[Update, July 15, 2013: The Federal MDL court reached the opposite conclusion, denying summary judgment on liability and causation and on punitive damages. Those cases will now proceed to trial.]


In case you missed it, last week I had a guest post up at TortsProf lamenting how recent changes in civil procedure law have created a situation in which judges are frequently deciding complex cases by improperly deciding for themselves what the true facts were, in advance of a jury trial, and sometimes on nothing but the initial complaint.


Unfortunately, we just had another example: the recent order in the NuvaRing litigation consolidated in New Jersey state court dismissing all of the bellwether cases, primarily on causation grounds. It’s not the end of the game — motions for reconsideration will be filed, as will an appeal, and it doesn’t affect the federal court MDL — but it’s disappointing nonetheless. There’s much to complain about (and, on appeal, to reverse), but I’m going to focus on the “learned intermediary” part. First, a little bit of background.


The most popular form of hormonal contraceptives are combined hormonal contraceptives (“CHC”), which use an estrogen (typically ethinyl estradiol) to prevent ovulation and thicken cervical mucus. Estrogen use, however, is correlated with an increased risk of venous thromboembolism, such as deep vein thrombosis, and thus pulmonary embolisms (blood clots in the lungs) and cerebral venous thrombosis (blood clots in the brain), so CHCs add a progestin to counterbalance that risk.


NuvaRing uses desogestrel as its progestin (that’s different from Mirena, which uses levonorgestrel), making it a “third-generation” CHC. Since 1995 — six years before NuvaRing went on the market — it has been known that third-generation CHCs have a significantly higher risk of causing thrombosis and blood clots than second-generation CHCs. As the New Jersey court opinion recounts (based on the plaintiffs’ filings), by the time NuvaRing went on the market in 2001, 15 studies had examined that difference in risk, and 13 of those studies found an elevated risk, ranging between 1.4 times and 4 times greater risk of venous thromboembolism when using NuvaRing. Implants without a hormone added, like Essure, don’t have these same risks.


NuvaRing’s prescribing information and patient insert warned about the general risk of venous thromboembolism when using CHCs, but then hedged on the increased risk with vague, ambivalent language obviously written more for purposes of litigation than for informing patients and doctors about the risks of the product:


The use of combination oral contraceptives is associated with increased risks of several serious side effects, including blood clots, stroke, or heart attack. NuvaRing is not for women with a history of these conditions. The risk of getting blood clots may be greater with the type of progestin in NuvaRing than with some other progestins in certain low-dose birth control pills. It is unknown if the risk of blood clots is different with NuvaRing use than with the use of certain birth control pills.


The risk “may be greater” but “is unknown?” That wasn’t even accurate when NuvaRing was first put on the market, and now, more than a decade later, it is even less defensible: last year, a meta-study of 625 studies published between January 1995 and April 2010 found the risk of venous thromboembolism for CHCs that use desogestrel, like NuvaRing, was about double the risk of second generation CHCs. Yet, the manufacturer (Organon and Merck) have refused to update the label; perhaps it’s because they sell over $600 million worth of NuvaRings every year to over a million women.


If this was a law school exam, there really wouldn’t be much more to say: Organon and Merck plainly failed to warn women that NuvaRing was twice as likely (or more) to cause blood clotting injuries than other contraceptives that work just as well, despite ample evidence of the danger. Under general principles of strict liability, “where the product contains an ingredient … whose danger is not generally known [to the public] … the seller is required to give warning against it, if he has knowledge, or by the application of reasonable, developed human skill and foresight should have knowledge, of the presence of the ingredient and the danger.” Feldman v. Lederle Labs., 97 N.J. 429, 452, 479 A.2d 374, 386 (1984)(Quoting Second Restatement of Torts, Section 402A, comment j).


In a fair and rational legal system, the case would go to a jury to resolve three issues:


(1) whether NuvaRing really is more dangerous than other CHCs (Organon and Merck could point to the handful of studies showing no difference, and their own researchers continue to say there’s no difference);

(2) whether a better warning would have caused either the doctor not to prescribe it or the patient not to take it;

and, assuming the jury found for the plaintiff on both of those,

(3) the damages suffered by the plaintiff.


But this is pharmaceutical liability law, and so the rules are, sadly, often applied in absurdly skewed manner to benefit the defendant drug maker.


Take, for example, the case of Jackie Bozicev, one of the cases just dismissed. In April 2007, Jackie delivered her second child, was prescribed NuvaRing the next month, and then in December suffered a pulmonary embolism in front of her two-year-old son. She died the next day. The Court held that, under the odious “learned intermediary” doctrine, what Jackie knew about the risks was irrelevant —all that mattered was what her prescribing doctor knew about the risks of NuvaRing, as if Jackie had no say in the matter.


Her doctor testified that she knew NuvaRing had a risk of venous thromboembolism like all CHCs, and that NuvaRing sales representatives had come to speak with her (see an example of what that’s like here), but she couldn’t remember what they said. She said she knew NuvaRing may have a greater risk to it — essentially repeating the same erroneous information on the prescribing information — but she certainly didn’t say she knew NuvaRing was twice as dangerous as the alternatives, nor did she say she considered it  while prescribing Ms. Bozicev NuvaRing, nor did she disclose this risk.


Too bad. The Court held, “Ms. Bozicev’s treating physician was aware of the risk of VTE and ‘nonetheless determined NuvaRing’s use to be warranted. … [Plaintiff] has not presented evidence to the Court as to whether [her doctor] would have changed her decision to prescribe NuvaRing had the warning reflected a higher risk of injury.” Thus, Organon and Merck are off the hook because, though Jackie plainly didn’t know NuvaRing was more dangerous, and though Organon and Merck still vehemently deny it is more dangerous, her doctor somehow knew for sure that it was, despite her testimony that she didn’t.


How did the Court get to that hopelessly wrong answer? By relying on a completely different case, Strumph v. Schering Corp., 256 N.J. Super. 309 (App.Div.1992)(affirmed without opinion, 133 N.J. 33 (1993)), in which a person with paranoid schizophrenia was admitted to the hospital after a suicide attempt and was prescribed a drug (Trilafon) that warned about, and was generally known by psychiatrists to cause, in rare instances, a serious neurological disorder. In Strumph, both treating physicians testified quite clearly they were aware of the risks of the drug and that they conducted a risk-benefit analysis before prescribed it.


Indeed, as Judge Skillman’s dissent in the Appellate Division’s order in Strumph notes (that dissent is important because it’s what the New Jersey Supreme Court affirmed in its one-sentence opinion, and the dissent is what the NuvaRing court cites), “in the normal course of human behavior one would not expect a physician to admit that he was not well informed regarding a powerful chemical compound he was dispensing to a patient or to admit that the medicine he prescribed was not the most appropriate alternative.” Strumph, 256 N.J. Super. at 325, quoting Seley v. G.D. Searle & Co., 423 N.E.2d 831, 839 n. 5 (Ohio 1981). Judge Skillman also noted the doctors had “no relationship with the drug manufacturer which would be likely to generate any bias in its favor.”


The NuvaRing cases, like Jackie Bozicev’s case, have nothing at all in common with Strumph; the NuvaRing plaintiffs weren’t suicidal individuals with paranoid schizophrenia for whom other medications hadn’t worked, patients who were thus willing — or required by involuntarily commitment — to accept virtually any potential side effect to save their lives, and to whom the risk was disclosed by the medication’s prescribing information. The NuvaRing plaintiffs were healthy women who had myriad birth control methods available to them, including safer hormonal methods. There’s no reason for the Court to even consider making a definitive factual finding before trial that the increased risks were irrelevant to the patients, and thus dismissing the cases, but that’s just what it did.


The whole supposed argument behind the odious “learned intermediary” doctrine is the dubious claim that a pharmaceutical company need only warn a doctor, not a patient, about the drug’s risks, after which it’s up to the doctor completely. Maybe that argument makes sense in Strumph, where the risk was disclosed by the drug manufacturer and was well known by the treating physicians, but it makes no sense whatsoever in the NuvaRing cases. Yet, faced with clear evidence that even prescribing physicians didn’t know the real dangers of NuvaRing — not least because Organon and Merck deny those dangers to this very day — the Court somehow went exactly the wrong way.


Maybe the judge will truly reconsider his opinion, or maybe it’ll be straightened out on appeal. Either way, it’s a reminder of just how difficult it is for injured people to obtain civil justice against corporate America.