Pharmacovigilance Reporting and Analysis:
Product Liability Concerns
Diane P. Sullivan
Richard Hamilton III
Plaintiff lawyers have used or attempted to use adverse event reports in litigation
in several ways: (1) to prove notice of a risk and failure to warn; (2) to prove or buttress
the argument that the product causes a particular injury; and (3) to parade a recitation of
horrible events caused by a product in front of a jury to influence it on issues relating to
causation or company conduct. Post-sale marketing defect/failure-to-warn claims have
become increasingly important in modern product liability litigation. Central to
plaintiffs’ claims in many recent mass torts—including Baycol, Vioxx, and Zyprexa, to
name a few—has been the contention that even if the drug manufacturers were unaware
of the drugs’ risks at the time they initially marketed the product, subsequent
developments put the manufacturers on actual or constructive notice of those risks, giving
rise to a duty to warn or liability for failing to warn.
While any number of factors are relevant to the post-sale duty to warn, including
manufacturer-sponsored clinical trials, subsequent published medical literature, and o ff-
label promotion, this presentation focuses primarily on adverse drug experience reports
(AEs) and their potential role in litigation.
II. The Post-sale Duty to Warn
The majority of jurisdictions now impose some form of post-sale duty to warn in
the product liability context. See M. Stuart Madden, Modern Post-Sale Warnings and
Related Obligations, 27 Wm. Mitchell L. Rev. 33 (2000). Although the Second
Restatement was silent on the matter, section 10 of the Third Restatement recognizes a
cause of action for post-sale failure to warn when:
(1) the seller knows or reasonably should know that the
product poses a substantial risk of harm to persons or
(2) those to whom a warning might be provided can be
identified and can reasonably be assumed to be unaware of
the risk of harm; and
(3) a warning can be effectively communicated to and acted
on by those to whom a warning might be provided; and
(4) the risk of harm is sufficiently great to justify the
burden of providing a warning.
Restatement (Third) of Torts: Products Liability § 10(b) (1998). This cause of action is
independent of any time-of-sale defect action. See id. cmt. j. Comment c provides that
for most products, there is no continuing duty to test and monitor the product after sale
and the post-sale duty to warn arises only ―when new information is brought to the
attention of the seller.‖ However, ―[w]ith regard to one class of products, prescription
drugs and devices, courts traditionally impose a continuing duty of reasonable care to t est
and monitor after sale to discover product-related risks.‖ Drug manufacturers therefore
have a ―continuous duty to keep abreast of scientific developments touching upon the . . .
product and to notify the medical profession of additional side effects discovered from its
use.‖ Id., citing Wooderson v. Ortho Pharm. Corp., 681 P.2d 1038, 1050-51 (Kan.1984);
Schene-beck v. Sterling Drug, Inc., 423 F.2d 919, 922 (8th Cir.1970). This duty stems
both from the manufacturer’s actual knowledge of its own research and safety monitoring
and from its constructive knowledge of the current state of science. Id. Under the
Restatement, this duty runs to the medical profession under the learned intermediary
doctrine. Id. 1
III. Adve rse Drug Experience Reports
One method by which drug manufacturers are obligated to conduct post-
marketing surveillance is through adverse drug experience reports (AEs). The FDA
requires manufacturers to report ―[a]ny adverse event associated with the use of a drug in
humans, whether or not considered drug related,‖ to the Division of Pharmacovigilance
and Epidemiology at the Center for Drug Evaluation and Research. 21 C.F.R. §
314.80(a) (2007). The Federal Register notes that AEs do not necessarily constitute
conclusions by either manufacturers or the FDA that the drug caused the adverse
experience, and expressly states that the manufacturer ―need not admit, and may deny‖
that the report constitutes an admission of causation. Id. § 314.80(k).
Because the FDA requires manufacturers to report even adverse experiences that
are not related to the drug, and because reporting rates vary, the FDA has stated that
―[a]ccumulated [AE] cases may not be used to calculate incidences or estimates of drug
risk.‖2 Indeed, one public comment to an FDA concept paper on pharmacovigilance
noted that while the FDA had once routinely done causality assessments of AEs, these
were dropped in 1983 because they were of ―dubious quality,‖ relied on no standard
There may be an exception to the learned intermediary doctrine in the case of vaccines
administered by non-physicians. See, e.g., Graham v. Wyeth Lab., 666 F. Supp. 1483,
1498 n.8 (1987).
Annual Adverse Drug Experience Report: 1996 (Oct. 30, 1997), available at:
method, and were contributing to backlogging. 3 Although it is possible that a given
adverse experience was caused by the drug, it could also be caused by another drug, the
disease for which the drug was prescribed, or any other factor. 4 Some AEs nevertheless
still contain causality assessments (see below).
This is not to say that AEs are entirely useless. Large numbers of similar but
unexpected AEs may suggest an association that merits further study, and they may help
generate hypotheses. How, and whether, the manufacturer is to investigate a risk ―signal‖
based on AEs alone is a difficult question. Potential associations and hypotheses should
not be mistaken for causation, and do not, on their own, suggest drastic action.
IV. Use of Adverse Drug Expe rience Reports in Litigation
Although AEs may sometimes help manufacturers and regulators detect risk
signals, they are usually inappropriate in the courtroom. One court, in excluding an
expert’s testimony, noted that the expert’s significant reliance on AEs showed that he
―follow[ed] more of a federal agency risk analysis approach, rather than a courtroom
causation analysis. It also show[ed] that he relied on data that lacks the indicia of
scientific reliability.‖ McClain v. Metabolife Int’l, Inc., 401 F.3d 1233, 1250 (11th Cir.
Nevertheless, plaintiffs often seek to use AEs in litigation. AEs can provide
compelling evidence for plaintiffs prosecuting failure-to-warn claims. Because a
Pharmaceutical Safety Assessments Analysis (Docket no. 02N-0528; April 11, 2003),
available at: http://www.fda.gov/ohrms/dockets/dailys/03/apr03/041803/02N-0528_emc-
See Gerald Faich, Adverse Drug Reaction Monitoring, 314 New Eng. J. Med. 1589,
manufacturer is required to submit AEs even for events that were not considered related
to a drug, and because AE data is readily available online in electronic format, 5 plaintiffs
can typically dig up piles of AEs documenting an alleged side effect’s coincidence with a
drug that long predate any label change warning of that side effect. Plaintiffs can thus
claim that the manufacturer knew, or should have known, about the side effect’s relation
to the drug. Plaintiffs even attempt to use AEs to prove causation.
Fortunately, numerous courts have found that AEs are legally insufficient to prove
causation. The court in Soldo v. Sandoz Pharmaceuticals Corp., 244 F. Supp. 2d 434
(W.D. Pa. 2003) held that AEs ―themselves say that causation has not been proven,
reliance on the case reports is per se unscientific.‖ Id. at 537-38. Court holding that AEs
are inadmissible have variously noted that AEs ―make little attempt to screen out
alternative causes for a patient’s condition,‖ 6 ―are merely accounts of medical events‖
that ―reflect only reported data, not scientific methodology[,]‖ 7 ―do not demonstrate a
causal link sufficient for admission to a finder of fact in court,‖ 8 ―represent anecdotal
information of chance associations, do not purport to assess cause and effect and have no
epidemiological significance.‖ 9
Glastetter v. Novartis Pharm. Corp., 252 F.3d 986, 989-90 (8th Cir. 2001).
Rider v. Sandoz Pharm. Corp., 295 F.3d 1194, 1199 (11th Cir. 2002).
Nelson v. American Home Products Corp., 92 F. Supp. 2d 954, 969 (W.D. Mo. 2000).
Wade-Greaux v. Whitehall Laboratories, Inc., 874 F. Supp. 1441, 1481 (D.V.I. 1994).
Accord Hollander v. Sandoz Pharmaceuticals Corp., 289 F.3d 1193, 1211 (10th Cir.
2002); In re Meridia Products Liability Litigation, 328 F. Supp.2d 791, 808 (N.D. Ohio
2004); Dunn v. Sandoz Pharmaceuticals Corp., 275 F. Supp. 2d 672, 682 (M.D.N.C.
2003); Cloud v. Pfizer, Inc., 198 F. Supp. 2d 1118, 1133-34 (D. Ariz. 2001); Caraker v.
Sandoz Pharmaceuticals Corp., 172 F. Supp. 2d 1046, 1050 (S.D. Ill. 2001); Brumbaugh
This approach is not universal, and a few courts have allowed causation testimony
based in part on AEs. See, e.g., In re Phenylpropanolamine Prods. Liab. Litig., 289 F.
Supp. 2d 1230, 1242 (W.D. Wa. 2003) (allowing expert testimony on general causation
that was based in part on AEs due to ―the sheer volume of case reports, case series and
spontaneous reports associating PPA with hemorrhagic stroke to women.‖); In re
Richardson-Merrell, Inc. (“Bendectin” Products Liability Litigation), 624 F. Supp. 1212,
1233 (S.D. Ohio 1985) (allowing expert to testify about adverse experience reports).
Plaintiffs have had mixed results in attempting to use AEs to prove notice. In
Benedi v. McNeil-P.P.C., Inc., 66 F.3d 1378, 1385 (4th Cir. 1995), the court affirmed the
admission of AEs to prove notice, 10 but noted that district courts had discretion to admit
or exclude such reports under Rule 403. In theory, this approach has some merit. One of
the goals of post- marketing surveillance is to detect previously unknown risks or side
effects. And it stands to reason that AEs could provide a signal, particularly if a company
were inundated with a particular type of AE. Such a signal would of course not prove
causation, but might alert a pharmaceutical or device manufacturer to the possibility of an
association warranting further study. For the scientist, this is (hopefully) where the
v. Sandoz Pharmaceuticals Corp., 77 F. Supp. 2d 1153, 1156-57 (D. Mont. 1999); Saari
v. Merck & Co., 961 F. Supp. 387, 394 (N.D.N.Y. 1997); Haggerty v. Upjohn Co., 950 F.
Supp. 1160, 1164 (S.D. Fla. 1996), aff’d, 158 F.3d 588 (11th Cir. 1998); Hagaman v.
Merrell Dow Pharmaceuticals, 1987 WL 342949, at *8 (D. Kan. June 26, 1987);
Reynolds v. Warthan, 896 S.W.2d 823 (Tex. App. 1995); Heckstall v. Pincus, 797
N.Y.S.2d 445, 447 (N.Y.A.D. 2005); Pauley v. Bayer Corp., 2006 WL 463866, at *2 (Pa.
C.P. Jan. 26, 2006).
The court implied that had the AEs been offered for the truth of their contents, they
would have been hearsay, noting that ―the district judge properly gave a limiting
instruction to the jury that it could only consider the [AEs] and case summaries as
evidence of notice to McNeil, and not for the truth of the matter contained in them.‖
Benedi, 66 F.3d at 1386.
inferences end. 11 But a jury, particularly in the hands of a skilled plaintiffs’ lawyer,
might well make the extra leap and infer causation, reasoning that surely so many reports
must indicate the likelihood of causation. This is where proper application of Federal
Rule of Evidence 403 or its state law equivalent is absolutely critical for defendants. In
such a case, the AEs’ probative value in suggesting notice would be far outweighed by
the prejudice arising from the unwarranted and impermissible inference of causation. See
Soldo, 244 F. Supp. 2d at 546 (excluding AE causality assessments because the
likelihood of grossly misleading the jury greatly outweighed any probative value); see
also Accutane (below).
V. Recent Decisions: Baycol
Two recent decisions involving AEs merit discussion, and provide strong support
for exclusion of AE evidence in personal injury trials. In the Baycol MDL, the court
recently excluded plaintiffs’ expert opinions on the comp arative safety of Baycol versus
other drugs. See In re Baycol Prod. Liab. Litig., 495 F. Supp. 2d 977 (D. Minn. July 9,
2007). Several experts purported to use meta-analyses of AEs to determine Baycol’s
relative safety versus other cholesterol- lowering drugs. The court excluded those
experts’ testimony, finding that the data and methodologies employed did not satisfy
Daubert. The court stated that ―given the limitations inherent in [AE] data, there is
insufficient evidence before the Court as to the known or potential rate of error,‖ and
there was no evidence that these types of analyses were generally accepted. Id. at 988.
The court noted that numerous other courts had taken the same approach to expert reports
―[A]n association is not equivalent to causation.‖ Reference Manual on Scientific
Evidence at 336 (2d. Ed. Federal Judicial Center 2000).
based on AEs and observed that the FDA itself had cautioned that AE ―[r]eporting rates
can by no means be considered incidence rates, for either absolute or comparative
However, the District court refused to hold that AEs are generally inadmissible in
all contexts. The court, citing the PPA decision (discussed above) noted that the AE data
―presented a very strong signal concerning Baycol and its association with
rhabdomyolysis, and such evidence may be relevant at trial‖ (presumably to prove notice,
although the court did not specify). Id. at 990. The court therefore declined to rule,
leaving the ultimate decisions to individual trial courts on remand.
VI. Recent Decisions: Accutane
The second decision is In re Accutane Prods. Liab. Litig., 2007 U.S. Dist. LEXIS
32236 (M.D. Fla. 2007). Although the FDA no longer requires causality assessments in
AEs, some foreign regulatory agencies do; defendants therefore included them in their
AEs. Plaintiffs argued that some of these causality assessments constituted (1)
admissions that Accutane cause inflammatory bowel disease (IBD) and (2) notice of the
same to defendants. Defendants sought to exclude this evidence on Rules 402, 403, and
The court excluded the evidence. First, the court noted that the causality
assessments did not even purport to establish causation—at most, they flagged a potential
relationship. Id. at *20. Second, they did ―not contain data which is reliable and upon
which an expert opinion of causality can be based[,]‖ because there were based on
various reporters’ subjective beliefs and did not take into account patients’ medical
histories or other drugs. Id. at *22. The court found plaintiffs’ expert’s testimony
speculative and conclusory insofar as it was based on the AEs. Id. at *25.
Finally, with respect to notice, the court noted that so long as defendants admitted
that they received notice that some individuals believed that Accutane caused their
injuries, the AEs would be unnecessary. The court also held that the AEs were
inadmissible under Rule 403:
While this Court has attained an understanding of the
purpose and use of causality assessments, it is likely a lay
juror would have difficulty distinguishing that the term
―causality assessment,‖ as the term relates to safety
surveillance, is not the same as ―causation.‖ . . . In fact, it is
highly probable that a juror would perceive the company's
―yes‖ response in the causality assessment field as an
admission by Defendants’ physicians that Accutane did in
fact cause the adverse events reported. Therefore, the
potential prejudice outweighs the probative value of the
Id. at *26-27. The court realized that while scientists and highly trained product liability
attorneys may be able to distinguish between ―causality‖ and ―causation,‖ it may be
hoping too much to expect the same of a jury. The court reaffirmed this decision in a
recent order. In re Accutane Prods. Liab. Litig., MDL No. 1626, slip op. (M.D. Fla. Aug.
This decision stands in stark contrast to McCarrell v. Hoffmann-La Roche Inc.,
No. ATL-L-1951-03 (N.J. Law Div. 2007), in which the court admitted these same AE
causality assessments both for notice and causation. McCarrell was the first New Jersey
case to be tried against Roche, and approximately 400 more await their day in court.
During the four week trial, plaintiff’s counsel used the causality assessments as a
centerpiece of their case and won a $2.6M verdict.
Because they are so easy to manipulate—particularly when they contain causality
assessments—plaintiffs in pharmaceutical cases are likely to continue their attempts to
use AEs in litigation. But most recent decisions applying Daubert to AEs have held that
AEs are inadmissible to prove causation; Rule 702 motions to exclude on these grounds
therefore seem to have a high likelihood of success. And while the doctrinal problems
with admitting AEs to show notice are less severe, the potential for abuse makes AEs
prime targets for Rule 403 motions.
Early action is important, and there is no need to wait until trial to begin your
campaign against AE evidence. In addition to a motion in limine citing the growing body
of case law excluding AEs and causality assessments, use other methods to convince your
trial judge that this material is not appropriate evidence. In the mass tort context, some
judges with a consolidated docket or a major class action invest time in ―science days‖ –
a forum for attorneys on both sides to educate the Court about the specific technical
issues relevant to the litigation. Preview your themes regarding AEs, especially the
FDA’s insistence that pharmacosurveillance techniques are not to be used to infer
causation, and arm the Court with the grounds to later exclude plaintiff’s evidence. If
plaintiffs intend to call a regulatory witness at trial, ensure that your deposition
investigates her reliance on AEs in generating her opinion, and at the very least attempt to
have the expert acknowledge the limitations of AE evidence to provide you with damning
quotes for your future briefing. Motions to exclude plaintiff’s causation expert also
provide a perfect forum to attack AEs, and also to have the ―first word‖ on the issue in
your moving papers. Finally, if your regulatory expert has reason to appear before the
Court prior to trial, spend a few minutes educating the Court on pharmacosurveillance
and the fundamental disconnect between AEs and causation.
As will be discussed in the presentations, manufacturers can use the
pharmacosurveillance process to limit litigation or litigation exposure in the first place.