858 F.3d 787
3rd Cir.2017Background
- MDL product-liability litigation: plaintiffs allege Pfizer’s SSRI Zoloft causes cardiac birth defects when taken in early pregnancy; plaintiffs relied on expert statistician Dr. Nicholas Jewell to opine on general causation.
- District Court held a Daubert hearing and excluded testimony of multiple plaintiff experts (including Dr. Anick Bérard) for unreliable methods; subsequently excluded all of Dr. Jewell’s testimony and granted summary judgment for Pfizer.
- Dr. Jewell used a weight-of-evidence/Bradford Hill approach and applied techniques including trend analysis of non-significant results, selective meta-analysis, and reanalysis of individual studies to control for confounding by indication.
- District Court criticized Jewell for inconsistent application of methods (e.g., failing to perform quantitative tests he referenced, selective use/exclusion of studies, unclear reanalyses) and for not adequately addressing larger, later studies (e.g., Furu) that failed to replicate earlier significant findings.
- Third Circuit reviewed whether the District Court imposed a legal rule requiring replicated, statistically significant epidemiological results and whether excluding Jewell was an abuse of discretion.
Issues
| Issue | Plaintiff's Argument | Defendant's Argument | Held |
|---|---|---|---|
| Whether district court required replicated, statistically significant studies as a legal threshold for admissibility | District Court erroneously imposed a legal rule that replicated, significant observational studies are required for reliability | District Court merely made a factual finding about prevailing teratology practice; no legal bright-line rule | Court held no legal standard was created; district court made a factual finding and did not categorically require significance before applying Bradford Hill |
| Whether Dr. Jewell’s use of trend analysis (including reliance on non-significant results) was reliable | Jewell: non-significant but congruent results can bolster inference of causation; trend methods and combined p-value techniques support his conclusions | Pfizer: trend analysis and selective emphasis on non-significant results are unreliable and conclusion-driven | Court held Jewell unreliably applied trend analysis (failed to perform quantitative tests, selectively emphasized results), so testimony properly excluded |
| Whether Jewell’s selective meta-analyses and study selection were admissible | Jewell: combining certain studies and reanalyzing others tested robustness and addressed heterogeneity | Pfizer: selection/exclusion of studies and inconsistent meta-analytic choices were unexplained and biased | Court held selective, unexplained meta-analytic choices rendered application unreliable and admissibility properly denied |
| Whether Jewell adequately applied Bradford Hill/weight-of-evidence to support general causation | Jewell: Bradford Hill factors support causation when weighed together despite some non-significant findings | Pfizer: Jewell failed to explain how criteria were measured/weighed and did not refute contrary higher-powered studies | Court held Jewell did not reliably apply or explain how Bradford Hill factors were weighed; exclusion was not an abuse of discretion |
Key Cases Cited
- Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (gatekeeping factors for expert admissibility)
- In re Paoli R.R. Yard PCB Litig., 35 F.3d 717 (3d Cir.) (requirement to reliably apply flexible diagnostic/weight-of-evidence methods)
- Wade-Greaux v. Whitehall Labs., Inc., 874 F. Supp. 1441 (D.V.I.) (teratology practice favoring replicated, significant epidemiological results)
- Magistrini v. One Hour Martinizing Dry Cleaning, 180 F. Supp. 2d 584 (D.N.J.) (weight-of-evidence/Bradford Hill application must explain scientific weighting)
- Milward v. Acuity Specialty Prods. Grp., Inc., 639 F.3d 11 (1st Cir.) (caution against atomistic use of Bradford Hill components)