Sergeants Benevolent Ass'n Health & Welfare Fund v. Louisiana
2015 U.S. App. LEXIS 19797
| 2d Cir. | 2015Background
- Plaintiffs are three health-benefit plans (HBPs) that sued Sanofi‑Aventis under RICO and state laws, alleging mail fraud based on Aventis’s alleged concealment of safety risks of the antibiotic Ketek (telithromycin). Plaintiffs sought class certification for all HBPs who paid for Ketek prescriptions (Apr 2004–Feb 2007).
- Ketek was approved by the FDA in 2004 for three respiratory indications despite a large Phase 3014 trial later found to have widespread fraud and unreliable data; the FDA relied heavily on foreign post‑marketing data.
- Post‑approval safety signals (reports of serious hepatic events, a 2006 public health advisory, label changes, and a 2007 withdrawal of two indications and a black‑box warning) coincided with a sharp decline in U.S. Ketek prescriptions.
- Plaintiffs’ causation theory: doctors prescribed Ketek because Aventis concealed material safety information, so HBPs paid for prescriptions that would not have been written if risks had been disclosed. Plaintiffs relied mainly on aggregate sales decline and expert testimony (no regression analysis isolating causes).
- The district court denied class certification (relying on this Court’s Zyprexa decision) and later granted summary judgment to Aventis because plaintiffs could not prove causation by generalized proof and offered no individualized proof. The Second Circuit affirmed.
Issues
| Issue | Plaintiff's Argument | Defendant's Argument | Held |
|---|---|---|---|
| Whether common issues predominate under Rule 23(b)(3) for a RICO mail‑fraud class | Plaintiffs: physicians’ prescribing choices were primarily one‑dimensional (safety-driven); a steep sales decline after safety disclosures shows doctors relied on Aventis’s concealment, so generalized proof can show class‑wide causation | Aventis: physicians’ prescribing decisions are individualized and multi‑factorial; correlation in sales does not prove causation for every HBP; generalized proof therefore cannot show class‑wide causation | Denied class certification — generalized proof insufficient; Zyprexa controls |
| Whether plaintiffs can prove RICO causation (but‑for and proximate) for the named HBPs at summary judgment | Plaintiffs: aggregate evidence (sales drop, expert testimony) suffices to show their own injury caused by fraud | Aventis: plaintiffs offer no individualized or robust aggregate causal analysis; correlation is inadequate; summary judgment proper | Summary judgment for Aventis — plaintiffs failed to prove causation for their individual claims |
| Whether Study 3014 or FDA actions establish material concealment making Ketek inherently so dangerous that reliance can be inferred | Plaintiffs: Study 3014 (and withheld data) showed much higher hepatic risk; FDA later withdrew indications and added warnings, supporting inference that no physician would have prescribed Ketek if risks were known | Aventis: Study 3014 data are unreliable (fraudulent); FDA votes cited effectiveness concerns as well as safety; withdrawals/warnings do not prove that every physician would have stopped prescribing | Court: record does not support the ‘‘extremely dangerous’’ drug exception; Study 3014 unreliable; withdrawals do not establish the necessary one‑dimensional reliance inference |
| Whether failure to present a regression or stronger aggregate analysis is fatal to class and individual claims | Plaintiffs: did not perform regression; relied on expert correlation opinion | Aventis: absence of regression or other robust econometric analysis leaves confounding variables unaddressed (seasonality, generic entry, marketing changes) | Court: lack of regression/robust causal analysis fatally weakens generalized proof; plaintiffs’ evidence akin to Zyprexa correlation evidence |
Key Cases Cited
- UFCW Local 1776 v. Eli Lilly & Co., 620 F.3d 121 (2d Cir. 2010) (class certification denial where physicians’ individualized prescribing decisions defeat generalized proof of RICO causation)
- Bridge v. Phoenix Bond & Indem. Co., 553 U.S. 639 (2008) (mail‑fraud predicate and causation/reliance principles)
- McLaughlin v. Am. Tobacco Co., 522 F.3d 215 (2d Cir. 2008) (individualized proof required where purchasers may have acted for diverse reasons)
- In re Neurontin Mktg. & Sales Practices Litig., 712 F.3d 21 (1st Cir. 2013) (robust regression analysis held sufficient to support causation finding for individual HBPs)
- Klay v. Humana, Inc., 382 F.3d 1241 (11th Cir. 2004) (circumstantial proof can support class‑wide inference of reliance where contract terms are central to assent)
- In re U.S. Foodservice Inc. Pricing Litig., 729 F.3d 108 (2d Cir. 2013) (payment of an inflated invoice can be circumstantial proof of reliance for class treatment)
- Holmes v. Sec. Inv. Prot. Corp., 503 U.S. 258 (1992) (RICO proximate causation requirement)
- Anderson v. Liberty Lobby, Inc., 477 U.S. 242 (1986) (summary judgment standard; speculation insufficient to defeat SJ)