To some extent this little run was the product of defendants doing a better job of picking their spots: Three of the five cases were suicide cases involving SSRI (that’s “selective serotonin reuptake inhibitor,” if you really want to know) anti-depressants, which is probably the industry’s strongest fact pattern for arguing preemption, another was a fraud on the FDA claim in a precedent-bound court (Michigan’s in the Sixth Circuit), and the fourth involved an unusual claim that, despite the FDA’s approval of the drug, there was no possible warning that could have made the drug “safe” under the purported common-law standard. All these were excellent preemption fodder.
Also these five opinions were all in federal court. Preemption – being the displacement of state by federal law – has naturally always been a tougher sell in state as opposed to federal courts. We’re not just talking about preemption in drug and medical device cases, but in all types of cases. The same decisional pattern exists in other types of preemption that we’ve worked on, airbags, tobacco, etc.
Well, to the point of that post, our little win streak just got snapped. The Honorable Allan L Tereshko of the Philadelphia Court of Common Pleas issued a ruling a little over a week ago finding no preemption in another suicidality case. For those of you with access to online libraries, the decision is also at Collins v. Smithkline Beecham Corp., 2008 Phila. Com. Pl. Lexis 57 (Pa. C.P. March 11, 2008) (as of yesterday Westlaw didn’t have it).
Yeah, we know that ten days is like an eternity in the blogosphere. Ten days ago, very few people had heard of the Reverend Wright’s stuff, for example. But we once again wish to remind our readers – we do not claim to be operating an unbiased reporting service. We’re defense lawyers, and in every legally-related thing we do, including this blog, we hold the interests of our clients paramount. That’s our ethical responsibility. You’ll just have to live with that.
Anyway, we knew about Collins the day after it was handed down. We (belatedly) thank Jeff Andrus at HarrisMartin Publishing for being the first of several readers to pass the Collins decision along to us. With respect to trial court decisions that are adverse to our clients, however, our policy is not to identify them until after they have been added to one or the other of the major online legal libraries (Lexis and Westlaw). We know good and well (thanks to Google Analytics) that people working at both of these services regularly read this blog, and frankly we’re not in business to make adverse decisions more widely known and available.
Sometimes it works out, sometimes it doesn’t.
With Collins it didn’t, and now that the decision’s on Lexis anyway, we can blog about it.
First things first. Generally we like Judge Tereshko (or at least that’s the word from Bexis, who practices in Philadelphia). He’s been head of the Complex Litigation center in the Philadelphia Common Pleas Court for more than a year, now, and not only is he fair to both sides, but he’s also a hard worker. In other words, he decides cases and writes opinions. Judges like Judge Tereshko are why the Philadelphia court is no longer counted as one of ATRA’s “judicial hellholes,” after making regular appearances on that list for quite a few years.
But just because we like Judge Tereshko doesn’t mean we agree with everything he decides. Not surprisingly we have our differences with the reasoning in Collins. What did you expect? We’re defense lawyers.
So now we’re tell you why.
We think that the court got off on the wrong foot when it refused to consider the FDA’s amicus briefs filed in other cases. As best we can tell, that happened because of a Pennsylvania principle that only “admissible evidence” can be used in summary judgment motions. Collins found the FDA’s briefs to be inadmissible hearsay. Slip op. at 5. We think the court misunderstood the purpose of the briefs. I don’t think anybody thinks agency (or anybody else’s) briefs should be evidence to be heard by the jury. We sure don’t. As we recently discussed in another context, the judge and only the judge instructs the jury on the law.
But that wasn’t the point of the FDA briefs in the preemption motion for summary judgment. Their purpose was purely legal – to provide the judge with information on the position of the relevant administrative agency about a matter that affected the application and construction of the Agency’s regulations. The “admissible evidence” rule doesn’t – or shouldn’t – apply to matters relevant only to questions of law, and not to issues of fact. Moreover, FDA amicus briefs count as “advisory opinions” of the Agency under 21 C.F.R. §10.85. See McNellis v. Pfizer, Inc., 2006 WL 2819046, at *3 (D.N.J. Sept. 29, 2006). Under §10.85(j), FDA advisory opinions “may be used in administrative or court proceedings” although they cannot be urged as “legal requirement[s].”
So we think, as an initial matter of procedure, that the court in Collins was incorrect when it refused to consider the FDA’s views as stated in its amicus briefs.
As is the case with just about any decision rejecting preemption in the prescription drug area, the court turns almost immediately to the “presumption against preemption.” Slip op. at 6. We’ve already discussed at length why we think that, in preemption cases arising by direct operation of the Supremacy Clause – those involving actual conflict – as opposed to preemption by statutory construction (express preemption) or preemption by implication (field preemption), there shouldn’t be any presumption one way or the other, and the scope of preemption should simply be decided by the scope of the conflict. Our earlier post demonstrated that the Supreme Court (as opposed to lots of lower courts) has not applied any such presumption in its conflict preemption cases. We hope this is an issue that the Supreme Court will resolve in the pending Levine matter. We know one of the amici briefed this issue in Kent, and we hope that happens again when the briefs are filed in Levine.
So as expected, Collins’ “bedrock principle,” slip op. at 7, is a “presumption” of dubious applicability – and indeed both of the two presumption cases that the court cited, Schemerhorn and Metropolitan Life, involved express, not implied, preemption. “Gravel” – or even “shifting sands” – would be a better description of the status of a presumption against preemption in implied conflict preemption cases.
Then the court goes off on an unusual tangent. It seems to hold that, because Congress did not include a private right of action under the FDCA, it’s “intent” must have been to preserve any and all tort claims against regulated manufacturers – even those that directly conflict with FDA decisions about what’s supposed to be in the labeling. Slip op. at 7. The support for that somewhat breathtaking proposition? A characterization of a prior Congress’ deletion of a private right of action provision by a law review article:
The draft version of the Act included a provision for creating a Federal cause of action. This provision was rejected because, “a common law right of action exists.” See Adler v. [sic] Mann, Preemption and Medical Devices: The Courts Run Amok, 59 Mo. L. Rev. 985, 924 and supporting cite at N. 130.Slip op. at 7.
The problem is (as the title of the article shows) Adler and Mann weren’t exactly neutral observers. The actual history of the FDCA doesn’t in fact say what their article said it did. Another problem is that the source material is so obscure that nobody ever checks it out. Well we have. Defendants should never sit by silently when Adler and Mann are cited for what turns out to be a very questionable proposition.
The actual legislative history of the FDCA establishes that the initial 1933 draft legislation – proposed five years before the FDCA became law – contained a specific provision allowing a private action “for injury or death proximately caused by a violation of this Act.” See S. 1944 §24, 73rd Cong., 1st Sess. (1933) (reprinted in C. Dunn, Federal Food, Drug, and Cosmetic Act, A Statement of Its Legislative Record, at 49 (G.E. Steichert & Co. 1938). The 1934 substitute version of the bill deleted this provision, see generally, S. 2000, 73rd Cong., 2d Sess. (1934) (reprinted in Dunn, FDCA Legislative Record, at 52-87). There’s almost no mention of this deletion, but a 1933 exchange between the bill’s sponsor, Sen. Royal S. Copeland (D-NY), and a witness reveals that Sen. Copeland felt private actions were “gratuitous” under existing law.
Senator COPELAND. Let me ask you about section 24, on page 31. Is that a little gratuitous?Hearing Before a Subcommittee of the Senate Committee on Congress, Statement of Walter G. Campbell, Chief, Food & Drug Administration, Department of Agriculture (Dec. 7, 1933) (reprinted in Dunn, FDCA Legislative Record, at 1106). That’s it. That’s the sum total of all the “legislative history” there is on this point. Adler and Mann’s characterization of it is, at best, misleadingly overreaching.
Mr. CAMPBELL. That is a statement of legal rights.
Senator COPELAND. They have that power now, if they will ever get it?
Mr. CAMPBELL. Right.
“Gratuitous” does not necessarily mean “duplicative” – and certainly not Adler and Mann’s claim that, “a common law right of action exists.” In fact, there was no precedent (we’ve looked), as of 1933, holding that common-law litigants could already assert violations of federal food and drug law. Moreover, a single senator’s passing comment in a hearing years before the legislation finally became law – that could be interpreted any number of ways – is hardly a valid use of legislative history. Indeed, it illustrates the proposition that the statements of individual legislators are not properly considered “legislative history” at all.
So that’s a rather long-winded discussion of the second issue we have with Collins. If the lack of a private right of action is somehow evidence that (along with the questionable presumption against preemption) is enough to establish that no conflict preemption exists, then there’s no preemption in any case, anywhere – because there’s no private right of action under the FDCA, in any case, anywhere.
The next point the court states is “critical to our analysis” in Collins is the FDA’s “changes being effected” (or “CBE”) regulation, 21 C.F.R. §314.70. See Slip op. at 8. Buying a frequently made, but quite erroneous, view of the FDA's regulations (while refusing to acknowledge the Agency's contrary view) the court interprets the CBE regulation for drugs as meaning that “a possessor of an existing NDA shall supplement the label on its product when it has reasonable evidence of a hazard, a causal relationship need not be proved.” Slip op. at 9 (quoting regulation) (emphasis original).
This statement is erroneous for several reasons. First of all, it ignores entirely that the FDA has to approve, albeit after the fact, all CBE requests – and in SSRI suicide cases the FDA specifically considered and specifically rejected the exact warning that the plaintiffs sought at least six times.
Second, it ignores the administrative history of §314.70, which – far more persuasively than the snippet the court borrowed from Adler and Mann – demonstrates that CBE supplements are intended only to be used for “new” information – an emergency situation – and not for something like the purported suicide risk, which the FDA had already evaluated several times. We demonstrated this administrative history here, and, lo and behold, the FDA itself later agreed with us, and has proposed a change to these regulations precisely to make clearer that they are intended to provide a limited, emergency exception to the general rule of Agency preapproval. That change should be effective, we hope, in the next six months or so, and at least one bogus argument will go bye-bye.
So there are lots of reasons why it’s not a good idea for a court to ignore what the FDA (or any agency) has to say about its own regulations.
Third, the minimal standard “a causal relationship need not be proved” is also a misconstruction of the FDA’s regulations. The standard that the FDA says should apply is the “substantial evidence” standard of 21 C.F.R. §201.57(c). 73 Fed. Reg. 2848, 2850 (FDA Jan. 16, 2008). The court cites 21 C.F.R. §201.80(e), but that has to do with putting a warning on the label with prior FDA approval. That’s an entirely different situation from a label change made without so much as a “by your leave.” If somebody’s going to take the extraordinary step of acting without prior FDA approval, there ought to be a really good reason for doing so. Thus the FDA states (and we agree) that “substantial evidence” must be required. If there’s no proven causal relationship, then the label change isn’t urgent enough to bypass normal FDA channels.
By refusing to consider the FDA’s positions, it may be that the court in Collins has created a bigger conflict with federal law than existed previously.
At the end, the Collins opinion circles back to the issue of no private FDCA right of action. Slip op. at 12-13. The court basically holds that, if the feds aren’t going to provide a federal claim, then preemption can’t preclude the states from providing a remedy. Id. The court offers a really long block quote from an older Supreme Court opinion called Silkwood for this proposition. Id. at 13.
The Supreme Court in Riegel, however, disposed of the Silkwood argument:
The dissent would narrow the pre-emptive scope of the term “requirement” on the grounds that it is “difficult to believe that Congress would, without comment, remove all means of judicial recourse” for consumers injured by FDA-approved devices. [citation to minority quotation from Silkwood omitted] But, as we have explained, this is exactly what a pre-emption clause for medical devices does by its terms. The operation of a law enacted by Congress need not be seconded by a committee report on pain of judicial nullification. It is not our job to speculate upon congressional motives.
Riegel v. Medtronic, Inc., 128 S. Ct. 999, 1008-09 (U.S. 2008) (emphasis added). The Supreme Court’s observation in Riegel is even more apt in a conflict preemption situation, where the conflict, rather than congressional intent, is the main issue. The reason that preemption exists is precisely to preclude states from creating common-law remedies that are contrary to federal law. The existence of a federal cause of action is, at bottom, a red herring – irrelevant to conflict preemption analysis.
There are two lessons that jump out at us from the Collins decision. The first is that it’s not a very good idea for the courts to ignore the FDA. The regulatory scheme is quite complicated, and courts acting on their own – guided only by advocates for one position or another – are likely to come to erroneous conclusions.
The second lesson is that the old Silkwood analysis never made any sense, and it’s a good thing that the Supreme Court majority in Riegel jettisoned it. The proposition that Congress wouldn’t preempt state law without making some sort of federal arrangement for would-be state law claimants is simply an excuse to eliminate implied preemption altogether – because Congress rarely creates remedies for state tort plaintiffs and when it does it specifies the preemptive scope, as with the Vaccine Act.
Elimination of conflict preemption obviously isn’t federal law, and as demonstrated by the Pennsylvania Supreme Court’s decision in Cellucci v. General Motors Corp., 706 A.2d 806, 811-12 (Pa. 1998), holding “no airbag” claims impliedly preempted due to their conflict with federal agency action even though there was no federal remedy, it isn’t Pennsylvania state law either.
So we’ll be looking for an appeal at some point. But we realize we might not get it. That’s because, Collins or no Collins, Judge Tereshko is still a good judge in our books. He’ll probably throw the case out on some other grounds (there are lots of state-law reasons why suicide cases have poor prospects) – or the defendants will win it at trial, which they’ve tended to do lately when one of these cases escapes preemption.