Photo of Bexis

In our two prior posts, Loper Bright Likely Lays Lohr Low, and Could Loper Bright Finally Do in FDA’s Rickety Off-Label Speech Ban?, we focused on ways that Supreme Court’s holding in Loper Bright Enterprises v. Raimondo, 144 S. Ct. 2244 (2024), that courts were not to defer to administrative agency views in evaluating such agencies’ interpretations of their organic statutes, could be used by our clients in a positive fashion in the defense of prescription medical product liability.

Today we’re addressing the flip side.  How do we defend against the other side’s attempts to use Loper Bright for nefarious purposes?

Understanding the limitations of Loper Bright is a good start.

Keep in mind that Loper Bright solely deals with the process of statutory interpretation.  The Court held that administrative agencies have “no special competence on or solving statutory ambiguity” 144 S. Ct. at 2266.  “[T]he interpretation of the meaning of statutes, as applied to justiciable controversies, [i]s exclusively a judicial function.”  Id. at 2258 (citation and quotation marks omitted).  Loper Bright relied on the Administrative Procedure Act, enacted in 1946, which provided, that in reviewing agency decisions, a “court shall decide all relevant questions of law, [and] interpret constitutional and statutory provisions.”  5 U.S.C. § 706.

Loper Bright contrasted judicial review of statutory interpretation with judicial review of other aspects of agency decisionmaking:

[The APA] specifies that courts, not agencies, will decide “all relevant questions of law” . . . − even those involving ambiguous laws − and set aside any such action inconsistent with the law as they interpret it.  And it prescribes no deferential standard for courts to employ in answering those legal questions.  That omission is telling, because Section 706 does mandate that judicial review of agency policymaking and factfinding be deferential.  See §706(2)(A) (agency action to be set aside if “arbitrary, capricious, [or] an abuse of discretion”); §706(2)(E) (agency factfinding in formal proceedings to be set aside if “unsupported by substantial evidence”).

144 S. Ct. at 2261 (emphasis original).  The APA thus “makes clear that agency interpretations of statutes − like agency interpretations of the Constitution − are not entitled to deference.”  Id. (emphasis original).  Each statute “must have a single, best meaning” and “if it is not the best, it is not permissible.”  Id. at 2266.

Loper Bright further recognizes that a “statute’s meaning may well be that the agency is authorized to exercise a degree of discretion,” and that such statutes are common.  Id. at 2263.  In such cases:

When the best reading of a statute is that it delegates discretionary authority to an agency, the role of the reviewing court under the APA is . . . to independently interpret the statute and effectuate the will of Congress subject to constitutional limits.  The court fulfills that role by recognizing constitutional delegations, fixing the boundaries of the delegated authority, and ensuring the agency has engaged in reasoned decisionmaking within those boundaries.

Id. (citations and quotation marks omitted).

To us, these are the key points to emphasize in preventing the other side from claiming that, after Loper Bright, everything that the FDA does is fair game for de novo judicial re-examination, whether in mass tort litigation directly or in satellite APA litigation.  First, Loper Bright does not reduce the extant standard of review for either “agency policymaking” or “agency factfinding.”  These remain subject to “deferential” review standards:  either “arbitrary and capricious,” abuse of discretion” or “substantial evidence,” depending on the context.  Second, Loper Bright recognizes, explicitly, that statutes, such as the FDCA with the FDA, confer agency “discretion,” and that when they do, the role of a reviewing court is to “recognize” that type of “delegation,” determine its “boundaries,” and then apply the aforesaid “abuse of discretion” standard of “reasoned decisionmaking.”

“Boundaries.”  What are the boundaries of FDA discretionary power?

According to Loper Bright, those same boundaries determine what the courts must leave to the FDA’s discretionary functions.  “Discretionary function” is a good place to look, since there is a lot of law on the subject.  “Discretionary function” happens to be an exception to liability under the Federal Torts Claim Act.  28 U.S.C. §2680(a) (liability does “not apply” to “the exercise or performance or the failure to exercise or perform a discretionary function . . ., whether or not the discretion involved be abused”). 

Berkovitz v. United States, 486 U.S. 531 (1988), a pre-Vaccine Act vaccine product liability case, directly addressed the boundaries of FDA discretion, after accepting as true the complaint’s dubious (and pre-TwIqbal) allegations that the FDA violated its own regulations in allowing the release of the vaccine in question.  Id. at 540.  In general, the FDA lacks discretion “when a federal statute, regulation, or policy specifically prescribes a course of action” that must be followed.  Id. at 536.  If the administrative action “cannot appropriately be the product of judgment or choice, then there is no discretion.”  Id. (citation omitted).  Berkovitz distinguished a prior decision, United States v. Varig Airlines, 467 U.S. 797 (1984), that had found the Federal Aviation Administration immune from suit over its discretionary decision “certifying certain airplanes for operation,” because the agency’s certification “system” was a discretionary “policy determination as to how best to “accommodat[e] the goal of air transportation safety and the reality of finite agency resources.”.  486 U.S. at 537-58 (quoting Varig, 467 U.S. at 820).

In Berkovitz, though, plaintiffs alleged that the vaccine licensing process was statutorily and regulatorily contingent “only upon examination of the product and upon a determination that the product complies with the standards prescribed in the regulations,” and that the agency “issued a product license without first receiving data that the manufacturer must submit.”  Id. at 542 (citation omitted).  That was not a discretionary function:

[The agency] has no discretion to issue a license without first receiving the required test data; to do so would violate a specific statutory and regulatory directive.  Accordingly, to the extent that petitioners’ licensing claim is based on a decision . . . to issue a license without having received the required test data, the discretionary function exception imposes no bar.

Id. at 542-43.  But while an “agency has no discretion to deviate from [a] mandated procedure,” plaintiffs could not simply denounce an agency decision to approve a product as “incorrect.”  Id. at 544-45.  That type of allegation “hinges on whether the agency officials making that determination permissibly exercise policy choice.”  Id. at 545.  Here, the plaintiffs’ allegations were a moving target and it was unclear if they were even bringing such a claim, so the issue was remanded.  Id.

A second prong of Berkovitz involved the plaintiffs’ attack on the agency decision to release the particular vaccine lot, as opposed to its regulation of the vaccine generally.  The same dichotomy emerged.  The regulatory scheme at the time “allow[ed] the [FDA] to determine the appropriate manner in which to regulate the release of vaccine lots, rather than mandating certain kinds of agency action.”  Id. at 546.  That approach was “substantially similar” to what the Varig case had held was discretionary.  Id.

[T]he discretionary function exception bars any claims that challenge the [FDA’s] formulation of policy as to the appropriate way in which to regulate the release of vaccine lots.  In addition, if the policies and programs formulated by the [agency] allow room for implementing officials to make independent policy judgments, the discretionary function exception protects the acts taken by those officials in the exercise of this discretion.

Id. (citation omitted).  On the other hand, allegations that FDA personnel “knowingly approved the release of a lot that did not comply with safety standards . . . involved no policy discretion” and would not be barred as a discretionary function.  Id. at 547.

Taking Berkovitz as a starting point, courts have unanimously applied the discretionary function exception to preclude claims attacking FDA approval of drugs and medical devices.  Starting with Bexis’ favorite MDL, the plaintiffs in the Bone Screw litigation tried precisely that and drew back a nub.  They sued the FDA over its “substantial equivalence” clearance of §510(k) devices that they claimed should have been subjected to pre-market approval.  In re Orthopedic Bone Screw Products Liability Litigation, 1998 WL 964498, at *2 (E.D. Pa. Nov. 3, 1998).  They attacked FDA clearance claiming that the devices were not, in fact, substantially equivalent.  Id. at *3.  But whether and how to clear §510(k)  devices for labeling and marketing was well within the FDA’s discretion:

A review of the FDA’s responsibilities under the 510(k) clearance procedures reveals that the decisions made thereunder require the exercise of judgment and are therefore a matter of choice rather than particularly prescribed activity.  Specifically, the decisions which the FDA may be required to make in the 510(k) process include [four determinations]. . . .  Plaintiffs cite no statute or regulation which narrowly prescribes the scope of review beyond the standard articulated in that statute. . . .  Even that language makes review of clinical data discretionary upon decision of the Secretary or accredited person.  On its face, the statute requires the FDA to exercise decisionmaking and discretion in evaluating medical devices for 510(k) clearance, particularly in regard to determining the sufficiency of the scientific data supporting its findings.

Id. at *5 (citations omitted).  Under Berkovitz, agency discretion extends to both “formulation of policy” and actions that “involve[] the permissible exercise of policy judgment.”  Id.  Further, “the function involved in this §510(k) process also includes the determination of the nature and scope of the scientific data the agency should evaluate.”  Id.

This decision became final in In re Orthopedic Bone Screw Products Liability Litigation, 2000 WL 1578494, at *1 (E.D. Pa. Oct. 12, 2000), and was affirmed by Orthopedic Bone Screw Products Liability Litigation, 264 F.3d 344 (3d Cir. 2001).  The Third Circuit wholeheartedly agreed that §510(k) device clearance was an FDA discretionary function:

The government argues that the statutory and regulatory provisions grant it discretion to decide what information is relevant in the § 510(k) process.  Moreover, it points out that no mandatory authority dictates how it should go about determining whether devices have the same intended use, same technological characteristics, or similar performance in safety and efficacy.  Accordingly, the government contends that the §510(k) process requires the FDA to make discretionary choices utilizing its expertise.  We agree.

When §510(k) applications are brought before the FDA, regulators must decide what data and other information is relevant, what is reliable, and how much is sufficient.  Certainly in weighing evidence and comparing medical devices in this manner, the FDA utilizes judgment and choice.  Appellants’ suggestion that the FDA violated statutory and regulatory provisions is, in reality, a claim that the FDA’s judgment is wrong.  Because substantial equivalence determinations as well as the manner in which those decisions get made are functions committed to the discretion of the FDA, we will not second guess their outcomes.

Id. at 364.

The §510(k) process requires judgment regarding what evidence is relevant, how well that evidence demonstrates safety and efficacy, and what weight should be given conflicting evidence and opinions.  Implicit in this judgment is the balancing of values such as safety and cost.  Decisions made in this context reflect policy choices and cannot be categorized as ministerial.

Id.

Similarly, in King v. U.S. FDA, 35 F. Appx. 511 (9th Cir. 2002) (affirming King v. United States, 2001 WL 761184, (N.D. Cal. July 2, 2001)), the court determined that “the FDA’s pre-market approval of [a medical device] fell within the discretionary function exception.”  Id. at 513.  Although, the plaintiff alleged “numerous” agency violations, “she d[id] not point to any specific regulation or statute that mandates certain conduct with which the FDA did not comply.”  Id.  Device approvals were a discretionary function that Congress delegated to the FDA:

Congress empowered the FDA to subject Class III devices . . . “to premarket approval to provide reasonable assurance of [their] safety and effectiveness.” 21 U.S.C. §360c(a)(1)(C).  It seems clear to us that FDA decisions concerning pre-market approval of medical devices . . . “involve[] the permissible exercise of policy judgment” that the discretionary function exception exists to protect.    The district court was therefore correct that the FDA’s acts were discretionary.

Id. at 514 (Berkovitz citation omitted).

The same is true of drug approvals.  Forsyth v. Eli Lilly & Co., 904 F. Supp. 1153, 1157 (D. Haw. 1995), dismissed an attack on the FDA’s approval of a prescription drug.  Given the statutory and regulatory requirements for new drug applications:

agency employees have discretion to determine what methods of testing are reasonably applicable for a given drug, whether the testing performed by the manufacturer was adequate and whether adequate methods and facilities are used for manufacturing the drugs.  Because the regulation does not provide any guidance, agency employees must use their own judgment in making these determinations.  They must also determine whether the information submitted by the manufacturer or otherwise available to the agency is sufficient to guide them. The regulation governing withdrawal of the agency’s approval of a drug likewise grants agency employees broad discretion.

Id. at 1160 (citations omitted).  FDCA “prohibited acts,” particularly misbranding, “do not limit agency discretion in determining whether the statutes are complied with.”  Id.  The misbranding provision did not “sufficiently define criteria to limit agency discretion in determining whether a drug’s label is false or misleading,” since it “simply states that a drug shall be deemed to be misbranded ‘[i]f its labeling is false or misleading in any particular.’”  Id. (quoting 21 U.S.C. §352(a)).  An FDA drug approval therefore “fall[s] squarely within the discretionary function.”  Id.  See also Zammit v. Shire US, Inc., 2005 WL 8155121, at *1 (E.D. Mich. May 11, 2005) (plaintiff cannot sue the FDA; “the law provides only a limited avenue for challenging injuries allegedly arising from the FDA’s approval of (or failure to withdraw a prior approval of) an allegedly dangerous drug”).  See Bailey v. Eli Lilly Co., 607 F. Supp. 660, 662 (M.D. Pa. 1985) (dismissing claim against the FDA due to “the judgmental or discretionary function of the FDA in reviewing data submitted by drug companies for determination by FDA of whether or not a drug is safe for use”) (pre-Berkovitz, but still widely cited); Gray v. United States, 445 F. Supp. 337, 340 (S.D. Tex. 1978) (drug approval is discretionary; “[t]he FDA was given a general statutory mandate to assure the public that a marketed drug is safe for use”) (pre-Berkovitz, but still widely cited).

Similarly, Owen v. FDA Office of Generic Drugs, 2021 WL 3883112 (W.D.N.C. Aug. 27, 2021), held that an agency determination that a generic drug met the FDCA’s equivalence standards was a discretionary function.

While the FDA follows regulations when determining whether to approve a particular drug, those regulations do not specifically prescribe a course of action for [the FDA] to follow.  As such, the application of those regulations still involves the exercise of discretion by the FDA.”  The Plaintiff does not cite any statutes or regulations which eliminate the FDA’s discretion in the drug approval process.  Accordingly, the Court concludes that the FDA’s decision to approve certain drugs, including [a generic drug], involves a choice or judgment by the agency.

Id. at *4 (citations and quotation marks omitted).

The converse is also true.  An FDA decision whether to remove a product from the market, or not, is also discretionary.  The plaintiff claimed in Wiley v. United States, Dep’t of HHS, 2013 WL 537529 (D. Nev. Sept. 23, 2013), that the FDA negligently “failing to shut [a regulated firm] down” during an investigation of regulatory violations.  Id. at *2.  The FDA’s enforcement power was not subject to any “specific and mandatory regulation or statute” that required it “to force a drug manufacturer to close its operation.”  Id. at *4.  Thus, “the FDA appears to have . . . discretion to determine what steps to take during the [its] investigation.”  Id.  In particular:

The FDA’s decision to send [a warning] letter regarding the [claimed] violations, rather than causing [the firm being investigated] to close while it complied, was a discretionary decision that involved protected policy judgments, including the balancing of public policy decisions such as the public’s need for [these] drugs and the public’s interest in safe and effective drugs.

Id. at *5.  See Gelley v. Astra Pharmaceutical Products, Inc., 466 F. Supp. 182, 186 (D. Minn. 1979) (“As the gravamen of this allegation is that the FDA failed to withdraw its approval of [a drug] . . ., it seems clear that discretionary policy judgments are implicated.”) (pre-Berkovitz, but still widely cited), aff’d, 610 F.2d 558, 563 (8th Cir. 1979) (“we agree with the district court’s determination that the “discretionary function” exception also applies”) (citation omitted).

The FDA also regulates products other than prescription medical products, such as food and cosmetics.  In Fisher Brothers Sales, Inc. v. United States, 46 F.3d 279 (3d Cir. 1995) (en banc), an FDA decision to seize a food shipment was determined to be discretionary, despite the plaintiff’s allegations that the agency’s testing had negligently deviated from the agency’s written procedures.  Fisher rejected this characterization of the claim, because “[t]he reality here is that the injuries of which the plaintiffs complain were caused by the [FDA’s] decisions and, as a matter of law, their claims are therefore ‘based upon’ those decisions.”  Id. at 286.  FDA actions are discretionary when “based upon the exercise of a discretionary function whenever the immediate cause of the plaintiff’s injury is a decision which is susceptible of policy analysis and which is made by an official legally authorized to make it.”  Id. at 285-86.  A discretionary seizure did not lose its discretionary nature “whether the plaintiff purports to be attacking the data base on which the policy is founded or acknowledges outright that he or she is challenging the policy itself.”  Id. at 286.

If plaintiffs injured by regulatory policy decisions were permitted to prosecute damage actions by challenging the manner in which the underlying data was collected, federal courts, of necessity, would be required to examine in detail the decisionmaking process of the policymaker to determine what role the challenged data played . . . and what the policymaker’s decision would have been if he or she had received the unchallenged data but not the challenged data (or had received other data in lieu of the challenged data). . . .  [T]his is precisely the kind of inquiry that the Supreme Court sought to foreclose.

Id.  Since “every policy decision involves an exercise of the policymaker’s judgment about the reliability, adequacy, and significance of the information available,” a discretionary decision does not become non-discretionary by virtue of attacks on “the reliability of the available data.”  Id. at 287.  Accord Seaside Farm, Inc. v. United States, 842 F.3d 853, 860 (4th Cir. 2016) (FDA food “contamination warnings − in both timing and content − are a prototypical discretionary function”; rejecting reliance on FDA “guidance” as limiting agency discretion).

The only claim against the FDA since Berkovitz that was held not to involve FDA discretionary functions was far removed from FDA product approval or recall actions.  See Jerome Stevens Pharmaceuticals, Inc. v. FDA, 402 F.3d 1249, 1254-55 (D.C. Cir. 2005) (allegations improper FDA disclosure of an applicant’s confidential information).

In sum, Loper Bright expressly recognizes that it is entirely appropriate for Congress to grant administrative discretion to agencies such as the FDA.  Thus, nothing in Loper Bright should affect the FDA’s power to exercise such discretion where it exists.  As to what the boundaries of such FDA discretion are, numerous courts applying the analogous “discretionary function” exception to the FTCA have placed product-specific decisions about whether prescription medical products should marketed, or not, and what such products’ labels should state are within the scope of FDA discretion.  Attempts by plaintiffs to second-guess these sorts of FDA decisions thus should fail after Loper Bright, as they have failed previously.