PREP Act Immunity Reaches Contract Claims for Vaccine-Related “Loss” and Supports Collateral-Order Appeals

I. Introduction

In Dressen v. AstraZeneca AB (10th Cir. July 23, 2026), the Tenth Circuit addressed a question that sits at the intersection of pandemic-response policy and ordinary private-law remedies: whether the Public Readiness and Emergency Preparedness Act (the PREP Act) immunizes a vaccine manufacturer not only from tort suits, but also from breach-of-contract claims seeking damages for injuries alleged to have been caused by a covered countermeasure.

The dispute arose from Brianne Dressen’s participation in a 2020 AstraZeneca COVID-19 vaccine clinical trial. Before receiving the investigational vaccine, she signed an informed-consent form (the ICF) describing compensation and medical-treatment assurances and warning that federal law “may limit” her right to sue for vaccine-related harms, while pointing to the Countermeasures Injury Compensation Program (CICP). After Dressen suffered serious injuries and AstraZeneca declined her requests for care/compensation, she sued for breach of contract (including breach of the duty of good faith and fair dealing). AstraZeneca moved to dismiss, asserting PREP Act immunity.

The district court denied the motion, reasoning that PREP Act immunity covers tort claims but not contract claims. The Tenth Circuit reversed on the scope of immunity, held it had interlocutory jurisdiction under the collateral-order doctrine, and remanded for the district court to consider whether AstraZeneca waived PREP Act immunity through the ICF’s language.

II. Summary of the Opinion

  • Jurisdiction: The denial of PREP Act immunity is immediately appealable under the collateral-order doctrine because the statute grants “immun[ity] from suit and liability,” and forcing covered persons to litigate would irreparably defeat the “immunity from suit” Congress provided.
  • Merits—Scope of “loss” and “claims for loss”: PREP Act immunity is not limited to tort causes of action. It reaches all claims seeking recovery for statutorily defined “loss,” including contract claims, so long as the loss bears the required causal relationship to administration or use of a covered countermeasure.
  • Causation: The court treated the Act’s causation language (“caused by, arising out of, relating to, or resulting from”) as broad and, at minimum, satisfied by traditional but-for causation. Dressen’s damages claim would not exist without the vaccine administration and resulting injury; multiple but-for causes can coexist.
  • Remand: The case returns to the district court to decide whether PREP Act immunity can be waived and, if so, whether the ICF effected a waiver. The panel also declined to reach Dressen’s equitable-estoppel theory because it was not preserved below.

III. Analysis

A. Precedents Cited

1. Collateral-order doctrine and “immunity from suit”

The court’s jurisdictional holding draws heavily from Supreme Court collateral-order doctrine cases emphasizing both the doctrine’s narrowness and its special accommodation for immunities that include a right not to be tried.

  • GEO Grp., Inc. v. Menocal (2026): Quoted for the three “non-negotiable conditions” for collateral-order review and the doctrine’s “narrow, stringent” scope. This frames the Tenth Circuit’s caution and category-based approach.
  • Mohawk Indus., Inc. v. Carpenter (2009): Used to emphasize that the third prong turns on whether delay imperils a “substantial public interest” or a “value of a high order,” and to reinforce the Court’s preference for rulemaking over ad hoc expansion of interlocutory appeals.
  • Digital Equipment Corp. v. Desktop Direct, Inc. (1994): The pivotal case for the panel’s analysis. The opinion uses Digital Equipment’s skepticism toward generalized “right not to stand trial” rhetoric, while also relying on its key carve-out: when Congress confers “immunity from suit” in positive law, courts should not “gainsay” its importance.
  • Mitchell v. Forsyth (1985): Cited for the principle that certain immunities are lost if the defendant must undergo trial, supporting the idea that immunity from suit warrants immediate review.
  • Will v. Hallock (2006) and Van Cauwenberghe v. Biard (1988): Cited to distinguish between mere interests in avoiding trial and high-order immunities that justify interlocutory review; Will’s “substantial public interest” formulation is central to the third prong.
  • United States ex rel. Fiorisce, LLC v. Colorado Technical University, Inc. (10th Cir. 2025): Distinguished because the False Claims Act public-disclosure bar does not expressly provide “immunity from suit.” The panel treats Fiorisce as consistent with Digital Equipment’s framework rather than a limitation on statutory “immunity from suit” appeals.
  • Hampton v. California (9th Cir. 2023) and Goins v. Saint Elizabeth Med. Ctr. (6th Cir. Jan. 22, 2024) (unpublished): Cited as persuasive circuit authority recognizing collateral-order review of PREP Act immunity denials because PREP Act immunity includes immunity from suit.
  • Additional doctrinal framing comes from Mohamed v. Jones (10th Cir. 2024) (importance/independence rationale), Johnson v. Jones (1995) (separateness from merits), and Swint v. Chambers Cnty. Comm'n (1995) (rulemaking context).

Doctrinal contribution: Although other circuits had already allowed collateral-order appeals in PREP Act cases, the Tenth Circuit’s decision squarely anchors that result in the statute’s express “immune from suit” language and Digital Equipment’s “rare” category, giving district courts a clearer signal that denials of PREP Act immunity will be subject to immediate appellate correction in this circuit.

2. Statutory interpretation and breadth of “all claims for loss”

On the merits, the court uses mainstream textualist tools: start with the statutory words, read them in context, and apply interpretive canons cautiously.

  • Robinson v. Shell Oil Co. (1997) and Smith v. Bd. of Governors of the Fed. Rsrv. Sys. (10th Cir. 2023): Cited for the basic sequence—text first, context and whole-statute meaning, and stop if unambiguous.
  • Ali v. Fed. Bureau of Prisons (2008), Allen v. Env't Restoration, LLC (10th Cir. 2022), and Cohen v. JP Morgan Chase & Co. (2d Cir. 2007): Used to support the expansive ordinary meaning of “any” and “all.”
  • Norfolk & Western Railway Co. v. American Train Dispatchers' Ass'n (1991): The key analogy for breadth. There, “all other law” covered contractual obligations when necessary to effect an approved transaction; here, “all claims for loss” is limited by causation, not by the legal label (tort vs. contract).
  • Yates v. United States (2015) (plurality) and Marx v. Gen. Revenue Corp. (2013): Invoked in rejecting Dressen’s canon-based arguments (ejusdem generis, noscitur a sociis, expressio unius).
  • Lackey v. Stinnie (2025): Used to emphasize that where Congress knows how to draft an exception (here, contract carveouts in PLCAA), courts should not add an atextual limitation.
  • Lewis v. United States (1998) and Small v. United States (2005): Cited for the general caution that “any” may sometimes yield to purpose where literalism would detach text from statutory function—then used to reject the notion that including some contract claims creates such a mismatch here.
  • Griffin v. Oceanic Contractors, Inc. (1982) and Robbins v. Chronister (10th Cir. 2006): Provide the high bar for the absurdity doctrine—only truly “unthinkable” results justify departing from clear text.
  • United States v. Davis (2019), Hernandez-Carrera v. Carlson (10th Cir. 2008), and Bondi v. VanDerStok (2025): Used to limit constitutional avoidance to genuinely ambiguous statutes and to deny it any role where “text, context, and structure” resolve the case.
  • Connolly v. Pension Ben. Guar. Corp. (1986), Concrete Pipe & Prods. of Calif. v. Constr. Laborers Pension Tr. for S. Calif. (1993), Omnia Com. Co. v. United States (1923), and Duke Power Co. v. Carolina Env't Study Grp., Inc. (1978): These cases defeat the Takings Clause-based avoidance argument by recognizing broad congressional authority to adjust economic burdens—even if doing so disrupts private contractual expectations—and by emphasizing that limiting or eliminating liability is commonplace.

3. Causation baseline

  • Bostock v. Clayton County (2020) and Comcast Corp. v. Nat'l Ass'n of Afr. Am.-Owned Media (2020): Supply the court’s but-for causation framing and the notion that injuries can have multiple but-for causes.

B. Legal Reasoning

1. “All claims for loss” is claim-type neutral

The central move is simple and consequential: the PREP Act does not immunize against “torts”; it immunizes against “all claims for loss” meeting the statute’s causal nexus. The court refuses to smuggle a cause-of-action limitation into a damages-category definition.

The statutory definition of “loss” (“any type of loss, including …”) is treated as expansive and illustrative, not exhaustive or tort-specific. The court further uses a practical example—business interruption loss—to show that the enumerated categories often arise in non-tort settings (such as insurance contracts), undermining the argument that “loss” is inherently “tort-based.”

2. Canon warfare is rejected as mismatched to the text’s structure

Dressen’s principal interpretive strategy was to use canons (ejusdem generis, expressio unius, noscitur a sociis, and surplusage) to narrow “loss” to tort harms. The panel responds that these canons do not apply in the way she deploys them, and that the phrase “any type of loss, including” is a strong textual signal of breadth that resists narrowing.

3. The statute’s remedial structure does not imply a contract-claim carveout

Dressen argued that because the CICP’s compensation scheme is limited (e.g., “serious physical injury or death”), Congress could not have intended to foreclose contract remedies for other harms. The court treats that as a non sequitur: the compensation program’s limits are evidence that Congress intentionally drew sharp lines about what is compensable, while the immunity provision still broadly forecloses litigation to prevent “the specter of financial ruin” from deterring emergency countermeasure development and deployment.

4. Causation is satisfied where the countermeasure is a necessary condition of the claimed damages

The opinion reads the Act’s causation formulations—“caused by,” “arising out of,” “relating to,” “resulting from”—as deliberately comprehensive. It holds that, at minimum, but-for causation suffices on the arguments presented.

That leads to a decisive rebuttal of the “the breach, not the vaccine, caused the loss” framing. The panel explains that Dressen’s alleged contract damages depend on vaccine injury: absent the administration of the vaccine and resulting harm, there would be no claimed entitlement to payment for “research injuries.” The breach may be a but-for cause as well, but multiple but-for causes can exist; the Act requires only that the loss have the requisite causal relationship to administration or use.

5. Constitutional avoidance and the Takings Clause do not narrow the statute

The panel first finds no ambiguity that could trigger avoidance. It then explains that Congress may alter legal liability regimes even if doing so nullifies contractual expectations, and that such adjustments are typically evaluated under deferential standards. By emphasizing that the government “has taken nothing for its own use” and that liability limitations are commonplace, the court shuts down the effort to recharacterize PREP Act immunity as an uncompensated taking.

6. Waiver is left open—and becomes the key battleground on remand

Having held that the PREP Act reaches contract claims, the court does not treat that as the end of the matter. Instead, it remands for the district court to decide: (1) whether PREP Act immunity can be waived, and (2) whether AstraZeneca waived it in the ICF by promising to cover costs of research injuries and to facilitate treatment/referrals.

This preserves an important doctrinal safety valve: even if Congress created broad immunity to encourage countermeasure deployment, private actors’ own trial documents may still matter—if waiver is legally recognized and factually supported.

C. Impact

1. PREP Act immunity in the Tenth Circuit is now explicitly claim-type agnostic

The most direct precedential effect is that, in the Tenth Circuit, PREP Act immunity is not confined to tort theories. Plaintiffs cannot avoid immunity merely by pleading a covered “loss” (e.g., injury, medical expenses, emotional harms, property losses) under a contract label. This is especially relevant to litigation arising from:

  • clinical trials (informed-consent forms, reimbursement provisions, “research injury” payment language),
  • vaccine and therapeutics distribution agreements and downstream service relationships, and
  • attempts to repackage product-liability or malpractice narratives as implied-warranty or breach-of-promise claims.

2. Interlocutory appeals will become more common in PREP Act cases

By holding denials of PREP Act immunity immediately appealable, the court increases the likelihood that immunity questions will be resolved early and uniformly. Practically, this can:

  • reduce settlement pressure created by discovery burdens in high-stakes injury cases,
  • centralize doctrinal development at the appellate level, and
  • shift litigation strategy toward threshold motion practice and jurisdictional positioning.

3. Drafting and risk communication: waiver language becomes pivotal

The remand invites renewed focus on the ICF’s wording. If courts recognize waiver of PREP Act immunity, manufacturers and trial administrators will face pressure to clarify whether “research injury” payment language is:

  • a non-binding description of potential assistance,
  • a conditional undertaking limited by the PREP Act/CICP, or
  • a contractual commitment intended to be enforceable notwithstanding immunity.

Conversely, plaintiffs may target specific language as an express or implied waiver of an otherwise broad statutory shield.

4. Compensation policy tension remains—and is highlighted rather than resolved

The opinion underscores Congress’s design: broad immunity paired with a narrow compensation program (CICP) and a narrow willful-misconduct exception. Whatever one thinks of that balance, the court treats it as a legislative choice, not a judicially adjustable feature.

IV. Complex Concepts Simplified

  • PREP Act immunity (“immune from suit and liability”): More than a defense to paying damages; it can be a defense to being sued at all, which is why interlocutory review matters.
  • “All claims for loss”: The statute focuses on the kind of harm (“loss”) rather than the legal theory (tort vs. contract). If the plaintiff’s damages are for a listed type of “loss,” the label on the complaint does not control.
  • Causation (“caused by / arising out of / relating to / resulting from”): These are broad connecting phrases. The court held the connection exists where, at minimum, the loss would not have happened without the covered countermeasure’s administration or use.
  • Collateral-order doctrine: A narrow exception allowing immediate appeal of certain non-final orders—especially those denying immunities from suit— because waiting until the end would destroy the very right the immunity protects.
  • Interpretive canons: Tools like ejusdem generis and noscitur a sociis help resolve ambiguity, but they cannot be used to rewrite clear breadth signaled by phrases like “any type of loss, including.”
  • Constitutional avoidance: Courts only narrow statutes to avoid constitutional problems when the statute is genuinely ambiguous; clear text defeats avoidance. And limiting liability typically does not equal a “taking” of property.

V. Conclusion

Dressen v. AstraZeneca AB establishes (for the Tenth Circuit) a clear rule: PREP Act immunity extends to all claims for statutorily defined “loss”—including breach-of-contract claims—so long as the loss bears the statute’s causal relationship to administration or use of a covered countermeasure. It also confirms that denials of PREP Act immunity are immediately appealable under the collateral-order doctrine because the Act confers immunity “from suit.”

The opinion’s practical significance is twofold: it forecloses a major pleading workaround (relabeling vaccine-injury damages as contractual loss), while leaving open an equally significant question for future development—whether and how PREP Act immunity may be waived through clinical-trial documents and related promises. The remand ensures that the next phase of PREP Act litigation will increasingly focus on waiver, drafting, and the boundary between statutory public-emergency protections and private undertakings.