People have long been fascinated by robots. Way before the term was coined in a 1920 play or Isaac Asimov popularized it, there were stories about machines that acted like living things. The droids of Star Wars universe are famed for the likeability and pluck. However, there is still the specter that some of those droids, and certainly the cyborgs of the Terminator series or androids of various works of fiction, would turn on their creators. Thus, robots can be scary. Not as scary as clowns, of course, but still scary.
In defending drug and device product liability litigation, recalls are scary. Judges and juries have been known to tilt the playing field in favor of the plaintiff when the medical product at issue has been recalled. Fine issues of defect and causation can be glossed over when the recalled product gets tagged as “bad.”
It looks like that is a big part of what was going on with the summary judgment ruling in Mendoza v. Intuitive Surgical, Inc., No. 18-CV-06414-LHK, 2020 WL 3078178 (N.D. Cal. June 10, 2020). This case involved an injury allegedly sustained in connection with a hysterectomy performed with a recalled model of an instrument used with a robotic surgical device. The instrument, Hot Shears Monopolar Curved Scissors (called “MCS” in the decision instead of the catchier “Hot Shears”), was used to lyse adhesions and cauterize wounds in surgeries performed with the robot. (Think of R2D2’s zapper more than a blaster IG-11 might use in The Mandalorian.) Soon after her procedure, which involved lysing extensive adhesions with the MCS, the plaintiff had complications requiring a bowel resection. About eighteen months later, the MCS was recalled because of the risk of potential micro-cracks that could cause unintended burns from leaked electrosurgical energy. Id. at *2. More than five years after the recall, plaintiff sued the company that manufactured the robot and the MCS. We are not sure why the two-year statute of limitations in either Indiana or California was not a bar, but we do see the impact of Supreme Court personal jurisdiction decisions here. The Indiana resident, whose surgery, injury, and care occurred in her home state, sued the defendant in federal court in its home state. After some wrangling on pleadings and about a year of discovery, the case was teed up for summary judgment.
Along the way, plaintiff came forward with two experts and defendant’s motion to exclude them on Daubert was denied. Plaintiff asserted her claims under California law—there was no choice of law discussion at all—and proceeded on fairly standard negligence and strict liability claims focused on alleged defects in the MCS and a tip cap accessory (“TCA”) used with it. Plaintiff did not contest summary judgment on any claims based on the TCA or conduct related to it, so we can dispense with that part of the decision. The court started with a cursory discussion of summary judgment standards, including the burden standards from Celotex. Id. That was about it for burden in the decision, which should be seen as foreshadowing, like when a robot starts to show signs of emerging self-awareness well before overt rebellion against the human overlords. As the court started to walk through the evidence and argument on the three species of possible defect, this is where the bias against recalled products—if not robots—started to creep into the analysis. We do not want to sound like defense hacks who think the manufacturing of medical devices should always win, but we do think a summary judgment analysis should be rigorous, with testing of plaintiff’s evidence in support of her claims, even if the device was recalled.
As to manufacturing defect, the court recognized that California law requires proof that the plaintiff’s device “differs from the manufacturer’s intended result or from other ostensibly identical units of the same product line.” Id. at *4 (citation omitted). The purported defect was micro-cracks on the MCS used in plaintiff’s surgery and plaintiff conceded that he had no direct evidence of it. While the court agreed with plaintiff that circumstantial evidence could be used to show a manufacturing defect under California law, there was at least one hole in the court’s analysis of what evidence plaintiff had. Plaintiff pointed to one of her experts, who opined that she had determined through “differential diagnosis” that there must have been a “thermal injury to the bowel as a result of the da Vinci System,” and the court pointed out that micro-cracks can only be seen under a microscope. Id. However, that does not add up to sufficient proof to carry plaintiff’s burden to show a manufacturing defect in the MCS that hurt plaintiff, even if we ignored the issues with the use of a “differential diagnosis” in this fashion and the court’s failure to consider burden. Where was the proof, circumstantial or otherwise, that plaintiff’s MCS was outside of specifications? How can an opinion about a “thermal injury to the bowel as a result of the da Vinci system” suffice to show the assumed manufacturing defect caused an injury when the MCS is designed to produce heat (and burns to plaintiff’s bowel was not documented)? The answer to these questions may be in the court’s recitation that “it is undisputed that the MCS model used in Plaintiff’s surgery was later subject to a recall due to concerns about a manufacturing defect.” Id. at *5. It may be that the court did not know the difference between a manufacturing defect and a design defect. It may also be that the court did not think that matters for a recalled device.
The court’s analysis on design defect was even more conclusory. Basically, plaintiff had experts who said something about design defect and Daubert motions as to them had been denied previously. That was enough, but along the way the court did provide something useful. The court endorsed that California applies comment k to the Restatement (Second) of Torts applies to all prescription medical devices. Id. This implicitly rejected the cases that limit comment k treatment to implantable medical devices, which the robot and MCS are not.
The conclusion that strict liability did not apply meant that negligence did and plaintiff had to show a failure to warn to recover for a design defect. Rather than address plaintiff’s evidence on those issues, the court merely concluded in a footnote that summary judgment for negligence was inappropriate “necessarily” because the court had “rejected Defendant’s argument that evidence of a defect in the MCS is lacking.” Id. at *6 n.4. As to showing a failure to warn, the court apparently wrapped that into its consideration of whether Defendant had disproven a failure to warn.
On that issue, the court found a question of fact, citing no record evidence (e.g., an expert saying the missing warnings were required at the time of plaintiff’s surgery) but pointing to plaintiff’s assertion in a brief that she did contend warnings should have been given. Turning burden on its head, the court stated “To the extent that Defendant argues that no such warning was in fact required, Defendant raises a question of fact for the jury.” Id. at *6. As to proximate cause for failure to warn, it should have been a slam dunk because the prescribing physician was not deposed. Yet, the court rejected the argument with a call to “a plaintiff may carry his burden of proving causation by indirect evidence” without identifying any evidence that plaintiff had in the record to support that a different warning about the MCS would have changed anything in plaintiff’s case. Id. That is not how summary judgment is supposed to work, at least since Celotex and the rest of the Supreme Court’s 1986 trilogy on the subject.
The court also mangled its consideration of causation, which it dealt with as a single requirement instead of something separately required for each claim (even though there had been scattered mentions of causation along the way). Plaintiff was required to come forward with evidence that a manufacturing defect, design defect, and/or warnings defect proximately caused her injuries. The court did not attempt to test the sufficiency of plaintiff’s evidence on those issues. Instead, it merely looked to what one of plaintiff’s experts opined about medical causation (with the “system” as opposed to the MCS or a defect in it) and the possibility of an undocumented thermal injury. Id. at *7. That the court found genuine issues of material fact about whether there was a thermal injury and whether plaintiff had an expert to say thermal injury might have caused plaintiff’s injury does not matter much, because those were not the right questions to ask. We could try to bring that back to robots, but what could possibly go wrong if the right questions are not asked by those programming robots?