Bartlett v. Mutual, a Stevens-Johnson/Toxic Epidermal Necrolysis (SJS/TEN) generic drug case that produced a $21 million judgment upheld on appeal, gave us writer's block.

Upon first reading of the opinion we thought Bartlett told the following story:

  1. New Hampshire has adopted the controversial third aspect of the risk/utility doctrine so that plaintiff needed only to prove that the risk posed by the drug (sulindac, a non-steroidal anti-inflammatory (NSAID)) outweighed its utility (it's often prescribed for the sort of shoulder pain suffered by plaintiff as well as certain forms of arthritis).
  2. In such cases "liability does not run through the warning label" so Pliva v. Mensing doesn't apply and there is no pre-emption per Wyeth v. Levine.
  3. Mutual decided to put on no defense at all - not even to prove that its product had some utility.
  4. Mutual got hammered.
  5. Mutual thereafter sought to escape the consequence of its trial tactic with a kitchen sink appeal.

So there was lots to blog about. Can a jury, presented with only one salient data point, a woman devastated by SJS/TEN, do a better job of judging whether sulindac is safe and effective than the FDA? What other states allow jurors to decide whether a product should be marketed at all so that Pliva can be avoided in generic drug cases? Under what circumstances should a defendant consider placing all its chips on plaintiff's burden of proof? Why are the damage awards in SJS/TEN cases so high? Should appeals be rifle shots or shotgun blasts? But something about the narrative just didn't make sense. We couldn't figure out what it was so we thought maybe we could write something about the peripheral issues related to Bartlett.

For example, the court recognized that sulindac "has a minuscule risk of causing SJS/TEN" yet found that fact of no moment as Mutual hadn't proved its utility. So we could write about whether there ought to be liability for de minimis risks when we already live in a world of inevitable risks. Then there's the question of causation. Plaintiff had less than 100 cases to point to over a quarter century and nothing to tie them to sulindac other than that prescription preceded onset. Or we could write about the irony of holding that it's sensible to let a jury decide that sulindac should not be sold in New Hampshire at a time when sulindac is generating lots of excitement about its ability to fight cancer of the ovary, breast and colon while preventing new cases via anti-inflammatory / pro-apoptotic mechanisms. We even considered musing about why there seem to be more more SJS/TEN cases and who's responsible for the hypersensitivity reaction since it's starting to look as though the blame falls on early gut microbiome disruption precipitated by antibiotics.

But our minds kept wandering back to the opinion. Why doesn't it make sense? Exactly what is wrong with Bartlett? Finally we remembered Irwin Mainway and his Bag O'Glass.

What would be the reason for talking about a product's warning when the claim is that no warning could ever render it more useful than risky? I can't think of a good one. Nevertheless, the trial court, on the day before trial, "ruled that warning-related evidence would be admissible, to a limited extent, with regard to the issue of whether sulindac was unreasonably dangerous." Hmmmm. Isn't that reasoning circular? Think about it: the label is inadequate because the product is unreasonably dangerous and the product is unreasonably dangerous because the label is inadequate. That's what the trial and appellate court are saying. That's how Bartlett turned an untenable marketing defect claim into a design defect claim against which Mutual had no real defense.

Here's how it worked. Originally plaintiff claimed that the sulindac warning was defective because the label didn't mention SJS/TEN. However, the warning label did say: "Hypersensitivity. Rarely, fever and other evidence of hypersensitivity (see ADVERSE REACTIONS) including ... severe skin reactions have occurred during therapy with sulindac. Fatalaties have occurred in these patients." Furthermore, according to the trial court "[t]he label expressly listed SJS/TEN as potential adverse reactions in its 'Adverse Reactions' section." Things got worse for plaintiff once the treating doctor was deposed.

Bartlett's treating physician "admitted at his deposition that he never reviewed Mutual's sulindac label before treating Bartlett and that 'nothing about it influenced [his] prescribing of the drug' or what he told Bartlett about it." Because '[c]ause-in-fact requires the plaintiff to show that the injury would not have occurred but for the negligent conduct" which in turn, "requires the plaintiff to establish that the negligent conduct was a substantial factor in bringing about the harm ...  Bartlett 'must prove that had the learned intermediary [i.e., her doctor] been warned adequately, the drug would not have been used, or would have been used differently.' 5 Louis R. Frumer & Melvin I. Friedman, Products Liability § 50.05[4], at 50-84 (2010)." Since the label was never read its contents could not have affected the outcome. The court thus held "Mutual is therefore entitled to summary judgment on this issue."

So Bartlett needed a cause of action that would allow her to bypass the warning, but it had to sound in strict liability because she couldn't prove proximate cause either. Since molecules can't be redesigned and since the sulindac had not been mismanufactured the only avenue left was marketing defect where the product of proximate cause's foreseeability inquiry would necessarily manifest itself as a  warning - which led right back to the problem that the label had never been read. With some help from the trial court she seized upon the idea that some products, like Bag O'Glass, "are too risky to even be marketed." But how could she prove it was too risky to ever be marketed?

Bartlett used the sulindac warning, a warning that was neither the proximate nor even the "but for" cause of her injury, to prove that sulindac was too risky. She said what made it too risky was that there was no reference to SJS/TEN in the warning section of the label. And by what authority did she claim to be able to do that? Restatement (Second) section 402A's comment k, added to ensure that negligence principles would be applied in claims over unavoidably unsafe drugs. She argued that risky drugs need "proper directions and warning" and thus drugs lacking them are necessarily too risky! Then she turned around and argued that too risky drugs are ones that should never be marketed because no warning can sufficiently modulate their risk. She thereby avoided the impact of the label's never having been read and Pliva all at once. Better yet, the "comment k defense" was now turned into a trap for Mutual so that almost any evidence it put on would support Bartlett's circular claim. When Mutual decided not to set a foot in the snare the court blamed Mutual's loss on trial tactics gone awry. What a nightmare.

Well, that's that. Now we know what was bugging us - bootstrapping.

Bag O'Glass was too risky because of its innate properties and not because the warning, "Kid! Be Careful. Broken Glass!", made it riskier. Irwin Mainway's argument that a warning could alter the risk posed by a too risky product elicited derisive laughter from the SNL audience. In the 1st Circuit the argument elicited an affirmation of the conclusion derived from it.