Have you ever seen a superbly qualified expert give an opinion that is utterly unsupported? That is what happened last week in In re Lipitor (Hempstead v. Pfizer), No. 2:14-cv-1879, 2015 WL 9165589 (D.S.C. Dec. 11, 2015), and the district court did an admirably methodical job putting the kibosh on it. The expert was a former Rhodes Scholar, trained at top universities, and was the director of the endocrinology and metabolism division at a prestigious teaching hospital on the west coast. Yet, she seemingly allowed plaintiffs’ lawyers to lure her into placing her reputation at jeopardy by giving a medical causation opinion that was obviously wrong and boiled down to little more than "because I said so."  

Of course, we are oversimplifying, but not by much. The product was Lipitor, a statin, which is one of the most useful and health-promoting drugs in history. Even Consumer Reports, which is run by consumer advocates, has said that statins reduces the risk of heart attack, heart disease, and death and that statins are quite safe overall. But Lipitor is also one of the best selling drugs in history, which has attracted reams of litigation. The allegation du jour against Lipitor is that it causes Type-2 diabetes, which is curious since adult weight gain is a major risk factor both for Type-2 diabetes and high cholesterol. Do you think maybe people taking statins also show an increased prevalence of Type-2 diabetes because they are overweight or obese? Hmm.

Anyway, back to our wayward expert, Dr. Elizabeth Murphy. The plaintiff in Hempstead took Lipitor off and on starting in 1999 and was diagnosed with new-onset diabetes five years later. 2015 WL 9165589, § II (the order was just posted to Westlaw, so there is no star pagination yet.) Naturally, she sued Lipitor’s manufacturer and had the good fortune of being selected as a "bellwether" case.

Or maybe it was good fortune for the defendants, judging by the poor quality of Dr. Murphy’s opinions. Dr. Murphy opined that Lipitor caused the plaintiff’s Type-2 diabetes because (1) the relative risk of developing diabetes while on Lipitor is 1.6 and (2) the plaintiff developed diabetes while taking Lipitor. The plaintiff’s attorneys tried to dress up the opinion to make it look more scientific, but Dr. Murphy admitted time and again in her deposition that there was no more to it, testifying for example that "if the patient was taking the Lipitor and they developed diabetes while on it, . . . I would think that it would be a contributing factor" and "Lipitor has an increased risk, and that is why my conclusion is that it was a contributing cause." Id. at § IV.A.

It’s the epitome of an ipse dixit opinion, and the reason you ought to review the order is to see how the district court dismantled it. Start with the so-called methodology. Dr. Murphy offered a "commonly used" five-step test to come to her conclusion, but she could not identify any professional organization or peer-reviewed text that used it. Nor could she identify anyone in the medical profession who used the test to determine the cause of Type 2 diabetes, other than other litigation experts, and she had never used the test in her own practice treating diabetic patients. Id.Strike one.

Next, having identified a so-called methodology, Dr. Murphy did not apply it. She made it only through the first three steps: (1) were there reports in the literature or reliable scientific data of new-onset diabetes with Lipitor; (2) was causation biologically plausible; and (3) did new onset diabetes appear after Lipitor was given? According to Dr. Murphy, because literature showed a relative risk of 1.6 and because the patient developed diabetes after taking Lipitor, the drug must have been a cause.

This opinion did not hold up for a number of reasons. First, the court did a very nice job of explaining how a relative risk of 1.6 was not sufficient to draw conclusions on causation. As the court explained,

A relative risk ratio in this range suggests that while some people exposed to the substance developed the disease due to exposure . . . , most would have developed the disease anyway. . . . Thus, to establish specific causation with a relative risk between 1.0 and 2.0, an expert . . . must opine to a reasonable degree of medical certainty that the particular plaintiff is in the minority of those that developed the disease due to exposure to a particular drug or substance rather than in the majority that would have developed the disease regardless.

Id. at § III (emphasis added). In other words, a relative risk of 1.6 goes only so far—it tends to show that it is possible for a substance to cause a disease, but only in a minority of cases and certainly not to a reasonable degree of medical certainty in any particular case.

Second, all that Dr. Murphy had to offer in addition to relative risk was a temporal relationship, i.e., that the plaintiff had developed Type-2 diabetes after starting Lipitor. But the temporal relationship was weak—the plaintiff took Lipitor for five years before the onset of diabetes. And in any event, a temporal relationship alone is rarely sufficient to prove specific causation. This is Scientific Evidence 101, and the court called Dr. Murphy’s logic "flawed," observing that "[t]he fact that exposure to [a substance] may be a risk factor for [a disease] does not make it an actual cause simply because [the disease] developed." Id. § IV.A (quoting Guinn v. AstraZeneca Pharm. LP, 602 F.3d 1245, 1255 (11th Cir. 2010)). Or as the court put it in a very thorough discussion of temporal relationships as evidence of causation, which we commend to you, "[T]emporal proximity is generally not a reliable indicator of a causal relationship." Id. § V.

Third, the fourth factor in Dr. Murphy’s proffered five-part test was other possible causes, but she failed to take other risk factors into account. The most astounding was adult weight gain—the plaintiff’s weight had increased by nearly 50 percent over 25 years and by 15 percent over the 10 years leading to the onset of diabetes. Id. § IV.B.2. The plaintiff’s total weight gain made her12 times more likely to have diabetes, yet Dr. Murphy did not consider this risk factor in her analysis. She also did not consider the plaintiff’s positive family history of diabetes, her advancing age, her hypertension, or factors suggesting that she suffered from metabolic syndrome. Id. § IV.B.3-6. This is a major problem.

Fourth, the last factor in the five-part test was how likely it is that Lipitor caused Type-2 diabetes in this patient. But Dr. Murphy did not attempt to answer this question either. She admitted that she has seen other patients with the same risk factors develop diabetes without taking statins and that the plaintiff may have developed diabetes in the future even without taking Lipitor. Id.§ IV.B.7. With regard to the plaintiff’s fluctuating glucose levels, Dr. Murphy testified that "I’m not sure that . . . I can explain . . . what was going [on]. There’s [sic] many potential causative factors." Id. She kept coming back to the increased relative risk, which this court had long since dispensed with, concluding that even taking a relative risk of 1.6 at face value, "nearly two-thirds of the patients (63%) who develop diabetes while taking Lipitor would have developed diabetes regardless." Id. It is difficult to argue with those numbers.

Finally, the plaintiff took refuge in that diagnostic method whose abuse we have long bemoaned—the differential diagnosis. Differential diagnoses are not all bad, but an expert’s opinions based on a differential diagnosis still must be scientifically reliable. That is where Dr. Murphy failed, as it was her obligation to "rule in" Lipitor as a cause and "to offer an explanation as to why . . . other recognized causes, alone, are not responsible for the disease." Id. § V. On this score, Dr. Murphy "offers no data or facts to make the leap from a possibility to a probability that Lipitor was a substantial contributing factor," and she did not rule out other potential causes whose risks "greatly exceed" Lipitor. Id.As the court observed, merely labeling analysis as a "differential diagnosis" does not mean that the court has to accept it. Id.Amen.

Coming back to our opening point, why would a reputable physician stick her neck out in this fashion? We, of course, blame the lawyers, who apparently could do no better with this plaintiff and her particular medical history. But still, the court concluded that Dr. Murphy’s method was "invalid on its face." Id. This expert probably wishes she had not signed up.