August signals different things to different people. When it comes to football, two August events come to mind: 1) NFL preseason games and 2) high school “two-a-day” practices. The former are not like real NFL games, but they are better than nothing, particularly if you like to see how rookies are doing, guess at final rosters, or look for insight into compiling fantasy teams. The latter have lost some of their mysticism due to school restrictions driven by safety concerns. Long gone are the days of no water during practices (see Bear Bryant at Texas A&M in 1954) and the tradition of two practices per day in full pads seems to be disappearing. There are still vomit-inducing gassers and thigh-burning hills. There are also still numerous drills focused on the most basic of football plays, the run up the middle. Good offenses need to be able to get the crucial yard with a dive on third and 1 and good defenses need to be able to stop the offense cold in such situations. This has not changed. There was a time, however, when an end run, a run to the outside, was considered a bit sneaky, particularly if there was an accompanying fake dive. The “end run” has lived on in our vernacular as a less than straightforward way to accomplish a goal.
When it comes to prescription drug product liability cases, the most direct—and we would maintain, appropriate—way to establish liability cases is with a failure to warn claim. The vast majority of jurisdictions recognize that the duty to warn in such cases runs to the prescribing physician (or other health care provider). We have gone on many times about what the learned intermediary doctrine really means. One thing most courts to come down clearly on the issue of proximate cause in prescription drug failure to warn claims have recognized is that a plaintiff bears the burden of establishing that a different warning (at the time, as to the condition plaintiff claims to have developed) would have changed the prescriber’s behavior such that the injury would not have happened. Overwhelmingly, proximate cause in such cases turns on the testimony of the prescribing physician that she likely would or would not have still prescribed the same drug to the plaintiff even if the drug’s label (and other labeling like Dear Doctor Letters) had been changed to what plaintiff contends it should have. We think this is appropriate.
When the prescriber testifies essentially, “no, that different warning you propose, Ms. Plaintiffcounsel, would not have changed my decision to prescribe this drug to my patient, Mr. Plaintiffpatient,” the plaintiffs often go to an end run. Plaintiffs usually testify, with the full benefit of hindsight and perhaps some profit motive, that they never would have agreed to take the medication their trusted doctor prescribed for their debilitating or life-threatening condition if they had known there was a risk of what they are suing over years later. With such testimony in hand, the plaintiff can argue that the focus for proximate cause should be on their own decision making not on that of the learned intermediary to whom the duty to warn runs. The opinion in Gilliland v. Novartis Pharms. Corp., No. 1-12-cv-00029, 2014 U.S. Dist. LEXIS 106734 (S.D. Iowa July 25, 2014), addressed just such an argument. From our perspective, it did so poorly, misunderstanding the logic behind the learned intermediary doctrine and ignoring an essential link in the causation chain.
Gilliland is yet another in a long line of Aredia/Zometa decisions we have discussed here and it also addressed a number of issues we will gloss over. It included a criticism of Rule 56 from an article in the “Green Bag” journal, which has the full name of “The Green Bag: An Entertaining Journal of Law.” Id. at **7-8. It rejected the bulk of a Daubert challenge on specific causation—implausibly precluding the challenged expert’s dose-effect opinion “based on his personal observations” but allowing him to read into the record statements on the same topic from a medical group’s position paper. Id. at **13-24. It had partial summary judgment walkovers on negligent manufacturing and express warranty. Id.at **42-44. It let plaintiff file another brief as to whether she could make out a strict liability manufacturing defect even though Rule 56 required her to have come forward with such evidence already. Id. at **44-47. It also denied partial summary judgment as to implied warranty simply because it had done so as to negligent failure to warn. Id. at **47-49.
For the meat of the opinion, the court recognized that it was deciding an issue of first impression under Iowa law and was supposed to be predicting what the Iowa Supreme Court would do if confronted with the issue. Id. at **31-32. In reaching its prediction, the court did not cite a single decision from the Iowa Supreme Court or any other Iowa state court. The only decision under Iowa law it cited was a federal court saying the “intermediary defense is still viable under Iowa law.” Id. at *25. While how Iowa courts address proximate cause in negligence cases, maybe even negligent failure to warn cases, might have been instructive, it also would have been nice if the cited case under Iowa law had understood the difference between a doctrine (or rule) and a defense. As we have said many times, the burden is on plaintiff to prove that the warning was inadequate and that its inadequacy proximately caused the plaintiff’s injury; there is no burden on the manufacturer to prove its warning was adequate or that a different warning would not have mattered. In Gilliland, it appears that all four oncologists who prescribed Zometa to the plaintiff testified that they would have prescribed the drug to plaintiff even with the warnings she urged for osteonecrosis of the jaw (whatever those warnings were supposed to have been) because “the benefits of Zometa far outweigh its risks.” Id. at **29-30. Keep in mind that this was a drug for cancer and, here, the plaintiff was on it for almost four years, where the average life expectancy for patients with untreated multiple myeloma like hers was six months. Id. at *38. So, plaintiff could not have carried her burden on proximate cause without an end run.
She did claim she never would have taken Zometa if she had known of the risk of ONJ. This led the court to its analysis of the “threshold question” of whether Iowa would ever allow proximate cause to be established based on such testimony. However, there was never any evidence discussed that made this testimony potentially relevant. There was no discussion of what the label said at any point in time, what the plaintiff (or, rather, her expert) said the label should have said before plaintiff started on the drug or by the time her ONJ had developed, what the prescribers told plaintiff about the risks of the drug, or what the prescribers said they would have told the plaintiff about the risks of the drug if the label had said what plaintiff wanted. Without evidence that the prescribers did not tell plaintiff about the risk of ONJ but likely would have told her something about it if the label had been changed, there is no predicate for the hypothetical decision of the plaintiff.
The only discussion of whether plaintiff could raise an issue that the label should have said something different about ONJ did not identify any evidence going to the adequacy of the warning. First, the court stated “a reasonable jury could infer that neither Dr. Renno nor Dr. Verdirame was aware of the risk of ONJ when they prescribed Zometa to Gilliland because had they been aware of such a risk, they would have required that Gilliland undergo a dental evaluation prior to starting Zometa therapy, and she would have complied.” Id. at **27-28. This says nothing about whether the label was adequate. In addition, Dr. Renno was the first prescriber of Zometa, but plaintiff could only point to testimony from Dr. Verdirame, one of three other prescribers, that he would have sent plaintiff for a pre-Zometa dental exam if he had been aware of an ONJ risk. Id. at *27. We do not see how Dr. Verdirame’s testimony says anything about whether the warnings were adequate when Dr. Renno started prescribing it. Second, the court shifted the burden to defendant to establish that the Dear Doctor Letter on ONJ it had sent to these doctors ten months before plaintiff was first prescribed Zometa adequately warned of the risk of ONJ. Id. at **28-29. Holding the defendant to the standard that “no reasonable jury could conclude that [defendant’s] warning of the risk of ONJ is anything but adequate” and identifying no evidence from plaintiff that the warning wasinadequate, the court ended its analysis of adequacy. We have no idea from the opinion what plaintiff contended the label or DDL should have said or whether there was evidence that such a warning would have had any impact on one or more of the prescribers in terms of the risks they disclosed to the patient. Even when the court later turned to whether plaintiff “generated a triable issue of fact as to causation,” we never get any evidence that plaintiff ever would have received different information on ONJ.
Regardless, the court entertained the question of whether Iowa law would allow plaintiff to carry her causation burden by pointing to her own testimony on what she would have done if she had received different information on ONJ. It took up the question in the reverse, though, as the defendant’s suggestion that it adopt a position “shared by many courts throughout the country: that whether the plaintiff-patient would have refused treatment if a better warning had been provided is irrelevant.” Id.at *30. That is clearly not the question that the court should have been asking, but it was also clear that most courts have kept the proximate cause focus on the learned intermediary’s decision whether to prescribe. Somehow, the court considered the majority position to be an “extension of the learned intermediary rule,” where it really was plaintiff who sought to make new law. As far as we can tell, the court did what it said and looked only at the cases the defendant cited—and not any cases plaintiff cited or any cases from Iowa courts—to decide that the Iowa Supreme Court would not “extend” the law as defendant proposed. This, of course, is all backwards.
The court brushed aside cases it considered to be premised on plaintiff testimony that warning them would not have changed their use of the drug. We think these cases can also be categorized as ones where the evidence, expanded to look beyond the prescriber to the plaintiff, still did not establish proximate cause for failure to warn. Gilliland, as we have said, could be categorized with these cases based on the lack of evidence that any prescriber would have said something different to plaintiff based on a different warning. The court characterized the other cases as being premised on patients generally following physician recommendations. This characterization seems to be plucked from a few words in each case rather than the overall logic of the learned intermediary doctrine. The opinion's logic seems thin:
Even assuming that the premise [that patients generally follow their treating physician’s recommendations] is sound, the legal effect of such premise should, at most, raise a rebuttable presumption that plaintiff-patients can attempt to overcome. Therefore, assuming that the Iowa Supreme Court would adopt the premise at issue, Iowa’s learned intermediary jurisprudence suggests that the State’s highest court would conclude that the legal effect of such premise is, at most, to raise a rebuttable presumption. Indeed, to conclude otherwise, i.e., that patients should always be deemed to undergo the treatment therapy recommended by their physicians, regardless of the risks associated with such therapy, would frustrate the purpose of the learned intermediary rule, which is to enable patients to make informed and intelligent decisions whether to undergo a recommended therapy by balancing the probable risks against the probable benefits of the course of treatment proposed by their physicians. See Third Products Restatement § 6(d) cmt. b (“The rationale supporting this ‘learned intermediary’ rule is that only health-care professionals are in a position to understand the significance of the risks involved and to assess the relative advantages and disadvantages of a given form of prescription-based therapy. The duty then devolves on the health-care provider to supply to the patient such information as is deemed appropriate under the circumstances so that the patient can make an informed choice as to therapy.”). The learned intermediary doctrine certainly does not allow health care professionals to substitute their judgment for that of their patients. Nor does it obviate the need to consider whether the plaintiff-patient’s decision concerning her recommended course of treatment would have been different, assuming that the warning at issue had been more adequate. Rather, the “learned intermediary” rule merely recognizes that only health care professionals are able to fully understand and appreciate the risks associated with a given prescription medication. Accordingly, the law makes it their duty to provide to their patients whatever information they deem pertinent so that the patients can make informed decisions. See Third Products Restatement § 6(d) cmt. b (stating that health care providers have a duty “to supply to the patient such information as is deemed appropriate under the circumstances so that the patient can make an informed choice as to therapy”).
Id. at **32-34. Aside from footnotes that did not cite any cases, that was the court’s reasoning for why proximate cause for a prescription drug failure to warn claim, in a jurisdiction with the learned intermediary doctrine, can be established with testimony from the plaintiff about what she would have done with different information.
We think this analysis was based on a fundamental misunderstanding of the rationale for the learned intermediary doctrine. The quotes from the Restatement are right, but incomplete. Because the doctor has a duty to inform her patient and is in the best position to do so, the state law duty of the prescription drug manufacturer is to provide information about the drug to doctors, not to patients. Requiring the manufacturer to warn patients directly does not make sense in this scheme. If the manufacturer satisfies its duty to warn prescribers, then it is not liable for failure to warn regardless of what information the physician relays to the patient to whom she prescribes the drug. Consistent with the duty to warn running to prescribers, cases have identified situations where there is no proximate cause for failure to warn like where the prescriber did not read the label or where the prescriber had independent knowledge of the risk that was insufficiently disclosed in the label. In this scheme, there would need to be a good reason to expand the focus of inquiry on proximate cause for failure to warn beyond the prescriber. Other than to note—in the section on what evidence plaintiff could muster—three cases (one of which we hammered here) that reached a similar conclusion, the court never provided much of a reason.
The analysis of whether plaintiff had evidence to establish proximate cause under newly expanded Iowa law was similarly lacking. As we have said, there was no evidence linking a different warning on ONJ to a doctor telling plaintiff something new, so what plaintiff said she would have done with new information should not have mattered. Not surprisingly, the court found that plaintiff’s own testimony—and her eventual unexplained decision to stop the drug that ostensibly helped keep her alive—was enough to create a genuine issue of material fact and preclude partial summary judgment. Id. at **38-39. Along the way, the court made one of the most befuddling statements we have seen: “Although there can be no doubt that the testimony at issue is hypothetical and self-serving, so is the testimony by Gilliland’s treating oncologists that they would have still prescribed Zometa to her, even if Novartis had provided a better warning of the risk of ONJ. If, as [defendant] argues, the testimony of these physicians is relevant, so too must be Gilliland’s testimony in question.” Id. at *38 n.23. Talk about an end run.