Photo of Eric Hudson

This is my first post as a new member of the Drug & Device Law team.  Like many of you, I’ve been a consistent reader of the blog for years and I rely on it regularly.  I remember talking with Jim Beck and Mark Herrmann many years ago when the blog was just getting started (pretty sure it was 2006).  Now here we are.  I’m a partner with Butler Snow LLP’s Pharmaceutical, Medical Device and Healthcare practice, and I’ve focused my practice on mass torts for almost my entire career.  I’m excited about contributing to the blog and welcome any comments. Now on to business. 

MDL discovery is inevitably burdensome on defendants, and the challenges defendants face in attempting to limit the scope of MDL discovery are exponentially more difficult than in single plaintiff cases. Throw a few hundred cases together in an MDL, and courts seem much more willing to view the scope of permissible discovery through a vastly broader lens.  So we read with interest decisions that limit discovery, place some of the burden on the plaintiffs, or—even better—implement cost shifting and require the plaintiffs to pay for some or all of what they seek in discovery.

We previously blogged about the Tasigna MDL here and were pleased to note the court’s willingness to require plaintiff side social media discovery.  So we thought it worth looking back at In re Tasigna (Nilotinib) Products Liability Litigation, 2023 WL 3563615 (Mag. M.D. Fla. March 31, 2023).  The Tasigna decision addresses the defendant’s request that plaintiffs pay the costs of anonymizing clinical trial data prior to production. Plaintiffs claimed they needed the extensive clinical trial data so one of their statistician experts could utilize the data for signal analysis. The defendant opposed the production on numerous grounds, but the court permitted it.  The defendant then moved for a protective order requesting that the plaintiffs pay the costs of anonymizing the data, and the court denied that request without prejudice – finding it premature and noting that the utility of the data had yet to be determined.  So far so good.

After the defendant incurred costs of approximately $335,000 in hiring a third party vendor to anonymize the clinical trial data and making the production, it moved to allocate those costs to the plaintiffs.  The defendant’s two main points for cost shifting were: (1) plaintiffs premised their request for the clinical trial data on the need for one of their experts to conduct signal analyses, but plaintiffs withdrew that expert before producing a report, and (2) plaintiffs didn’t pose any questions about the data to any witnesses during depositions.  Those points supported the strong argument that, after the significant expense and plaintiffs’ repeated arguments about the importance of the data, the cost, expense, time and effort were for nothing.

Plaintiffs contended that because of the way the anonymization was done (they submitted an affidavit from their statistician asserting that the anonymization of specific dates of adverse events rendered the data useless for signal detection), their expert was unable to use the data.  But since the defendant produced the data in rolling productions over many months, the plaintiffs had ample opportunity to identify any alleged problems with the anonymization.  Plaintiffs claimed they didn’t discover the alleged problem with the anonymization until after they received all the data—meaning they did nothing with it as they received the rolling productions.

Both sides argued that the seven-factor test for cost shifting under Zubulake v. UBS Warburg LLC, 217 F.R.D. 3099 (S.D.N.Y. 2003) supported their positions.  The court opted for a case-specific approach instead of analyzing the seven factors, citing Zubulake’s general direction that “there is no talismanic guidance as to which cases call for cost shifting nor a formula for determining the amount to be allocated.”  Tasigna, 2023 WL 3563615, *2.  The court found fault with the plaintiffs for failing to investigate the utility of the data as it was produced.  But it also faulted both sides for failing to engage in any meaningful meet and confer about the specifics of the anonymization and the costs.  Based on those considerations, the court concluded that:

Because Plaintiffs did not take available steps to determine the suitability of the data or the costs at earlier stages, they bear some responsibility for the unhappy result, sufficient to require that they bear some of the out-of-pocket costs.

Id. The court tempered this finding with its view that the discovery was not sought in “bad faith” and the defendant had control of information about the on-going costs.  Without additional explanation or details about its calculations, the court ordered that the plaintiffs should pay $75,000 of the approximately $335,000 in vendor costs incurred by the defendant (or about 22%).

While the total amount is somewhat disappointing, this decision is a useful tool for supporting cost shifting arguments.  In MDLs, any cost shifting is a victory.