Full Opinion

United States Court of Appeals FOR THE DISTRICT OF COLUMBIA CIRCUIT Argued November 5, 2025 Decided August 14, 2026 No. 24-7135 RAMONA MATOS RODRIGUEZ, ET AL., APPELLEES v. PAN AMERICAN HEALTH ORGANIZATION, APPELLANT JOAQUIN MOLINA, ET AL., APPELLEES Appeal from the United States District Court for the District of Columbia (No. 1:20-cv-00928) Daniel S. Volchok argued the cause for appellant. With him on the briefs was David W. Bowker. Samuel J. Dubbin argued the cause for appellees. With him on the brief were Charles J. Cooper, Michael W. Kirk, Haley N. Proctor, and Joseph O. Masterman. Andrew Kim was on the brief for amicus curiae Human Trafficking Legal Center in support of appellees. 2 Dennis H. Hranitzky and Alex H. Loomis were on the brief for amici curiae Professor William S. Dodge, et al. in support of appellees. Before: MILLETT, PAN, and GARCIA, Circuit Judges. Opinion for the Court filed by Circuit Judge GARCIA. GARCIA, Circuit Judge: In 2018, a group of Cuban doctors sued the Pan American Health Organization (PAHO), alleging that it cooperated in a human trafficking scheme. PAHO moved to dismiss, arguing that even if the complaint’s allegations were true, it was immune from suit under the International Organizations Immunities Act (IOIA). The district court denied that motion. Most defendants cannot appeal the denial of a motion to dismiss. Instead, they must litigate the case until a final judgment is entered and then, if necessary, seek appellate review. But because IOIA immunity serves in part to protect defendants from the burdens of litigating the merits of a suit, PAHO was entitled to seek immediate appellate review under the collateral order doctrine. PAHO did so, and we affirmed, agreeing with the district court that the doctors adequately pleaded a theory that, if proven, would pierce PAHO’s immunity. On remand, PAHO filed a second motion to dismiss, which remains pending before the district court. Unlike the first, this motion contests the complaint’s allegations with evidence to support PAHO’s claim to immunity. If the district court denies that motion, PAHO will be entitled to a second round of immediate appellate review before it is required to litigate any merits issues. But before resolving that motion— indeed, to help the court properly resolve it—the district court granted the plaintiffs’ request for discovery in response to 3 PAHO’s evidentiary submissions. Rather than comply, PAHO sought immediate review of that discovery order too. We conclude that we lack jurisdiction over the appeal. Holding otherwise would give defendants in PAHO’s situation not just two opportunities for appellate review as to immunity, but a boundless entitlement to appeal any number of district court orders aimed at resolving that threshold issue. We decline PAHO’s invitation to extend our appellate jurisdiction that far. I In 2013, Brazil launched Mais Médicos (Portuguese for “More Doctors”), a public health program aimed at increasing access to medical care. The program recruited healthcare professionals from around the world to serve Brazilian communities that otherwise lacked reliable healthcare resources. Cuba offered to assign employees from its own Ministry of Health to Mais Médicos placements in Brazil. To facilitate the arrangement, PAHO entered agreements with the two countries. PAHO is the world’s oldest international public health organization and is headquartered in Washington, D.C. The agreements gave PAHO a central role in the payment structure for Cuba’s participation in Mais Médicos. Brazil paid for the services of thousands of Cuban healthcare professionals by transferring funds to PAHO. PAHO retained five percent of those funds for itself and sent the remainder to Cuba. According to PAHO, the 5% “offset” enabled the organization to recoup indirect costs associated with the technical and administrative support it provided to the program. Appellant’s Brief 8. Concerns arose that Cuba’s “medical missions,” including its partnership with Mais Médicos, were part of a human- 4 trafficking scheme. The U.S. Department of State noted that the Cuban government was collecting immense sums from such programs but would often withhold wages from the physicians on the ground. See Compl. ¶ 31 (quoting U.S. Dep’t of State, Trafficking in Persons Report 162–64 (2019)). Some physicians in these programs reported substandard working and living conditions. Id. ¶ 30 (quoting U.S. Dep’t of State, Trafficking in Persons Report 4 (2017)). As one amicus noted, a complaint filed with the International Criminal Court alleged that 75% of participants in these missions had not joined the programs voluntarily; 75% had been threatened or witnessed coworkers being threatened; and 40% were separated from their children as a consequence for defecting. Human Trafficking Legal Center Amicus Brief 9 n.4 (citing U.S. Dep’t of State, Trafficking in Persons and Cuba’s Labor Export Program (Jan. 20, 2025), https://perma.cc/GW7H- GTQP). In 2018, four Cuban doctors who allege they were coerced to participate in Mais Médicos and who are now living in the United States sued PAHO for “enabling, managing, and enforcing illegal human trafficking of Cuban medical professionals in Brazil.” Compl. ¶ 1. The doctors alleged, among other things, that the Cuban government restricted their access to family members, see id. ¶ 102; limited their freedom of movement, see id. ¶¶ 88–89; and compensated them with only a fraction of the sums Brazil paid for their work, see id. ¶ 96, all under threat of “harassment, intimidation, and retaliation,” id. ¶ 79. They asserted claims under the Trafficking Victims Protection Act (TVPA), which forbids various forms of human trafficking, see 18 U.S.C. § 1590, and empowers trafficking victims to sue and recover damages from “whoever knowingly benefits . . . from participation in a venture” that the defendant “knew or should have known” involved such a violation, id. § 1595(a). The doctors sued on 5 behalf of a class of about 3,500 similarly situated Cuban medical professionals. Compl. ¶ 118. Nearly eight years into the litigation, the parties remain enmeshed in a threshold dispute over PAHO’s claim to immunity under the IOIA. Enacted in 1945, the IOIA affords “[i]nternational organizations . . . the same immunity from suit and every form of judicial process as is enjoyed by foreign governments.” 22 U.S.C. § 288a(b). In 1960, President Eisenhower designated PAHO an international organization under the Act. See Rodriguez v. PAHO, 29 F.4th 706, 712 (D.C. Cir. 2022) (PAHO I). Because IOIA immunity “develops in tandem” with “the law of foreign sovereign immunity,” it is coextensive with immunity under the Foreign Sovereign Immunities Act (FSIA). Jam v. Int’l Fin. Corp., 586 U.S. 199, 210 (2019). The FSIA, in turn, provides foreign sovereigns with immunity “from the jurisdiction of the courts of the United States,” 28 U.S.C. § 1604, subject to several enumerated exceptions, see id. § 1605. One of those exceptions—the “commercial activity” exception—covers claims “based upon a commercial activity carried on in the United States by the foreign state.” Id. § 1605(a)(2). A defendant seeking to raise an IOIA or FSIA immunity defense before litigating the merits has two procedural options for challenging the court’s subject matter jurisdiction and can pursue either or both. The defendant can file what we will call a “legal” motion to dismiss, which accepts “the plaintiff’s factual allegations as true” and argues that those allegations do not “bring the case within any of the exceptions to immunity invoked by the plaintiff.” Phoenix Consulting Inc. v. Republic of Angola, 216 F.3d 36, 40 (D.C. Cir. 2000). Or the defendant can file a “factual” motion to dismiss, challenging the 6 plaintiff’s jurisdictional allegations with its own evidence. See id. The “factual” motion requires the district court to “resolve any disputed issues of fact” necessary to decide whether the defendant is entitled to immunity and thus whether the court may exercise jurisdiction over the action. Id. In resolving such disputes, the district court “must give the plaintiff ample opportunity to secure and present evidence relevant to the existence of jurisdiction,” and so may allow “jurisdictional discovery” that is “carefully controlled and limited.” Id. (quotation marks omitted). PAHO pursued both paths. It first filed a “legal” motion to dismiss. The district court denied it, concluding that the doctors had alleged facts that, if true, would trigger the commercial activity exception to IOIA immunity. Rodriguez v. PAHO, 502 F. Supp. 3d 200, 214 (D.D.C. 2020). We affirmed. PAHO I, 29 F.4th at 716–17. The physicians alleged in their complaint that PAHO engaged in unlawful financial activity within the United States, including by acting as a financial intermediary “moving money for a fee” between Brazil and Cuba using a “Washington, D.C. bank account.” Id. We held that those allegations, if proven, would suffice to establish the “commercial activity” exception to immunity. Id. at 717. We then remanded to the district court for further proceedings. Id. at 720.1 PAHO then filed a second, “factual” motion to dismiss. This time, PAHO argued that the doctors’ factual allegations were mistaken. Contrary to the complaint, PAHO urged, none of the Mais Médicos money ever entered the United States or a U.S. bank account, and the collection of a 5% “fee” was more 1 In ruling on PAHO’s first motion, the district court held that PAHO was immune as to two other TVPA claims and a claim under the Racketeer Influenced and Corrupt Organizations Act. PAHO I, 29 F.4th at 710. 7 akin to sovereign activity than commercial activity. See Factual Mot. to Dismiss 1–3, 34, Dkt. No. 80. In response, the doctors sought discovery to further develop the details of PAHO’s conduct and clarify whether that conduct was “commercial activity carried on in the United States” such that the court could exercise jurisdiction. See Rodriguez v. PAHO, 2024 WL 4251808, at *1 (D.D.C. Aug. 12, 2024) (PAHO II). PAHO voluntarily produced various documents in support of its motion, including declarations and bank statements, but opposed the doctors’ requests for jurisdictional discovery. For nearly two years, the parties attempted to negotiate an agreement regarding the jurisdictional discovery dispute. When those efforts stalled, the district court ordered PAHO to comply with ten of the plaintiffs’ requests for production, two interrogatories, and one deposition request. Id. at *3. PAHO appealed again. II Before we can consider the merits of PAHO’s challenge to the discovery order, “[w]e must begin with the question of our [appellate] jurisdiction.” Process & Indus. Devs. Ltd. v. Fed. Republic of Nigeria, 962 F.3d 576, 580 (D.C. Cir. 2020) (P&ID). We conclude that we lack jurisdiction because the discovery order is not effectively “final” under the collateral order doctrine. We thus dismiss the appeal and remand to the district court for further proceedings. A PAHO invokes this court’s jurisdiction to review district court “final decisions” under 28 U.S.C. § 1291. Typically, such decisions are those that “end[] the litigation” and “leave[] nothing for the court to do but execute the judgment.” Geo Grp., Inc. v. Menocal, 607 U.S. 438, 443 (2026). Section 1291, however, is construed “practical[ly].” Id. at 444. As a 8 result, the collateral order doctrine permits courts to hear immediate appeals from a “small class” of “non-terminal” decisions that are effectively “final.” Id. To qualify for immediate review under this doctrine, the type of order at issue must “(1) conclusively determine the disputed question, (2) resolve an important issue completely separate from the merits of the action, and (3) be effectively unreviewable on appeal from a final judgment.” Id. Derived from Cohen v. Beneficial Industrial Loan Corp., 337 U.S. 541 (1949), this test identifies orders that “finally determine claims of right separable from, and collateral to, rights asserted in the action” and that are “too important to be denied review and too independent of the cause itself to require that appellate consideration be deferred.” Id. at 546. But, as the Supreme Court has repeatedly and recently reiterated, the collateral order doctrine is “narrow,” “stringent,” and of “modest scope.” Geo Grp., 607 U.S. at 444 (first quoting Digital Equipment Corp. v. Desktop Direct, Inc., 511 U.S. 863, 868 (1994); and then quoting Will v. Hallock, 546 U.S. 345, 350 (2006)). Ordinarily, the denial of a motion to dismiss is not immediately appealable. See Van Cauwenberghe v. Biard, 486 U.S. 517, 527 (1988). The collateral order doctrine, however, generally permits appeals from the denial of a motion to dismiss based on an immunity from suit. See Geo Group, 607 U.S. at 444. That category includes motions to dismiss based on FSIA or IOIA immunity. P&ID, 962 F.3d at 583; Phoenix Consulting, 216 F.3d at 39; Foremost-McKesson, Inc. v. Islamic Republic of Iran, 905 F.2d 438, 443 (D.C. Cir. 1990). Whether a defendant is entitled to such immunity presents an important question about the court’s subject matter jurisdiction, which is separate from the merits. And because IOIA immunity is an “immunity from litigation burdens as well as from the entry of adverse judgments,” its improper rejection 9 is “effectively unreviewable” at the conclusion of the litigation. P&ID, 962 F.3d at 582, 584. “[I]f the immunity is wrongly denied at the outset of a case, it cannot be vindicated after final judgment,” because the defendant would already have been forced to litigate the merits of the suit. Id. at 581. Immediate appeal enables immune defendants to promptly seek redress from errors that would otherwise “frustrate the significance and benefit of entitlement to immunity from suit.” Phoenix Consulting, 216 F.3d at 39 (quotation marks omitted). Notably, this analysis applies to both “legal” and “factual” motions to dismiss. See, e.g., Princz v. Fed. Republic of Germany, 26 F.3d 1166, 1168–69 (D.C. Cir. 1994) (legal); Phoenix Consulting, 216 F.3d at 39 (factual). To avoid piecemeal appeals, a defendant may assert both legal and factual arguments in support of dismissal as part of one motion, and obtain immediate appellate view of an adverse ruling. Where circumstances necessitate filing separate “legal” and “factual” motions to dismiss, the defendant may obtain immediate appellate review of adverse rulings on both motions. This case requires us to decide whether, and when, defendants are also entitled to appeal jurisdictional discovery orders targeted at resolving their entitlement to IOIA immunity. Like the denial of a motion to dismiss, pretrial discovery orders generally are not immediately appealable. See Mohawk Indus., Inc. v. Carpenter, 558 U.S. 100, 108 (2009). But in Nyambal v. International Monetary Fund, 772 F.3d 277 (D.C. Cir. 2014), we permitted collateral order review of one type of jurisdictional discovery order. In that case, the plaintiff sued the International Monetary Fund (IMF), which moved to dismiss based on IOIA immunity, attaching affidavits and evidence to support that defense. Id. at 279. The plaintiff sought discovery to investigate whether the IMF had waived its immunity. Id. The district court granted that 10 request without determining whether the plaintiff had made any well-founded allegations that IOIA immunity did not apply. Id. We exercised jurisdiction over the IMF’s appeal, explaining that the discovery order “denie[d] an international organization protection” from burdens “similar” to “the burden of a lawsuit.” Id. at 280 (citing Foremost-McKesson, 905 F.2d at 443); see id. at 279, 281 (noting that the district court “authorized jurisdictional discovery” without confirming the “plausibility” of the plaintiff’s theory). PAHO argues that Nyambal governs here. We disagree. For the reasons below, we conclude that the Cohen analysis differs when a district court issues a jurisdictional discovery order in circumstances like those present here. We focus on Cohen’s third requirement: that the order be effectively unreviewable after final judgment. The “crucial question” for that requirement is “whether deferring review until final judgment so imperils” an “interest” at stake in the decision “as to justify the cost of allowing immediate appeal of the entire class of relevant orders.” Mohawk, 558 U.S. at 108. We are not persuaded that the “interest” at stake in a case like this one “justif[ies]” the cost of permitting immediate appeal for an entire class of similar orders. Specifically, we hold that immediate review of a jurisdictional discovery order is not available under the collateral order doctrine at least where (1) the order is issued after a judicial determination that the complaint adequately pleads an immunity exception, (2) the defendant files a factual motion to dismiss disputing those factual allegations, and (3) the discovery addresses the theories the court has found adequately pleaded by the complaint. To be clear, we do not hold that interlocutory appeal is available whenever one of these factors is absent. Today we hold only that the confluence of all three renders immediate review inappropriate in this case and others like it. 11 There is a stark difference between the “interest[s]” that are “imperil[ed]” when a district court denies a motion to dismiss based on immunity grounds and when, as here, the district court orders discovery for the purpose of deciding the immunity question itself. Id. In the motion-to-dismiss context, our cases emphasize that deferring review until after final judgment would necessarily require a potentially immune defendant “to defend litigation on the merits despite an unresolved assertion of immunity.” P&ID, 962 F.3d at 582 (emphasis added); see also id. (distinguishing the writ of mandamus granted in In re Papandreou, 139 F.3d 247 (D.C. Cir. 1998), which addressed a jurisdictional discovery order, because “the collateral-order doctrine would not have applied [in Papandreou] under the reasoning that we adopt here”). “[C]onsideration of the merits [is] itself an infringement on foreign sovereign immunity.” Id. at 584 (quoting Republic of Philippines v. Pimentel, 553 U.S. 851, 864 (2008)). The burden that PAHO faces here is materially different. It is true that responding to jurisdictional discovery is itself a burden, as Nyambal recognized. Nyambal, 772 F.3d at 280. But that burden, in circumstances like these, is not comparable to the burden of being forced to litigate the merits of a suit. In a case like this one, the defendant faces the burden of answering discovery related to factual questions the defendant itself placed in dispute to support its immunity claim, after a court determined that the plaintiff adequately pleaded an exception to IOIA immunity—a determination that is itself appealable (and was appealed here). Once the plaintiff presents that adequately pleaded theory, the defendant “bears the burden of proving” the claim is not “within a statutory exception to immunity.” Phoenix Consulting, 216 F.3d at 40. And when a defendant seeks to meet that burden by disputing the factual allegations in a complaint regarding immunity, resolving that dispute may naturally “require[] jurisdictional 12 discovery and factual resolution of immunity questions to take place.” P&ID, 962 F.3d at 584. Indeed, we have explained that district courts “must give the plaintiff ample opportunity to secure and present evidence relevant to the existence of jurisdiction.” Phoenix Consulting, 216 F.3d at 40 (quotation marks omitted). And the immunity question must be resolved “before the [defendant] is required to defend the merits.” P&ID, 962 F.3d at 584 (citing Phoenix Consulting, 216 F.3d at 39–40). In short, when a defendant disputes a complaint’s well-pleaded factual allegations with its own evidence aimed at establishing its immunity, answering discovery targeted to resolving that dispute is part of the bargain. Our conclusion is further supported by the idea that immediate appellate review of a category of orders is unlikely to have much utility when the type of order is unlikely to be reversed on appeal. As part of the Cohen analysis, the Supreme Court has explained that if “interlocutory appeals [of a category of order] are less likely to bring error-correcting benefits,” this fact counts against immediate appealability. Johnson v. Jones, 515 U.S. 304, 316 (1995); see also id. (citing and quoting Richardson-Merrell, Inc. v. Koller, 472 U.S. 416, 434 (1985), for the following proposition: “[T]he fact that ‘[m]ost pretrial orders [of the kind there at issue] are ultimately affirmed by appellate courts’ militate[s] against immediate appealability.”). Similarly, other circuits have cited the fact that a type of order is subject only to abuse-of-discretion review as a mark against immediate appeal because it suggests “the likelihood of reversal is too negligible to justify” the costs of immediate appellate review. Donlon Indus., Inc. v. Forte, 402 F.2d 935, 937 (2d Cir. 1968); accord James v. Jacobson, 6 F.3d 233, 237–38 (4th Cir. 1993); Midway Mfg. Co. v. Omni Video Games, Inc., 668 F.2d 70, 72 (1st Cir. 1981) (“[W]hen we are asked to review not a question of controlling law but an exercise of a judge’s discretion that depended upon the 13 particular facts before him, this factor mitigates against immediate review.”). That approach makes sense. If a “class of relevant orders” is, based on its objective characteristics, inherently less likely to be reversed after appellate review, then the “cost[s]” of appellate review are harder to “justify” in the aggregate. Mohawk, 558 U.S. at 108. Here, the first feature of this case identified above—a prior judicial determination that the complaint adequately pleads an exception to IOIA immunity—diminishes the relative likelihood that an error will be “correct[ed]” via reversal on appeal. Johnson, 515 U.S. at 316; see Donlon, 402 F.2d at 937. That determination makes jurisdictional discovery more likely to be appropriate and immediate appellate review less necessary. Unlike in Nyambal, the plaintiffs here sought discovery after the district court (and this court) squarely determined that the complaint adequately pleads an exception to IOIA immunity. That determination establishes that the assertion of jurisdiction over PAHO is at least “plausible” given the complaint’s allegations. PAHO I, 29 F.4th at 711. The second and third features have a similar effect. When a defendant then files a factual motion to dismiss disputing the complaint’s well-pleaded jurisdictional allegations, subsequent inquiry is further concentrated on those factual issues placed in contention by the defendant. And when the discovery order investigates the theories of immunity exceptions the court has already found adequately pleaded, we can have further confidence that the discovery granted will be necessary to resolve the jurisdictional dispute. This procedural posture, in short, provides objective guardrails that reduce the likelihood that the ordered discovery is improper. So the risks posed by deferring review are less likely to “justify” the institutional costs of our immediate intervention. Mohawk, 558 U.S. at 112; see Johnson, 515 U.S. at 316. 14 PAHO primarily responds by emphasizing our decision in Nyambal. According to PAHO, that case, in its single paragraph of analysis, held that every jurisdictional discovery order involving a defendant invoking IOIA or FSIA immunity is immediately appealable no matter the circumstances. We are not persuaded. Nyambal necessarily addressed only the circumstances presented in that case. The district court there ordered discovery without any prior assessment of whether the plaintiff had adequately pleaded an exception to immunity. See Nyambal, 772 F.3d at 279. That case simply did not involve a situation like this one, where a court had determined that the complaint adequately pleaded an exception to immunity; the defendant disputed the factual allegations related to that theory, then provided the district court with evidence of its own selection supporting its view of the facts; and the trial court ordered discovery related to that theory. For the reasons we have given, those distinctions matter for the Cohen analysis.2 We are also wary of the consequences of PAHO’s broad reading of Nyambal, which seemingly has no limit. PAHO argues that immediate review is necessary because “subjecting a foreign sovereign to litigation burdens”—any litigation burdens—necessarily “triggers the . . . third prong[] of the collateral-order doctrine.” Appellant’s Brief 23 (quoting P&ID, 962 F.3d at 582). Taken literally, that approach would 2 Because Nyambal is distinguishable, we need not resolve whether Jam v. International Finance Corp., 586 U.S. 199 (2019), effectively overruled Nyambal, see Appellee’s Brief 11, 26, or whether Nyambal conflicts with prior circuit precedent, see Dodge et al. Amicus Brief 5, 11, 13 (discussing the availability of mandamus in our cases Papandreou, 139 F.3d 247, and Beecham v. Socialist People’s Libyan Arab Jamahiriya, 424 F.3d 1109 (D.C. Cir. 2005), to address improper jurisdictional discovery). 15 suggest not just that every jurisdictional discovery order is immediately appealable, but also that any order of any type entered against a potentially immune defendant is immediately appealable unless and until the immunity issue has been finally resolved. After all, any order imposing any burden on such a defendant could not be retroactively addressed by later review. But this court has rejected such a sweeping and categorical approach in cases involving sovereign immunity. See Beecham v. Socialist People’s Libyan Arab Jamahiriya, 424 F.3d 1109, 1111 (D.C. Cir. 2005) (dismissing appeal from an order requiring a foreign sovereign to participate in a discovery conference before immunity was resolved). The Cohen analysis is sensitive to the risk that “the blunt, categorical instrument of § 1291 collateral order appeal” will “unduly delay the resolution of district court litigation and needlessly burden the courts of appeals.” Mohawk, 558 U.S. at 112. That risk is certainly present here. Defendants claiming IOIA or FSIA immunity are already guaranteed a path to appellate review of a challenge to the court’s subject matter jurisdiction. The holding PAHO requests—that all jurisdictional discovery orders are appealable for putatively immune defendants—would allow seriatim interlocutory appeals and substantial delay over every interrogatory, document request, or deposition. This case illustrates the problem: Nearly eight years have passed, and the question of immunity has yet to be resolved. PAHO seeks further “piecemeal, prejudgment appeals,” which “undermine[] efficient judicial administration and encroach[] upon the prerogatives of district court judges, who play a special role in 16 managing ongoing litigation.” Id. at 106 (quotation marks omitted). We reject that request.3 Our approach does create the possibility that potentially immune defendants will be required to comply with some jurisdictional discovery orders that might be pared back if this court engaged in direct review of all such orders. But the possibility that an error may be “imperfectly reparable” in a later appeal does not inescapably require immediate review. Digital Equipment, 511 U.S. at 872. Nor are potentially immune defendants without recourse against overbroad discovery orders. If a district court orders jurisdictional discovery without first identifying well-pleaded allegations that an exception to immunity applies, leading to the type of fishing expedition at issue in Nyambal, immediate review is available. See Nyambal, 772 F.3d at 280. Further, in appropriate circumstances, a defendant “may ask the district court to certify, and the court of appeals to accept, an interlocutory appeal pursuant to 28 U.S.C. § 1292(b).” Mohawk, 558 U.S. at 110. And mandamus can provide relief 3 The Tenth Circuit has adopted a similar approach. See Hansen v. PT Bank Negara Indonesia (Persero), TBK, 601 F.3d 1059 (10th Cir. 2010). In Hansen, the Tenth Circuit concluded that it had jurisdiction over an appeal from a jurisdictional discovery order “only if the district court’s order did not adequately limit permissible discovery to the question of . . . immunity.” Id. at 1064. Because the defendant had not demonstrated discovery was unrelated to factual issues bearing on jurisdictional immunity, the order was not appealable. Id.; see also Maxey ex rel. Maxey v. Fulton, 890 F.2d 279, 283 (10th Cir. 1989) (discovery orders “narrowly tailored” to facts bearing on qualified immunity are not appealable); Asante-Chioke v. Dowdle, 103 F.4th 1126, 1131 (5th Cir. 2024) (same); 15A Wright & Miller’s Federal Practice & Procedure § 3914.10.5 (3d ed. 2026) (“For the most part . . . the strong policies against discovery appeals overcomes the theory of immunity appeals.”). 17 from particularly egregious discovery orders. See Papandreou, 139 F.3d at 256. B Applying this analysis, we lack jurisdiction over PAHO’s appeal. The discovery order was issued after a judicial determination that the complaint adequately alleges facts that would trigger the commercial activity exception to IOIA immunity. The district court and this court held that the complaint adequately pleads that PAHO was “moving money for a fee” as a “financial intermediary” for U.S.-involved transactions in the alleged trafficking scheme, and that those allegations would trigger the commercial activity exception. PAHO I, 29 F.4th at 715–17. PAHO then filed a factual motion to dismiss contesting those allegations with its own evidence. And consistent with both the court’s ruling and PAHO’s motion, all of the ordered discovery relates to transactions involving Mais Médicos funds in PAHO bank accounts, the 5% “fee” PAHO collected for its role in the program, and the connection of both to the United States. The district court “approve[d]” three requests for production to investigate “correspondence touching on the use of a U.S. bank to facilitate Mais M[é]dicos.” PAHO II, 2024 WL 4251808, at *3. It authorized another four requests for production along with the request to depose Gerald Anderson, Director of Administration for PAHO, because they were “relevant to Plaintiff’s theory that the administration of the program” through certain “transactions” was “carried on” in Washington, D.C. Id. The court justified three additional requests for production to clarify “terms” in documents “relat[ed] to payments or the movement of money”; to address “the question of how the 5% fee was ultimately used (and specifically whether it was used to administer the program in any way)”; and to probe “the commercial nature of PAHO’s 18 activity.” Id. And it permitted two interrogatories as “relevant” to the theory these activities were “carried-on” in the United States, and, again, to “showing the commercial nature of PAHO’s activity.” Id. The discovery requests, in short, are focused on “corroborat[ing] that the 5% fee ended up in the United States.” Id. at *1. PAHO’s primary argument on this front is that the district court “erroneously ordered discovery on what it expressly recognized was a new theory of jurisdiction—one that neither that court nor this Court ever approved.” Appellant’s Brief 28. We disagree. The district court explicitly stated that it was not ordering discovery related to any new theories plaintiffs raised. See PAHO II, 2024 WL 4251808, at *2. And we are unpersuaded by PAHO’s effort to show that the discovery is in fact unrelated to the “carried-on” theory to which the district court directed its order. Take for example Interrogatory 1, which requests that PAHO “identify all people . . . who participated from the United States in the planning, implementation, management, and oversight of [Mais Médicos] operations and finances.” Id. at *3. PAHO insists this interrogatory has “nothing to do with the alleged use of U.S. bank accounts to move money.” Appellant’s Brief 36. But answers to that interrogatory would identify exactly those people likely to have information regarding any “use of U.S. bank accounts to move money.” That is, the request is certainly “relevant” to plaintiffs’ theory. PAHO II, 2024 WL 4251808, at *2. Beyond confirming that the discovery order aims to resolve a factual issue relevant to the theory the court found well-pleaded, we do not engage in a more granular assessment of the propriety of this discovery. Our determination that the discovery targets that theory confirms that immediate appeal is inappropriate. 19 Further, there is no plausible argument that the district court has in effect denied PAHO’s claim to immunity or is otherwise requiring PAHO to litigate on the merits before resolving that threshold, jurisdictional issue. All agree that this order, at most, indicates more work must be done to decide whether IOIA immunity applies. III The appeal is dismissed for lack of jurisdiction, and the case is remanded to the district court for further proceedings. So ordered.