Smith v. Burwell , 209 F. Supp. 3d 98 ( 2016 )

UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF COLUMBIA
_________________________________________
)
Cheryl Kort, et al., )
)
Plaintiffs, )
)
v. ) Civil No. 1:14-cv-01519 (APM)
)
Sylvia M. Burwell, )
Secretary of the U.S. Department of Health )
and Human Services, et al., )
)
Defendants. )
_________________________________________ )
MEMORANDUM OPINION
I. INTRODUCTION
Plaintiffs Cheryl Kort, Lorraine Kovnat, and George Hepburn all have exhibited symptoms
of cognitive impairment. None, however, have a firm diagnosis for their respective illnesses.
Diseases associated with cognitive impairment—such as Alzheimer’s Disease, frontotemporal
dementia, and Parkinson’s Disease—are notoriously difficult to diagnose because they share
common symptoms. Doctors attempt to diagnose such diseases by eliminating potential causes, a
process known as “differential diagnosis.” Plaintiffs believe that a recently developed imaging
procedure—a beta-amyloid positron emission tomography scan (“BA Scan”)—can assist their
physicians in eliminating a potential cause of their symptoms and thereby identifying the precise
condition from which each suffers.
The Food and Drug Administration has approved the use of BA Scans for adults with
cognitive impairment who are being evaluated for Alzheimer’s Disease. A negative BA Scan is
inconsistent with a diagnosis of Alzheimer’s and thus potentially could be used to exclude
Alzheimer’s as the cause of a patient’s cognitive impairment. Stated differently, while a BA Scan
cannot be used to definitively diagnose a patient with Alzheimer’s, a negative scan potentially
could rule it out as a cause.
The Medicare program, however, does not cover the costs of BA Scans, except for limited
use in certain clinical studies. That is because, in September 2012, the Centers for Medicare and
Medicaid Services, or CMS, which administers the Medicare program, determined that the then-
existing medical and scientific evidence did not support a finding that BA Scans are reasonable
and necessary for the diagnosis of an illness. Key to CMS’ decision was its finding that the
evidence did not show that BA Scans improved health outcomes of patients exhibiting cognitive
impairment or informed the management of such patients’ diseases.
Plaintiffs filed this suit against CMS, the Department of Health and Human Services, and
the Department’s Secretary Sylvia M. Burwell, arguing that the factors Defendants considered in
evaluating BA Scans and their ultimate denial of Medicare coverage for such scans violated the
Administrative Procedure Act. Specifically, Plaintiffs claim that Defendants’ consideration of
health outcomes and disease management in determining whether to cover BA Scans is contrary
to the plain language of the Medicare Act and inconsistent with similarly-situated coverage
determinations. Further, Plaintiffs argue that even if the “coverage standard” Defendants applied
was proper, the denial of coverage for BA Scans cannot be reconciled with Defendants’ statements
indicating that the procedure has diagnostic value. Defendants counter that the coverage standard
they employed fell well within the broad authority granted to them by Congress and that their
coverage decision was supported by then-existing scientific evidence (or the lack thereof). They
also assert that their coverage decision is congruent with their past actions.
2
Plaintiffs and Defendants have cross-moved for summary judgment. Their Motions are
now before this court. Upon consideration of the parties’ filings and the Administrative Record,
the court finds that Defendants’ denial of Medicare coverage to BA Scans did not violate the
Administrative Procedure Act, except in one respect: the failure to adequately explain why
Medicare covers a different test that relies on similar technology—known as FDG PET scans—
but not BA Scans, for patients who have exhibited symptoms of cognitive decline but whose
diagnosis remains uncertain. The court therefore grants in part and denies in part both Plaintiffs’
and Defendants’ Cross-Motions for Summary Judgment. Further, it remands the Decision Memo
for further proceedings consistent with this Memorandum Opinion.
II. BACKGROUND
A. Regulatory Framework
Medicare is a federally funded health insurance program for the elderly and disabled.
Title XVIII of the Social Security Act, 42 U.S.C. § 1395 et seq.—commonly known as the
Medicare Act—tasks the Secretary of the Department of Health and Human Services (the
“Secretary”) with administering Medicare. The Secretary does so through the Centers for
Medicare and Medicaid Services (“CMS”). See 46 Fed. Reg. 56,911, 56,911-34 (Nov. 19, 1981)
(establishing the Health Care Financing Administration, which was later renamed CMS). CMS,
in turn, contracts with private entities to which healthcare providers and suppliers submit their
claims for reimbursement. See 42 U.S.C. § 1395kk-1; id. § 1395u; 42 C.F.R. § 421.200.
Medicare Part B is one of the program’s four segments. It provides insurance coverage for
outpatient services, including “diagnostic services which are—(i) furnished to an individual as an
outpatient by a hospital or by others under arrangements with them made by a hospital, and
(ii) ordinarily furnished by such hospital (or by others under such arrangements) to its outpatients
3
for the purpose of diagnostic study.” 42 U.S.C. § 1395x(s)(2)(C); see also 42 C.F.R. § 410.28(“Medicare Part B pays for hospital or [other] diagnostic services furnished to outpatients,
including drugs and biologicals required in the performance of the services[.]”). Part B does not
guarantee coverage for all diagnostic services, however. Section 1395y(a)(1)(A) of the Medicare
Act establishes that “no payment may be made under . . . part B . . . for any expenses incurred for
items or services” that “are not reasonable and necessary for the diagnosis or treatment of illness
or injury or to improve the functioning of a malformed body member.” 42 U.S.C.
§ 1395y(a)(1)(A) (emphasis added). That provision lies at the center of this dispute.
Among the ways that the Secretary, through CMS, informs Medicare contractors and
providers of the items and services that she has deemed “reasonable and necessary for the diagnosis
or treatment of illness and injury,” id., are National Coverage Determinations (“NCDs”). An NCD
is a decision “with respect to whether or not a particular item or service is covered nationally.”
See 42 U.S.C. § 1395ff(f)(1)(B); see also 68 Fed. Reg. 55,634, 55,635 (Sept. 26, 2003) (“In
general, an NCD is a national policy statement granting, limiting, or excluding Medicare coverage
for a specific medical item or service.”); 42 C.F.R. § 405.1060(a)(1) (“An NCD is a determination
by the Secretary of whether a particular item or service is covered nationally under Medicare.”).
If an NCD concludes that a particular use of an item or service shall not be covered under Medicare,
such decision is binding on all entities and persons that implement the Medicare program. See 42
C.F.R. § 405.1060(a)(4) (“An NCD is binding on fiscal intermediaries, carriers, . . . [administrative
law judges], and the [Medicare Appeals Council],” among others); see also 42 U.S.C.
§ 1395ff(c)(3)(B)(ii)(I) (“If the Secretary has made a[n NCD] . . . such determination shall be
binding on the qualified independent contractor in making a decision with respect to a
reconsideration [of an initial coverage determination].”). Plaintiffs here ask the court to overturn
4
an NCD that barred Medicare reimbursement for all BA Scans, except for those provided to
patients participating in certain clinical studies.
B. Factual Background
1. PET Scans
BA Scans are within a family of imaging procedures known as Positron Emission
Tomography (“PET”) scans. In scientific terms, a PET scan “is a minimally invasive diagnostic
imaging procedure used to evaluate normal tissues as well as diseased tissues in conditions such
as cancer, ischemic heart disease, and some neurologic disorders,” in which a “radiopharmaceutical
(or ‘tracer’) [that] emits positrons when it decays” is “injected” into a patient and “a positron
camera (tomograph) [is used] to measure the decay of [the radiopharmaceutical] within human
tissue.” See J.A. of Administrative Record, ECF Nos. 36 & 36-1 [hereinafter AR], at 4719. In
terms understandable to the rest of us, a patient receiving the procedure has a substance—the
“radiopharmaceutical” or “tracer”—injected into his or her body, which the PET scan allows a
physician to view. What that substance does in the patient’s body—principally, the rate at which
it decays—may assist a physician in identifying the medical condition from which the patient is
suffering.
PET scans first were developed in the 1970s, but Medicare did not cover any form of the
procedure until 1995. See Defs.’ Cross-Mot. for Summ. J., ECF No. 13 [hereinafter Defs.’ Mot.],
at 13.1 In the two decades since, CMS has expanded Medicare coverage of PET scans to include
a range of scans that employ several different radiopharmaceuticals, which are used to diagnose
1
Citing CMS, Decision Memo for PET (FDG) (CAG-0065N) (Dec. 15, 2000) available at
https://www.cms.gov/medicare-coverage-database/details/nca-decision-memo.aspx?NCAId=85&fromdb=true
[hereinafter 2000 FDG PET Decision], at 4.
5
many different conditions. See id. at 13-14.2 But Medicare does not provide blanket-coverage for
all PET scans. Rather, the NCD Manual, which “lists all Medicare-covered uses of PET scans,”
makes clear that “a particular use of PET scans is not covered unless th[e] manual specifically
provides that such use is covered.” NCD Manual § 220.6 (emphasis added).
2. Alzheimer’s Disease and Beta-Amyloid Plaques
“Dementia is a syndrome involving cognitive and behavioral impairment in an otherwise
alert patient.” AR at 4715. Alzheimer’s Disease, one of several conditions that cause the
syndrome, “is an irreversible dementia characterized by progressive, relentless cognitive and
functional decline.” Id. at 4716. Although Alzheimer’s “is the number one cause of dementia in
older Americans (age 65 and over), contributing to 60-80% of cases,” id., frontotemporal
dementia, cerebrovascular dementia, dementia with Lewey bodies, Parkinson’s Disease, and
Creutzfeldt-Jakob disease, among others, also cause dementia and share many of Alzheimer’s
symptoms. Id. at 4718. It is precisely “because several other neurological diseases can mimic the
dementia seen in” Alzheimer’s that the accuracy of “clinical diagnosis” of the disease is “poor.”
Id.; see also id. at 4717 (“[D]espite being the ‘cornerstone’ of diagnosis, clinical assessment of
[Alzheimer’s] remains poor.”). And it is in large part because clinical diagnosis is poor that
Plaintiffs believe they will benefit from BA Scans. See, e.g., Pls.’ Mot. for Summ. J., ECF No.
10-1 [hereinafter Pls.’ Mot.], at 18 (“Plaintiffs suffer from symptoms of cognitive impairment that
preclude a reliable diagnosis based solely on a clinical assessment. Therefore, each Plaintiff
would benefit from a [BA Scan].” (citations omitted)).
The build-up of beta-amyloid plaques in the brain is a physiological “hallmark[]” of
Alzheimer’s. AR at 4718. Such build-up, however, is “seen in other diseases” as well, including
2
Citing Medicare NCD Manual [NCD Manual], § 220.6, available at https://www.cms.gov/Regulations-and-
Guidance/Guidance/Manuals/downloads/ncd103c1_part4.pdf.
6
several of the dementia-causing conditions with symptoms similar to Alzheimer’s. Id. It also can
“be detected in cognitively normal older adults.” Id. Indeed, “[a]utopsy studies demonstrate that
approximately 33% of older individuals . . . who are cognitively normal have amyloid
accumulation at levels consistent with” Alzheimer’s. Id. Thus, as noted by CMS’ “Decision
Memo for [BA Scans] in Dementia and Neurogenerative Disease” (the “Decision Memo”)—
which accompanies and explains the NCD at issue in this case—while it “is widely accepted that
the presence of amyloid plaques in the human brain is virtually necessary for the diagnosis of”
Alzheimer’s, “there are competing views” as to what exactly that means. Id. The National
Institute on Aging’s Alzheimer’s Association has “conclude[d] that at this point, it remains unclear
whether it is meaningful or feasible to make the distinction between [beta-amyloid plaques] as a
risk factor for developing the clinical syndrome of [Alzheimer’s] versus [beta-amyloid plaque]
accumulation as an early detectable stage of [Alzheimer’s] because current evidence suggests that
both concepts are plausible.” Id. at 4719 (citation and quotation marks omitted).
3. BA Scans
a. FDA Approval of Amyvid
Historically, beta-amyloid plaques “could only be identified upon autopsy.” Pls.’ Mot. at
3. That has changed in recent years with the development of BA Scans, PET scans that allow
physician’s to view beta-amyloid plaques in the body. See AR at 20-21. BA Scans “became
clinically available for the first time in April 2012,” when the Food and Drug Administration
(“FDA”) approved a radiopharmaceutical to be used in connection with such scans. Pls.’ Mot. at
12. That radiopharmaceutical—Amyvid—is manufactured by Lilly USA, LLC (“Lilly”) and was
approved specifically for use with a PET Scan “of the brain to estimate [beta]-amyloid . . . plaque
density in adult patients with cognitive impairment who are being evaluated for Alzheimer’s . . .
7
and other causes of cognitive decline.” AR at 29. In approving Amyvid, the FDA noted that
while a “negative Amyvid scan indicates sparse to no amyloid plaques and is inconsistent with a
. . . diagnosis of [Alzheimer’s]” and a “positive Amyvid scan indicates moderate to frequent
amyloid . . . plaques” which “is present in patients with [Alzheimer’s],” “Amyvid is an adjunct to
other diagnostic evaluations” and “does not establish a diagnosis of [Alzheimer’s] or other
cognitive disorder.” AR at 13.
b. Request for Medicare Coverage of BA Scans
FDA approval of a drug does not automatically entitle it to coverage under Medicare.
Although “[b]oth CMS and the FDA review scientific evidence, and may review the same
evidence,” the agencies operate “under different statutory standards and different delegated
authority.” 68 Fed. Reg. at 55,636. “Whereas the FDA must determine that a product is safe and
effective as a condition of approval, CMS must determine that the product is reasonable and
necessary as a condition of [Medicare] coverage.” Id. Thus, on June 29, 2012, after receiving
FDA approval, Lilly submitted a letter to CMS seeking Medicare coverage for BA Scans that
would use Amyvid. AR at 10-28.
Lilly’s letter to CMS formally requested that the agency “reopen and revise Section 220.6
of the” NCD Manual—which denies Medicare coverage to all PET Scans not specifically listed
in that Section, see NCD Manual § 220.6—to guarantee coverage for BA Scans. Id. at 10. Lilly
argued that BA Scans would “provide physicians with accurate and reliable diagnostic information
with which to evaluate patients within the Medicare population suffering from cognitive
impairment and being evaluated for Alzheimer’s disease and other causes of cognitive decline.”
Id. Lilly urged CMS to grant nationwide coverage for the procedure through the issuance of an
NCD. Id.
8
c. CMS’ Scientific Review of BA Scans
Lilly’s letter prompted CMS to initiate a substantial scientific review process, which
included several public comment periods, the convening of an expert panel, the release of a
proposed NCD, and ultimately, the issuance of an NCD. The process commenced with a public
comment period that began on October 9, 2012, and concluded on November 8, 2012. See AR at
4736. During that period, CMS received 27 comments, 26 of which “supported Medicare
coverage of [BA S]cans in the diagnostic context of suspected dementia.” Id.
Thereafter, on January 30, 2013, the Medicare Evidence Development and Coverage
Advisory Committee (“MEDCAC”)—which was “established to provide independent guidance
and expert advice to CMS on specific clinical topics” and to assist CMS “in making decisions
based upon the reasoned application of scientific evidence,”3—assembled a panel of 16 M.D.- or
Ph.D.-holding experts. See AR at 1000, 1003-04. The panel’s task was “to review available
evidence and hear public testimony on the use of [BA Scans] for the management of dementia and
neurogenerative disease.” Id. at 999, 1003-04. The panel received a presentation from CMS;
heard testimony from five “guest speakers” and public comments from 13 others; participated in
a question-and-answer session with those who presented; and engaged in a discussion of the
evidence before it. See id. at 1000-01, 1004-05; see generally id. at 1464-1906 (MEDCAC
Meeting Transcript).
At the meeting’s conclusion, the panel was asked to vote, using a 5-point scale “with a
score of 1 being low or no confidence, and 5 representing high confidence,” on two questions
relevant to the case at hand:
1. How confident are you that there is adequate evidence to determine whether
or not PET imaging of brain beta[-]amyloid changes health outcomes
3
CMS, Factors CMS Considers in Referring Topics to MEDCAC, at 1 (Dec. 12, 2006), available at
https://www.cms.gov/medicare-coverage-database/details/medicare-coverage-document-details.aspx?MCDId=10.
9
(improved, equivalent or worsened) in patients who display early symptoms
or signs of cognitive dysfunction?,” and
2. How confident are you that these conclusions are generalizable to the
Medicare beneficiary population?”
Id. at 1002. The average vote for Question 1 was 2.17, meaning the panel had “medium-low
confidence” that then-existing evidence was sufficient to evaluate the effects of BA Scans on
health outcomes. Id. at 4741; see also id. at 1853. The average vote for Question 2 was 4.25,
meaning that the panel had “high confidence” that its finding as to Question 1 was generalizable
across all Medicare beneficiaries. Id. at 1875.
d. CMS’ Decision
On July 3, 2013, CMS issued the Proposed Decision Memo for Lilly’s requested coverage
of BA Scans. See generally id. at 1908-74. While an NCD itself may be relatively succinct, see,
e.g., id. at 1-6, such a determination is generally accompanied by a thorough memorandum that
explains CMS’ conclusion, see, e.g., id. at 4709-92. In this case, a 68-page Proposed Decision
Memo—the final version of which would accompany the NCD for BA Scans—provided
background information on PET scans, BA Scans, Alzheimer’s, and other cognitive conditions,
see id. at 1912-18; “summarized the published literature [and scientific studies] on whether [BA
Scans are] beneficial to patients with symptoms of [Alzheimer’s],” id. at 1919; see also id. at
1920-29, 1934-37; discussed the MEDCAC panel, id. at 1930; and presented CMS’ lengthy
analysis of the information before it, id. at 1933-56.
CMS expressly centered its analysis on two “Key Questions,” similar to those posed to the
MEDCAC panel:
1. Is the evidence adequate to conclude that [BA Scans] improve[] meaningful
health outcomes in beneficiaries who display signs and symptoms of
[Alzheimer’s]?
10
2. Is the evidence adequate to conclude that [BA Scan] results inform the
treating physician’s management of the beneficiary to improve meaningful
health outcomes? Those outcomes may include reasonably considered
beneficial therapeutic management or the avoidance of unnecessary,
burdensome interventions.
Id. at 1939.
CMS answered “‘No’ to both questions.” Id. The Proposed Decision Memo found “that
the evidence is insufficient to conclude that the use of [BA Scans] improves health outcomes for
Medicare beneficiaries with dementia or neurogenerative disease, and thus [BA Scans are] not
reasonable and necessary.” Id. at 1955. CMS did, however, propose reimbursement for “one [BA
Scan] per patient through coverage with evidence development . . . in clinical studies that meet
[certain] criteria” set forth by the agency. Id. In other words, CMS proposed limited coverage of
BA Scans for certain clinical trials.
A second public comment period, lasting 30 days, followed the release of the Proposed
Decision Memo. Id. at 4737. CMS received 202 comments. Id. Among the 202 comments was
a letter from Lilly that raised many of the challenges to CMS’ decision that Plaintiffs have raised
here. See id. at 4171-79.
Nearly two months after the close of the second public comment period, on September 27,
2013, CMS released its final “[NCD] for [BA Scans] in Dementia and Neurogenerative Disease,”
id. at 1-3, and the accompanying Decision Memo, id. at 4709-92. Both the NCD and the Decision
Memo mirrored the findings of the MEDCAC panel and the conclusions contained in the Proposed
Decision Memo. Over 84 pages, the Decision Memo explained how and why CMS reached its
final determination. See id. The NCD stated: “[CMS] has determined that the evidence is
insufficient to conclude that the use of [BA Scans] is reasonable and necessary for the diagnosis
or treatment of illness or injury . . . for Medicare beneficiaries with dementia or neurodegenerative
11
diseases, and thus [BA Scans are] not covered under [the Medicare Act].” Id. at 1. It did, however,
provide for limited coverage of BA Scans in certain clinical studies. See id.
III. LEGAL STANDARD
A. Cross-Motions For Summary Judgment on Administrative Procedure
Act Claims
Plaintiffs and Defendants have filed Cross-Motions for Summary Judgment as to Plaintiffs’
claims brought under the Administrative Procedure Act (the “APA”), 5 U.S.C. § 701 et seq. Cross-
motions for summary judgment ordinarily are reviewed under the standard set forth in Federal
Rule of Civil Procedure 56, which requires a court to grant summary judgment when the pleadings
and the evidence demonstrate that “there is no genuine issue as to any material fact and the movant
is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a). However, in cases such as this
one that involve the review of a final agency action, the Rule 56 standard does not apply.
See Stuttering Found. of Amer. v. Springer, 498 F. Supp. 2d 203, 207 (D.D.C. 2007). Instead, “the
district judge sits as an appellate tribunal” and “[t]he ‘entire case’ on review is a question of law.”
Am. Biosci. Inc. v. Thompson, 269 F.3d 1077, 1083 (D.C. Cir. 2001) (citing cases). “[T]he court’s
review is limited to the administrative record,” Fund for Animals v. Babbitt, 903 F. Supp. 96, 105
(D.D.C. 1995) (citing Camp v. Pitts, 411 U.S. 138, 142 (1973)), and its role is limited to
“determin[ing] whether or not as a matter of law the evidence in the administrative record
permitted the agency to make the decision it did,” see Sierra Club v. Mainella, 459 F. Supp. 2d
76, 90 (D.D.C. 2006) (citation and internal quotation marks omitted).
B. Chevron Review
As will be discussed below, Plaintiffs’ primary contention is that Defendants’ reliance on
a coverage standard that considers health outcomes and disease management is inconsistent with
Section 1395y(a)(1)(A) of the Medicare Act. This argument implicates the two-step formula set
12
forth in Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842-44
(1984). Where, as here, “Congress gives an agency authority to interpret a statue, [courts] review
the agency’s interpretation” using the Chevron framework. Council for Urological Interests v.
Burwell, 790 F.3d 212, 219 (D.C. Cir. 2015). Under Chevron’s first step, courts must determine
whether Congress has “directly spoken to the precise question at issue.” Chevron, 467 U.S. at 842.
“If the intent of Congress is clear, that is the end of the matter; for the court, as well as the agency,
must give effect to the unambiguously expressed intent of Congress.” Id. at 842-43. This inquiry
“begins with the text” of the statute, W. Minnesota Mun. Power Agency v. Fed. Energy Regulatory
Comm’n, 806 F.3d 588, 591 (D.C. Cir. 2015) (citing Engine Mfrs. Ass’n v. S. Coast Air Quality
Mgmt. Dist., 541 U.S. 246, 252 (2004)), but may also involve consideration of “the specific context
in which th[e] language [at issue] is used, and the broader context of the statute as a whole.”
Robinson v. Shell Oil Co., 519 U.S. 337, 341 (1997). “Because the judiciary functions as the final
authority on issues of statutory construction, [a]n agency is given no deference at all on the
question whether a statute is ambiguous.” Wells Fargo Bank, N.A., v. Fed. Deposit Ins. Co., 310
F.3d 202, 205-06 (D.C. Cir. 2002) (citation and internal quotation marks omitted).
Where a court determines that Congress has not directly spoken to the precise question at
issue, it must move to Chevron’s second step, which asks whether the agency’s interpretation of
the language in controversy is “based on a permissible construction of the statute.” Chevron, 467
U.S. at 843. An interpretation is permissible if it is a “‘reasonable’ explanation of how an agency’s
interpretation serves the statute’s objectives.” Northpoint Tech., Ltd. v. FCC, 412 F.3d 145, 151
(D.C. Cir. 2005); see also D.C. v. Dep’t of Labor, 819 F.3d 444, 449 (D.C. Cir. 2016) (“[U]nder
Chevron . . . the fundamental question is not whether we think the [agency]’s interpretation is
correct, but whether the [agency]’s interpretation of the Act is at least reasonable in light of any
13
ambiguities in the statute.”). “If the agency’s construction is reasonable, [courts] defer.” Burwell,
790 F.3d at 219 (citing Chevron, 467 U.S. at 842-43).
When applying Chevron’s second step, courts must defer to an agency’s interpretations of
the statute under which it operates. As the Supreme Court stated in Chevron, “[w]e have long
recognized that considerable weight should be accorded to an executive department’s construction
of a statutory scheme it is entrusted to administer, and the principle of deference to administrative
interpretations.” 467 U.S. at 844. And where, as here, a court is tasked with evaluating the
Secretary’s interpretation of a provision of the Medicare Act, such deference is heightened. See,
e.g., Cty. of Los Angeles v. Shalala, 192 F.3d 1005, 1016 (D.C. Cir. 1999) (“In marking off the
metes and bounds of our review under the second step of Chevron, we accord particular deference
to the Secretary’s interpretation of [the Medicare Act] ‘given the tremendous complexity of the
Medicare statute.’” (quoting Appalachian Reg’l Healthcare, Inc. v. Shalala, 131 F.3d 1050, 1054
(D.C. Cir. 1997))); Methodist Hosp. of Sacramento v. Shalala, 38 F.3d 1225, 1229 (D.C. Cir.
1994)) (“[I[n framing the scope of review, the court takes special note of the tremendous
complexity of the Medicare statute. That complexity adds to the deference which is due to the
Secretary’s decision.”); Appalachian Reg’l Healthcare, 131 F.3d at 1054 (noting that the “case
would be easier had the Board [within the Department of Health and Human Services] provided a
more thorough explanation for its decision,” but “nevertheless” finding that the Board’s “reading
[is one] we can accept, especially in light of the particular deference we afford the Secretary given
the tremendous complexity of the Medicare statute.”). This court must bear that deference in mind
in analyzing the action challenged here.
14
C. The “Arbitrary and Capricious” Standard of Review
Plaintiffs also argue that Defendants’ denial of coverage for BA Scans violated the APA
as agency action that was “arbitrary [and] capricious.” 5 U.S.C. § 706(2)(A). Application of the
“arbitrary and capricious” standard and “[t]he analysis of disputed agency action under Chevron
Step Two . . . is often ‘the same, because under Chevron step two, [the court asks] whether an
agency interpretation is arbitrary or capricious in substance.’” Agape Church, Inc. v. FCC, 738
F.3d 397, 410 (D.C. Cir. 2013) (citing Judulang v. Holder, 132 S. Ct. 476, 483 n.7 (2011)). Thus,
similar to Chevron’s second step, the arbitrary and capricious standard of review requires courts
to determine whether the action at issue was based on “reasoned analysis.” Motor Vehicle Mfrs.
Ass’n of U.S. Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 56-57 (1983).
Generally, an agency has engaged in reasoned analysis when the administrative record
indicates it “examine[d] the relevant data and articulate[d] a satisfactory explanation for its action
including a ‘rational connection between the facts found and the choice made.’” Id. at 43 (quoting
Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168 (1962)). Where, however, the
administrative record indicates that an agency “relied on factors which Congress has not intended
it to consider, entirely failed to consider an important aspect of the problem, offered an explanation
for its decision that runs counter to the evidence before the agency, or [made a decision that] is so
implausible that it could not be ascribed to a difference in view or the product of agency expertise,”
it has acted in an arbitrary and capricious manner. Id. Although this standard is not “particularly
demanding,” Pub. Citizen, Inc. v. FAA, 988 F.2d 186, 197 (D.C. Cir. 1993), and a reviewing court
may “uphold a decision of less than ideal clarity if the agency’s path may reasonably be discerned,”
Bowman Transp., Inc. v. Ark.-Best Freight Sys., Inc., 419 U.S. 281, 286 (1974) (citation omitted),
a court is not to “supply a reasoned basis for the agency’s action that the agency itself has not
15
given,” State Farm, 463 U.S. at 43 (citation and internal quotation marks omitted). And, as with
Chevron step two, the arbitrary and capricious standard “appropriately encourages courts to defer
to the agency’s expertise.” Ark Initiative v. Tidwell, 64 F. Supp. 3d 81, 90 (D.D.C. 2014), aff’d,816
F.3d 119 (D.C. Cir. 2016).
IV. DISCUSSION
The court first addresses Plaintiffs’ arguments under Chevron that the coverage standard
applied to BA Scans contravenes the will of Congress and is an impermissible interpretation of the
Medicare Act. In evaluating the coverage standard under Chevron’s second step, the court
simultaneously addresses Plaintiff’s assertion that the coverage standard is arbitrary and
capricious. The court then turns to Plaintiffs’ other claims under the arbitrary and capricious
standard, namely that the Decision Memo is inconsistent both with itself and with a CMS
regulation.
A. Chevron Step One
In applying the first step of Chevron, the court “begin[s], as always, with the plain language
of the statute in question.” Citizens Coal Council v. Norton, 330 F.3d 478, 482 (D.C. Cir. 2003).
Here, the statutory provision at issue reads as follows:
[N]o payment may be made under part A or part B of this subchapter for any
expenses incurred for items or services—(1)(A) which . . . are not reasonable and
necessary for the diagnosis or treatment of illness or injury or to improve the
functioning of a malformed body member.
42 U.S.C. § 1395y(a)(1)(A) (emphasis added).
Plaintiffs contend that the “statute’s use of the disjunctive term ‘or’ to separate items and
services that are reasonable and necessary for ‘diagnosis’ from those that are reasonable and
necessary for ‘treatment’ is critically important.” Pls.’ Mot. at 4-5. “[I]t is a commonsense
‘[c]anon[] of construction,’” Plaintiffs argue, “that when Congress uses the word ‘or’ to connect
16
two terms, it intends to ‘give[] independent significance’ to each.” Pls.’ Mot. at 25 (quoting Reiter
v. Sonotone Corp., 442 U.S. 330, 338-39 (1979)). Relying on definitions contained in dictionaries
and a Medicare manual, Plaintiffs assert that “‘[d]iagnosis’ and ‘treatment’ ordinarily are
understood to have separate meanings,” id. at 23, and “[b]y using these ‘distinct’ terms in
juxtaposition—separated by the disjunctive ‘or’—Congress unambiguously required Defendants
to cover diagnostic services based on their diagnostic value alone,” id. at 20 (emphasis added).
Thus, Plaintiffs claim, the “coverage standard Defendants have imposed on [BA Scans]”—which
considered the effect of the procedure on patients’ health outcomes and disease management—“is
flatly inconsistent with the plain text of the Medicare Act.” Pls.’ Mot. at 22; see also id. at 23
(“This case can be resolved at Chevron’s first step because the Medicare Act expressly requires
Defendants to cover a medical service that is reasonable and necessary to obtaining a ‘diagnosis,’
regardless of its treatment impact[, and] Defendants’ contrary coverage standard cannot be
reconciled with this statutory mandate.”). Stated differently, Plaintiffs argue that a coverage
standard that takes into account treatment-related factors—such as health outcomes and disease
management—to evaluate whether a test is reasonable and necessary for diagnosis, is improper.
Defendants counter that “Plaintiffs’ singular focus on the word ‘or’ leads them to ignore
the critical words ‘reasonable and necessary.’” Defs.’ Mot. at 27. They disagree that the canon of
construction on which Plaintiffs rely is an “absolute rule,” see Reply Mem. in Supp. of Defs.’ Cross-
Mot. for Summ. J., ECF No. 34 [hereinafter Defs.’ Reply], at 4 (quoting Marx v. Gen. Rev. Corp.,
133 S. Ct. 1166, 1177 (2013)), and argue that “[e]ven if Plaintiffs were correct that ‘or’ must be
read in the disjunctive, the statute would still bar payment for items and services not ‘reasonable
and necessary for . . . diagnosis’ of a disease,” Defs.’ Mot. at 27 (quoting 42 U.S.C.
§ 1395y(a)(1)(A) (emphasis added)). Citing the Court of Appeal’s decision in Williams Natural
17
Gas Company v. FERC, 943 F.2d 1320 (D.C. Cir. 1991), Defendants assert that the coverage
standard they employed was not barred under Chevron’s first step. Defs.’ Mot. at 28 (citing
Williams Natural Gas Co., 943 F.2d at 1331 (stating that where a statute “provides neither an
implicit nor an explicit definition of the terms ‘reasonable’ or ‘necessary,’ . . . Congress has left
the task of defining th[ose] term[s] to” the agency) (citation and internal quotation marks omitted)).
The court agrees with Defendants’ reading of the Medicare statute. Plaintiffs have
expended much breath attempting to direct the court’s attention to Section 1395y(a)(1)(A)’s use
of the word “or.” But it is the phrase “reasonable and necessary” that merits greatest regard. That
phrase undoubtedly has a limiting effect on the scope of Medicare coverage, “preclud[ing]
reimbursement for any items or services . . . which are not reasonable and necessary for the
diagnosis or treatment of illness or injury.” Heckler v. Ringer, 466 U.S. 602, 605 (1984) (citation
and internal quotation marks omitted); see also United Seniors Ass’n, Inc. v. Shalala, 182 F.3d
965, 967 (D.C. Cir. 1999) (“If a service is deemed not to have been reasonable and necessary,
Medicare will not make [a] payment[.]”); Power Mobility Coal. v. Leavitt, 404 F. Supp. 2d 190,
194 (D.D.C. 2005) (“All Medicare coverage is limited to services that are medically ‘reasonable
and necessary’ for the diagnosis or treatment of illness.” (emphasis added)). But it also has the
effect of vesting substantial authority in the Secretary. “[T]he words ‘reasonable’ and ‘necessary’
are among the broadest in the congressional lexicon of delegation.” Williams Nat. Gas, 943 F.2d
at 1331. And, Congress’ use of those words means that the “Secretary’s decision as to whether a
particular medical service is ‘reasonable and necessary’ . . . [is] clearly discretionary.” Heckler,
466 U.S. at 617 (emphasis added). To accept Plaintiffs’ argument that the Secretary, and thus
CMS, unlawfully considered health outcomes and disease management here, the court would have
to accept the novel principle that diagnosis and treatment occupy mutually-exclusive spheres in
18
which nothing related to one may be contemplated when assessing the other. The court declines
to read the statue to so substantially restrict the Secretary’s discretion.
Plaintiffs’ reading also suffers from a textual defect. The assertion that “or” creates a hard-
and-fast distinction between “treatment” and “diagnosis” is untenable because the same cannot be
said of the statute’s use of “or” in other contexts. See Sorenson v. Sec’y of Treasury of U.S., 475
U.S. 851, 860 (1986) (“The normal rule of statutory construction assumes that identical words
used in different parts of the same act are intended to have the same meaning.” (citations and
internal quotation marks omitted)). For instance, in the very same sentence, the phrase “illness or
injury” appears. 42 USC § 1395y(a)(1)(A). Those two words cannot reasonably be said to be
mutually-exclusive. Additionally, in the very same section are the phrases “palliation or
management,” id. § 1395y(a)(1)(C), and “cutting or removal,” id. § 1395y(a)(13)(C), among
others. See Hr’g Tr. at 7:7-8:3, Apr. 21, 2016 (draft). The use of the word “or” in those phrases
likewise cannot be said to connote categorically separate and distinct concepts. As these examples
demonstrate, Congress plainly did not intend that “or” in Section 1395y(a) would establish
concepts that are mutually-exclusive and non-overlapping in every respect.
Plaintiffs cite Hays v. Sebelius, 589 F.3d 1279 (D.C. Cir. 2009), in both their Motion and
reply brief for the proposition that “courts have not hesitated to invalidate the Secretary’s
interpretations of the Medicare Act that conflict with its statutory text, including the Secretary’s
interpretations of the same Coverage Provision at issue here.” Pls.’ Mot. at 29; see also Pls.’ Reply
at 12 (“The D.C. Circuit’s decision in Hays v. Sebelius . . . is a classic example of a similar effort
by Defendants to avoid the limits that Congress set in the statute.”). Plaintiffs, however, overstate
the relevance of Hays. Although that case did involve Section 1395y(a)(1)(A), it addressed a very
different question than the one at issue here: “whether ‘reasonable and necessary’ modifies
19
‘expenses’ . . . or ‘items and services.’” Hays, 589 F.3d at 1281. The Court of Appeals determined
that “reasonable and necessary” modifies the latter. And in so holding, it highlighted the great
latitude provided by the phrase “reasonable and necessary.” Id. (“If the Secretary is correct [that
‘reasonable and necessary’ modifies ‘expenses’], then Medicare may, as it has here, partially cover
an item or service, declining to reimburse expenses associated with the marginal difference in price
between a prescribed item or service and its least costly and medically appropriate alternative.”).
In sum, Hays does not alter the court’s conclusion that, under Chevron’s first step, the plain text
of the Medicare Act does not preclude Defendants’ reliance on health outcomes and patient
management to determine what is “reasonable and necessary.”
B. Chevron Step Two and Arbitrary and Capricious Review
Having rejected Plaintiffs’ argument at Chevron step one, the court “consider[s] whether
the content that the agency has given” the words “reasonable and necessary” “is ‘permissible,’” as
the Court used that term in Chevron. Williams Nat. Gas, 943 F.2d at 1331. Plaintiffs do not claim
that Defendants violated the APA by failing to “examine the relevant data”; nor do they argue that
Defendants failed to “articulate a satisfactory explanation for its” consideration of a BA Scan’s
effect on health outcomes and disease management 4 State Farm, 463 U.S. at 43. Instead, under
the combined Chevron step two and “arbitrary and capricious” framework, Plaintiffs argue that
the “coverage standard Defendants have imposed on [BA Scans] squarely conflicts with the
coverage positions they consistently have applied, and continue to apply, to highly similar medical
services.” Pls.’ Mot. at 30.
4
Nor does not the court read Plaintiffs’ Motion as making the more general, and traditional, Chevron step two
argument that Defendants’ interpretation of the Medicare Act is unreasonable. See D.C. v. Dep’t of Labor, 819 F.3d
at 449 (stating that “the fundamental question” under Chevron step two is “whether the [agency]’s interpretation of
the Act is at least reasonable in light of any ambiguities in the statute”). Even if Plaintiffs did make that argument,
the court would reject it for the reasons discussed in Part IV.B.2. below. Specifically, the substantial scientific and
medical literature discussed and/or cited in the Decision Memo supports the consideration of health outcomes and
disease management in assessing the utility of a diagnostic test.
20
1. Treatment of Similar Situations
“[F]oolish consistency” may be the “hobgoblin of little minds,”5 but consistency—whether
foolish or not—is a well-rooted concept in APA jurisprudence. “‘A long line of precedent has
established that an agency action is arbitrary when the agency offer[s] insufficient reasons for
treating similar situations differently.’” Cty. of Los Angeles, 192 F.3d at 1022 (quoting Transactive
Corp. v. U.S., 91 F.3d 232, 237 (D.C. Cir. 1996)); see also Petroleum Commc’ns, Inc. v. F.C.C.,
22 F.3d 1164, 1172 (D.C. Cir. 1994) (“We have long held that an agency must provide adequate
explanation before it treats similarly situated parties differently.”); Local 777, Democratic Union
Org. Comm., Seafarers Int’l Union of N. Am., AFL-CIO v. N. L. R. B., 603 F.2d 862, 872 (D.C. Cir.
1978) (stating that agencies may not “arbitrarily treat similar situations dissimilarly”). To be
certain, the actions of “[r]egulatory agencies do not establish rules of conduct to last forever,”
Am. Trucking Ass’ns, Inc. v. Atchison, T. & S.F.R. Co., 387 U.S. 397, 416 (1967), and “an
administrative agency is not disqualified from changing its mind,” Good Samaritan Hosp. v.
Shalala, 508 U.S. 402, 417 (1993) (citation and internal quotation marks omitted). However,
“[w]here an agency applies different standards to similarly situated entities and fails to support
this disparate treatment with a reasoned explanation and substantial evidence in the record, its
action is arbitrary and capricious and cannot be upheld.” Burlington N. & Santa Fe Ry. Co. v.
Surface Transp. Bd., 403 F.3d 771, 776-77 (D.C. Cir. 2005); see also Lilliputian Sys., Inc. v.
Pipeline & Hazardous Materials Safety Admin., 741 F.3d 1309, 1313 (D.C. Cir. 2014) (quoting
Burlington N.).
5
Ralph Waldo Emerson, Essays: Self-Reliance (1841).
21
2. The Coverage Standard
Before turning to Plaintiffs’ specific allegation that Defendants have applied the “health
outcomes” and “patient management” coverage standard inconsistently, the court provides
background on the standard itself. The Decision Memo directly addresses why Defendants
considered health outcomes and disease management in evaluating BA Scans. Responding to
“several” comments “that evidence of ‘improved health outcomes’ should not be a factor for a
coverage determination on [BA Scans],” AR at 4737, the Decision Memo summarized scientific
literature that explains why health outcomes and patient management are relevant to determining
whether a diagnostic test is reasonable and necessary.
The Decision Memo first quotes a 2003 publication about diagnostic testing in general,
identified as “Mol and colleagues (2003).” Id. at 4738. It quotes the authors as writing: “Whether
or not patients are better off from undergoing a diagnostic test will depend on how test information
is used to guide subsequent decisions on starting, stopping, or modifying treatment. Consequently,
the practical value of a diagnostic test can only be assessed by taking into account subsequent
health outcomes.” Id. (emphasis added).
The Decision Memo then cites a study authored by “Fryback and Thornbury (1991),”
which addresses how to assess the value of a diagnostic test. Id. In their paper entitled “The
Efficacy of Diagnostic Testing,” see AR at 840-46—which Lilly itself cited in its coverage request
letter, see id. at 55—the authors conceived of a “hierarchical model of efficiency” “as an
organizing structure for appraisal of the literature on efficacy of imaging,” id. at 840; see also id.
at 841 (“[A] more global analysis reveals diagnostic radiology to be part of a larger system whose
goal is to treat patients effectively and efficiently.”). The model, which consists of five “levels,”
illustrates the relationship between the efficacy of a diagnostic imaging test, on the one hand, and
22
health outcomes and patient management, on the other. The lowest level of the hierarchy—
Level 1—is a diagnostic imaging test that has achieved what the authors describe as “technical
efficiency,” that is, clarity in image quality. Id. One level up, at Level 2, is what the authors
describe as “Diagnostic Accuracy Efficacy,” that is, effectiveness in measuring the “performance
of the image[] for the purpose of making diagnoses[.]” Id. Only at higher levels of Fryback and
Thornbury’s model does a diagnostic imaging test attain “therapeutic” (Level 4) and “patient
outcome” (Level 5) efficacy. Id. at 843-44. At those levels, such a test “may influence a patient’s
diagnostic thinking and yet may have no impact on patient treatment” (Level 4) or may advance
the “ultimate goal of medical care . . . to improve, or return to normal, the health of the patient”
(Level 5). Id. Fryback and Thornbury’s article provides ample support for Defendants’ reliance
on health outcomes and patient management in assessing whether a test is “reasonable and
necessary.”6
And, so too, does other literature cited in the Decision Memo and included in the
Administrative Record. See, e.g., id. at 2914 (study by Laforce et al. which states that “to be
widely adopted a diagnostic test needs to have a significant impact on patient management and
outcomes”); id. at 2630-31 (study by Grundman et al. noting that a “remaining question is whether
clinical care that includes [BA Scans] will translate into better outcomes” and that “[a]dditional .
6
Countering Defendants’ reliance on Fryback and Thornbury, Plaintiffs write:
Although a diagnostic test may offer added special benefits when it leads to a health improvement,
Defendants offer no evidence that Congress intended that this was a mandatory requirement of
diagnostic coverage. Indeed, since the statutory coverage language was enacted in 1965, nearly
three decades before [Fryback and Thornbury] published their [study], it is an historical
impossibility that it was a basis for the statute.
Pls.’ Reply at 13-14. The court is not persuaded by this argument. Defendants need not offer evidence that Congress
mandated that CMS consider whether a “diagnostic test . . . leads to health improvement”; they need only demonstrate
that Congress did not foreclose such consideration by speaking directly to the issue, and that such consideration is
reasonable. And Defendants certainly need not present evidence that Congress considered Fryback and Thornbury’s
article when drafting the Medicare Act.
23
. . studies would be required . . . to explicitly quantify the relationship between [BA Scans] and
patient outcomes”); see also id. at 4781, 4785 (inclusion of Grundman et al. and Laforce et al. in
the Decision Memo’s bibliography). In sum, the Decision Memo and Administrative Record
explain why Defendants considered health outcomes and patient management, and leave little
doubt about either factors’ relevancy.
3. “Similarly-Situated” Items and Services
The court now turns to whether, as Plaintiffs have argued, Defendants’ denial of Medicare
coverage to BA Scans was inconsistent with their approval of coverage for similarly-situated items
and services.
a. FDG PET
Plaintiffs focus much of their argument on Defendants’ decision to cover a particular type
of PET scan known as FDG PET. That diagnostic test uses a different radiopharmaceutical—
fluorodeoxyglucose—to differentiate between patients suffering from frontotemporal dementia
and Alzheimer’s. Like BA Scans, FDG PET’s primary utility is assisting in the differentiation
between diseases that might cause dementia; it does not affirmatively diagnose any particular
disease. Although CMS did not at first grant Medicare coverage for FDG PET, see generally
2000 FDG PET Decision, it reversed course in 2004. The decision memo accompanying the NCD
that approved FDG PET stated:
The evidence is adequate to conclude that [FDG PET scans are] reasonable and
necessary in patients with documented cognitive decline of at least six months and
a recently established diagnosis of dementia who meet diagnostic criteria for both
Alzheimer’s . . . and frontotemporal dementia . . . who have been evaluated for
specific alternate neurodegenerative diseases or causative factors, and for whom
the cause of the clinical symptoms remains uncertain.[7]
7
CMS, Decision Memo for Positron Emission Tomography (FDG) and Other Neuroimaging Devices for Suspected
Dementia (CAG-00088R) (Sept. 15, 2004) available at https://www.cms.gov/medicare-coverage-
database/details/nca-decision-
24
Critically, for present purposes, CMS approved FDG PET even though “[n]o published studies
evaluated whether FDG PET can alter clinical decision-making and improve patient outcomes.”
Id. at 35. “Nonetheless,” CMS explained,
an expert consensus articulated in a report published by the [Alzheimer’s
Association] and confirmed by a panel convened by [the National Institute of
Aging] suggests that the addition of FDG PET may be warranted . . . in patients
with documented cognitive decline of at least six months who meet diagnostic
criteria for both [Alzheimer’s] and [frontotemporal dementia] and for whom the
subtype of neurodegenerative disease remains uncertain.
Id. On that basis, CMS found the “evidence adequate to conclude that FDG PET improves net
health outcomes” in patients who meet the above-referenced criteria. Id.Plaintiffs contend that “[i]t is arbitrary and capricious for Defendants to refuse to ever cover
[BA Scans] on the ground that they cannot ‘improve health outcomes’ or alter ‘disease
management’ when Defendants have not held FDG PET scans to this standard,” Pls.’ Mot. at 32,
adding that “Defendants fail to provide any plausible justification for the disparity,” Pls.’ Reply at
24. They argue, essentially, that Defendants’ 2004 FDG PET Decision set a precedent that
Defendants, without explanation, failed to follow when making the decision at issue here.
Defendants disagree. They point to the Decision Memo itself, in which CMS explained
why it approved FDG PET but not BA Scans: “FDG PET is a fundamentally different—not a
similar—technology [to a BA Scan]. FDG PET measures the physiological process of metabolism,
while [a BA Scan] looks at the anatomical burden of amyloid plaques.” AR at 4739. Emphasizing
the significance of this distinction, Defendants argue in their Motion that while a BA Scan “uses a
radiopharmaceutical that preferentially binds to [beta-amyloid] plaques, and thus measures
amyloid density in the brain,” “FDG PET uses a different radiopharmaceutical . . . that is
memo.aspx?NCAId=104&NcaName=Positron+Emission+Tomography+(FDG)+and+Other+Neuroimaging+Device
s+for+Suspected+Dementia+(1st+Recon)&bc=AiAAAAAAEAAA& [hereinafter 2004 FDG PET Decision], at 3.
25
preferentially metabolized by certain tissues.” Defs.’ Mot. at 36. Defendants argue that this
“difference is enough to distinguish the two tests.” Id. The court disagrees.
The similarities between FDG PET and BA Scans are manifest. Both are diagnostic tests
that involve the use of a PET scan and a radiopharmaceutical tracer. Both are indicated for use on
overlapping patient populations exhibiting symptoms of cognitive impairment. And, although
neither test can affirmatively diagnose a disease, both have diagnostic value as a tool for
differentially diagnosing patients who exhibit symptoms associated with several different diseases.
Compare 2004 FDG PET Decision at 3 (approved for patients with cognitive decline who may
have Alzheimer’s or frontotemporal dementia but for whom “the cause of the clinical symptoms
remains uncertain”) with AR at 29 (FDA approval of BA Scans for “adult patients with cognitive
impairment who are being evaluated for Alzheimer’s . . . and other causes of cognitive decline”);
see also id. at 4740 (stating that it “may or may not be true—the evidence is not clear” “that [a BA
Scan] is a better tool than FDG PET for differentiating [frontotemporal dementia] from
[Alzheimer’s]”).
And both are the same in another critical respect—neither test has demonstrated
effectiveness in terms of improving health outcomes or altering patient management. Defendants
approved FDG PET even when “[n]o published studies evaluated whether FDG PET can alter
clinical decision-making and improve patient outcomes.” 2004 FDG PET Decision at 35. Yet,
the absence of such studies with respect to BA Scans led Defendants to reach the opposite
conclusion—denying coverage for the procedure.
The court cannot reconcile these different conclusions on the present record and finds that
Defendants’ failure to provide a cogent explanation for the disparate outcomes was arbitrary and
capricious, in violation of the APA. See, e.g., Cty. Of Los Angeles, 192 F.3d at 1022-23 (finding
26
that the Secretary’s “sole justification” for why certain data was “suitable for one significant
calculation but unreliable for another” was insufficient and concluding that the “Secretary’s
proffered distinction is thus not reasonable”). Several commenters challenged Defendants’
disparate treatment of FDG PET and BA Scans. See AR at 4739. Yet, the only explanation that
Defendants offered to distinguish the two was that “FDG PET is a fundamentally different—not a
similar—technology.” Id. That may be so, but that does not explain why Defendants agreed to
cover FDG PET despite the complete lack of published studies evaluating the procedure’s effect
on health outcomes and patient management, 2004 FDG PET Decision at 35, when similarly
insufficient clinical evidence was the basis for Defendants’ denial of coverage for BA Scans. The
“post hoc” explanation Defendants offer in their reply brief—that an “expert consensus” existed
as to the benefits of FDG PET that was absent with regard to BA Scans, see Defs.’ Reply at 15-
16—was not offered in the Decision Memo itself and, therefore, cannot cure its deficiency. See
Citizens to Pres. Overton Park, Inc. v. Volpe, 401 U.S. 402, 419 (1971), abrogated on other
grounds by Califano v. Sanders, 430 U.S. 99 (1977) (“‘[P]ost hoc’ rationalizations . . . have
traditionally been found to be an inadequate basis for review.”).
Although the court concludes that Defendants’ failure to adequately distinguish between
its treatment of FDG PET and BA Scans was arbitrary and capricious, it disagrees with Plaintiffs’
assertion that “[h]aving conceded coverage for FDG-PET patients, [Defendants] must provide it
for appropriate [BA Scan] patients.” Pls.’ Mot. at 33. Instead, the court is compelled to remand
the Decision Memo to CMS so that the agency can evaluate in the first instance whether its
coverage decisions can be reconciled. See PPG Indus., Inc. v. United States, 52 F.3d 363, 365
(D.C. Cir. 1995) (“Under settled principles of administrative law, when a court reviewing agency
action determines that an agency made an error of law, the court’s inquiry is at an end: the case
27
must be remanded to the agency for further action consistent with the corrected legal standards.”);
see also Cty. Of Los Angeles, 192 F.3d at 1023. Remand, without vacating the NCD, is the
appropriate relief here. See Allied-Signal, Inc. v. United States Nuclear Regulatory Comm’n, 988
F.2d 146, 150-51 (D.C. Cir. 1993) (“The decision to vacate depends on the ‘the seriousness of the
order’s deficiencies (and thus the extent of doubt whether the agency chose correctly) and the
disruptive consequences of an interim change that may itself be changed.’” (citation omitted)).
b. Alzheimer’s Treatment and Diagnostic Tests
The above conclusion, however, does not end the matter. Plaintiffs advance the additional
argument that Defendants’ decision not to extend Medicare coverage to BA Scans “cannot be
reconciled with their consistent and longstanding practice of providing coverage for a host of
Alzheimer’s-related treatment services.” Pls.’ Mot. at 33. Specifically, Plaintiffs assert that
“Defendants consistently reimburse Alzheimer’s patients for the costs of visits to physicians’
offices” and “for the prescription of memantine and cholinesterase inhibitors, two drugs used to
mitigate Alzheimer’s symptoms,” id. at 34, as well as “cover counseling services for Alzheimer’s
patients and their families [and] mental health services for appropriate Alzheimer’s patients
struggling to cope with the implications of their disease,” id. Because none of these items and
services “prevent, stabilize or reverse” a condition, but are nonetheless covered under Medicare,
id., Plaintiffs posit, they “should not, under Defendants’ theory, be covered any more than [BA
Scans],” id. at 33.
Plaintiffs make a similar argument as to Defendants’ coverage of diagnostic services for
conditions that, like Alzheimer’s, are incurable. They assert that Defendants “have covered and
continue to cover services that aid in the diagnosis of these conditions, even though those services
cannot ‘improve health outcomes’ or alter ‘disease management.’” Pls.’ Mot. at 35. Plaintiffs cite
28
as examples “MRI scans used to diagnose multiple sclerosis,” id. at 35-36; “diagnostic
electromyography/nerve conduction studies for patients with symptoms of amyotrophic lateral
sclerosis,” also known as ALS, id. at 36; “electromyography/nerve conduction testing to aid in the
diagnosis of patients experiencing symptoms of muscular dystrophy, a progressive
musculoskeletal disease,” id.; “services that aid in the diagnosis of Cruetzfeldt-Jakob disease, a
rapidly progressive and uniformly fatal neurological condition,” id.; and “genetic testing used in
the diagnosis of Huntington’s disease, a neurodegenerative genetic disorder that leads to cognitive
decline and behavioral problems,” id. at 37. Plaintiffs contend that these diagnostic services, like
the Alzheimer’s-related items and services referenced above, do not “prevent, stabilize or reverse”
any condition, and thus, it “is arbitrary and capricious for Defendants to cover the[m] on the basis
of their diagnostic value while refusing to apply the same standard to [BA Scans].” Id.
The court cannot, based on the record before it, find Defendants’ coverage of any of the
cited Alzheimer’s-related items or services or diagnostic tests for incurable conditions to be
inconsistent with the denial of coverage for BA Scans or the coverage standard applied thereto.
At the outset, the court observes that these items and services are fundamentally different from a
BA Scan. None uses PET technology. None is used to determine the cause of dementia symptoms.
None counts, as its primary recipients, patients who may have, but have not yet been diagnosed
with Alzheimer’s. It is thus more challenging for the court to evaluate the consistency of
Defendants’ decision with respect to BA Scans with Defendants’ decisions as to these items and
services than it is for the court to do so with regard to FDG PET.
Such evaluation is made even more challenging by the fact that the record before the court
does not allow it to compare BA Scans to the allegedly similarly-situated items and services in the
same manner that it permits the court to do so with FDG PET. The NCD for FDG PET is included
29
in the Administrative Record, see AR at 4-8; the accompanying decision memo is cited in the
parties’ pleadings, see, e.g., Defs.’ Mot. at 13; Pls.’ Reply at 22; and the Decision Memo itself
specifically addressed FDG PET, even directly comparing it to BA Scans, see AR 4739-40.
Contrastingly, the other items and services Plaintiffs raise are not discussed in the Decision Memo.
Further, Plaintiffs have not put before the court the actual decision memo or the NCD, to the extent
such exist, for any of these items or services. Nor have Plaintiffs cited in their briefs the decision
memos for any of these items or services, and cite the NCD of only one—MRI scans used to
diagnose multiple sclerosis. See Pls.’ Mot. at 36, n.18. As the party asserting a violation of the
APA, it was Plaintiffs’ burden to put forward a record from which the court could fully evaluate
their arguments. See San Luis Obispo Mothers for Peace v. U.S. Nuclear Regulatory Comm’n,
789 F.2d 26, 37 (D.C. Cir. 1986) (“[T]he party challenging an agency’s action as arbitrary and
capricious bears the burden of proof.”). Given that “the court’s review is limited to the
administrative record,” Fund for Animals, 903 F. Supp. at 105 (citing Camp, 411 U.S. at 142), it
is not able, as Plaintiffs urge, to determine that Defendants “treat[ed] similarly situated entities
differently,” Lilliputian Sys., 741 F.3d at 1313, by granting Medicare coverage for the cited items
and services while denying coverage to BA Scans.
Finally, Plaintiffs argue in their reply brief—for the first time—that Defendants’ coverage
of BA Scans is inconsistent with their coverage of PET scans for the diagnosis of conditions other
than Alzheimer’s, such as lung cancer, extranodal metastasis, and neck tumors. See Pls.’ Reply at
27-28 (quoting 2000 FDG PET Decision at 11, 21, 23). Plaintiffs did not, however, raise that
argument in their initial Motion. The court will not break with the “well-settled prudential doctrine
that courts generally will not entertain new arguments first raised in a reply,” Aleutian Pribilof
30
Islands Ass’n, Inc. v. Kempthorne, 537 F.Supp.2d 1, 12 n. 5 (D.D.C. 2008) (citing Herbert v. Nat’l
Acad. of Scis., 974 F.2d 192, 196 (D.C. Cir. 1992)), by entertaining that argument here.
* * *
In summary, applying the Chevron framework, the court concludes that the plain text of
the Medicare Act did not preclude Defendants’ consideration of health outcomes and disease
management in determining whether to afford coverage for BA Scans as “reasonable and necessary
for diagnosis.” Yet the court also concludes that the NCD did not adequately explain why that
standard resulted in different coverage outcomes for BA Scans and FDG PET, even though the
two tests are similarly situated and both lacked studies demonstrating an effect on health outcomes
or disease management. The agency’s failure to offer a sufficient rationale for treating these two
tests differently was arbitrary and capricious under the APA.
C. The Decision Memo’s Internal Inconsistency
Plaintiffs next argue that the Decision Memo is arbitrary and capricious because it is
“internally inconsistent.” That is, “Defendants’ refusal to cover [BA S]cans on the ground that
such scans cannot ‘improve health outcomes’ or alter ‘disease management’ is irreconcilable with
their own concessions in the administrative record,” specifically the concession “that a negative
[BA Scan] ‘virtually excludes’ an Alzheimer’s diagnosis.” Pls.’ Mot. at 40-41 (citing AR at 4758,
4766). “Once this is conceded,” Plaintiffs contend, “it is inescapable that such diagnostic
information can improve patient health and impact disease management.” Id. at 41.
Plaintiffs focus on “four harms,” Pls.’ Reply at 35, which they argue will be avoided if a
patient receives a negative scan: “depriv[ation] of the opportunity to be diagnosed with, and
treated for, the actual cause of his or her symptoms[], . . . an especially severe injury on the subset
of those patients whose symptoms derive from a potentially reversible condition,” Pls.’ Mot. at 41;
31
for patients suffering from frontotemporal dementia, the “potentially toxic” effects of “memantine
and cholinesterase inhibitors, the two drugs typically prescribed to Alzheimer’s patients to mitigate
their symptoms,” id. at 42; for all patients, the “harmful side effects” of “memantine and
cholinesterase inhibitors,” including “fatigue, dizziness, headache, sleepiness, constipation,
vomiting, back pain, coughing, shortness of breath, and hallucination,” id. at 42-43; and the
subjecting of patients to “unnecessary medical procedures and other interventions, such as MRI
and CT scans,” for which “[t]here is inherent harm,” id. at 43.
The court does not question the legitimacy of the harms of misdiagnosis that Plaintiffs have
identified. Defendants—aside from remarking that the “side effects of [memantine and
cholinesterase inhibitors] are ‘typically mild and transient,’” Defs.’ Mot. at 41 (quoting AR at
3396)—do not either, see id. at 39-41. Indeed, Defendants, in the Decision Memo, examined each
one of the harms Plaintiffs have set forth here. They determined, however, that then-existing
evidence regarding BA Scans did not support a conclusion that the test would improve health
outcomes or disease management, see, e.g., AR at 4769-70, and thus, would not allow patients to
avoid the harms Plaintiffs have identified, see id. at 4760 (addressing the ability for a negative BA
Scan to “hasten the work up for other, potentially treatable diseases” and concluding that the
“evidence [in favor of] such arguments . . . is of limited persuasiveness, based almost entirely on
clinical vignettes and case studies, which carry unmitigated risk of methodological bias and
confounding, rather than clinical trials”); id. at 4759 (acknowledging that “in differentiating
frontotemporal dementia and [Alzheimer’s] . . . potential for harm appears to exist,” but noting
that “CMS covered FDG PET in 2004 for use specifically in the differential of [frontotemporal
dementia] and [Alzheimer’s]”); id. (discussing several studies involving memantine and
cholinesterase inhibitors and concluding that “[i]n these particular cases, no additional harm
32
appears to result from misdiagnosis that places patients on such dementia medications”); id. at
4760 (finding that a negative BA Scan would not avoid other interventions because, absent “a
convincing clinical picture, work up to exclude other, diagnosable and potentially treatable
diseases should proceed anyway (as it would if an amyloid scan were negative)”). The court
cannot and will not substitute its judgment for that of Defendants by invalidating those
conclusions.
Further, the court is not moved by the statements in the Decision Memo that Plaintiffs
characterize as “concessions” “irreconcilable” with Defendants’ denial of coverage for BA Scans.
Pls.’ Mot. at 40-41. Plaintiffs, on several occasions, point to the following statement: “In this
case, a negative scan virtually excludes [Alzheimer’s].” AR at 4758; see also Pls.’ Mot. at 3 (“As
Defendants concede, a negative [BA S]can ‘virtually excludes’ an Alzheimer’s diagnosis.”
(quoting AR at 4758) (emphasis added)); Pls.’ Mot. at 6 (“Once Defendants concede that a
negative scan ‘virtually excludes’ Alzheimer’s, AR 4758, . . . they cannot avoid the inevitable
logical consequences of that admission[.]”); Pls.’ Mot. at 22 (“Once Defendants concede that a
negative [BA S]can ‘virtually excludes’ Alzheimer’s, AR 4758 (id.), it follows unavoidably that
this diagnostic information can improve the health or alter the disease management [of certain
patients.]”). In quoting this statement, Plaintiffs ignore its context. The “case” that the statement
refers to is a hypothetical, used in the Decision Memo “for the sake of illustration . . . although [it]
has never been demonstrated.” AR at 4757. The statement, which is explicitly based on unproven
assumptions, cannot fairly be read as a “concession” or “admission” that negative BA Scans impact
outcomes and disease management. Nor can the other two statements Defendants cite to support
that proposition, as both are similarly qualified by conditional language. The statement that “a
negative [BA S]can could virtually exclude [Alzheimer’s] in many patients,” AR at 4758
33
(emphasis added), and another noting “the potential power of a negative scan to virtually exclude
significant brain beta amyloid deposition,” do not eclipse the Decision Memo’s detailed discussion
of the relevant evidence and its thoroughly explained conclusions.
D. Inconsistency with a CMS Regulation
Plaintiffs last argue that Defendants’ reliance on CMS regulation 42 C.F.R. § 410.32 to
supports its consideration of health outcomes and disease management is arbitrary and capricious.
Section 410.32 reads, in relevant part, as follows:
All diagnostic x-ray tests, diagnostic laboratory tests, and other diagnostic tests
must be ordered by the physician who is treating the beneficiary, that is, the
physician who furnishes a consultation or treats a beneficiary for a specific medical
problem and who uses the results in the management of the beneficiary’s specific
medical problem. Tests not ordered by the physician who is treating the beneficiary
are not reasonable and necessary.
42 C.F.R. § 410.32(a). The Decision Memo quotes a portion of Section 410.32 and states that the
“expectation that a medical test inform physician management . . . is . . . consistent with [the
regulation].” AR at 4752 (emphasis added). Plaintiffs contend that “Section 410.32(a) simply
identifies the type of health care provider who is authorized to order a diagnostic test; it does not
require the test to ‘improve health outcomes’ or alter ‘disease management.’” Pls.’ Mot. at 38; see
also Pls.’ Reply at 31 (“In any event, the regulation does not require diagnostic services to improve
health outcomes or alter disease management, as Defendants contend.”). They claim that
“Defendants misinterpret their own regulation, once again pressing an interpretation that is
inconsistent with the plain text before them.” Pls.’ Mot. at 38. The court disagrees.
“[Courts] must give substantial deference to an agency’s interpretation of its own
regulations.” Thomas Jefferson Univ. v. Shalala, 512 U.S. 504, 512 (1994). Accordingly, such an
interpretation merits “controlling weight unless it is plainly erroneous or inconsistent with the
regulation.” Id. (internal quotation and citation omitted); see also Wyoming Outdoor Council v.
34
U.S. Forest Serv., 165 F.3d 43, 52 (D.C. Cir. 1999) (“So long as an agency’s interpretation of
ambiguous regulatory language is reasonable, it should be given effect.”). As the court has
concluded above, Defendants’ consideration of health outcomes and disease management was
permissible under the Medicare Act itself. Therefore, Defendants need not have relied on a
regulation such a Section 410.32(a) to justify their consideration of those factors.
In any event, Defendants’ interpretation of Section 410.32(a) is not unreasonable, plainly
erroneous, or inconsistent with the statute. A regulation that limits Medicare coverage of
diagnostic tests to those ordered by a “treating” physician “who uses the results in the management
of the beneficiary’s specific medical problem,” 42 C.F.R. § 410.32(a), can be reasonably read as
allowing CMS to consider health outcomes and patient management in deciding whether to cover
a diagnostic test.
IV. CONCLUSION
For the foregoing reasons, the court grants in part and denies in part Plaintiffs’ and
Defendants’ Cross-Motions for Summary Judgment. This matter is remanded to the agency for
further proceedings consistent with this Memorandum Opinion. A separate Order accompanies
this Memorandum Opinion.
Dated: July 19, 2016 Amit P. Mehta
United States District Judge
35