Photo of Stephen McConnell

The Monday curse continues! Another bad case that makes us feel so disappointed, so cheated, so sad, so … Monday.
We have blogged on the Bartlett case before. We did it here. And here. And here. And here. And that’s not all. It’s been the gift that keeps on giving. Though, as was the case over the holidays, some of those gifts were items we’d just as soon return. Make no mistake about it, Judge LaPlante works hard and writes lengthy, scholarly opinions. He seems thoughtful and diligent. Some of what he wrote we liked a lot. Some — not so much. Being thoughtful and diligent doesn’t always make you right. (And sometimes one can be thoughtless and lazy and still be utterly right. Every day we bank on that possibility.)
In the Bartlett case the plaintiff suffered truly horrific injuries, claiming they were the result of a generic anti-inflammatory drug called Sulindac. The details are in our earlier posts, so let’s get right to the bottom line: the jury returned a verdict of $21 million. The defendant filed a motion for judgment as a matter of law and, in the alternative for a new trial. Last week, Judge LaPlante denied those motions. Bartlett v. Mutual Pharmaceutical Co., 2010 WL 5494053 (D. N.H. Jan. 5, 2010).
The coverage we’ve seen so far of this decision emphasizes the defendant’s claims that the trial was unfair because the plaintiff lawyer inserted all sorts of prejudicial stuff (including putting up on the screen a precluded question about the company’s net sales — supposedly an accident due to a computer “glitch”) as well as the defendant’s claim that the judge seemed biased because he said a few things that defendant construed as rooting for the plaintiff, and also appeared to get teary-eyed during emotional testimony (the Judge told the jury the next day that it was only an allergy).
We have some familiarity with at least one of the plaintiff’s lawyers in the case and he is — how do we put this? — aggressive. When we read the allegations of inflammatory conduct by plaintiff counsel we were nodding our heads. For example, “Bartlett’s counsel mentioned their own experience in the United States military, which obviously was not relevant.” 2010 WL 5494053 at *30. Let’s just say we could have predicted that tactic, as well as much else that happened. None of it is spontaneous. It’s in the playbook. Even the judge stated that “Bartlett’s counsel made this court’s job at trial more difficult than it needed to be, repeatedly testing the limits of this court’s rulings (not to mention it’s patience).” Id. at *25. So the judge issued limiting instructions, etc. But the fact is that the plaintiff’s tactics paid off. We don’t attribute that to any bias on the judge’s part. But something went wrong.
Being the DDL nerds we are though, we are more interested in some of the thorny legal issues. At first blush, it looks like Judge LaPlante took care of most of those thorny issues nicely. He dismissed every one of plaintiff’s claim save one: design defect. That means the judge dismissed claims for failure to warn and negligence. But the problem is that failure to warn never really left the case entirely. That’s a big deal, because failure to warn theories have a lot of jury appeal for plaintiffs. People might get a design wrong or might mess up manufacturing, but failing to warn starts to feel like lying (that’s certainly the way any plaintiff lawyer will couch it). It would be thoroughly understandable if Mutual thought it crucial to keep the warning theory out of the case. Sadly, New Hampshire law, according to Judge LaPlante, and earlier according to a First Circuit case called Brochu (but never according to any New Hampshire state court – a problem in and of itself), put defendant in a bind. With only the design defect claim left, the obvious defense is that no safer design is possible. But the judge held that this was an affirmative defense, and that defendant would need to prove the drug is unavoidably unsafe and had an adequate safety warning. You say goodbye, and I say hello. Hello-Goodbye warning. Or is it Goodbye-Hello?
Mutual decided to withdraw what the judge called the “comment k defense.” And yet the judge still ruled “that warning-related evidence would be admissible, to a limited extent, with regard to the issue of whether Sulindac was unreasonably dangerous.” Id. at *15. A bad deal got way, way worse. That “to a limited extent” isn’t so limited. Or should one be grateful the verdict was only $21 million and not $21 billion? The judge instructed the jury that “Mutual’s ‘conduct in … responding’ to Sulindac’s safety risks, which included any failure to change its warning, was ‘not relevant to this case, and you should put [it] out of your mind.” Id. at *20. Good luck. Here is what the opinion says about that: “Mutual protests that the warning was still a part of the case, which is true, but not in the sense that Mutual could have been held liable for failing to change it.” Id. No, only in the sense that the allegedly inadequate warning shows that the product was defective. Three-card monte, anyone? The judge can say that defendant chose to waive its comment k defense and that, having made its bed, it should lie in it, but it looks very much as if somebody else short-sheeted the bed.
Step back for a moment. It simply doesn’t make any sense for a jury to be in the business of deciding that an FDA-approved drug is unreasonably dangerous. Call it preemption or call it common sense, but it seems crazy to allow a jury to second-guess the FDA’s basic approval of the product, especially in the context of a trial with a very sympathetic plaintiff and a very aggressive plaintiff lawyer. We’re pleased as punch to dwell, and mostly practice, in a jurisdiction that simply doesn’t allow design defect claims against an FDA-approved drug – and even if it did would require an alternative feasible design – not that the product wasn’t aspirin, or some other different product. Nowhere in the opinion resides a satisfactory explanation why that rule should not apply here. The opinion cites Wyeth v. Levine, and only grudgingly notes that it was a failure-to-warn, not a design defect claim. Id. at *19. And here comes some unfortunate sophistry: “Mutual was not held liable for failing to change Sulindac’s design; it was held liable for selling an unreasonably dangerous product, with greater risks than benefits. Federal law did not require Mutual to sell Sulindac.” Id. Yikes. That’s how we get around conflict preemption? You aren’t required to sell the drug?! The FDA said “yes,” but you don’t have to market? That sounds a whole lot like saying “no.”
Maybe it wasn’t sophistry; maybe it was honesty. We can’t help but think that’s what the trial ended up being about — whether Sulindac, a drug that had been FDA-approved for 32 years, ever should have been sold. It turns out that if you look at the press coverage of the jury verdict, that’s exactly what plaintiff’s counsel said: the drug never should have been sold. But the policy question, which is (or should be) answered by straightforward preemption principles, is whether a jury, rather than the FDA, gets to make that decision.
There are many reasons why the answer to that question should be No, but let’s end with one more example from the Bartlett case. The court permitted plaintiff to introduce evidence of adverse event reports. The court concluded that if hotshot experts thought it okay to rely on AER’s, and if the FDA itself uses AER’s, then there’s no reason why the jury shouldn’t be able to consider AER’s. 2010 WL 5494053 at *10-11. Now we’ve beaten to death why AER’s should not be used in a civil trial – why the FDA can act in the public interest on a lower quantum of evidence than a jury must have to give the defendant’s money to the plaintiff. We’ve put together an Adverse Event Report Cheat Sheet filled with cases laying out compelling reasons for excluding AER’s. We also include links to FDA documents that make clear that AER’s are low-quality, biased (sometimes litigation-generated), limited, and may not be used to calculate incidences or estimates of drug risk. AER’s are not the stuff of good science. Consequently it’s utterly misleading for plaintiffs to trot out AER evidence, and then wrap them in the alleged bona fides of FDA use. And it’s grotesquely misleading for anybody to say in this case that 90% of adverse events aren’t reported. Not this kind of adverse event. To say it’s misleading is to say that juries are, inevitably, misled. And that brings us back to the fundamental point: jurors will too often arrive at the wrong answer to a question they should never have been asked.
We expect we’re not finished with Bartlett yet. Let’s see what the First Circuit does with it.