Ozempic MDL Daubert Hearing Closes With 4,000 GLP-1 Cases Riding on Judge Marston's Gatekeeping

The Ozempic MDL Daubert hearing closes today with no ruling and 4,022 cases in the balance

The Eastern District of Pennsylvania reserved nine days — 10 to 18 September 2026 — for the Rule 702 hearing in MDL No. 3094, and that window closes today with the question under advisement rather than decided from the bench. The Ozempic MDL Daubert hearing asks Judge Karen Spencer Marston to decide whether the experts who would tell a jury that semaglutide and tirzepatide are capable of causing gastroparesis, ileus and intestinal obstruction may testify at all; roughly 4,022 personal injury actions were pending in the Multidistrict Litigation (MDL) as of 1 September 2026, and substantially all of them depend on the answer.

That is an unusual amount of litigation to hang on an evidentiary motion, and it is worth being precise about why. This is not a Bellwether Trial. No jury has been impaneled in MDL 3094 and none is scheduled. What happened over the last nine days is the gatekeeping stage: argument and expert examination on whether the plaintiffs' General Causation opinions satisfy Federal Rule of Evidence 702 as amended in December 2023. If they do, the litigation moves toward Specific Causation, bellwether selection and the ordinary pressure of trial dates. If they do not, the MDL does not shrink — it ends, in the way the Zantac MDL ended, through summary judgment entered against a plaintiff population that can no longer prove the first element of its case.

What is the Ozempic MDL Daubert hearing, and what was argued from 10 to 18 September?

MDL No. 3094 is captioned In re: Glucagon-Like Peptide-1 Receptor Agonists (GLP-1 RAs) Products Liability Litigation. The Judicial Panel on Multidistrict Litigation (JPML) centralized it in the Eastern District of Pennsylvania on 2 February 2024, transferring an initial 55 actions. The defendants are Novo Nordisk, which markets semaglutide as Ozempic, Wegovy and Rybelsus, and Eli Lilly, which markets tirzepatide as Mounjaro and dulaglutide as Trulicity. The claims are Failure to Warn claims: plaintiffs say the labels did not adequately disclose the risk of severe, sometimes persistent gastrointestinal injury, and that the manufacturers marketed the drugs — particularly for weight loss — without conveying it.

The MDL was originally assigned to Judge Gene E.K. Pratter, who died on 17 May 2024; the JPML selected Judge Marston to take over the pretrial proceedings on 6 June 2024. Judge Marston's case management approach has been to refuse to run the litigation on parallel tracks. Rather than let general and specific causation, diagnosis, warnings and preemption develop together across thousands of individual files, she identified three cross-cutting issues capable of disposing of the whole docket and ordered them resolved first.

The Ozempic MDL Daubert hearing is where the second and third of those issues met the record. Rule 702 motions were due on 28 April 2026, after expert depositions closed earlier that month, and the court set aside 10 to 18 September 2026 to hear them alongside the associated summary judgment arguments on warning adequacy and Preemption (Products). In substance, three questions were on the table across those nine days:

• General causation. Can the plaintiffs' epidemiologists, gastroenterologists and pharmacologists reliably conclude that GLP-1 receptor agonists are capable of causing gastroparesis, ileus and bowel obstruction in the general population, using methods the field accepts and applying them to the facts of this case?
• Warning adequacy. Were the labels in force at the relevant times adequate as a matter of law, given what the manufacturers knew?
• Preemption. Even if a state-law duty to warn differently existed, does federal drug regulation displace it, because there was no newly acquired information that would have let the manufacturers change the label unilaterally?

Those are separable questions with separable consequences, and a practitioner reading the eventual opinion should expect them to be answered separately. A plaintiff population can survive Rule 702 and still lose on preemption; it can beat preemption and still lose on Rule 702. The commentary that treats the hearing as a single up-or-down vote on "the Ozempic lawsuits" is compressing three distinct rulings into one.

How MDL 3094 got here: three cross-cutting issues and one already decided

The most important thing to understand about this hearing is that it is the second act. The first cross-cutting issue has already been decided, and it was decided against the plaintiffs.

What did Judge Marston decide about gastroparesis diagnosis in the GLP-1 MDL?

On 15 August 2025 Judge Marston excluded the plaintiffs' diagnostic expert opinions under Rule 702 and held that a plaintiff alleging drug-induced gastroparesis must show a gastric emptying study performed at the time of diagnosis and demonstrating delayed gastric emptying. Three modalities qualify: gastric emptying scintigraphy, a stable isotope breath test, or a wireless motility capsule. A clinical diagnosis resting on symptom presentation alone — nausea, vomiting, early satiety, abdominal pain — will not support the claim, however confidently a treating physician recorded it.

Before that ruling, symptom-based diagnosis was the ordinary route into the litigation, and a large share of the intake that built the docket had been assembled on exactly that basis. The court was explicit about the trade-off, reasoning that it would be more unjust to expose defendants to liability on the strength of an unreliable diagnosis than to bar claims that cannot be objectively substantiated. Defense-side commentary read it as an objective-testing threshold with the practical effect of a screening order; plaintiffs' firms read it as a records problem to be worked through case by case. Both readings are right. The ruling did not dismiss a defined set of cases on a date certain. It changed what a viable Plaintiff Fact Sheet (PFS) has to contain, and it did so after tens of thousands of retainers had already been signed.

Two features of that August 2025 decision matter for predicting what comes next. First, it shows this judge willing to exclude expert testimony in a mass tort where exclusion has immediate docket-wide consequences — she has done it once already in this very litigation. Second, it shows her doing so on methodological grounds rather than on a view of who is more persuasive: the objection was that the diagnostic method did not reliably establish the condition, not that the experts were unconvincing. That is the distinction Rule 702 draws, and it is the distinction that will govern the general causation ruling.

Stage	Date	What happened
JPML centralization	2 February 2024	55 actions transferred to the Eastern District of Pennsylvania; MDL No. 3094 created
Reassignment	6 June 2024	Docket passes from Judge Gene E.K. Pratter, who died on 17 May 2024, to Judge Karen Spencer Marston
Cross-cutting issue one decided	15 August 2025	Plaintiffs' diagnostic opinions excluded under Rule 702; gastric emptying study required at diagnosis
Scheduling order for issues two and three	21 March 2025, amended thereafter	Expert disclosure, deposition and briefing schedule set for causation, warnings and preemption
Expert depositions close	April 2026	General causation experts deposed on both sides
Rule 702 and summary judgment motions	28 April 2026	Cross-motions filed on causation, warning adequacy and preemption
Pending caseload	1 July 2026	3,848 actions pending
Pending caseload	1 September 2026	4,022 actions pending
Ozempic MDL Daubert hearing	10-18 September 2026	Nine-day Rule 702 and summary judgment hearing; closes today, under advisement
Ruling	not scheduled	No deadline binds the court; comparable opinions have taken three to nine months

The Rule 702 standard Judge Marston is applying

The Ozempic MDL Daubert hearing is being argued under a version of Rule 702 that is materially stricter than the one most mass tort practitioners internalized in the two decades after Daubert v. Merrell Dow Pharmaceuticals. The December 2023 amendment made two changes that matter here.

The first is the burden. The amended rule states that the proponent must demonstrate to the court that it is more likely than not that the admissibility requirements are met. That language was added because a substantial body of case law had drifted into treating reliability as a question of weight for the jury, and had admitted expert testimony on a showing considerably thinner than a preponderance. The rule now forecloses that drift on its face.

The second is the application requirement. Rule 702(d) now asks whether the expert's opinion "reflects a reliable application of the principles and methods to the facts of the case." It is no longer sufficient to identify a generally accepted methodology — a Bradford Hill analysis, a weight-of-the-evidence synthesis, a meta-analysis — and assert that it was followed. The court is directed to examine whether it actually was followed, in this case, on this evidence.

For a Daubert Challenge in a pharmaceutical MDL, that translates into a familiar set of pressure points, and the defendants will have pressed all of them across the nine days:

• Whether the epidemiology shows an association strong enough and consistent enough to support a causal inference, or whether the observational studies are confounded by the underlying conditions the drugs treat.
• Whether the experts accounted for known confounders. Type 2 diabetes is itself a leading cause of gastroparesis, and obesity, prior abdominal surgery, connective tissue disease and a range of medications are independently associated with delayed gastric emptying. An expert who does not engage with those alternative explanations is exposed under Rule 702(d), and that precise failing was decisive in the acetaminophen litigation at the district court level.
• Whether the experts' litigation opinions match what they have written and said outside litigation.
• Whether regulatory conclusions cut against them. The FDA added ileus to the postmarketing experience section of the Ozempic label in September 2023 after receiving adverse event reports; a postmarketing listing is not an agency finding of causation, and defendants will read it as the absence of one.

The regulatory record both sides are reading

The label history is short, contested, and central to two of the three issues argued this month. Ozempic was approved for type 2 diabetes in 2017 and Wegovy for chronic weight management in 2021; gastrointestinal adverse reactions have been on the labels throughout, because nausea, vomiting and diarrhea are the most common side effects of the class and were evident in the registration trials. What changed later, and what the litigation is actually about, is the treatment of the severe end of that spectrum. In September 2023 the FDA added ileus to the postmarketing experience section of the Ozempic label following adverse event reports, and comparable entries appear in the prescribing information for Wegovy and for Eli Lilly's tirzepatide product Mounjaro.

Each side reads that sequence as supporting its position, which is why it will appear in the opinion. For the defendants it is the record of a regulator that reviewed the same signal, required a postmarketing listing, and declined to make a causal finding — which supports both the adequacy of the warning and the argument that there was no newly acquired information sufficient to permit a unilateral label change under the changes-being-effected regulation, the predicate for Preemption (Products). For the plaintiffs it is the record of warnings that arrived years after the manufacturers had the underlying reports, in a section of the label that is read less often and carries less weight than a contraindication or a boxed warning, and while the drugs were being marketed to a weight-loss population without diabetes that had no reason to weigh a gastrointestinal risk against a glycemic benefit. The Learned Intermediary Doctrine sharpens the dispute rather than resolving it: the question becomes what a reasonably prudent prescriber would have done with a postmarketing listing, and that is not a question a postmarketing listing answers on its own.

The plaintiffs' answer, in outline, is that general causation does not require a regulatory imprimatur, that mechanism evidence here is unusually direct because delayed gastric emptying is the drugs' intended pharmacological action rather than an inferred side effect, and that Rule 702 asks whether the method is reliable rather than whether the judge finds the conclusion persuasive.

Will the Ozempic lawsuits be dismissed if the plaintiffs' experts are excluded?

Yes — and that is the honest answer, not a hedged one. If Judge Marston excludes the plaintiffs' general causation experts in full, the ordinary consequence is summary judgment for the defendants across the MDL, because General Causation is an element of every claim in it and expert testimony is the only way to prove it. That is what happened in Zantac, and it is why the hearing that closed today is a larger event than most bellwether trials.

The realistic outcomes are wider than that binary, however, and the intermediate ones are more likely than either extreme:

• Full exclusion. All general causation opinions excluded; summary judgment follows; the federal docket is resolved subject to appeal to the Third Circuit. State court filings and unfiled claims remain, but their value collapses.
• Injury-specific exclusion. The court admits opinions on one injury and excludes them on another. Gastroparesis, ileus and bowel obstruction rest on different evidence bases, and a ruling that splits among them is entirely plausible. This is the outcome that would do the most to reshape intake criteria without ending the litigation.
• Partial exclusion by expert or by opinion. Particular witnesses or particular opinions are excluded; the litigation continues on a narrowed expert record, with the defendants better positioned at trial.
• Denial on causation, judgment on preemption. The experts survive, but the Failure to Warn claims fail as preempted. The effect on the docket is the same as exclusion; the reasoning, and the appellate posture, are entirely different.
• Denial across the board. Motions denied, and the litigation moves to specific causation, Plaintiff Fact Sheet (PFS) enforcement and bellwether selection — at which point the settlement conversation begins in earnest for the first time.

There is no basis in the record for putting probabilities on these. What can be said is that the 15 August 2025 ruling establishes that this court will exclude when it thinks the method fails, and that the same ruling has already removed from the docket the weakest diagnostic profiles — which is a double-edged fact, because the cases that remain are, on average, better documented than the ones the intake machine originally produced.

What a general causation ruling has done to other pharma MDLs

Three recent federal pharmaceutical MDLs bracket the range of outcomes, and none of the competing coverage of this hearing sets them side by side.

Litigation	Gatekeeping ruling	Immediate effect	Where it stands
Zantac / ranitidine, MDL No. 2924 (S.D. Fla., Judge Rosenberg)	6 December 2022: all ten of the plaintiffs' general causation experts excluded as methodologically unreliable	Summary judgment for GSK, Pfizer, Sanofi and Boehringer Ingelheim; roughly 50,000 federal claims resolved	Final judgment entered September 2023; appealed to the Eleventh Circuit; the state court inventory, not the MDL, drove the eventual settlements
Acetaminophen and autism/ADHD, MDL No. 3043 (S.D.N.Y., Judge Cote)	December 2023: plaintiffs' general causation experts excluded under amended Rule 702, principally for failure to account for confounders	Federal docket effectively terminated	Vacated and remanded by the Second Circuit on 13 July 2026 in Rutledge v. Walgreen Co.; roughly 550 cases revived
GLP-1 receptor agonists, MDL No. 3094 (E.D. Pa., Judge Marston)	15 August 2025: diagnostic opinions excluded, gastric emptying study required. General causation, warnings and preemption argued 10-18 September 2026	Intake criteria reset; docket continued to grow to 4,022 actions	Under advisement as of 18 September 2026

Two lessons follow. The first is that a general causation exclusion is not a case management event but a terminating one; there is no version of this in which an adverse ruling merely slows the GLP-1 litigation down. The second is that the first ruling is no longer the last word. For roughly two years after the Zantac order, the working assumption in the mass tort bar was that a well-reasoned Rule 702 exclusion in an MDL was close to unreviewable in practice. That assumption is now demonstrably wrong.

How the Second Circuit's decision in Rutledge v. Walgreen changed the stakes

On 13 July 2026 the Second Circuit vacated the acetaminophen exclusion and remanded, in a precedential opinion. The court's stated concern was the line between gatekeeping and deciding: a district court must take a hard look at how an expert applied an accepted method, but it may not resolve a Rule 702 motion by substituting its own judgment of how persuasive the expert is. The panel directed courts to a middle ground between those two propositions, and about 550 cases returned to the docket as a result.

Rutledge is not binding in the Eastern District of Pennsylvania, which sits in the Third Circuit. It matters anyway, for three reasons.

First, it is the most recent federal appellate statement on how the amended Rule 702 works in a mass tort, and it is directly about general causation in a pharmaceutical MDL. A district judge writing an exclusion opinion in September 2026 writes it knowing that opinion will be read against Rutledge on appeal, and writes accordingly — which in practice means a longer opinion, more granular engagement with each expert's actual reasoning, and conspicuous care to characterize every ruling as methodological rather than evaluative.

Second, it changes the settlement arithmetic on the defense side. Before Rutledge, a defendant holding a Rule 702 win in an MDL held something close to a final result. After it, that win carries visible appellate risk, and the discount a defendant should apply when valuing a post-exclusion settlement is correspondingly larger. This is the quiet reason the acetaminophen reversal is relevant to a GLP-1 hearing: it re-priced what a gatekeeping victory is worth.

Third, it changes the arithmetic for the funders. Third-Party Litigation Funding in pharmaceutical Mass Tort portfolios is underwritten on assumptions about how binary and how terminal the Rule 702 stage is. A world in which exclusions are reversed at a non-trivial rate is a world with a longer tail and a higher variance, and portfolios priced on the pre-Rutledge assumption are mispriced in both directions.

Does the Ozempic MDL Daubert hearing affect the NAION vision-loss cases in MDL 3163?

No, and the conflation is widespread enough to be worth correcting directly. A great deal of the coverage published this month describes the 10-18 September window as though it covered the GLP-1 vision-loss claims. It does not.

The NAION claims — alleging that GLP-1 receptor agonists cause non-arteritic anterior ischemic optic neuropathy, an optic nerve injury that can produce sudden and permanent vision loss — were centralized separately. The JPML created MDL No. 3163 by order of 15 December 2025 and sent it to the Eastern District of Pennsylvania, where it was assigned to Judge Marston as well. Same judge, same courthouse, different docket, different schedule, different science, different experts.

	MDL No. 3094	MDL No. 3163
Injuries	Gastroparesis, ileus, intestinal obstruction	Non-arteritic anterior ischemic optic neuropathy (NAION)
Created	2 February 2024	15 December 2025
Judge	Karen Spencer Marston, E.D. Pa.	Karen Spencer Marston, E.D. Pa.
Structure	Three cross-cutting issues frontloaded	Cross-cutting issues frontloaded by case management order of 2 July 2026
Gatekeeping stage	Argued 10-18 September 2026; under advisement	Expert challenges and summary judgment motions due 2 July 2027
Scale	4,022 actions pending at 1 September 2026	Substantially smaller; growing

The structural point is that Judge Marston is running the same playbook twice, roughly a year apart. In MDL 3163 she directed the parties in July 2026 to frontload the questions capable of resolving the whole litigation — general causation and preemption — over the plaintiffs' objection that the sequencing would slow things down, and answered that objection by letting marketing discovery and plaintiff fact sheets proceed in parallel so that bellwether selection can begin quickly if the early motions do not dispose of the case. Whatever she writes in MDL 3094 will therefore be read twice: once as the disposition of the gastrointestinal docket, and once as a preview of the standard the vision-loss docket will face in 2027.

There is a further consequence that has had little attention. The GI ruling will land roughly nine months before the NAION expert deadline. Plaintiffs' experts in MDL 3163 will write their reports with the MDL 3094 opinion in front of them, and will draft to it. That is an advantage the GI plaintiffs did not have, and it is a reason to expect the NAION expert record to be visibly more defensive in its treatment of confounders — age, hypertension, sleep apnea, diabetic vasculopathy and crowded optic disc anatomy — than the record Judge Marston has just spent nine days examining.

What it means for defendants, claimants' counsel, funders and PI intake firms

For the manufacturers and their counsel. The value of a full exclusion here is no longer the value of a final judgment. It is the value of a judgment that must survive the Third Circuit, on an issue the Second Circuit has just spoken to in the other direction. Two practical consequences follow: preemption is worth pressing as hard as causation, because a preemption win rests on a different body of law and does not carry the same appellate exposure; and any settlement structure negotiated after a favorable ruling should be priced with reversal risk in it rather than treated as a cleanup exercise.

For claimants' counsel. The two rulings in this MDL are converging on the same demand, which is documentary. The August 2025 order made an objective test a precondition at the individual level; a general causation opinion, whichever way it goes, will define what the class of provable injuries actually is. Inventories assembled against a symptom-based or advertising-driven intake standard contain a proportion of cases that will not clear either bar, and the cost of carrying them is real. The Statute of Limitations analysis in the states where unfiled claims sit is the other live issue, and it does not pause while a judge deliberates.

For litigation funders. A single-event binary this large is the hardest thing to underwrite in a mass tort portfolio, and the usual mitigation — diversification across dockets — fails here, because the GI and NAION dockets share a judge, a defendant group, a drug class and now a methodological standard. Correlation across those two positions is much higher than a naive portfolio view implies. The Rutledge reversal cuts the other way, lengthening the tail on positions that would previously have been written off at the moment of exclusion.

For personal injury intake firms. GLP-1 advertising spend has been among the heaviest in the sector for two years, and a material share of the resulting signed cases was generated before August 2025 against criteria that no longer qualify. The cases that clear both the diagnostic threshold and whatever survives this hearing are worth materially more per file than the undifferentiated inventory; the ones that clear neither are a cost. Firms that have not re-run their books against the gastric emptying study requirement are carrying an unmeasured liability, and the ruling that follows this hearing will make it measurable.

Frequently asked questions

What happened at the Ozempic MDL Daubert hearing in September 2026?

The Eastern District of Pennsylvania reserved 10 to 18 September 2026 for argument and expert examination on whether the plaintiffs' general causation opinions in MDL 3094 satisfy Federal Rule of Evidence 702, together with the defendants' summary judgment motions on warning adequacy and preemption. The window closes today. Judge Marston did not rule from the bench, and as of 18 September 2026 no decision has issued.

How many Ozempic lawsuits are pending in MDL 3094?

Roughly 4,022 actions were pending in MDL No. 3094 in the Eastern District of Pennsylvania as of 1 September 2026, up from 3,848 on 1 July 2026 and from the 55 actions the JPML centralized on 2 February 2024. Those figures count the federal MDL only; they do not include state court filings or unfiled claims held in firm inventories, which are widely understood to be larger still.

When will Judge Marston rule on the GLP-1 general causation experts?

No deadline binds her, and no ruling date has been set. Comparable general causation opinions in pharmaceutical MDLs have issued between roughly three and nine months after argument — Judge Rosenberg's Zantac opinion ran to several hundred pages — so a decision between December 2026 and mid-2027 is the reasonable planning assumption. An opinion written in the shadow of Rutledge v. Walgreen Co. is likely to sit at the longer end of that range.

What happens next in the Ozempic MDL after the Rule 702 hearing?

If the plaintiffs' experts are excluded in whole, summary judgment for Novo Nordisk and Eli Lilly follows and the federal docket is resolved subject to appeal to the Third Circuit. If they survive, the litigation moves to specific causation, plaintiff fact sheet enforcement and the selection of a bellwether pool — the first point at which the MDL would have a trial date and therefore genuine settlement pressure. A split ruling by injury or by expert would narrow rather than end the litigation, and would reset intake criteria a second time.

Does the ruling affect the GLP-1 vision-loss cases?

No. The NAION claims sit in MDL No. 3163, created by the JPML on 15 December 2025 and assigned to the same judge on its own schedule, with expert challenges and summary judgment motions due 2 July 2027. The GI ruling will shape how those cases are worked up, because it will state the standard, but it does not dispose of them.

Does a Rule 702 loss end an individual plaintiff's case?

In the federal MDL, effectively yes: general causation is an element of every claim, and if no admissible expert can establish it, summary judgment follows regardless of how well documented the individual injury is. What survives an exclusion is the possibility of appeal, claims pending in state courts that are not bound by the federal ruling, and — in principle — a later filing on a new and stronger scientific record, subject to the applicable Statute of Limitations. None of those is a substitute for admissible causation testimony.