United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued February 16, 2018 Decided May 25, 2018
No. 17-1043
ERIC FRIEDMAN,
PETITIONER
v.
FEDERAL AVIATION ADMINISTRATION,
RESPONDENT
On Petition for Review of an Order
of the Federal Aviation Administration
Z.W. Julius Chen argued the cause for petitioner. With him
on the briefs was Gregory S. Walden. Pratik A. Shah entered
an appearance.
Peter G. Dickos and Allison Jones Rushing were on the
brief for amicus curiae American Diabetes Association in
support of petitioner.
Casey E. Gardner, Attorney, Federal Aviation
Administration, argued the cause and filed the brief for
respondent.
Before: HENDERSON and TATEL, Circuit Judges, and
WILLIAMS, Senior Circuit Judge.
2
Opinion for the Court filed by Circuit Judge TATEL.
TATEL, Circuit Judge: For the second time, Eric Friedman,
a type-one diabetic and aspiring commercial pilot, challenges
the Federal Aviation Administration’s refusal to grant him a
medical certificate required for commercial flight. Because
Friedman uses insulin to manage his blood sugar, the FAA
required him to submit data from a relatively new method of
blood-glucose testing known as continuous glucose monitoring
(CGM), which Friedman declined to provide. In a prior
decision, our court held that the FAA “ha[d] not borne its
burden of justification” for requiring CGM data. Friedman v.
FAA (Friedman I), 841 F.3d 537, 544 (D.C. Cir. 2016). On
remand, the FAA explained that it needed the data because
CGM is able to detect hypoglycemic episodes often missed by
more traditional monitoring, and it supported that explanation
with medical studies in the administrative record. Because that
explanation satisfies our remand order, we deny the petition for
review.
I.
The opinion in Friedman I describes the background of
this case. Friedman I, 841 F.3d at 539–41. Here, we provide a
brief summary and then describe the FAA’s actions on remand.
Congress has directed the FAA to promulgate regulations
that “promote safe flight of civil aircraft,” 49 U.S.C.
§ 44701(a), and to discharge this obligation “in a way that best
tends to reduce or eliminate . . . accidents in air transportation,”
id. § 44701(c). Specifically, Congress has instructed the FAA
to ensure that pilots are “physically able to perform the duties
related to[] [their] position.” Id. § 44703(a). To implement this
directive, the FAA requires pilots to hold medical certificates
of one of three classes: pilots who fly commercial airliners
3
must have first-class medical certificates, 14 C.F.R.
§ 61.23(a)(1), pilots who otherwise fly commercially must
have second-class certificates, id. § 61.23(a)(2), and pilots who
fly privately must have third-class certificates, id.
§ 61.23(a)(3). In keeping with Congress’s instruction that the
FAA consider “the duty of [a commercial] air carrier to provide
service with the highest possible degree of safety in the public
interest,” 49 U.S.C. § 44701(d)(1), the Federal Air Surgeon, to
whom the FAA Administrator has delegated his medical
certification authority, 14 C.F.R. § 67.407, has adopted more
rigorous medical standards for pilots who fly commercially
(first- and second-class) than for those who fly privately (third-
class), see id. §§ 61.23(a), 67.101–67.315.
Under FAA regulations, some medical conditions,
including diabetes treated with insulin, are presumptively
disqualifying for any class of medical certificate. Id.
§§ 67.113(a) (first-class), 67.213(a) (second-class), 67.313(a)
(third-class). The primary danger of flying with insulin-treated
diabetes, the FAA has explained, is the risk of hypoglycemia,
a state of abnormally low blood-glucose levels, which is
common among diabetics, especially those who use insulin,
and which can impair a range of cognitive functions, including
reaction time, memory, problem solving, and spatial reasoning.
Special Issuance of Third-Class Airman Medical Certificates
to Insulin-Treated Diabetic Airman Applicants, 61 Fed. Reg.
59,282, 59,282 (Nov. 21, 1996) (explaining that the FAA’s
disqualification policy was based in large part on concerns
about hypoglycemia).
Notwithstanding this presumptive disqualification, FAA
regulations provide that, “[a]t the discretion of the Federal Air
Surgeon,” the agency may grant a “Special Issuance” medical
certificate to any applicant who shows an ability to safely
perform all authorized duties “to the satisfaction of the . . . Air
4
Surgeon.” 14 C.F.R. § 67.401(a). Pursuant to this authority, the
Air Surgeon has awarded special-issuance third-class medical
certificates to some insulin-treated diabetics—a practice that
has led to “no medically related accident, incident, or inflight
incapacitation, from any cause.” Friedman I, 841 F.3d at 540.
Although the Air Surgeon has never granted a special-
issuance first- or second-class medical certificate to an insulin-
treated diabetic, the FAA announced in 2013 that it “hope[d]
that in the future [it] may be able to . . . safely certificate . . . a
subset of [insulin-treated diabetics]” at the first- and second-
class level, and asked the American Diabetes Association
(ADA) to form an expert panel “to investigate the possibility
of identifying” such diabetics. Letter from Michael P. Huerta,
Administrator, FAA, to Robert Ratner, Chief Scientific &
Medical Officer, ADA (July 15, 2013), Joint Appendix (J.A)
416. In its report, the panel recommended a protocol for
identifying low-risk insulin-treated diabetics and concluded
that for such pilots “there is no incremental risk of permitting
[them] to fly.” ADA, Expert Panel Recommendations for Pilots
with Insulin-Treated Diabetes 6 (2015), J.A. 368.
The FAA declined to adopt the recommended protocol,
explaining that it found the panel’s recommendations lacking
in “evidentiary support” and neither “operationally feasible”
nor sensitive to commercial pilots’ “real-world challenges.”
Letter from James R. Fraser, Federal Air Surgeon, FAA, to
Robert Ratner, Chief Scientific & Medical Officer, ADA (Apr.
9, 2015), J.A. 447. The FAA also believed that the panel’s
protocol was too lax, as it would have allowed certification of
pilots with blood-glucose levels “outside the normal glycemic
range at least 20 percent of the time.” Id. Despite declining to
adopt this (or any) protocol for special issuance of first- or
second-class medical certificates to insulin-treated diabetics,
the FAA emphasized that it did not have a “blanket ban,”
5
stating it was open to issuing certificates on an ad hoc basis.
Letter from James R. Fraser, Federal Air Surgeon, FAA, to Eric
Friedman (Mar. 9, 2015), J.A. 361.
In April 2015, Petitioner, Eric Friedman, an insulin-treated
diabetic who currently holds a special-issuance third-class
medical certificate, applied to upgrade his certificate to first
class. After receiving Friedman’s application, the FAA
requested blood-glucose testing results either from finger-stick
testing, which involves pricking a finger and using a
glucometer to test extracted blood, or “if applicable,” from
CGM, a newer and less commonly used method of blood-
glucose testing that measures glucose levels continuously using
a subcutaneously inserted sensor and a small, wearable device.
Letter from James R. De Voll, Medical Appeals Manager,
FAA, to Eric Friedman (Apr. 30, 2015), J.A. 73.
In response, Friedman provided only finger-stick data,
explaining that he does not use CGM. Over the next several
months, the FAA again suggested—and later demanded—that
Friedman provide CGM data for a minimum of 90 days,
eventually warning that failure to do so would result in denial
of his application. Friedman reiterated that he does not use a
CGM device “because neither of [his] treating physicians . . .
has found . . . [the device] is clinically indicated or medically
necessary.” Letter from Eric Friedman to James R. De Voll,
Medical Appeals Manager, FAA (Nov. 6, 2015), J.A. 42.
Friedman also pointed out that the ADA expert panel had
recommended against CGM’s use due to its tendency to
“record transient postprandial spikes in glucose levels,” which
are “of no clinical significance.” ADA, Expert Panel
Recommendations for Pilots with Insulin-Treated Diabetes 7
n.ii, J.A. 369. The panel itself wrote in support of Friedman’s
application, stating that although CGM is “useful in identifying
trends” in blood-glucose levels, it is less accurate than finger-
6
stick monitoring and thus neither necessary nor appropriate for
medical certification decisions. Letter from Daniel Lorber, on
behalf of ADA Expert Panel, to James Fraser, Federal Air
Surgeon, FAA (Nov. 6, 2015), J.A. 65–66. Unconvinced, the
FAA informed Friedman that it was “unable to proceed with”
his application “until [it] receive[d] the information previously
requested.” Letter from James R. De Voll, Medical Appeals
Manager, FAA, to Eric Friedman (Dec. 1, 2015), J.A. 53.
Friedman then filed a petition for review in this court, in
which he claimed that the FAA’s denial violated the
Administrative Procedure Act (APA), 5 U.S.C. § 706(2)(A),
and the Pilot’s Bill of Rights, Pub. L. No. 112-153, 126 Stat.
1159 (2012) (codified at 49 U.S.C. § 44703 note). Our court
granted the petition, explaining that “[t]he FAA’s letters
communicating its demand for CGM data to Friedman . . .
fail[ed] to articulate any rationale for consideration of the
additional information.” Friedman I, 841 F.3d at 544.
Although, in its brief before the court, the FAA relied on the
expert panel’s acknowledgment of CGM’s value, we found this
argument unconvincing given the panel’s unequivocal
opposition to the use of CGM for medical-certification
decisions. Id. The agency, we emphasized, was unable to
“identify any FAA statements that could be construed as
explaining its denial of Friedman’s application.” Id. at 545.
Because the FAA had “not borne its burden of justification,”
id. at 544, we remanded for it to “offer reasons for its denial of
Friedman’s application,” id. at 545.
Two months later, on January 27, 2017, the FAA sent
Friedman a letter “in response to the D.C. Circuit’s decision.”
Letter from Michael A. Berry, Federal Air Surgeon, FAA, to
Eric Friedman (Jan. 27, 2017), J.A. 439 (“January 27 Letter”).
The FAA began by explaining that although it “ha[s] not
allowed special issuance first- or second-class medical
7
certification for [insulin-treated diabetics],” its “goal” was to
change that, and “[t]o that end,” it was “working to develop an
evidence-based framework.” Id. Turning to Friedman’s
application, the FAA acknowledged his “assertion, supported
by [his] treating physicians, that [he] demonstrate[s] excellent
glycemic management as evidenced by self-monitoring with a
traditional ‘finger-stick’ recording glucometer.” Id. That said,
the FAA went on to explain why it nonetheless demanded that
Friedman submit CGM data in order to obtain a medical
certificate. Because the adequacy of that explanation is central
to this appeal, we quote it in full:
Current studies of the safety and efficacy of CGM
devices . . . show . . . that . . . hypoglycemia remains
common and frequently goes unrecognized by
traditional finger-stick testing. Thus, self-monitoring
using finger-stick testing alone is not an adequate
mitigation strategy for operations requiring a first- or
second-class medical certificate.
Because it is impossible to know the true extent
of glycemic variability through self-monitoring with
traditional “finger-stick” tests, we have determined
that a fixed period (90 days) of CGM is necessary in
order to consider your eligibility for an Authorization
for special issuance of a first-class medical certificate.
The CGM data that we have requested will be
reviewed for evidence of glycemic control and
stability, as well as to evaluate the potential use of
CGM as risk mitigation during operations requiring a
first- or second-class medical certificate.
Id., J.A. 439–40. Since Friedman had “not provided the
requested CGM report and related data,” the FAA concluded,
8
it was “unable to consider [him] for a special issuance first-
class medical certificate.” Id., J.A. 440.
In this petition for review, Friedman argues that the FAA
“still fails to articulate a defensible rationale for denying [his]
application.” Pet’r’s Br. 2. Again arguing that the FAA has
violated both the APA and the Pilot’s Bill of Rights, Friedman
urges us to “remand for the FAA to (finally) grant [him] a first-
class medical certificate.” Id. at 3. The ADA has filed an
amicus brief in support of Friedman.
II.
At the outset, we emphasize the narrowness of the issue
before us. Because the FAA sent its January 27 letter in
response to this court’s remand, the only question we must
answer is whether it has “fill[ed] the analytical gap identified
in [our] opinion.” Heartland Regional Medical Center v.
Leavitt, 415 F.3d 24, 29 (D.C. Cir. 2005). Here, that gap is the
FAA’s failure to “offer reasons for its denial of Friedman’s
application.” Friedman I, 841 F.3d at 545. In determining
whether the FAA has satisfactorily provided such reasons, we
are guided by well-settled principles of administrative law. We
look to see whether the agency has “examine[d] the relevant
data and articulate[d] a satisfactory explanation for its action,”
which may not neglect any “important aspect of the problem,”
“run[] counter to the evidence,” or be “so implausible that it
could not be ascribed to a difference in view or the product of
agency expertise.” Motor Vehicle Manufacturers Ass’n of the
United States, Inc. v. State Farm Mutual Automobile Insurance
Co., 463 U.S. 29, 43 (1983). Although we will not, as in
Friedman I, “supply a reasoned basis for the agency’s action
that the agency itself has not given,” id. at 43 (citing SEC v.
Chenery Corp., 332 U.S. 194, 196 (1947)), we may—central to
this petition—“‘uphold a decision of less than ideal clarity if
the agency’s path may reasonably be discerned,’” id. (quoting
9
Bowman Transportation, Inc. v. Arkansas-Best Freight System,
Inc., 419 U.S. 281, 286 (1974)).
Friedman insists that the explanation the FAA provided in
its January 27 letter is “no better” than that considered in
Friedman I. Pet’r’s Br. 28. This is inaccurate. In Friedman I,
the FAA had simply demanded, without explanation, that
Friedman submit CGM data. Not until its brief before this court
did the FAA provide a reason, namely that the ADA panel had
acknowledged CGM’s value. But as we pointed out, the FAA
“overstate[d] the usefulness of this concession,” given the
panel’s conclusion that CGM was inappropriate for this
purpose. Friedman I, 841 F.3d at 544. The agency, in other
words, “[did] not identify any FAA statements that could be
construed as explaining its denial of Friedman’s application.”
Id. at 545.
By contrast, in its January 27 letter, the FAA provided its
own, unequivocal medical explanation for requiring CGM
data: that such data is needed to detect hypoglycemic episodes
that could well be missed by traditional finger-stick
monitoring. “Current studies of the safety and efficacy of CGM
devices,” the FAA explained, show that “hypoglycemia
remains common and frequently goes unrecognized by
traditional finger-stick testing.” January 27 Letter, J.A. 439–
40. As a result, “self-monitoring using finger-stick testing
alone is not an adequate mitigation strategy for operations
requiring a first- or second-class medical certificate.” Id., J.A.
440. The FAA thus concluded that “CGM is necessary in order
to consider [Friedman’s] eligibility for” a special-issuance
first-class medical certificate. Id.
This explanation, moreover, finds support in the
administrative record. In contrast to Friedman I, where there
was a “complete absence of a relevant administrative record to
10
review,” Friedman I, 841 F.3d at 545, the administrative record
now before us includes “[c]urrent studies” finding that because
finger-stick monitoring involves only static measurements, it
often fails to detect hypoglycemic episodes occurring between
tests, see, e.g., Bruce W. Bode et al., Glycemic Characteristics
in Continuously Monitored Patients with Type 1 and Type 2
Diabetes, 28 Diabetes Care 2361, 2361 (2005), J.A. 680
(“Regardless of how often fingerstick blood glucose
measurements are undertaken, discrete results offer only a
static picture at any point and do not provide a sense of the
number, intensity, and duration of glycemic excursions.”);
Alberto Maran et al., Continuous Subcutaneous Glucose
Monitoring in Diabetic Patients: A Multicenter Analysis, 25
Diabetes Care 347, 347–352 (2002), J.A. 686–91 (noting that
“glucose fluctuations during the day are often missed with”
finger-stick testing, id. at 347, J.A. 686, and concluding that a
CGM device, tested on 70 diabetic patients, was accurate and
well-tolerated); see also Timothy S. Bailey et al., American
Association of Clinical Endocrinologists and American
College of Endocrinology 2016 Outpatient Glucose
Monitoring Consensus Statement, 22 Endocrine Practice 231,
232 (2016), J.A. 645 (“[P]atients’ [finger-stick] monitoring
may be infrequent or intermittent, their reports may be
inaccurate, and overnight glucose levels are seldom measured.
Given these limitations, episodes of hypo- and hyperglycemia
may be missed . . . .”). CGM, on the other hand, “provides a
much more rigorous measure of glycemic variability and
glycemic exposure.” Lisa Gilliam & Irl B. Hirsch, Practical
Aspects of Real-Time Continuous Glucose Monitoring, 11
Diabetes Tech. & Therapeutics S-75, S-76 (2009), J.A. 721.
Friedman argues that CGM is an “inappropriate” source of
data because it records clinically insignificant spikes in blood
glucose and tends to lag behind and differ from finger-stick
values. Pet’r’s Br. 22; see also id. at 9–10. He fails, however,
11
to explain how these limitations, acknowledged by the FAA,
undermine the agency’s intended use of CGM data: to
“augment” the finger-stick data that Friedman has already
provided “to confirm his assertion that he can recognize and
appropriately respond to hypoglycemic episodes.” Resp’t’s Br.
31–32; see also January 27 Letter, J.A. 440 (“[S]elf-monitoring
using finger-stick testing alone is not an adequate mitigation
strategy.” (emphasis added)).
Next, pointing out that “none [of the studies] addresses the
medical evaluation of pilots with [insulin-treated diabetes],”
Pet’r’s Br. 22, Friedman maintains that the only word on the
subject is that of the expert panel, which, he insists, was surely
aware of the studies the FAA cites, yet concluded that “CGM
data is unnecessary,” id. The FAA, however, was not bound by
the expert panel’s assessment: “When specialists express
conflicting views, an agency must have discretion to rely on the
reasonable opinions of its own qualified experts.” Marsh v.
Oregon Natural Resources Council, 490 U.S. 360, 378 (1989).
Moreover, the FAA did not, as Friedman asserts, ignore the
expert panel. Quite to the contrary, it reviewed the panel’s
recommendations and explained that it “require[d] more
evidentiary support” than the panel had provided, January 27
Letter, J.A. 440, given that the panel “did not ground its
recommendations in any evidence-based medical literature,”
Resp’t’s Br. 39.
Finally, Friedman claims that there is no reason to believe
that the FAA relied on the studies it cites, as it “did not identify
any medical studies in its terse January 2017 Letter.” Pet’r’s
Reply Br. 14. From our perspective, it certainly would have
been helpful for the FAA to have elaborated on its generic
reference to “current studies.” But aided by the FAA’s brief,
which cites and explains the relevant studies, we can “discern”
the basis for its decision. BellSouth Corp. v. FCC, 162 F.3d
12
1215, 1224 (D.C. Cir. 1999); see id. (“‘[T]he agency is not
required to author an essay for the disposition of each
application. It suffices, in the usual case, that we can discern
the why and wherefore.’” (quoting ICBC Corp. v. FCC, 716
F.2d 926, 929 (D.C. Cir. 1983))). Specifically, in response to
our remand, the FAA “offer[ed] reasons for its denial of
Friedman’s application.” Friedman I, 841 F.3d at 545. Put
another way, it can no longer be said that the agency has “not
identif[ied] any FAA statements” explaining why it needs
CGM data. Id. (emphasis omitted). Equally important, the FAA
offered its explanation in accordance with APA standards: it
“examine[d] the relevant data”—the medical studies and
Friedman’s medical record—“and articulate[d] a satisfactory
explanation for its action”—that CGM data is necessary to
detect hypoglycemic events that finger-stick monitoring
misses. State Farm, 463 U.S. at 43. Friedman, moreover, has
given us no basis for believing that the FAA either neglected
an “important aspect of the problem,” or that its position is “so
implausible that it could not be ascribed to a difference in view
or the product of agency expertise.” Id.
We can easily dispose of Friedman’s claims under the
Pilot’s Bill of Rights. That statute provides, in part, that the
“[g]oals of the [FAA]’s medical certification process are . . . to
give medical standards greater meaning by ensuring the
information requested aligns with present-day medical
judgment and practices” and “to ensure that . . . the application
of such medical standards provides an appropriate and fair
evaluation of an individual’s qualifications.” Pilot’s Bill of
Rights, Pub. L. No. 112-153, § 4(b), 126 Stat. 1159, 1163
(codified at 49 U.S.C. § 44703 note). According to Friedman,
this provision requires the FAA to “align[]” its requests with
“present-day medical judgment” and “provide[] an appropriate
and fair evaluation.” See Pet’r’s Br. 5 (internal quotation marks
omitted); see also id. at 21–22. That, of course, is precisely
13
what the FAA—as required both by our remand and the APA—
has now done, and we think it quite obvious that the Pilot’s Bill
of Rights’ aspirational, goal-setting language creates no
additional obligations. Cf. Rothe Development, Inc. v. United
States Department of Defense, 836 F.3d 57, 65 (D.C. Cir. 2016)
(describing a statute’s “goal[s]” as “aspirational”).
III.
For the foregoing reasons, we deny the petition for review.
In doing so, we have taken the FAA at its word: that it is
“working to develop an evidence-based framework” to allow
special issuance of first- and second-class medical certificates
to certain insulin-treated diabetics, January 27 Letter, J.A. 439,
and that “should Mr. Friedman’s CGM data corroborate his
assertion” that he “does not pose a risk while he is in the
cockpit,” he will receive a certificate, Oral Argument 31:55–
33:15, 38:32–40.
So ordered.