The Avandia MDL has been a long, winding, and ultimately meritless road.  The FDA approved the drug to treat Type II diabetes in 1999, and the MDL got going in 2007, after a widely publicized, but ultimately disproven meta-analysis purported to show that Avandia presented an increased risk of heart attacks. 

That was 18 years ago.  In that time, the FDA asked the manufacturer to add a black box warning on myocardial ischemia—only to remove it six years later when it determined that the science did not support it.  The FDA likewise imposed a REMS program that significantly restricted access to the product—only to end the program after long-term study results dispelled any increased risk.  In the meantime, the parties and the courts have expended a tremendous amount of public and private resources, and patients were deprived of a therapeutic option that undoubtedly would have benefited many. 

All because of a cardiac risk that the FDA has determined did not exist.  Heck, just last November, Bexis published a blogpost entitled, “Avandia Litigation – Is This Finally the End?” 

Well, it is not quite the end.  At least not yet.  The personal injury claims have long been resolved—either settled or dismissed (for example, here, here, and here).  But third party payers still have RICO and consumer fraud claims asserting that they paid for more Avandia prescriptions than they would have had the manufacturer accurately disclosed information on cardiac risks in real time.  We have serious doubts about the merits of these claims:  The payers got exactly what they paid for, and they do not allege that any patient was actually harmed.  Moreover, any price impact would have been dwarfed by the FDA’s requirement, and then removal, of the ultimately unnecessary boxed warning. 

Regardless, the district court has now certified a TPP class and is allowing them to proceed in a collective manner. In In re Avandia Marketing, Sales Practices, and Products Liability Litigation, No. 07-md-1871, 2025 U.S. Dist. Lexis 97465 (E.D. Pa. May 22, 2025), the district court certified a class of TPPs who purchased Avandia from January 1, 2005 through August 14, 2007.  The TPPs’ theory of liability has shifted over time, but it appears they have settled on arguing that the manufacturer marketed the drug as having better cardiovascular outcomes, but had data showing that was not true as early as 2005.  The truth allegedly came out, according to the TPPs, when the aforementioned meta-analysis was published in 2007.  (Although the court does not explain it, these allegations appear to frame the beginning and end of the class period.) 

In urging class certification, the TPPs argued that the manufacturer’s “marketing fraud” was a standardized campaign directed at the entire healthcare community with “market-wide impact,” thus making it susceptible to class-wide proof.  The manufacturer disagreed, mainly on the ground that that leads to denial of most TPP economic harm class actions—causation.  We see the manufacturer’s point.  To link the alleged wrongdoing to the alleged harm, the TPPs have to prove that the manufacturer misrepresented the benefits of Avandia, that a prescribing physician relied on those representations in prescribing Avandia, that the prescriber would not have prescribed the product but for the representations, and the prescriber would instead have prescribed a cheaper medicine or nothing at all.  These elements require individualized proof, especially reliance.  There are, after all, many reasons why physicians prescribed Avandia, and they may or may not have even seen or heard the alleged “marketing,” let alone relied on it.

Despite this, the court found that the TPPs could prove causation on a class-wide basis.  First, the court determined that it could infer reliance on the manufacturer’s marketing, and it distinguished cases holding otherwise on the basis that “every provider’s goal when considering whether to prescribe Avandia is largely the same—to treat a patient’s diabetes without otherwise causing harm.”  Id. at *20.  Based on this (oversimplified) view of prescribing decisions, the court concluded that “even though the decision to prescribe Avandia is not ‘one-dimensional,’ it is not so subjective that a provider’s reliance cannot be inferred.”  Id. at *20-*21. 

Second, the court found that the TPPs could prove class-wide reliance through statistical econometric models, even though the court already excluded the plaintiffs’ expert’s regression analysis purporting to show causation.  In its place, the court accepted the plaintiffs’ offer of the manufacturer’s own internal studies of how its marketing impacted sales and other “generalized” proof, such as “papers, internal corporate studies, and communications.”  Id. at *21-*22. 

The court therefore found that common issues predominated over the element of reliance.  That is not the correct outcome.  It was undisputed that providers prescribed Avandia for many reasons, and the manufacturer submitted testimony from individual providers contesting that they relied on the manufacturer’s marketing.  The court, however, ruled that this individualized evidence was minimal by comparison and that it would not engage in conjecture on what other evidence the manufacturer would be able to marshal at trial. 

The manufacturer offered other reasons why individual issues would swamp common issues.  RICO claims require proof of concrete financial loses, yet there were multiple scenarios under which putative class members suffered no aggregate loss at all.  Determining this element—which relates to liability, and not only damages—could be determined only be examining each class member.  The court, however, ruled that each TPP’s harm occurred at the time of each Avandia purchase, which means that any TPP who paid for even one Avandia prescription suffered an injury.  That logic erases individualized issues, but it clears the way for class-wide recovery for a class whose members may not have lost any money, or may even have come out ahead.  Indeed, the plaintiffs’ own expert concluded that 26 percent to 33 percent of TPPs in one dataset had “zero or negative damages.”  Id. at *28. 

The manufacturer also cited the plaintiffs’ reliance on oral statements, which by nature are inappropriate for class treatment.  The court rejected this argument too, on the basis that the manufacturers’ marketing messages and tactics did not vary by region and that “the oral component of the fraudulent sales presentations did not vary appreciably” from provider to provider. 

Finally, the court found that the class was ascertainable.  Under the TPPs’ proposal, they would use vendors to compile a list of every TPP, then each potential class member would submit sworn affidavits as claim forms following judgment.  The court accepted this method, relying on Third Circuit authority allowing “some level of verification” during the claims administration process.  The devil, however, is in the details, and when starting with an overly broad list of every TPP, the contemplated affidavits will be much more than mere “verification.” Moreover, the facts asserted will not be subject to cross-examination. The class members would essentially be expected to prove their claims on an individual basis after the fact, which is not what class actions should be. 

In certifying a TPP class, this order is in the distinct minority, and the difference is that these plaintiffs will be allowed to round over the edges of individualized inquires using “common” proof at trial.  If we are searching for a silver lining, we would point to the limited, two-and-a-half-year class period.  But that’s not saying much given the overall result.  We would not be surprised if the manufacturer seeks an interlocutory appeal, and the manufacturer also has a motion for summary judgment pending.  This may not be our last word on Avandia.