Whether additional warnings would have made a difference to learned intermediary physicians was the issue to be decided in two appellate decisions handed down this week. On facts that weren’t all that much different – at least as far as the prescribers were concerned (hold that thought for later) – the courts came to diametrically different conclusions.
On the one hand, it was lights out for a Georgia plaintiff, but a not-so-nifty Texas two-step not only let a plaintiff there hold on to a bloated verdict, at least for now, but also employed the same warning that should have cut off warning causation as proof of medical causation. Who said they aren’t slick down in the Lone Star State?
The Georgia case is yet another of these suicide cases brought against manufacturers of anti-depressant drugs. While we think these sorts of claims should be preempted, since the FDA didn’t believe that the scientific evidence justified a warning, these are lousy cases (for plaintiffs) for a variety of reasons. For one, the underlying depression, not the drug used to treat it, is the overwhelming medical cause of suicide – why suicide rates go up when drug use declines. For another, suicide is considered a superseding cause of death in many states. For yet another, suicide is not something particularly likely to be viewed with sympathy by a jury. Finally, and that means back to the point of this post, most psychiatrists are very careful about how, if at all, they broach the subject of suicide to their clinically depressed patients.
Thus the warning causation facts in Dietz v. SmithKline Beecham Corp., ___ F.3d ___, 2010 WL 744273, slip op. (11th Cir. March 5, 2010) – the Georgia case – began with an adult patient diagnosed with clinical depression. However, he told his prescriber that he had no suicidal thoughts, and there was no prior medical history of mental illness. 2010 WL 744273, at *1. The doctor prescribed Paxil and some other drugs, and referred him to a psychologist. Id. Needless to say, the poor man killed himself shortly thereafter. Id.
The Dietz opinion doesn’t indicate if the drug carried any suicide warning at the relevant time (2002). The prescriber testified that he thought, then and now, that the benefits of the drug outweighed its risks – including what the prescriber considered a “slightly increased” suicide rate in other population groups (minors). 2010 WL 744273, at *1. The doctor’s practice, with “adult patients he considers stable” was not to discuss suicide risk at all. Id. Why give them any ideas? “Moreover, [the prescriber] testified that, even in retrospect, he agreed with his decision to treat [the deceased] with Paxil and would do so again today under the same circumstances.” Id. He gave that testimony right after reviewing the current black-box Paxil label. Id. at *2.
Thus, it didn’t matter what the label said – for this patient, the prescriber in Deitz would not have given a suicide warning.
The warning causation facts in Centocor, Inc. v. Hamilton, ___ S.W.3d ___, 2010 WL 744212, slip op. (Tex. App. [Corpus Cristi] March 4, 2010), begin with a plaintiff who took Remicade to treat Crohn’s Disease, an autoimmune condition. The plaintiff later complained that the drug gave her quasi-lupus (“lupus-like syndrome”) – another autoimmune condition. The drug has to be infused intravenously for this use, so after writing a prescription the prescriber referred the plaintiff to an infusion clinic. Id. at *6.
The warning on Remicade at the time specifically discussed this quasi-lupus: “Treatment. . .may result. . .rarely, in the development of a lupus like syndrome. . . . If a patient develops symptoms suggestive of a lupus-like syndrome. . .treatment [with the drug] should be discontinued.” 2010 WL 744212, at *3 (quoting relevant warning). There’s a lot more language than just this quote. Id. The prescriber claimed that he discussed this risk with the plaintiff. The plaintiff, sn surprise, denied it. Id. at *6. There was no question, however, that the prescriber in fact knew about the risk. Thus at best, the plaintiff could establish that, notwithstanding his knowledge of the warned-of risk of quasi-lupus, the prescriber elected not to pass along the warning. Id. at *32. That one got money, lots of it.
So what happened? In Deitz, the court cited Georgia’s “long-recognized, unwavering use of the learned intermediary doctrine,” including the Georgia Supreme Court’s 2003 adoption of the rule. Deitz, 2010 WL 744273, at *2 & n.2. As the prescriber said he wouldn’t do anything different even with an adequate warning on the drug, no different warning could have made a difference:
[Plaintiff] cannot demonstrate that [defendant’s] alleged failure to warn [the prescriber] about increased suicide risks associated with Paxil proximately caused [the decedent] to commit suicide. The doctor provided explicit, uncontroverted testimony that, even when provided with the most current research and FDA mandated warnings, he still would have proscribed Paxil for [the decedent’s] depression. Pursuant to Georgia’s learned intermediary doctrine, this assertion severs any potential chain of causation through which [plaintiff] could seek relief, and [plaintiff’s] claims thus fail.
Id. at *3.
Short and sweet (and to be published). Just the way we like them.
The Texas two-step in Hamilton was necessarily a lot more involved. But it started out with the same basic rule. There, what’s more, the prescriber actually had a warning staring him in the face. At best – crediting the plaintiff’s self-serving denials – Hamilton then was a case where the prescriber chose not to pass along a warning that he had received. Yet the plaintiff walked away with almost $4 million for temporary injuries, and for fraud, no less. What happened?
If you’re a defendant, you know you’re in trouble when the court decides it’s on a first-name basis with the plaintiff. In Hamilton it was “Patricia” this and “Patricia” that throughout a 75-page opinion. That sort of sets the tone.
To save the verdict, the court decided it had to take the doctor out of the picture altogether. That’s the first step of this Texas two-step. It did so by adopting a “direct to consumer” exception to the learned intermediary rule. That started on the very first page, where the court, quoted Perez v. Wyeth Laboratories, 734 A.2d 1245 (N.J. 1999), for two propositions: (1) “[m]edical services are in large measure provided by managed care organizations,” 2010 WL 744212, at *1; (2) “[d]rug manufacturers now directly advertise products to consumers on the radio, television, the Internet, billboards on public transportation, and in magazines.” Id.
The first doesn’t support the exception, and the second, while true, is irrelevant to the case. As to the first, why would sophisticated institutions – the “managed care organizations” – be influenced by TV ads? But what we think is worse, there wasn’t any direct to consumer advertising in Hamilton; not the stuff Perez was complaining about, anyway. There’s no allegation in Hamilton that the plaintiff saw any advertisement for Remicade, was influenced by it, and then marched into her doctor’s office demanding the drug. The caricature presented by the Perez is just that, a caricature. It didn’t happen in Hamilton, and it rarely happens anywhere.
It’s very simple. Prescription drugs are available by prescription. Unless a doctor makes a medical judgment that your situation merits a particular drug, you’re not going to get it. For that reason, all prior Texas precedent had rejected any direct to consumer exception. In re Norplant Contraceptive Products Litigation, 165 F.3d 374, 379 (5th Cir. 1999) (rejecting “aggressive marketing’ to consumers as exception; “as long as a physician-patient relationship exists, the learned intermediary doctrine applies”) (applying Texas law); Ebel v. Eli Lilly & Co., 536 F. Supp.2d 767, 781-82 (S.D. Tex. 2008) (“First, Plaintiff provided no evidence of Defendant’s mass media distribution of information in 2002. Second, this Court does not predict that the Texas Supreme Court would adopt that exception to the learned intermediary doctrine”), aff’d in pertinent part, 321 Fed. Appx. 350, 356 (5th Cir. 2009).
But be that as it may. There were no ads in this case. Nothing in the opinion suggests that the plaintiff had ever heard of Remicade before she went to see her prescriber. What did happen is that after the drug had already been prescribed, a different doctor at the infusion clinic showed the plaintiff a videotape, and that tape included some patient testimonials. Hamilton, 2010 WL 744212, at *6-10. That’s right – the court devoted over five Westlaw pages just to describing the video.
But no amount of description can change the simple fact that this video wasn’t direct to consumer advertising. The plaintiff didn’t flip the switch on her TV. She didn’t open up a glossy magazine. No. The video was played specifically for her by a doctor, in a clinic, after an individualized treatment decision had already been made:
[The clinic doctor] testified that because the decision to take Remicade has already been made by the time a patient arrives at his infusion clinic, he does not typically warn patients about adverse side effects. . . . [He] testified that he did not warn [plaintiff] about lupus-like syndrome. [The clinic nurse] stated that she does not review drug package inserts with patients, and she did not review the package insert for Remicade with [plaintiff] at any time . . . . In addition. . ., [the nurse] showed patients a video produced by Centocor. On December 19, 2001, while [plaintiff] was receiving her first infusion, she was shown the Centocor video, which [the clinic doctor] referred to as a “treatment companion” kit. [He] testified that the purpose of the video was to show “some of the effects of the drug on certain people. . . . It’s very dramatic in some people.”
Hamilton, 2010 WL 744212, at *6-7 (emphasis added).
With all due respect, the “backdrop” of the Perez decision, with which the Hamilton opinion opens, 2010 WL 744212, at *1, doesn’t have anything to do with that case. It’s simply a smokescreen utilized by the court to uphold what is otherwise (and we’d argue still is) a fatally flawed verdict. What happened in Hamilton wasn’t direct to consumer advertising. There was no purported end run around the role of the prescribing physician as the gatekeeper both to the drug and to information about the drug. None of the things that the Hamilton court goes on to describe as pernicious about direct to consumer advertising, id. at 26-28, had anything to do with the course of this plaintiff’s medical treatment.
Instead, the video in Hamilton was the equivalent of those “tear off” patient warnings that have appeared on package inserts for decades. It was information, prepared by the manufacturer, that the prescriber (or some other physician) could elect to supply to the patient if the physician determined that it might be helpful. Plaintiff only saw the video because a medical professional decided that she should. She was given the video by her doctors. It was not something she could have gotten elsewhere, before any doctor had prescribed the drug.
That’s step one – knock out the learned intermediary rule. Step two of the Texas two-step was to create another avenue of causation. That was tough, in part because the video didn’t say anything false about quasi-lupus. It simply didn’t mention it. It was doubly tough because the drug actually helped the plaintiff; the drug was “indicated for the reduction in signs and symptoms of Crohn’s disease,” 2010 WL 744212, at *9, and it did just that: reducing her symptoms. Id. at *10 (“[i]nitially, Patricia had a good response to the Remicade therapy”). It was triply tough because, even after learning about quasi-lupus as a possible risk, plaintiff continued to take the drug, since it made her feel better. Id. at *11-14.
Interestingly, when the videotape was shipped new from the defendant, it came with some “brochures” and “with the package insert.” 2010 WL 744212, at *9. These were commonly kept in a plastic “sleeve.” Id. at *10. Predictably, plaintiff denied that the copy she saw, which nobody claimed was brand new, came with this written material. Id. The sleeve was missing from the copy she gave her lawyer. Id. On that, even we think it’s quite possible that she was telling the truth. But just because somebody else lost the sleeve (and thus the package insert), how does that make the defendant liable – let alone liable for fraud?
Plaintiff’s alternative causation theory, which the Hamilton court bought in toto, was to convert the warning on the package insert from the defendant’s shield into her sword. That’s because the plaintiff didn’t have any scientific proof of medical causation, so once she didn’t have to worry about her doctor’s knowledge sinking her case under the learned intermediary rule, she turned around and used that label, and her doctors’ own knowledge as a substitute:
For example, first, the package insert issued by [defendant] and approved by the FDA acknowledges that lupus-like syndrome is a risk associated with taking Remicade. The package insert itself described findings from [defendant’s] clinical trials prior to FDA approval that found that some patients may rarely suffer from lupus-like syndrome as a result of Remicade. . . . Second, [plaintiff’s] doctors repeatedly testified that they knew of the risks of Remicade treatment-specifically, the risk of lupus-like syndrome.
2010 WL 744212, at *30. And so, plaintiff established general causation without expert testimony.
That’s the second half of the two-step, and it’s an interesting little trick. Once the learned intermediary rule is knocked out, the fact that the defendant adequately warned the doctor can’t hurt the plaintiff – because, of course, the claim is now that the defendant should have warned the plaintiff directly. Even if that might have happened in Hamilton, it’s easy enough for the plaintiff simply to deny, deny, deny, and she did. It’s credibility, and of course a South Texas jury decided that.
So then the warning becomes the plaintiff’s friend. Who needs epidemiology when you can simply point at the defendant’s own label, and say that the defendant “itself” agrees. 2010 WL 744212, at *30.
There are a lot of other questionable aspects of Hamilton. For instance, there was evidence that the defendant encouraged doctors to do the infusion work themselves to make more money. 2010 WL 744212, at *4. But how could that be relevant, since this plaintiff saw the critical video at an off-premises infusion clinic run by independent physicians? Id. at *6. And there was evidence that the defendant took steps to counteract bad publicity from a 2003 article in the New England Journal of Medicine. Id. at *4. But the prescription here was in 2001, id. at *6, and more importantly, the prescriber already knew about the risk. It sounds like the plaintiff was allowed to dump in whatever derogatory information she could find, regardless of its having any factual connection to the case.
And how could a mere omission (at most) of the quasi-lupus risk from the video rise to the level of fraud? We’ve been under the impression that Texas follows the usual rule the arms-length transactions between buyers and sellers of products do not import any affirmative duty to speak – at least that’s what the law’s been before in drug cases. Stephanz v. Laird, 846 S.W.2d 895, 904 (Tex. App. [Houston] 1993) (“failure to disclose information is not fraudulent unless one has an affirmative duty to disclose, such as where a confidential or fiduciary relationship exists”); Conger v. Danek Medical, Inc., 27 F. Supp.2d 717, 722 (N.D. Tex. 1998) (applying Stephanz to prescription medical product liability litigation). But who knows, Texas is so big, it once had the right to become five separate states – so maybe the folks in South Texas believe the rules are different down there.
But all that’s really for another post – we hope on occasion of the Texas Supreme Court agreeing to review this case. For today’s purposes, limited to the learned intermediary rule and causation, it’s the Texas two-step in Hamilton that intrigues (and repulses) us: (1) knock out the warning to the doctor, and the doctor’s decision not to pass it along, by carving an exception to the learned intermediary rule that isn’t even presented by the facts, and (2) then use that same warning, and the same physician knowledge, to establish medical causation in the absence of any other evidence that would pass muster under Texas’s tough expert testimony rules.