American Clinical Laboratory v. Alex Azar, II

United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued April 23, 2019 Decided July 30, 2019

No. 18-5312

AMERICAN CLINICAL LABORATORY ASSOCIATION,
APPELLANT

v.

ALEX MICHAEL AZAR, II, IN HIS OFFICIAL CAPACITY AS
SECRETARY OF HEALTH AND HUMAN SERVICES,
APPELLEE

Appeal from the United States District Court
for the District of Columbia
(No. 1:17-cv-02645)

Ashley C. Parrish argued the cause for appellant. With
him on the briefs were Mark D. Polston, Elizabeth N. Swayne,
and Amelia G. Yowell.

R. Scott Caulkins and Jeffrey J. Sherrin were on the brief
for amicus curiae American Association of Bioanalysts in
support of plaintiffs-appellant.

David McAloon was on the brief for amici curiae The
College of American Pathologists, et al. in support of appellant.

Dennis Fan, Attorney, U.S. Department of Justice, argued
the cause for appellee. With him on the brief were Abby C.
 2
Wright, Attorney, Robert P. Charrow, General Counsel, U.S.
Department of Health & Human Services, Janice L. Hoffman,
Associate General Counsel, Susan Maxson Lyons, Deputy
Associate General Counsel, and Debra M. Laboschin,
Attorney. Alisa B. Klein, Attorney U.S. Department of Justice,
entered an appearance.

Before: GRIFFITH, MILLETT and PILLARD, Circuit Judges.

Opinion for the Court filed by Circuit Judge PILLARD.

PILLARD, Circuit Judge: The Protecting Access to
Medicare Act (PAMA, or the Act) seeks to align Medicare
reimbursement rates for laboratory tests with rates paid for
such tests in the private market. To enable the Secretary of
Health and Human Services (HHS) to ascertain the private
market’s reimbursement rates, PAMA requires “applicable
laboratories” to report private payor data to the Secretary that
the Medicare program then uses to set new, presumably lower,
Medicare reimbursement rates. The Secretary must implement
the statute’s definition of “applicable laboratory” before it can
be used to collect the requisite data. In 2016, the Secretary
issued a final rule doing so, and plaintiff American Clinical
Laboratory Association (ACLA) filed suit claiming the rule
unlawfully excluded most hospital laboratories from the Act’s
reporting requirements.

Based on PAMA’s prohibition of judicial review of “the
establishment of payment amounts,” the district court
dismissed ACLA’s complaint for lack of subject matter
jurisdiction. We conclude that the statutory provision stripping
jurisdiction to review payment amounts does not cover the
statute’s data-collection provision. We also reject ACLA’s
claim that the Secretary’s rule was ultra vires. We thus reverse
 3
and remand to the district court to consider in the first instance
whether the rule comports with the APA.

I. Background

The federal Medicare program, which pays for healthcare
for elderly and disabled individuals, see 42 U.S.C. § 1395 et
seq., is the nation’s largest purchaser of clinical laboratory
services. In 2013, the HHS Office of Inspector General
concluded that Medicare was paying 18 to 30 percent more
than private insurance companies for a range of common
laboratory tests. Congress responded by enacting the
Protecting Access to Medicare Act, Pub. L. No. 113-93, 128
Stat. 1040 (2014). A central goal of the Act is to set Medicare
reimbursement rates for laboratory tests at approximately the
price private insurers pay for the same tests. See 42 U.S.C.
§ 1395m-1(b)(1)(A).

To inform the Secretary’s rate setting, the statute requires
“applicable laborator[ies]” within the private sector to report
“private payor” data—both the price and volume of laboratory
tests—to HHS every three years. Id. § 1395m-1(a). The statute
defines the term “private payor” as a “health insurance issuer
and a group health plan,” a “Medicare Advantage plan,” or a
“medicaid managed care organization.” Id. § 1395m-1(a)(8).
It calls on the Secretary to establish parameters for data
collection through notice and comment rulemaking. Id.
§ 1395m-1(a)(12). Applicable laboratories that fail to report
accurate data face monetary penalties of up to $10,000 per day.
Id. § 1395m-1(a)(9).

In separate provisions, the statute explains how the
Secretary is to use private payor data on each laboratory test
already available in the market to calculate a “weighted
median” rate, which becomes Medicare’s reimbursement rate
for that test. Id. § 1395m-1(b). For new tests that do not have
 4
private payor data, the Secretary is to use a “gapfilling process”
and consult with an expert advisory panel to establish the new
test’s Medicare payment rate. Id. §§ 1395m-1(c)-(d), (f).
Because the market-approximating Medicare rates are likely to
be lower than existing Medicare rates, the statute allows for a
multi-year “[p]hase-in” process to transition to the market-
based rates. Id. § 1395m-1(b)(3). Finally, the statute declares
that “[t]here shall be no administrative or judicial review . . . of
the establishment of payment amounts under this section,” id.
§ 1395m-1(h)(1), thereby barring otherwise available review
by the Departmental Appeals Board and Provider
Reimbursement Board as well as the federal courts.

This appeal is about whether the Secretary’s
implementation of PAMA’s definition of “applicable
laboratories” is subject to review in response to a claim that it
unlawfully excludes hospital laboratories—which tend to
charge higher prices than standalone laboratories—from the
dataset used to determine new Medicare rates. See Medicare
Program; Medicare Clinical Diagnostic Laboratory Tests
Payment System, 81 Fed. Reg. 41,036 (June 23, 2016).
Laboratory tests are available to the public through three main
types of laboratories: physician-office laboratories (which in
2015 comprised 17 percent of all labs), independent
laboratories (50 percent of all labs), and hospital laboratories
(33 percent of all labs). The statute does not expressly discuss
those distinct types of institutional settings, instead generally
defining an applicable laboratory as “a laboratory that, with
respect to its revenues under this subchapter, a majority of such
revenues are from this section, section 1395l(h) of this title, or
section 1395w-4 of this title.” 42 U.S.C. § 1395m-1(a)(2). In
plain terms, that definition refers to a laboratory that receives
most of its overall Medicare funding from the Physician Fee
Schedule (PFS) or the Clinical Laboratory Fee Schedule
(CLFS). Those Medicare fee schedules are typically used to
 5
pay for laboratory services provided by independent
laboratories and physician-office laboratories. We refer to this
part of the statute’s definition of applicable labs as the
majority-payments test. 1

Applying the majority-payments test to hospital
laboratories has proved more complicated than for independent
laboratories and physician-office laboratories. Medicare
reimburses laboratory services provided by hospital
laboratories in a range of different ways, and it is not obvious
which, if any, are relevant to PAMA. When hospital
laboratories serve admitted inpatients and registered
outpatients, Medicare does not use the PFS or CLFS, but pays
for those services through distinct fee schedules that bundle the
laboratory testing with other hospital services. See Medicare
Program; Medicare Clinical Diagnostic Laboratory Tests
Payment System, 80 Fed. Reg. 59,386, 59,393 (Oct. 1, 2015).
A hospital laboratory serving only inpatients and outpatients
accordingly is not an applicable laboratory under PAMA
because it receives no Medicare reimbursements from the
applicable fee schedules.

However, some hospitals also provide “outreach
services”—that is, laboratory services for people who are
neither inpatients nor outpatients. Hospitals’ outreach services
may compete for such business with independent laboratories,
and Medicare reimburses hospitals for those services under the

1
The statute authorizes the Secretary to establish a “low volume or
low expenditure threshold” to exclude especially small laboratories.
42 U.S.C. § 1395m-1(a)(2). The Secretary has set the low
expenditure threshold at $12,500, meaning that labs receiving less
than $12,500 of Medicare revenue from the CLFS and PFS are
exempt from reporting requirements. ACLA does not challenge that
provision, which exempts approximately 95 percent of physician-
office laboratories.
 6
PFS and CLFS. Considered as a freestanding entity, a hospital
laboratory that offered outreach services could fit the statutory
definition of an applicable laboratory if it received most of its
Medicare revenue from the PFS and CLFS.

In October 2015, the Secretary proposed a rule to
implement the data reporting provision of PAMA. See 80 Fed.
Reg. at 59,386. The Secretary acknowledged that it was
“important” to “define laboratory broadly enough to
encompass every laboratory type that is subject to” the
applicable fee schedules, id. at 59,391, so proposed to include
any “entity that includes a laboratory” as well as any “entity
that itself is a laboratory,” id. at 59,392. But to make the
threshold identification of the relevant entity to be scrutinized
under the statutory majority-payments test, the proposed rule
defined “entity” as the institutional unit associated with a
distinct taxpayer identification number (TIN). Id. at 59,387. A
TIN is the institutional identifier Medicare service providers
use to report to the IRS tax-related information about all types
of Medicare payments. Id. at 59,421.

The practical effect of applying the majority-payments test
to an entity defined by its TIN would be that essentially no
hospital laboratory could qualify as an applicable laboratory.
See id. at 59,393. An independent laboratory that takes
Medicare payments has a TIN, but so does an entire hospital
that takes Medicare payments. Identifying the relevant entity
at the level of the TIN would mean that, for purposes of the
majority-payments test, Medicare reimbursements to a hospital
lab under the PFS and CLFS would be compared to the entire
hospital’s Medicare revenue rather than to its laboratory’s
Medicare revenue. Hospitals, in contrast to stand-alone
laboratories, receive Medicare reimbursements for a wide
range of services—everything from surgeries to room and
board—so even if the hospital laboratory received the majority
 7
of its Medicare revenue from the PFS and CLFS, the majority
of the hospital’s overall Medicare revenues would not come
from those fee schedules.

Insofar as hospitals’ Medicare revenue associated with
those fee schedules is dwarfed by their other types of Medicare
revenue, reliance on the TIN to define the relevant entity would
exclude all hospital-based laboratories from the “applicable
laboratories” reporting requirement at the threshold, without
subjecting the laboratories as such to the majority-payments
test. The Secretary anticipated that consequence, explaining:
“[W]e believe the statute intends to limit reporting primarily to
independent laboratories and physician offices . . . and not to
include other entities (such as hospitals, or other health care
providers) that do not receive the majority of their revenues
from PFS or CLFS services.” Id. at 59,393.

Many healthcare entities and other stakeholders opposed
that part of the proposed rule, arguing that reading the statute’s
definition of applicable laboratory to generally exclude
hospital laboratories violated PAMA. See, e.g., 81 Fed. Reg.
at 41,045. Why did that matter? A major laboratory services
provider asserted that hospital laboratories receive 1.5 to 4
times higher private payor reimbursement rates than
independent laboratories for the same test. (The record does
not specify whether that estimate isolated reimbursement
pursuant to the PFS or CLFS schedules, or included laboratory
tests provided—presumably at greater cost—as part of
inpatient and outpatient services). The American Hospital
Association advocated in favor of requiring hospital
laboratories to report their rate data, believing that would
“generally increase the weighted median,” and make the new
Medicare rates “more representative of overall market rates.”
Joint Appendix (J.A.) 595 (letter to Acting Administrator of
Centers for Medicare & Medicaid Services).
 8
Commenters advocated identifying entities by their
National Provider Identifier (NPI) numbers rather than their
TIN. Healthcare providers use NPI numbers to bill Medicare.
Because more hospital laboratories have NPI numbers distinct
from those of their associated hospitals, they reasoned,
identifying the relevant entity at the NPI level would mean
hospital laboratories with their own NPI numbers would be
evaluated for whether they, considered separately, meet the
majority-payments test defining “applicable laboratories.”

In the final rule issued in June 2016, the Secretary accepted
that suggestion and defined applicable laboratories at the NPI
level rather than the TIN level. 2 81 Fed. Reg. at 41,037. The
final rule reiterated the Secretary’s reading of the statute as
excluding the majority of hospital laboratories because those
laboratories receive most of their Medicare revenue from
bundled payments for inpatients and outpatients rather than
from the CLFS or PFS. Id. at 41,045. However, the Secretary
“agree[d] with commenters” that hospital outreach
laboratories—i.e., “laboratories that furnish laboratory tests for
patients that are not admitted hospital inpatients or registered
outpatients of the hospital” and that “are enrolled in Medicare
separately from the hospital of which they are a part”—“should
be accounted for” in the new payment rates. Id.

As it turns out, however, NPIs suffer from virtually the
same flaw as TINs because “[v]ery few hospitals have
laboratory-specific NPIs, and they generally submit claims
under the hospital’s NPI.” J.A. 271 (letter from ACLA to
Department of Health and Human Services Office of Inspector

2
The rule rejected commenters’ alternative suggestion that entities
be identified by their Clinical Laboratory Improvement Amendments
(CLIA) certificate, which is used to certify that a laboratory meets
health and safety regulations, because that certificate is not
associated with Medicare billing. 81 Fed. Reg. at 41,045-46.
 9
General). In 2016, approximately 2,000 out of 260,000 total
laboratories nationwide (0.7 percent) reported data to the
Secretary. Out of the approximately 7,000 hospital laboratories
in the United States, only 21 (0.3 percent) reported data,
comprising about one percent of all the reporting labs. Much
of the data collected by the Secretary came from the country’s
two largest independent laboratories—Quest and Labcorp—
which have lower cost structures than other laboratories. See
J.A. 70-71 (Declaration from Senior Vice President of Quest
Diagnostics). According to ACLA, that skewed the Medicare
reimbursement rates low. The government, for its part,
maintains that the data used to calculate the 2018 rates “was
sufficient and resulted in accurate weighted medians of private
payor rates.” See Medicare Program; Revisions to Payment
Policies Under the Physician Fee Schedule and Other
Revisions, 83 Fed. Reg. 59,452, 59,672 (Nov. 23, 2018).

In 2018, the Secretary finalized another rule—not at issue
here—that amended the implementation of “applicable
laboratories” in an effort to include more hospital-based
outreach laboratory services in the next set of data. Id. at
59,452. As HHS explained, “we are confident that our current
policy supports our collecting sufficient applicable information
. . . and that we received sufficient and reliable applicable
information with which we set [2018 rates],” but “we continue
to consider refinements to our policies that could lead to
including even more applicable information for the next data
reporting period.” Id. at 59,672. The 2018 rule requires
laboratories providing outreach services to report data using the
CMS–1450 14x TOB—a billing form used only by hospital
outreach laboratories. Id. at 59,673-75. The new rule, in effect,
categorizes as an applicable laboratory that portion of a
hospital laboratory that provides outreach services—even if
those services comprise only a minority of the laboratory’s
overall services. See id. at 59,673 (“[W]e believe that if we
 10
were to utilize such an approach in defining applicable
laboratory, all hospital outreach laboratories would meet the
majority of Medicare revenues threshold . . .”). Notably, this
approach was not encompassed by the relief that ACLA seeks
here. ACLA is emphatic that the statute requires the Secretary
to use the hospital laboratory as the denominator in the
majority-payments test. See Appellant’s Br. 65. The Secretary
will use the new data about hospital laboratories’ outreach
services in a revised fee schedule as of January 1, 2021. 80
Fed. Reg. at 59,667.

Plaintiff in this case, American Clinical Laboratory
Association, is a trade association of laboratories. It submitted
comments to the Secretary both before the 2015 proposed rule
and before the 2016 final rule. See J.A. 82. ACLA brought
suit in 2017 to challenge the 2016 rule’s implementation of
applicable laboratory as contrary to the statute and arbitrary
and capricious in violation of the APA. The district court
dismissed for want of jurisdiction. Am. Clinical Lab. Ass’n v.
Azar, 334 F. Supp. 3d 301, 309 (D.D.C. 2018). This Court
reviews de novo a district court’s legal determination as to
subject matter jurisdiction. See Piersall v. Winter, 435 F.3d
319, 321 (D.C. Cir. 2006).

II. Standing

Although the Secretary scarcely challenges standing on
appeal, we have an independent obligation to assure ourselves
that ACLA has standing to challenge the final rule. Steel Co.
v. Citizens for a Better Env’t, 523 U.S. 83, 94-95 (1998). The
“constitutional minimum” of standing requires that a plaintiff
have suffered a concrete and particularized injury that is fairly
traceable to the challenged conduct and is likely to be redressed
by a favorable decision. See Lujan v. Defs. of Wildlife, 504
U.S. 555, 560-61 (1992). In order to have associational
 11
standing, ACLA must demonstrate that “at least one of [its]
members satisfies” this test. See Am. Library Ass’n v. FCC,
401 F.3d 489, 492 (D.C. Cir. 2005). At the motion to dismiss
stage, “general factual allegations of injury resulting from the
defendant’s conduct may suffice” to establish standing. Lujan,
504 U.S. at 561. We hold that ACLA meets those familiar
requirements.

First, ACLA has established injury in fact. By adopting an
impermissible definition of “applicable laboratories” that
excludes virtually all hospital laboratories, ACLA asserts that
the rule harms its members in various ways. First, it
disproportionately burdens independent laboratories with the
cost of data-production obligations not borne by its hospital-
based competitor laboratories. ACLA submitted a declaration
from the Senior Vice President of Quest Diagnostics, an ACLA
member, asserting that “laboratories that reported private payor
information were significantly disadvantaged as compared to
other laboratories that, while required to report under PAMA,
were excused from that obligation by the Secretary.” J.A. 73.
Second, it artificially depresses the reimbursement rates by
excluding data from a portion of the market that receives
higher-than-average Medicare reimbursements for its
laboratory services. An affidavit from the Chief Executive
Officer of Joint Venture Hospital Laboratories, LLC, an ACLA
member, attests that the “elimination of . . . hospital[]
laboratories from the reporting requirements skews the data”
that is used to calculate the weighted median of commercial
payor rates and ultimately set the Medicare reimbursement
rate. J.A. 61. And, because “hospital laboratories typically
receive higher commercial rates than other types of
laboratories,” the Medicare reimbursement rates are lower than
they would be if the Secretary collected more data from
hospital laboratories. Id.; see also U.S. Gov’t Accountability
Office, GAO 19-67, Medicare Laboratory Tests;
 12
Implementation of New Rates May Lead to Billions in Excess
Payment 12 (2018), https://www.gao.gov/assets/
700/695756.pdf (hospital laboratories “typically receive
relatively higher private-payer rates . . . by leveraging the
market power of their affiliated hospital when negotiating rates
with private payers.”). ACLA has adequately shown that at
least one of its members is reimbursed by Medicare at a rate
lower than it would be if we were to rule in ACLA’s favor and
invalidate the challenged limitation in the definition of
“applicable laboratory.” That establishes injury in fact.

As for causation and redressability, ACLA has met its
burden at this stage. See Lujan, 504 U.S. at 561. According to
the Joint Venture Hospital Laboratories affidavit, excluding
data from hospital laboratories “significantly depress[es]” the
weighted median payment rates that are used to generate the
new Medicare fee schedule. J.A. 62. Another declaration, by
the President of Aculabs, Inc. (also a member of ACLA),
attests: “If the Secretary’s failure to implement Congress’s
directives is not corrected, the impact on Aculabs’ business will
be severe” because Aculabs “will not receive Medicare-derived
reimbursement sufficient to cover its costs.” J.A. 49.
Requiring the Secretary to collect more fully representative
market data and use it to calculate a new weighted median
appears sufficiently likely to increase Medicare reimbursement
rates to establish redressability at this stage.

The Secretary briefly protests that ACLA cannot claim
lower repayment rates as an injury for standing purposes
because the statute expressly prohibits challenging the rates.
See Appellee’s Br. 25-26. That argument conflates two issues.
It is true that ACLA cannot challenge the rates themselves
under the statute’s jurisdiction-stripping provision. See 42
U.S.C. § 1395m-1(h)(1). But that does mean the rates cannot
be the source of ACLA’s members’ injury in a challenge to the
 13
data-collection rule. What matters is that ACLA’s challenge
(here, to the definition of applicable laboratory for purposes of
data collection) is sufficiently linked to its asserted injury
(lower reimbursement rates). See Sierra Club v. EPA, 292 F.3d
895, 899 (D.C. Cir. 2002). We assess in the next section the
distinct issue whether ACLA’s challenge is an impermissible
back-door effort to challenge reimbursement rates in
circumvention of the statutory bar.

III. Jurisdiction Stripping

The primary question on appeal is whether PAMA’s
provision eliminating administrative and judicial review of the
“establishment of payment amounts,” 42 U.S.C. § 1395m-
1(h)(1), bars our review of the rule the Secretary promulgated
to implement the statute’s data-collection provision, 81 Fed.
Reg. 41,036. When deciding whether a statute bars judicial
review, we begin with “the strong presumption that Congress
intends judicial review of administrative action.” Smith v.
Berryhill, 139 S. Ct. 1765, 1776 (2019) (quoting Bowen v.
Mich. Acad. of Family Physicians, 476 U.S. 667, 670 (1986)).
Even where, as here, a statutory provision expressly prohibits
judicial review, the presumption applies to dictate that such a
provision be read narrowly. See Dart v. United States, 848
F.2d 217, 221 (D.C. Cir. 1988). “When a statute is ‘reasonably
susceptible to divergent interpretation, we adopt the reading
that accords with traditional understandings and basic
principles: that executive determinations generally are subject
to judicial review.’” Kucana v. Holder, 558 U.S. 233, 251
(2010) (quoting Gutierrez de Martinez v. Lamagno, 515 U.S.
417, 434 (1995)). “Whether and to what extent a particular
statute precludes judicial review is determined not only from
its express language, but also from the structure of the statutory
scheme, its objectives, its legislative history, and the nature of
 14
the administrative action involved.” Block v. Cmty. Nutrition
Inst., 467 U.S. 340, 345 (1984).

We start with the text: “There shall be no administrative
or judicial review under section 1395ff of this title, section
1395oo of this title, or otherwise, of the establishment of
payment amounts under this section.” 42 U.S.C. § 1395m-
1(h)(1). The header of subsection (a) of the statute—the part
that defines the parameters for data collection—cross-
references payment rates in announcing that it deals with
“[r]eporting of private sector payment rates for establishment
of medicare payment rates.” See id. § 1395m-1(a). The district
court concluded that the language of the jurisdiction-stripping
provision combined with the header of subsection (a) meant
that data collection was “part and parcel of ‘the establishment
of payment amounts under this section,’ which Congress
shielded from judicial review.” Am. Clinical Lab. Ass’n, 334
F. Supp. 3d at 307.

That conclusion is plausible, but the text does not compel
it. Several features of the statute suggest that Congress meant
to bar challenges to the “establishment of payment amounts”
but not to prevent review of the rule delineating the data
collection practices that precede and inform the setting of those
amounts. The jurisdiction-stripping provision itself bars
review “under section 1395ff of this title, section 1395oo of
this title, or otherwise.” 42 U.S.C. § 1395m-1(h)(1). The two
cross-referenced sections cover administrative appeals by
patients or providers who wish to contest a coverage
determination or reimbursement amount. See 42 U.S.C.
§§ 1395oo, 1395ff. That suggests Congress intended to
preclude review of the amounts of money paid in the ultimate
reimbursement decisions. Additionally, subsection (a)’s
reference to reporting private sector data for the establishment
of payment amounts suggests that the two are not one and the
 15
same, but rather that collecting data from the private sector is a
separate statutory duty preceding the establishment of
Medicare payment rates. See 42 U.S.C. § 1395m-1(a). Indeed,
the final rule on data collection is not an “establishment of
payment amounts,” but a blueprint for which laboratories must
report private payor data to the Secretary, how they must do so,
and what consequences they face for noncompliance. 81 Fed.
Reg. at 41,037.

The structure of PAMA bespeaks the separation between
data collection and pricing. In subsections (b), (c), and (d) on
existing, new, and new advanced diagnostic laboratory
services, Congress explained that it was directing the Secretary
henceforth to calculate “weighted median” market-based
prices for existing services and to “gapfill” and consult an
expert panel to determine prices for new and new advanced
services. In a separate provision, Congress detailed the
framework for data collection. Compare 42 U.S.C. § 1395m-
1(b)-(d) (establishing processes for determining Medicare
rates), with id. § 1395m-1(a) (establishing processes for
collecting private-payor data). Whereas the rate-setting
provisions affect reimbursements for Medicare services, the
data-collection part of the statute imposes new obligations on
private parties (applicable laboratories). The latter requires
reporting of confidential data about private market rates—data
not otherwise used in any government program—so that the
Secretary may analyze it to formulate Medicare rates. Id.
§ 1395m-1(a). The Secretary is required to “establish through
notice and comment rulemaking parameters for data collection
under this subsection.” Id. § 1395m-1(a)(12). An applicable
lab’s failure to provide the specified information exposes it to
significant monetary penalties. Id. § 1395m-1(a)(9).

Textual and structural analysis of jurisdiction-stripping
provisions in other statutes supports the Secretary’s position
 16
that Congress did not bar review of PAMA’s entire process for
collecting data on private-payor rates. See Bowen, 476 U.S. at
675-76 (holding that the provision limiting review of amounts
of benefit payments under the Medicare program did not bar
review of the method by which such amounts were computed).
In Texas Alliance for Home Care Services v. Sebelius, for
example, we concluded that a provision’s “broad and
unequivocally preclusive language” barring review of “the
awarding of contracts” under Medicare also barred review of
the “applicable financial standards” that medical equipment
suppliers must meet in order to be eligible for those contracts.
681 F.3d 402, 409-11 & n.9 (D.C. Cir. 2012). In that case, the
challenged “applicable financial standards” were listed in the
statute itself as a “[c]ondition[] for awarding contract[s].” 42
U.S.C. § 1395w-3(b). In other words, the statutory text made
clear that the challenged action (defining applicable financial
standards) was encompassed within the terms of unreviewable
action (awarding contracts). The same was true in Mercy
Hospital, Inc. v. Azar, where we held that a statute barring
judicial review of “prospective payment rates” covered a
challenge to the formula the Secretary used to calculate those
rates. 891 F.3d 1062, 1067 (D.C. Cir. 2018). Through internal
cross-references, the “language of the statute tie[d] together the
prospective payment rate and the statutory adjustments” used
to calculate that rate. Id.

Unlike the provisions at issue in Texas Alliance and Mercy
Hospital, the statutory text here does not subsume the data
collection process within the establishment of payment
amounts. On the contrary, Congress set out the process for data
collection in a separate and distinct subsection and with its own
set of rules. Congress also required that the parameters for that
data collection be established through notice and comment
rulemaking. See 42 U.S.C. § 1395m-1(a)(12). Neither of the
statutes at issue in Texas Alliance or Mercy Hospital had
 17
explicit notice and comment requirements. And, as our
precedent makes clear, part of the purpose of notice and
comment rulemaking is to ensure the parties develop a record
for judicial review. See Int’l Union, United Mine Workers of
Am. v. Mine Safety & Health Admin., 407 F.3d 1250, 1259
(D.C. Cir. 2005) (“[Rulemaking n]otice requirements are
designed . . . to give affected parties an opportunity to develop
evidence in the record to support their objections to the rule and
thereby enhance the quality of judicial review.”).

Because the gathering of data under PAMA is not
“inextricably intertwined” with the establishment of payment
rates, we lack a basis on which to infer that Congress, in
eliminating jurisdiction over the latter, clearly meant also to bar
review of the former. Cf. Florida Health Scis. Ctr., Inc. v.
Sec’y of Health & Human Servs., 830 F.3d 515, 521 (D.C. Cir.
2016). We held that we lacked jurisdiction in Florida Health
for reasons that appear at first blush to apply here. The statute
at issue in Florida Health required the Secretary to identify as
Disproportionate Share Hospitals (DSH) entitled to additional
federal compensation those hospitals serving high proportions
of poor patients. When the Affordable Care Act directed that
DSH status be based largely on the percentage of the nation’s
overall uncompensated care each hospital provides, the
Secretary chose to estimate that percentage by reference to the
number of days Medicaid and low-income Medicare patients
spent in a given hospital as a proportion of the national total of
such patients’ hospital days. Id. at 517; see 42 U.S.C.
§ 1395ww(r)(2)(C). In other words, the Secretary used a
hospital’s national share of Medicaid and low-income
Medicare patient care as a proxy for its share of uncompensated
care. See Florida Health, 830 F.3d at 517. The Secretary did
so because research supported the correlation between a
hospital’s national shares of uninsured patients and its poor-
insured, and the proxy data was already readily available and
 18
subject to audit. Id. The statute expressly precluded review of
the Secretary’s uncompensated care estimates, see id.
§ 1395ww(r)(3)(A), but plaintiffs sought review of those
estimates anyway by challenging the accuracy of the Medicaid
data on which they were based, Florida Health, 830 F.3d at
518. We held that the statutory bar on the Secretary’s
“estimate” of how much uncompensated care a hospital
provided also barred review of the proxy data on which the
Secretary relied. Id. at 518-19. We “could not review a
decision that was ‘indispensable’ or ‘integral’ to, or
‘inextricably intertwined,’ with the unreviewable agency
action.” Id. at 519 (quoting Texas Alliance, 681 F.3d at 409-
11).

Important distinctions between the issues in this case and
Florida Health show that the data collection process at issue
here is not “inextricably intertwined” with the unreviewable
establishment of payment amounts. Most importantly, unlike
PAMA, the statute in Florida Health did not have a separate
data-collection provision imposing new obligations on private
parties nor did it have a notice and comment requirement. It
simply directed the Secretary to estimate the amount of
uncompensated services using “appropriate data,” including
data that the Secretary may “determine[]” serves as an adequate
proxy for uncompensated care rates. 42 U.S.C.
§ 1395ww(r)(2)(C). The statute’s text and structure made clear
that choosing which data to use was part of the Secretary’s
unreviewable obligation to estimate uncompensated care rates.
Florida Health, 830 F.3d at 517, 519.

PAMA’s data collection provision, on the other hand, is
distinct from its rate-estimation provisions. For data collection,
the statute obligates clinical laboratories that participate in the
Medicare program to report distinct reimbursement rates they
receive from private insurers and requires the Secretary to
 19
establish the rules governing that reporting through notice and
comment rulemaking. To be sure, the results of that data
collection process are used to establish Medicare payment
amounts. But the statute’s bifurcated structure supports
ACLA’s view that the two provisions and the processes they
require are distinct. This case differs from DCH Regional
Medical Center v. Azar, 925 F.3d 503 (D.C. Cir. 2019), for
similar reasons. We held there that a “methodology” used to
generate uncompensated care estimates under the same statute
at issue in Florida Health was itself unreviewable because
there was “no textual basis for separating estimates from their
underlying methodology,” id. at 507; the data-collection
provision of PAMA, in contrast, is separate from the provisions
establishing payment amounts. That the statutory scheme
requires private laboratories to report non-Medicare and
generally confidential business information (private market
rates) to the government on pain of monetary penalties further
stands this statutory scheme in sharp contrast to others where
the challenged action was found to be intertwined with other
agency actions regarding which Congress had barred judicial
review.

The government argues that it would make scant sense for
Congress to have barred review only of “basic math” while
“permitting review of every discretionary step that preceded
that math.” Appellee Br. 33. But establishing payment
amounts sometimes involves more than rote math. For
established laboratory services, the Secretary must array
private payor data from thousands of laboratories and calculate
a weighted median for each separate laboratory service. 42
U.S.C. § 1395m-1(b)(2). For new tests, regarding which
private payor data does not yet exist, the Secretary must
consider more complicated criteria and consult an expert panel.
See id. § 1395m-1(c)-(d), (f). That process is a good deal more
complex and discretionary than rote math.
 20
In view of PAMA’s text, its structure, and the distinct
nature of the processes of data collection and establishment of
payment rates, we cannot conclude that the bar against
reviewing the “establishment of payment amounts” also
prevents our review of the rule setting up a new and detailed
process for collecting data on market rates that private insurers
pay to laboratories. Because the statute is “reasonably
susceptible” to this interpretation, we hold that it does not bar
judicial review of the Secretary’s rule establishing the
parameters of data collection under 42 U.S.C. § 1395m-1(a).
Gutierrez de Martinez, 515 U.S. at 434.

IV. Ultra Vires

ACLA also argues that, even if the jurisdictional limitation
of Section 1395m-1(h) applies, we should nonetheless review
the Secretary’s final rule because it “exceeds his statutory
authority and is ultra vires.” Appellant’s Br. 61. Although we
hold that the statute itself does not bar review, we nonetheless
consider ACLA’s ultra vires argument because, if valid, it
would not just open the courthouse door, but invalidate the rule
and obviate any need to remand to the district court for
consideration of the arbitrary-and-capricious challenge.

If an agency exceeds “its statutory bounds, judicial review
remains available” to curb the rogue action. SAS Inst., Inc. v.
Iancu, 138 S. Ct. 1348, 1359 (2018). To challenge agency
action on the ground that it is ultra vires, ACLA must show a
“patent violation of agency authority.” Indep. Cosmetic Mfrs.
& Distribs., Inc. v. U.S. Dep’t of Health, Educ. & Welfare, 574
F.2d 553, 555 (D.C. Cir. 1978). Ultra vires review “is intended
to be of extremely limited scope,” and it “represents a more
difficult course . . . than would review under the APA.”
Trudeau v. Fed. Trade Comm’n, 456 F.3d 178, 190 (D.C. Cir.
2006) (internal quotations omitted).
 21
Here, the statute says that applicable laboratory “means a
laboratory that, with respect to its revenues under this
subchapter, a majority of such revenues are from” the PFS and
CLFS. 42 U.S.C. § 1395m-1(a)(2) (emphasis added). ACLA
argues that choosing to compare a laboratory’s total revenues
from the PFS and CLFS against the “total Medicare revenues
of any entity with an NPI (of which the laboratory is often only
one component),” Appellant’s Br. 64, violates the statute’s
command that the reporting unit be the laboratory rather than a
broader entity. Again, the reporting unit matters because
comparing a hospital laboratory’s reimbursements from PFS
and CLFS to the entire hospital’s Medicare revenue (as
opposed to just the hospital laboratory’s Medicare revenue)
means that a hospital laboratory without its own, laboratory-
specific NPI will not qualify as an applicable laboratory under
the statute.

HHS did not clearly step so far outside the scope of the
task that Congress gave it as to have acted ultra vires. PAMA
does not define the term “laboratory,” and the Secretary’s
charge was to operationalize that important term despite its
ambiguity. After incorporating industry comments into the
final rule, the Secretary chose to identify laboratories by their
NPI numbers. See J.A. 563-64 (Florida Hospital Association
recommending HHS define applicable laboratory at the NPI
level rather than the TIN level); compare 80 Fed. Reg. at
59,387 (proposing use of TIN numbers), with 81 Fed. Reg. at
41,037 (deciding to use NPI numbers). Appellant’s objection
to the Secretary’s efforts, even if meritorious, does not
establish that the Secretary acted ultra vires. We leave to the
district court on remand to address in the first instance the
merits of petitioner’s arbitrary-and-capricious challenge.
 22
* * *

For the reasons discussed above, we reverse the district
court’s holding on subject matter jurisdiction and remand for
further proceedings consistent with this opinion.

So ordered.