002.” Id.
The FY 2010 Final Rule should have stifled the Hospitals’ double-
application argument. After all, the Secretary refuted the central premise of that
argument—that the normalized average diagnosis-related group weights were
inflated. Those weights, the Secretary had explained, were not inflated. What’s
more, he had clarified that the cumulative budget-neutrality adjustment was not
meant “to offset any increase or decrease in the . . . average [DRG] weight due to
recalibration.” Id. In other words, the Secretary had effectively refuted the
Hospitals’ later assumption in its briefing here that a budget adjustment of 1.4%
means that, “if the hospital’s average DRG weight was 1.0 in 1996, then its
average DRG weight for treating the same patient in 2002 would necessarily be 1.4
percent higher.” Aplts.’ Opening Br. at 43.
Nevertheless, the Hospitals are immovable in their belief that the Secretary
has everything “exactly backwards” and fails to understand his own methodology.
Id. Indeed, the Hospitals argue that the Secretary “acted arbitrarily and
capriciously” by straying from his “prior practice” and failing “to consider an
important aspect of the issue”—i.e., the ostensibly inflated DRG weight. Id. at 27.
48
Once again, however, an examination of the record belies this argument. 21 Though
the FY 2010 Final Rule itself is proof enough, looking to earlier rules confirms
that the Secretary has long understood his methodology. Take the FY 1994 Final
Rule. In that rule, the Secretary explained that the recalibrated “weights are
normalized by an adjustment factor, so that the average [DRG] weight after
recalibration is equal to the average . . . weight prior to recalibration.” 58 Fed.
Reg. at 46,291 (FY 1994 Final Rule). But normalization does not always “achieve
budget neutrality with respect to aggregate payments to hospitals because payment
to hospitals is affected by factors other than average [DRG] weight.” Id.
(emphasis added). “Therefore,” the Secretary explained, he applies “a budget
21
In their briefing, the Hospitals make hay out of a supposedly
erroneous example that government counsel offered the D.C. district court during
similar litigation. See, e.g., Aplts.’ Opening Br. at 19–21, 30, 34, 42. Whether
this example offered to a different court in a different case is indeed inaccurate is
irrelevant here. Absent a “strong showing of bad faith or improper behavior,” we
assess the Secretary’s actions against his “contemporaneous explanation in light
of the existing administrative record.” Dep’t of Commerce, 139 S. Ct. at 2573,
2574 (citations omitted). We do not retroactively attribute an alleged
misunderstanding by government counsel in a highly technical case to the
Secretary when he acted, most notably, in the FY 2010 Final Rule. Cf. Sorenson,
567 F.3d at 1221 (“The court must rely on the reasoning set forth in the
administrative record and disregard post hoc rationalizations of counsel.”); Lewis
v. Babbitt, 998 F.2d 880, 882 (10th Cir. 1993) (“Judicial review under these
[APA] standards is generally based on the administrative record that was before
the agency at the time of its decision, and reviewing courts may not rely on
litigation affidavits that provide post hoc rationalizations for the agency’s action.”
(citations omitted)). We therefore refuse to consider the supposedly erroneous
example from the D.C. district court proceedings.
49
neutrality adjustment to assure the requirement of [budget neutrality in aggregate
payments] is met.” Id. Simply put, since 1994, the Secretary has understood
that—contrary to the Hospitals’ claims now—normalization offsets any increase in
the recalibrated weights, but other factors necessitate a cumulative budget-
neutrality adjustment to ensure that the statutory mandate of budget neutrality in
aggregate payments is achieved. That explanation guts the assumptions on which
the Hospitals’ double-application argument rests. And the Secretary has reiterated
this same explanation over the years. 22
To be sure, imprecise language has sometimes plagued the Secretary. The
FY 2010 Proposed Rule, for example, included the unfortunate phrase “artificially
high.” 74 Fed. Reg. at 24,184 (FY 2010 Proposed Rule). And in the 2000
Instructions, the Secretary did not clarify whether he applied a cumulative budget-
neutrality adjustment to the new FY 1996 base year. See J.A. at 342 (2000
Instructions). But the Secretary rectified both semantic oversights in the FY 2010
Final Rule. As we have explained, that rule made clear that the normalized
diagnosis-related group weights are not artificially high—a position the Secretary
22
See, e.g., Changes to the Hospital Inpatient Prospective Payment
Systems and Fiscal Year 1996 Rates, 60 Fed. Reg. 45,778, 45,792 (Sept. 1, 1995)
(“FY 1996 Final Rule”); Changes to the Hospital Inpatient Prospective Payment
Systems and Fiscal Year 2001 Rates, 65 Fed. Reg. 47,054, 47,070 (Aug. 1, 2000)
(“FY 2001 Final Rule”); 70 Fed. Reg. at 47,322 (FY 2006 Final Rule).
50
had long held. See, e.g., 58 Fed. Reg. at 46,291 (FY 1994 Final Rule). And the
FY 2010 Final Rule clarified that “the instructions for implementing . . . the [new]
FY 1996” base year—i.e., the 2000 Instructions—did in fact apply the “cumulative
budget neutrality adjustment.” 74 Fed. Reg. at 43,895 (FY 2010 Final Rule). So,
despite these occasional imprecise word choices, the Secretary’s longstanding
position has been that the normalized diagnosis-related group weights are not
artificially high, and his current methodology is the same as that which he used the
first time Congress added a new base year. 23
With this understanding, one can see the 2006 Instructions and the initial
2008 Instructions for what they were—mistakes. Recall that in those instructions,
the Secretary did not apply prior years’ budget-neutrality adjustments to the new
base years. See J.A. at 328 (2006 Instructions); id. at 349 (2008 Instructions). But
23
Notably, the Hospitals concede that the 2000 Instructions “may have
contained the [cumulative budget-neutrality] adjustments from the three prior
years.” Aplts.’ Opening Br. at 11–12. They also do not expressly challenge the
correctness of the Secretary’s interpretation of those instructions as having in fact
applied a cumulative adjustment to the new FY 1996 base year. See generally id.
Thus, we have no reason to doubt that the Secretary did in fact apply a cumulative
budget-neutrality adjustment to the new FY 1996 base year when calculating the
hospital-specific rate. Cf. Adirondack III, 782 F.3d at 709 (“Prior to 2006, the
budget neutrality adjustments applied to the hospital-specific . . . rates in a
straightforward way: once a base year was chosen and the rate was calculated, the
Secretary applied every budget neutrality adjustment from 1993 . . . to the
present.”); Adirondack I, 935 F. Supp. 2d at 132 (explaining that the 2000
Instructions applied “cumulative DRG budget neutrality adjustment factors from
FY 1993 to FY 2000”).
51
just six weeks after issuing the 2008 Instructions, the Secretary corrected his error
in those instructions by including the budget-neutrality adjustments from the years
before the new base year. See id. at 351 (Joint Signature Mem.). The FY 2010
Proposed Rule did the same for the 2006 Instructions. 74 Fed. Reg. at 24,184 (FY
2010 Proposed Rule). In that proposed rule, the Secretary even explained the
reasons behind the corrections. Applying the cumulative budget-neutrality
adjustment to new base years, the Secretary noted, was consistent with his
“established policy” and necessary for “a meaningful comparison between
payments under the Federal rate, which is adjusted by the cumulative budget
neutrality factor, and payments based on the hospital-specific rate.” Id. And
admitting that correcting the past errors “would lower the hospital-specific rate”
for some hospitals going forward, the Secretary declined to retroactively reduce
payments that had already been made to hospitals under the erroneous instructions.
Id. at 24,185. The 2006 Instructions and the initial 2008 Instructions, then, were
aberrations, and the Secretary’s current methodology the norm. Cf. Adirondack
III, 782 F.3d at 709 (“Prior to 2006, the budget neutrality adjustments applied to
the hospital-specific . . . rates in a straightforward way: once a base year was
chosen and the rate was calculated, the Secretary applied every budget neutrality
adjustment from 1993 . . . to the present.”).
52
In summary, the Secretary does not apply the cumulative budget-neutrality
adjustment twice. The Hospitals’ argument to the contrary rests on flawed
assumptions. And the Secretary has long understood his methodology and
explained it to the public. Thus, the Hospitals’ first argument for why the
Secretary’s methodology is supposedly arbitrary and capricious fails.
B
The Hospitals next argue that the Secretary’s methodology is arbitrary and
capricious and contrary to Congress’s statutory command because it “yields a
different payment than had he budget-neutralized the weights themselves.” Aplts.’
Opening Br. at 6. The logic behind this argument is as follows. “The budget-
neutrality adjustments have one purpose: to budget-neutralize the DRG weights.”
Id. at 27. The Secretary’s methodology is valid only if it “yields the same payment
as had [he] merely budget-neutralized the DRG weights themselves (rather than the
payment rates).” Id. at 19. The Secretary’s “methodology does not yield the same
payment” as the one he would have had he “budget-neutralized the DRG weights
themselves.” Id. at 20. Therefore, the Secretary’s methodology is arbitrary and
capricious.
Largely for the reasons we already have discussed in Part III.A.2, supra, this
argument fails. Congress’s statutory directive to the Secretary is not to budget
neutralize the DRG weights per se but instead to ensure that the process of
53
recalibration of those weights is performed “in a manner that assures that the
aggregate payments” are budget neutral. See § 1395ww(d)(4)(C)(iii) (requiring
the Secretary to recalibrate “in a manner that assures that the aggregate payments
. . . are no greater or less than those that would have been made [pre-
recalibration]” (emphasis added)); accord Adirondack II, 29 F. Supp. 3d at 32.
And the Secretary has discretion on how to accomplish this statutory budget-
neutrality directive. See Adirondack III, 782 F.3d at 710 (noting “the wide
discretion afforded the Secretary to implement the Medicare reimbursement
formula, including determining how to meet Medicare’s budget neutrality
requirements”). Stated otherwise, the statutory text teaches that the budget-
neutrality adjustment’s purpose is to achieve budget neutrality in aggregate
payments systemwide, not to guarantee that the diagnosis-related group weights
themselves are budget neutral. And the text leaves the “manner” in which the
Secretary achieves budget neutrality up to him. Thus, the Hospitals’ assertion that
the “one purpose” of the budget-neutrality adjustment under Congress’s statutory
scheme is “to budget neutralize the DRG weights” is incorrect. Aplts.’ Opening
Br. at 27. Whether the Secretary’s methodology yields the same payment as would
a methodology applying the adjustment to the weights is irrelevant; 24 the Secretary
24
Once again, the Hospitals try to cover up the logical holes in their
argument by reverting to the government counsel’s alleged claim before the D.C.
district court that the Secretary’s “methodology would yield the same payment as
54
does not act arbitrarily or capriciously if his chosen methodology adheres to the
statutory command to achieve budget neutrality in aggregate payments. 25
Therefore, the Hospitals’ second argument meets the same fate as its first.
C
For their final argument, the Hospitals claim that the Secretary’s
methodology is arbitrary and capricious because, if applied to a new base year
itself, it would reimburse them for less than 100% of their actual base-year costs.
This argument relies on language from 42 U.S.C. § 1395ww(d)(5)(D)(i)(I). 26 That
would be yielded had he budget-neutralized the weights themselves.” Aplts.’
Opening Br. at 28. And once again, we decline to rely on statements outside of
the administrative record. See supra note 20; cf. Dep’t of Commerce, 139 S. Ct.
at 2573. Even if we could consider this extra-record statement, we would not do
so here. Notably, the government counsel later corrected the representation on
appeal, explaining “that the Secretary is ‘fully aware of the reduction in a base
year’s payment rate that occurs because of [his] methodology.’” Aplee.’s Resp.
Br. at 37 n.12 (quoting Gov’t’s Br. at 27, Adirondack III, No. 14-5122 (D.C. Cir.
Nov. 20, 2014)).
25
Notably, the Hospitals’ challenge is not about whether “the Secretary
has failed to maintain budget-neutrality,” Aplee.’s Resp. Br. at 21—that is, failed
to carry out his statutory mandate—but rather about “the precise methodology
used by the Secretary,” Adirondack III, 782 F.3d at 710. Though the Hospitals
push back on the notion that they have agreed that the Secretary has “successfully
achieve[d] the goal of budget neutrality,” id., they never directly say that he has
not but rather turn our gaze back to the Secretary’s ostensible failing in properly
handling the problem of “inflated” DRG weights, Aplts.’ Reply Br. at 15 n.12.
26
As explained in footnote 15, this provision applies only to sole
community hospitals. For our purposes, that distinction is irrelevant because the
Hospitals’ argument invoking that provision fails. In other words, given that not
even sole community hospitals are entitled to 100% of their actual base-year
costs, Medicare-dependent hospitals—to which this language does not strictly
55
provision directs the Secretary to calculate the hospital-specific rate using “an
amount based on 100 percent of the hospital’s target amount for the cost reporting
period [i.e., the base year], as defined in subsection (b)(3)(C).”
§ 1395ww(d)(5)(D)(i)(I). Subsection (b)(3)(C), in turn, defines “target amount” as
a hospital’s “allowable operating costs” during the base year. Id.
§ 1395ww(b)(3)(C)(i)(I). The Hospitals read this language as a command for the
Secretary to reimburse them for 100% of their actual base-year costs. And they
reason that if the Secretary’s methodology—when applied to the new base year
itself—fails to satisfy that command, it must be arbitrary and capricious.
The Hospitals, however, come up short again. To start, their argument
evinces a fundamental misunderstanding about the nature of the current Medicare-
reimbursement system. Under the original Medicare Act, the Secretary paid
hospitals for “the reasonable cost” of treating each Medicare patient. Social
Security Amendments of 1965 §§ 1814(b), 1861(v)(1). But this system quickly
became unwieldy; “[t]he more [hospitals] spent, the more they would receive” in
reimbursement. Billings Clinic v. Azar, 901 F.3d 301, 303 (D.C. Cir. 2018).
Consequently, Congress replaced the old system with the inpatient prospective
payment system. See Social Security Act Amendments of 1983 § 601. A central
premise of this current system is that hospitals “are reimbursed at a fixed amount
apply—necessarily are not entitled to such reimbursement.
56
per patient, regardless of the actual operating costs they incur in rendering these
services.” Auburn Reg’l Med. Ctr., 568 U.S. at 149 (emphasis added). Against
this backdrop, the Hospitals’ claim to 100% of their actual base-year costs
looks—at the very least—suspect.
Closer inspection of the Hospitals’ argument makes clear that it is flawed.
The whole exercise of inquiring as to whether a given methodology would
reimburse the Hospitals for 100% of their actual base-year costs is odd under the
inpatient prospective payment system. By statutory mandate, the Secretary never
uses the hospital-specific rate he calculates for a base year to reimburse a hospital
for discharges from that base year itself. Thus, for example, he never uses the rate
calculated for the FY 2002 base year to reimburse a hospital for its 2002
discharges. Rather, that rate applies only for “discharges occurring on or after
October 1, 2006.” 42 U.S.C. § 1395ww(b)(3)(K)(i). It is, after all, an inpatient
prospective payment system. So whether the rate the Secretary calculated for the
FY 2002 base year would reimburse a hospital for its full 2002 costs is seemingly
irrelevant.
Moreover, the text on which the Hospitals rely further discredits their
argument. The transitive verb “base” ordinarily means “use as a point from which
(something) can develop.” Base, N EW O XFORD A MERICAN D ICTIONARY 134 (2d
ed. 2005). And as a noun, “base” often means “a main or important element or
57
ingredient to which other things are added.” Id. So when Congress said that the
payment using the hospital-specific rate “shall be” an “amount based on 100
percent of the hospital’s target amount,” 42 U.S.C. § 1395ww(d)(5)(D)(i)(I), it was
telling the Secretary that 100% of the hospital’s target amount was the starting
point from which to develop the hospital-specific rate. And that is exactly what
the Secretary does: in simplified form, he starts with 100% of a hospital’s base-
year target amount, divides that amount by the normalized average diagnosis-
related group weight for that year, and then applies the budget-neutrality
adjustment. In hitching their wagon to the phrase “an amount based on 100
percent of the hospital’s target amount,” then, the Hospitals effectively undercut
their own position.
And contrary to the Hospitals’ protestations, reading “based on” according
to its ordinary meaning does not “render[] the phrase ‘100 percent’ surplusage.”
Aplts.’ Opening Br. at 39. Under this commonsense reading, the phrase “100
percent” clarifies that the starting point for the Secretary’s calculations must be
100%—not 90%, 50%, or some other percentage—of the hospital’s target amount.
For example, if Hospital X’s base-year target amount was $1,000, the Secretary
could not begin with $900, divide that amount by the normalized average
diagnosis-related group weight, and then apply the budget-neutrality adjustment.
Without the phrase “100 percent,” the provision would instruct—without any
58
specificity—the Secretary to arrive at the hospital-specific rate using “an amount
based on . . . the hospital’s target amount.” On that wording, it would be unclear
whether the Secretary must use an amount based on all or may use only a portion
of the hospital’s target amount. Cf. Anna Jaques Hosp. v. Sebelius, 583 F.3d 1, 5
(D.C. Cir. 2008) (agreeing that a provision instructing the Secretary to calculate
“the wage index . . . ‘on the basis of’” survey data was ambiguous because it was
“silent about whether she must use all of the survey data”). The phrase “100
percent” has the effect of clearing up the matter. And so the Hospitals are
mistaken in asserting that an interpretation of “based on” as meaning the starting
point renders “100 percent” surplusage. 27
Aside from its conceptual and linguistic shortcomings, the Hospitals’
argument is logically inconsistent with the Hospitals’ overall approach.
Throughout their briefing, the Hospitals stress that they agree “that the Secretary
should apply all budget-neutrality adjustments once.” Aplts.’ Reply Br. at 12 n.10.
But if the Secretary had to pay hospitals 100% of their actual base-year costs, he
27
The Hospitals’ reading, however, does create surplusage. Congress
directs the Secretary to calculate the hospital-specific rate using “an amount based
on 100 percent of the hospital’s target amount.” 42 U.S.C.
§ 1395ww(d)(5)(D)(i)(I). The Hospitals read that language as a “command to use
‘100 percent’ of base-year costs.” Aplts.’ Opening Br. at 37. But that reading
gives no meaning to the words “an amount based on.” As the Hospitals
themselves point out, “every statutory word must be given meaning.” Id. Yet,
they are content to give four statutory words no meaning.
59
could not apply any budget-neutrality adjustment at all. For example, say Hospital
X’s actual base-year costs were $1,000 and the budget-neutrality adjustment was
0.9. If the Secretary applied that adjustment to the base year, he would pay
Hospital X $900 for its base-year discharges. Under the logic of the Hospitals’
argument, the Secretary could not apply that adjustment at all because doing so
would be “an immediate reduction in their base-year costs.” Aplts.’ Opening Br.
at 39. Yet, the Hospitals have repeatedly asserted that the Secretary is not
prohibited from applying a budget-neutrality adjustment to the hospital-specific
rate for a new base year. Thus, if—as the Hospitals agree—the Secretary may
apply a budget-neutrality adjustment to the hospital-specific rate, then the
Secretary cannot also be required to pay a hospital 100% of its base-year operating
costs. Simply put, the Hospitals’ third argument—like its first two—fails to
establish that the Secretary’s methodology is arbitrary and capricious.
IV
For the foregoing reasons, we hold that the Secretary’s methodology is not
arbitrary and capricious and AFFIRM the district court’s judgment.
60