There are times that you can tell that a witness at a deposition just isn’t committed to his or her answers.  They get long-winded, creeping toward the ultimate answer, circling it, poking it, testing it, before ultimately offering it and then cautiously backing away.  It’s the reddest of red flags.  And, if it comes from a causation expert, you could be on the verge of hobbling or ending the plaintiff’s case.  Here is just that type of testimony:

Q: So, Doctor, would you be willing to testify to a jury Nexium [sic] can cause fractures?

A: You know, it would depend on the question posed to me in front of a jury. I would be honest in my interpretation of the data.  It would [sic] not take it any more than it is or less than it is.

I don’t think—if this is a question asked before a jury, I’ll be honest with them.  I don’t think that looking at all of this data any reasonable physician or scientist or even a person of the lay public reading this can say there’s absolutely no relationship between Nexium use and risk of fracture, absolutely none.

The weight of the evidence suggests otherwise,that there is a relationship, a causal relationship, can be inferred because of a number of studies that seem to point the same way.  And the evolution of those studies and that’s what I would testify that, yes, it seems from the weight of the scientific evidence we have that the long-term use of Nexium seems to be related to the risk of fractures.

In re Nexium (Esomeprazole) Prods. Liab. Litig., 2014 WL 5313871, at *2 (C.D. Cal. Sep. 30, 2014).  That answer is from a real expert in the Nexium litigation.  It’s loaded with just the type of hedging you get from an uncomfortable and uncertain witness: “it depends”; “I’ll be honest”; and “it seems from the weight of the evidence.”  That testimony would never convince a jury.

But you’re not in front of a jury.  Not yet. And you don’t ever want to be there. You want to end it now – on a Daubert motion.  To do that, you’ll need more.

So, it’s time to roll up your sleeves and dig deeper, time to scrutinize the expert’s qualifications and methodology.  What did he consider?  What did he miss?  What methodology did he use?  How did he apply that methodology and to what?  How can the methodology be tested? What support is there in the scientific community for that methodology?

It’s clear from the court’s opinion that the defense attorney did exactly that.  First, she placed the expert’s qualifications in question.  While the expert was a doctor, he was not an epidemiologist, even though his causation opinion relied considerably on epidemiology.  Now, this admission alone likely won’t win your Daubert motion.  But it sets the stage nicely for an attack on the expert’s methodology.  Here, the Nexium court seemed at least dubious of the expert’s qualifications: “Bal [the expert] is an orthopedic surgeon, not a gastroenterologist or an epidemiologist.  The actual disorder suffered by the Plaintiffs is orthopedic, so use of an orthopedic surgeon to opine on causation is not completely random.”  Id.

Next, the defense laid into the expert’s methodology.  The expert, it seems, applied little to no scientific methodology to conclude that the association between pain pump inhibitors (PPI) and osteoporotic and related fractures (OP) constituted “causation,” not just an association.  He did not, for instance, evaluate the nine Bradford-Hill criteria that are often evaluated by doctors or scientists attempting to determine causation: “The parties also broadly agree that the factors laid out in a well-known epidemiology paper by Austin Bradford Hill should be analyzed to determine whether causation can be inferred from a correlation between Nexium and OP.  Bal does not seriously evaluate any of these factors.”  Id.

Moreover, the analysis that he did undertake was shoddy, considering only two of the Bradford-Hill criteria: biological gradient (or dose response) and biological plausibility. Id. at *3.  And even then, he gave those two criteria only a surface-level evaluation: “As to biological gradient, there is no indication that Bal independently evaluated any of the evidence that purports to show that higher doses of PPIs for longer periods are more strongly correlated with OP. . . . As for plausibility, Bal discusses several hypothetical mechanisms for PPI causation of OP, but he does not present any evidence that Nexium behaves in the hypothesized way in the real world.”  Id.  The court called this evaluation “thin.”  Id.

The defense uncovered additional weaknesses in the expert’s methodology.  He didn’t consider whether the evidence of association across studies was consistent.  Id. In fact, he couldn’t have, because he failed to even search for studies that did not show an association: “Bal provides no indication that he searched for, and failed to find, other published studies showing no association between PPIs and OP.”  Id.  Not considering the totality of the evidence is inconsistent with applying a scientific methodology.

In reaching his causation opinion, the expert also did not consider the strength of the association he saw in the studies he did review: “Perhaps most importantly, there is no discussion of the strength of the purported correlation or any indication that Bal considered the strength of the reported associations in any way.”  Id.  Without assessing the strength of the association, it’s hard to fathom that a scientist could reliably conclude that the association is so strong as to constitute causation.

The defense painted a fairly complete picture of incompleteness.  And with that picture painted, suddenly the expert’s lack of qualifications as an epidemiologist becomes more important.  In turn, his causation opinion starts to look like a slip-shod effort by someone who lacks qualifications,
not the result of the rigorous application of a scientific methodology.  And, finally, the reasons for the expert’s earlier hedging and uncertain answer become clear.

So, after having done all this effective work at the deposition, the defense filed its Daubert motion.  Of course, it won.  Id. at *4.