As Bexis demonstrated yesterday, we can have strongly held views. When we do, we are not always subtle about letting our readers know what those views are. When a rant builds, it can pour like an avalanche coming down a mountain—with more vinegar than cinnamon or sugar. When we read In re Actiq Sales and Marketing Practices Litigation, No. 07-4492, 2014 U.S. Dist. LEXIS 984411 (E.D. Pa. July 21, 2014), we felt the rant coming on. The case involves Daubert in a case with outsourced representation and collective causation evidence related to alleged off-label promotion of a prescription drug, so a few things we care about were brewing. But it is now the first day of August and a tropical vacation is looming. School for the kids and other rituals signaling the transition from summer to fall will be here soon. Thus, we resolved to keep this post somewhat calm despite a bad result, questionable analysis, and a fundamental misunderstanding of the relationship between FDA regulation of drug companies and what doctors do with their own patients. Rest assured, though, that we would have many more snarky comments about the opinion if not for our resolve to stay “chill,” as much as we ever do.
Like most of the third party payor cases we see, this one involved some familiar plaintiff firms representing a state agency and some funds for groups of union members claiming they had overpaid for the prescription of a drug beyond its approved indication. The particular causes of action asserted or the nature of the class certification that had previously been granted were not discussed substantively, but plaintiffs wanted disgorgement of all the profit the defendant had made from all off-label prescriptions filled during a five year period as a result of an alleged illegal marketing scheme. Much like one of the case discussed here, the alleged scheme was about trying to get prescriptions from doctors and for patients beyond the approval for a painkiller indicated only for cancer patients with tolerance to other painkillers. Setting aside some fairly significant legal challenges not addressed here, to get to the hundreds of millions of dollars of damages that plaintiffs want, they needed at least one expert to tally up the actual profits from the alleged scheme. They settled on Dr. Meredith Rosenthal, a health economics professor and frequent expert for the plaintiffs in such cases, including the Neurontin cases we have, ahem, discussed before. The manufacturer defendant challenged her qualifications and the methodology and fit of her opinions.
We start with the observation, calmly made, that there did not seem to be any consideration of the fundamental—to us—issue of whether an economist can ever determine why prescriptions were written. To get to millions of dollars of revenue from prescriptions, many physicians have to prescribe the drug to many patients. Most courts looking intelligently at this issue have recognized that prescriptions involve individualized decision making of doctors about their patients. Even where there is evidence that an actual scheme to promote off-label use was being pursued, simply assuming that all off-label prescriptions, or any particular percentage of them, were written because the scheme influenced them makes no sense. This is because doctors write prescriptions for indications beyond the approved label. It can even be the standard of care to do so. For a variety of reasons, FDA pays a great deal of attention to the scope of the indications it approves for prescription drugs and limiting what drug companies can say about uses beyond the approved indications. With or without reasons someone might discern from their records, and without a sassy head bob, doctors pretty much do what they want, at least in terms of which drug they prescribe to which patient at any given point in time. The desire to tally up the profits from off-label promotion, as opposed to off-label use, does not change this dynamic. So, keeping with our rantless approach, we will just say that we find it hard to believe there could be aDaubert opinion on Dr. Rosenthal’s opinion that did not examine whether such an opinion could be given by any expert in her position.
The Actiq court did not, though, and started off with one of the more lenient statements of the requirements of Rule 702 and Daubert that we have ever seen. It did include the requirement that the proponent of the evidence prove qualification, reliability, and fit by a preponderance of the evidence. Id. at *8. Sadly, that was the opinion’s last mention of—or, as far as we could tell, the court’s consideration of—the plaintiffs’ burden. As is often the case with challenges to the qualifications of someone with a doctorate-level degree and a full professorship, the challenge to Dr. Rosenthal’s qualifications was rejected fairly easily. The court characterized the challenge as being premised on the idea that only a cost accounting expert could properly address the subject of Dr. Rosenthal’s testimony. It was enough for the court that Dr. Rosenthal was a qualified economist and claimed to have taken an economist’s approach. Id. at **11-15. The court did not require plaintiffs to prove that Dr. Rosenthal was qualified to provide opinions on profits deriving from off-label prescriptions written as a result of off-label promotion.
The issue of reliability—not whether plaintiff proved Dr. Rosenthal’s methodology was reliable, but a more amorphous inquiry—took up the bulk of the opinion. Most of our issues with the analysis are covered by what we have already pointed out-- plaintiffs were not held to their burden of demonstrating reliability. The court also never considered whether any evidence on which Dr. Rosenthal relied satisfied the requirements of Rule 703—once, excusing her reliance on a document with “uncertain circumstances surrounding” it—and did not analyze the Daubertfactors for reliability after setting them out. Id. at **16 & 21. Walking through the steps that Dr. Rosenthal said she took and rejecting the defendant’s criticism of each, generally without requiring anything beyond Dr. Rosenthal’s own word, the court finally got to the step we care most about—the step that involved determining the percentage of off-label prescriptions that allegedly resulted from off-label marketing.
Beyond the general implausibility of making such a calculation without analyzing the facts related to the prescriptions themselves, Dr. Rosenthal’s approach was pretty wacky. She calculated the percentage of off-label prescriptions by looking at IMS National Disease and Therapeutic Index, focusing apparently on the diagnosis codes included therein, and then applied a 15% discount because the plaintiffs’ lawyers decided 15% of off-label prescriptions would not be from off-label marketing. Id. at **18-19. Were we not staying calm, we would ask a question represented by a three letter texting acronym. We might even ask it a few times. Rather than look to whether the plaintiffs proved that this basic methodology was reliable—for instance, because the plaintiff lawyers made up a key part of it or that an economist should not be deciding which diagnosis codes indicate off-label use—the court focused primarily on issues with the IMS data itself. The fact that Dr. Rosenthal admitted that the IMS data was not representative of the entire population of physicians prescribing Actiq gave the court some pause. Id. at 41. Representativeness, of course, is the touchstone of any attempt to extrapolate from a sample. Here, the lack of representativeness was because certain classes of prescribing physicians—like pain specialists, one of two categories of physicians that plaintiffs contended were legitimate on-label prescribers—were not included. This should have been a big problem for Dr. Rosenthal, but the opinion found “the issues of whether certain specialties should (or could) have been included in Dr. Rosenthal’s calculation is ultimately more a question of the accuracy of her analysis, not the methodology she used in conducting her analysis.” Id. at *44. We beg to differ. Deciding to rely on an incomplete and nonrepresentative sample is a basic methodology issue. A reliable methodology might require that the expert decline to offer an opinion due to the absence of reliable data.
The rejection of the argument that IMS data does not allow a reliable opinion as to what percentage of a drug’s prescriptions, even within the sample itself, are off-label seems so inconsistent with court’s gatekeeper role that we simply repeat it here:
For the reasons previously outlined, the Court again declines to exclude Dr. Rosenthal’s declaration based on admitted deficiencies in the IMS coding. It simply defies reason for Cephalon to suggest that data which was deemed reliable for purposes of monitoring off-label prescriptions when Actiq was approved by the FDA is somehow “wholly unreliable” in estimating the percentage of off-label prescriptions for purposes of calculating damages. There again has been no suggestion by Cephalon or its experts that better data is available. Beyond this, [the corresponding defense expert’s] suggestion that every patient file should be reviewed individually presents obvious practical difficulties which the Court need not address.
Id. at *41.
After all of this, the part that really galls us, if we were allowing ourselves to be galled, is the court’s rejection of the argument that Dr. Rosenthal cannot just assume a causal relationship between 85% of her estimated off-label prescriptions and the alleged off-label promotion. The court did disagree with plaintiffs that Dr. Rosenthal could just assume causation as a damages expert. Id. at 56. Any assumption had to be based on a sound methodology. The plaintiff lawyers picked 85% for Dr. Rosenthal—and she accepted it—based on a provision in the RiskMAP for the Actiq that required the company to deter off-label prescriptions if “groups of physicians (such as a particular specialty) are identified as having prescribed Actiq inappropriately, and these prescriptions represent potential off-label usage greater than 15% of total quarterly Actiq prescriptions.” Id. at *48. The “identified as having prescribed Actiq inappropriately” in the first clause somehow got lost and the opinion just focused on the 15% in the second clause. The plaintiffs could not explain why 15% mattered, other than it was in the RiskMAP, or what it had to with making an assumption about prescription due to off-label promotion. Id. at *58.
Nonetheless, the Court cannot find that Dr. Rosenthal’s assumption “that there is a positive level of off-label prescribing that can result from clinical judgment by physicians, but this level of prescribing would not exceed 15% of total quarterly Actiq prescriptions absent the alleged misconduct” is so lacking in basis that her declaration and testimony must be excluded. Ideally, Dr. Rosenthal would have engaged in more rigorous examination in this final step of her damages calculation.
Id. Without much more, the court rejected the challenge to the reliability of Dr. Rosenthal’s opinion and then proceeded to reject the challenge to its fit to the issues in the case.
After all that, we still cannot grasp why an economist is allowed to assume that any specific percentage of purported off-label prescriptions have been caused by purposed off-label promotions. That is not something that can be assumed for a single prescription, let alone for thousands. It also has nothing to do with expertise in economics, except that this is vehicle to the plaintiffs making a claim for a big number with a dollar sign in front. We suspect we will see this case again. Next time, we may not be feeling so mellow.