Photo of Stephen McConnell

Few medical treatments suffer from an image problem as bad as electroconvulsive therapy (ECT).  We remember when it was called electro shock therapy.  We also remember its gruesome depiction in One Flew Over the Cuckoo’s Nest. But modern medicine has redeemed ECT.  It can be a genuinely effective treatment for maladies such as depression, and it is nowhere near as scary as what happened to Jack Nicholson’s character at the end of the Academy Award winning film.  We cannot think of a wittier comedian working today than Gary Gulman (see his bit about the committee that came up with two-letter postal abbreviations for the 50 states), and he has been frank and funny about how ECT helped him deal with debilitating depression.

And yet, apparently, it does not work well for everyone.  In O’Neil v. Somatics, LLC, 2022 WL 4611938 (D. N.H. Sept. 30, 2022), the plaintiff claimed that ECT did not alleviate her depression but, instead, caused her to suffer serious neurocognitive and other injuries.  Permanent memory loss (which is warned of and is well known) was one of the asserted injuries.   The complaint included causes of action for negligence, strict liability, breach of various warranties, violation of the New Hampshire consumer protection statute, and common law fraud.  The defendant moved for summary judgment.  The plaintiff gave up on her warranty claims, but otherwise opposed the motion.

The defendant argued that the negligence claims were barred by both implied and express preemption.  The court‘s discussion of these issues contains some interesting bits — before ending with a judicial whimper.  

To the extent that the plaintiff complained that the defendant violated federal law by failing to report adverse events to the FDA or failing to investigate such reports, Buckman preemption should block those claims.  Buckman preempts FDA-related claims when they would not exist if the FDCA did not exist.  The O’Neil court at first seemed to appreciate the force of this argument, even to the point of holding that the plaintiff’s invocation of New Hampshire law on negligence per se could not escape Buckman.  That negligence per se generally exists does not save FDCA-based negligence per se from preemption because the claim was wholly based on alleged disregard of the FDCA and FDA regulations.  

So far, so good. 

Then the plaintiff endeavored to escape Buckman preemption by contending that the adverse event failures were violations of state law, not just federal law.  And here is where the O’Neil court wilts.  Does New Hampshire law independently require companies to report or investigate adverse events?  The court doesn’t know.  The New Hampshire Supreme Court has not addressed this issue.  Confronted with this situation, most courts would try to make a prediction about the state law.  Erie modesty should tilt the scale against invention of new duties. But the O’Neil court concluded that it could not grant summary judgment because the defendant had “not sufficiently developed the legal theory on which its argument depends.”  Is that appropriate judicial modesty?  Is it paralysis?  Is it laziness?  We know of several judges who would not hesitate to do their own research and analysis if they thought the parties had not done enough work.  More to the point, absent pretty clear proof that New Hampshire would apply state law to require adverse event reporting, the plaintiff’s claim should be dismissed.

Where is Posner when you need him?

The express preemption issue also runs aground on the granite rock of judicial agnosticism.  The defendant premised its express preemption argument on the ECT device’s 2018 down classification from class III to class II.  The court again faults the defendant for insufficient briefing, particularly on the point of how a 2018 down classification can be relevant to the plaintiff’s treatment in 2016.  Again, ‘dunno’ means denial of summary judgment.  But the court misses the point, and it is no excuse for the court if the parties did, too.  At the time of the plaintiff’s use, the device was Class III, and thus at least potentially protected by PMA preemption, since §510(k) is not available for Class III devices.  That should be the end of the story and the end of the claim.

It is not clear whether the O’Neil court did not like preemption, or did not understand preemption, or did not like summary judgment.  

The court did little better on non-preemption issues.  On the failure to warn claim, the O’Neil court at least had the courage and clarity to predict that New Hampshire would adopt the learned intermediary rule.  But beyond that, the decision again disappoints.  The court states that the “fatal flaw” in the defendant’s arguments “is that they turn on material facts that remain in genuine dispute.”  What facts?  The court does not say.  Are we alone in espying something ironic or weird here?  First, the court criticizes parties for inadequately explaining their arguments.  Then the court announces a conclusion without any real explanation.  

The O’Neil court also denied summary judgment on the strict liability liability design defect claim.  The defendant argued that the plaintiff had not adequately set forth the alleged defects. (Are we seeing a theme in the O’Neil case?  Is it like a Jane Austen novel, where nobody says what they mean?). The plaintiff alleged that the ECT device administered a dosage that was arbitrary or too high.  The plaintiff also decried the ECT device’s “allowance for bilateral” application, as opposed to only unilateral.  That sounds more like a practice of medicine issue than a product defect, but it was good enough for the O’Neil court.

There was one area where the O’Neil court’s insistence on detail worked in favor of the defendant:  the consumer fraud and common law fraud claims.  The court held that the fraud claims flunked the specificity requirement of Fed. R. Civ. P. 9(b).  The plaintiff mentioned only one alleged misrepresentation, and it was from 14 years before the plaintiff’s treatment.  Moreover, the complaint failed to allege specific facts evincing scienter or reliance.  Goodbye, fraud.

The O’Neil decision is an example of how not to decide is to decide.  In this case, the refusal to wrestle with some (admittedly not easy) preemption issues led to some bad decisions.  Legally deficient theories live to litigate another day. We cannot decide whether the O’Neil court’s bad decisions are more depressing or shocking.