Jones v. Novartis Pharmaceuticals Company

Filing 195

MEMORANDUM OPINION AND ORDER DENYING 108 MOTION to Strike Expert Testimony of Dr. Suzanne Parisian to the extent that it seeks to exclude Dr. Parisian's testimony in its entirety. However, Novartis' Motion is due to be GRANTED IN PA RT and DENIED IN PART as to the particular areas of Dr. Parisian's testimony that it seeks to exclude. GRANTING 112 MOTION to Strike Expert Testimony of Dr. William B. Hinshaw, GRANTING 116 MOTION to Strike Expert Testimony of Wayn e A. Taylor, Ph.D., GRANTING 118 MOTION to Strike Expert Testimony of Non-retained Experts to the extent that either expert intends to offer an opinion on causation and is otherwise DENIED. Signed by Judge Virginia Emerson Hopkins on 1/26/2017. (JLC)

FILED
2017 Jan-26 PM 02:36
U.S. DISTRICT COURT
N.D. OF ALABAMA

IN THE UNITED STATES DISTRICT COURT
FOR THE NORTHERN DISTRICT OF ALABAMA
SOUTHERN DIVISION
ERNESTEEN JONES,

)
)
Plaintiff,
)
)
v.
) Case No.: 2:13-CV-624-VEH
)
NOVARTIS PHARMACEUTICALS )
CORPORATION,
)
)
Defendant.
)

MEMORANDUM OPINION AND ORDER
I.

INTRODUCTION
This case comes before the court on Defendant Novartis Pharmaceutical

Corporation (“Novartis” or “NPC”)’s Motions To Strike Expert Testimony. Novartis
has moved to exclude the testimony of the following experts:
•

Dr. Suzanne Parisian (“Dr. Parisian”), the “Parisian Motion”
(doc. 108);1

•

Dr. William B. Hinshaw (“Dr. Hinshaw”), the “Hinshaw
Motion” (doc. 112);

•

Dr. Wayne A. Taylor (“Dr. Taylor”), the “Taylor Motion” (doc.
116);

1

system.

All page references to (Doc. __) correspond with the court’s CM/ECF numbering

•

II.

Dr. James Worthen (“Dr. Worthen”) and Dr. Timothy Mark
Ricketts (“Dr. Ricketts”), collectively in the “Non-Retained
Experts Motion” (doc. 118).

PROCEDURAL HISTORY
Plaintiff Ernesteen Jones (“Jones”) initiated this lawsuit against Novartis on

April 4, 2013 (doc. 1), alleging that she developed atypical femur fractures
(“AFF”)2 as a result of taking Novartis’ medication Reclast, which is a type of
bisphosphonate (“BP”) drug. Jones was prescribed Reclast by Dr. Thomas Traylor,
her treating physician for her osteoporosis. (Doc. 54) at 2, ¶ 9.3 She was
administered an annual five milligram Reclast injection, as prescribed, on
February 10, 2009, March 16, 2010, and March 17, 2011. Id. at 2, ¶ 8.
On October 26, 2011, Jones’s right femur fractured, requiring surgery. Id. at
3, ¶¶ 13-14. In early 2012, Jones began experiencing pain in her left thigh. Id. at 3,
¶ 16. After a bone scan revealed a stress fraction on her left femur, she had surgery
on her left femur to prevent a complete fracture. Id. at 3, ¶¶ 17-18.
Jones has asserted the following claims against Novartis: violations of the

2

An atypical femur fracture is an “atraumatic or low-trauma fracture[] located in the
subtrochanteric region or femoral shaft” that presents with specific “major” and “minor” clinical
features. (Doc. 125-19) at 3 n. 1(citing Shane E. et al., Atypical Subtrochanteric and Diaphyseal
Femoral Fractures: Second Report of a Task Force of the American Society for Bone and Mineral
Research, J. Bone Miner Res. 2, 13 (2013)(doc. 117-1)).
3

Jones filed a Third Amended Complaint on February 6, 2015. (Doc. 54). This statement
of alleged facts and background derives from that Amended Complaint.
2

Alabama Extended Manufacturer’s Liability Doctrine (“AEMLD”) (Count 1, id. at
6-9); failure to warn (Count II, id. at 10); negligence and wantonness (Count III,
id. at 10-12); and breach of warranty of merchantability. (Count IV, id. at 13).
III.

STANDARD FOR THE ADMISSIBILITY OF EXPERT TESTIMONY
A.

General Requirements - Judge as Gatekeeper

Regarding expert testimony, the Federal Rules of Evidence provide:
A witness who is qualified as an expert by knowledge, skill, experience,
training, or education may testify in the form of an opinion or otherwise
if:
(a) the expert's scientific, technical, or other specialized
knowledge will help the trier of fact to understand the evidence
or to determine a fact in issue;
(b) the testimony is based on sufficient facts or data;
(c) the testimony is the product of reliable principles and
methods; and
(d) the expert has reliably applied the principles and methods to
the facts of the case.
FED. R. EVID. 702 (2011). Rule 702 must be read in conjunction with three
seminal decisions by the Supreme Court related to expert testimony: Daubert v.
Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S. Ct. 2786, 125 L. Ed. 2d
469 (1993); Gen. Elec. Co. v. Joiner, 522 U.S. 136, 118 S. Ct. 512, 139 L. Ed. 2d
508 (1997); and Kumho Tire Co. v. Carmichael, 526 U.S. 137, 119 S. Ct. 1167,
3

143 L. Ed. 2d 238 (1999).
All rulings on Daubert motions are reviewed under an abuse of discretion
standard. See, e.g., Joiner, 522 U.S. at 141, 118 S. Ct. at 517 (“All evidentiary
decisions are reviewed under an abuse-of-discretion standard.”). “An abuse of
discretion can occur where the district court applies the wrong law, follows the
wrong procedure, bases its decision on clearly erroneous facts, or commits a clear
error in judgment.” United States v. Estelan, 156 F. App’x 185, 196 (11th Cir.
2005) (citing United States v. Brown, 415 F.3d 1257, 1266 (11th Cir.2005)).
In Daubert, the Supreme Court established that district judges act as
“gatekeepers” for expert testimony. 509 U.S. at 592–93, 113 S. Ct. at 2796. The
district court judge must assess the proffered testimony and make a preliminary
determination about the scientific validity of the expert's reasoning and
methodology. Id.
As another district court in this Circuit has stated,
Federal Rule of Evidence 702, read together with the trilogy of Supreme
Court opinions that led to the Rule's revision in 2011, compels the
district courts to perform a “gatekeeping” function when determining the
admissibility of expert scientific and technical evidence. See, e.g.,
United States v. Abreu, 406 F.3d 1304, 1306 (11th Cir. 2005) (quoting
United States v. Frazier, 387 F.3d 1244, 1260 (11th Cir. 2004)). “This
function inherently requires the trial court to conduct an exacting
analysis of the foundations of expert opinions to ensure they meet the
standards for admissibility under Rule 702.” Id. (internal quotation
4

omitted).
Broussard-Wadkins v. Maples, 895 F. Supp. 2d 1159, 1165 (N.D. Ala. 2012), aff'd
sub nom. Broussard v. Maples, 535 F. App'x 825 (11th Cir. 2013).
The burden under Rule 702 rests squarely with the proponent of the expert
witness:
The proponent of the expert testimony carries a substantial burden under
Rule 702. “The burden of laying the proper foundation for the admission
of the expert testimony is on the party offering the expert, and
admissibility must be shown by a preponderance of the evidence.”
Allison v. McGhan Med. Corp., 184 F.3d 1300, 1306 (11th Cir.1999)
(citing Daubert, 509 U.S. at 592 n. 10, 113 S. Ct. 2786). Thus, the
proponent must demonstrate that the witness is qualified to testify
competently, that his opinions are based on sound methodology, and that
his testimony will be helpful to the trier of fact. See, e.g., Frazier, 387
F.3d at 1260 (“The burden of establishing qualification, reliability, and
helpfulness rests on the proponent of the expert opinion . . . .”);
McCorvey v. Baxter Healthcare Corp., 298 F.3d 1253, 1257 (11th Cir.
2002); Maiz, 253 F.3d at 664.
See Cook ex rel. Estate of Tessier v. Sheriff of Monroe Cty., Fla., 402 F.3d 1092,
1107 (11th Cir. 2005).
B.

The Eleventh Circuit Test for Admissibility

The Eleventh Circuit has established a three-part inquiry for district courts
to follow in performing their gatekeeper role. For evidence to be admissible under
Rule 702, the district court must find that:
(1) the expert is qualified to testify competently regarding the matters he
5

intends to address;
(2) the methodology by which the expert reaches his conclusions is
sufficiently reliable as determined by the sort of inquiry mandated in
Daubert; and
(3) the testimony [will] assist[ ] the trier of fact, through the application
of scientific, technical, or specialized expertise, to understand the
evidence or to determine a fact in issue.
Hendrix ex rel. G.P. v. Evenflo Co., Inc., 609 F.3d 1183, 1194 (11th Cir. 2010)
(citing United States v. Frazier, 387 F.3d 1244, 1260 (11th Cir. 2004)). The party
offering the testimony must meet each prong by a preponderance of the evidence.
Id.
1.

Prong One: The Expert Must Be Qualified To Testify to the
Relevant Issue

To meet Prong One, a party must show that the expert has sufficient
“knowledge, skill, experience, training, or education” to form a reliable opinion
about the relevant issue. Hendrix, 609 F.3d at 1193. Experience in a particular
field is not enough to qualify an expert; the expert must have experience with the
issue before the court. See id. at 1201.
The Sixth Circuit, in a similar case, concluded that a district court did not
abuse its discretion in excluding the testimony of the plaintiff's expert. See
Thomas v. Novartis Pharms. Corp., 443 F. App’x 58, 63 (6th Cir. 2011). The

6

expert in Thomas was “an experienced maxillofacial surgeon who ha[d] treated
several patients suffering from osteonecrosis of the jaw.” Id. However, the expert
had not established his credentials to “diagnose the cause of [the plaintiff's]
osteonecrosis . . . which [was] the salient issue his opinion was offered to
establish.” Id.
2.

Prong Two: The Expert’s Opinion Must Be Sufficiently
Reliable

To meet Prong Two, the party proffering the expert’s testimony must show
that the expert’s opinion is sufficiently reliable. A district court has substantial
discretion in deciding how to test the reliability of an expert's testimony. Rink v.
Cheminova, Inc., 400 F.3d 1286, 1292 (11th Cir. 2005). “This deferential abuse of
discretion standard is applied stringently, even if a decision on expert testimony is
‘outcome determinative.’” Chapman v. Proctor & Gamble Distrib., LLC, 766 F.3d
1296 (11th Cir. 2014) (citing Joiner, 522 U.S. at 142-43, 118 S. Ct. at 517).
Pursuant to the second Daubert prong, the court should consider the
following factors: “(1) whether the expert's methodology can be tested; (2)
whether the expert's scientific technique has been subjected to peer review and
publication; (3) whether the method has a known rate of error; and (4) whether the
technique is generally accepted by the scientific community.” Rink, 400 F.3d at

7

1292 (citing Quiet Tech. DC–8, Inc. v. Hurel–Dubois UK Ltd., 326 F.3d 1333,
1341 (11th Cir. 2003)). However, these factors are not exhaustive and a court
“should consider any additional factors that may advance its Rule 702 analysis.”
Quiet Tech, 326 F.3d at 1341.
C.

Legal Standard for Causation Experts

In the Eleventh Circuit, other than in a small number of cases where the
medically community generally recognizes and agrees upon the toxicity of the
substance at issue to the injury alleged, both general and specific causation must
be established through expert testimony if the Plaintiff’s claims require proof of
causation:
For analyzing cases involving alleged toxic substances, we have
delineated two categories. McClain v. Metabolife Int'l, Inc., 401 F.3d
1233, 1239 (11th Cir. 2005). The first category consists of “cases in
which the medical community generally recognizes the toxicity of the
[substance] at issue” to “caus[e] the injury plaintiff alleges.” Id.;
Hendrix ex rel. G.P. v. Evenflo Co., 609 F.3d 1183, 1196 (11th Cir.
2010). In this category are “toxins like asbestos, which causes asbestosis
and mesothelioma; silica, which causes silicosis; and cigarette smoke,
which causes cancer.” McClain, 401 F.3d at 1239 . . . [i]n cases where
the cause and effect or resulting diagnosis has been proved and accepted
by the medical community, federal judges “need not undertake an
extensive Daubert analysis on the general toxicity question.” Id. at
1239.
In contrast, the second category contains cases, where the medical
community generally does not recognize the substance in question as
being toxic and having caused plaintiff's alleged injury. Id. These cases
8

require a two-part Daubert analysis, comprised of (1) general causation,
“whether the [substance] can cause the harm plaintiff alleges,” id., and
(2) specific causation, whether experts' methodology determines the
substance “caused the plaintiff's specific injury,” Hendrix, 609 F.3d at
1196 (citing McClain, 401 F.3d at 1239). For cases in category two, a
district judge “must assess the reliability of the expert's opinion on
general, as well as specific, causation.” Id. (first emphasis added).
Chapman, 766 F.3d at 1303-04. This case clearly falls into the latter category.
Accordingly, reliable expert testimony must establish both general and specific
causation.
1.

General Causation Standard

General causation refers to the “general issue of whether a substance has the
potential to cause the plaintiff’s injury.” Chapman, 766 F.3d at 1307 (citing Guinn
v. AstraZeneca Pharm. LP, 602 F.3d 1245, 1248 n.1 (11th Cir. 2010) (per
curiam)); see also McClain v. Metabolife Intern., Inc., 401 F.3d 1233, 1239 (11th
Cir. 2005) (emphasis in original) (“In the second category of toxic tort cases, the
Daubert analysis covers not only the expert's methodology for the
plaintiff-specific questions about individual causation but also the general
question of whether the drug or chemical can cause the harm plaintiff alleges. This
is called general causation.”). General causation exists “when a substance is
capable of causing a given disease.” Restatement (Third) of Torts: Liability for
Physical and Emotional Harm § 28 cmt. c(3) (2010) (“Restatement”).
9

2.

Specific Causation Standard

Specific causation refers to “the issue of whether the plaintiff has
demonstrated that the substance actually caused injury in her particular case.
Specific causation is often distinguished from general causation, which refers to
the more general issue of whether a substance has the potential to cause the
plaintiff's injury.” Guinn, 602 F.3d at 1248 n. 1; Chapman, 766 F.3d at 1308; see
also Restatement § 28 cmt. c(3) (“Scientists who conduct group studies do not
examine specific causation in their research. No scientific methodology exists for
assessing specific causation for an individual based on group studies.
Nevertheless, courts have reasoned from the preponderance-of-the-evidence
standard to determine the sufficiency of scientific evidence on specific causation
when group-based studies are involved.”).
The Eleventh Circuit in McClain laid out four criteria, based on an article
published by the Federal Judiciary Center, for proving causation between exposure
to a chemical and a particular illness in an individual:
1. “[T]he toxic substance in question must have been demonstrated to
cause the type of illness or disease in question . . . this focuses on the
issue of general causation;”
2. “[T]he individual must have been exposed to a sufficient amount of
the substance in question to elicit the health effect in question . . . this
focuses on the issue of individual causation;”
10

3. “[T]he chronological relationship between exposure and effect
must be biologically plausible . . . this also focuses on individual
causation;”
4. “[T]he likelihood that the chemical caused the disease or illness in
an individual should be considered in the context of other known
causes . . . this refers to the background risk of a specific disease.”
401 F.3d at 1242-43 (citing David Eaton, Scientific Judgment and Toxic Torts–A
Primer in Toxicology for Judges and Lawyers, 12 J.L. & POL’Y 1, 38-40 (2003)).
D.

Lack of Necessity of a Daubert Hearing

Whether a Daubert hearing is necessary is a decision within the sound
discretion of a district court. Cook, 402 F.3d at 1113. The abuse of discretion
standard “applies as much to the trial court’s decisions about how to determine
reliability as to its ultimate conclusion . . . [i]ndeed, the Rules seek to avoid
unjustifiable expense and delay as part of their search for truth and the just
determination of proceedings.” Kumho, 526 U.S. at 139, 152-53 (internal citations
omitted). There is no requirement that a Daubert hearing always be held. See
United States v. Hansen, 262 F.3d 1217, 1234 (11th Cir. 2001); Frazier, 387 F.3d
at 1264.
In this case, after extensively reviewing the parties’ briefings and exhibits,
the court determined that a Daubert hearing was not necessary. Furthermore, the
parties agreed to submit their Daubert-related Motions on their respective briefs.
11

(See Doc. 194). Accordingly, the court made a determination as to the
admissibility of testimony by Dr. Parisian, Dr. Hinshaw, Dr. Taylor, Dr. Worthen,
and Dr. Ricketts based on the briefing submitted by the parties.
IV.

ANALYSIS
A.

Dr. Suzanne Parisian

Jones offers the opinions of Dr. Parisian into evidence in this case. Dr.
Parisian has offered an expert report, dated August 31, 2015. (Doc. 125-22, the
“Parisian Report”). Dr. Parisian also offered a supplemental expert report, dated
May 2, 2016. (Doc. 125-23, the “Parisian Supplemental Report”). Dr. Parisian was
deposed on June 14, 2016, and the deposition transcript was filed into the record.
(Doc. 125-13, the “Parisian Deposition”).
On August 15, 2016, Novartis filed a Motion To Strike Dr. Parisian’s expert
testimony (Doc. 108). A day later, Novartis filed a brief in support of its Motion
(doc. 125-15, the “Parisian Brief”). On September 23, 2016, Jones filed a
Response opposing Novartis’ Motion To Strike. (Doc. 166-39, the “Parisian
Response”). On October 14, 2016, Novartis filed a reply brief in support of its
Motion To Strike. (Doc. 175, the “Parisian Reply”).
1.

Dr. Parisian’s Qualifications

Dr. Parisian has been offered by Jones to address “regulatory requirements
12

applicable to pharmaceutical manufacturers and drug labeling.” Parisian Response,
(Doc. 166-39) at 4. She formerly served as a Medical Officer at the Food and Drug
Administration (“FDA”). Parisian Curriculum Vitae, (Doc. 109-5) at 1. She is
board-certified in anatomic and clinical pathology and holds a masters degree in
biology. Id. at 16; Parisian Report, (Doc. 125-22) at 2, ¶2. From 1991 to 1995, Dr.
Parisian served as Lieutenant Commander in the Unites States Public Health
Service and was assigned to the FDA. Id. at 2, ¶4. From 1991 to 1995, she was
also employed by the FDA in several roles, including as Medical Officer in the
Office of Health Affairs, Chief Medical Officer in the Office of Device
Evaluation, and as an instructor in the FDA’s staff college. Id. at 2, ¶5.
While at the FDA, her responsibilities included assessing the health hazard
and health risk of products; assessing safety alerts, reviewing adverse event
reports; drafting and review of product labeling; promotions, advertising, and
corporate records; pre-market evaluation of new product applications and clinical
trials; review of marketing applications for safety and efficacy; and training of
FDA reviewers regarding the design and evaluation of clinical data for proposed
products. Id. at 2-3, ¶ 5.
As a Medical Officer in the Office of Device Evaluation (“ODE”), Dr.
Parisian participated in the review of proposed clinical trials and pre-marketing
13

applications. Id. at 4, ¶9. She also trained new medical officers and scientific
reviewers in application, clinical trial, and labeling evaluations as well as methods
for health risk assessments, health hazard evaluations, annual report requirements,
adverse event reporting, and labeling review. Id. As an instructor in the FDA’s
staff college, she had primary responsibility for review of marketing applications
and labeling and instruction on evaluation and review of product marketing. Id. at
5, ¶ 11. She also presided over 162 health risk assessments convened to advise the
FDA on health risk issues for the public. Id. at 3, ¶6. Dr. Parisian advised and
trained other FDA employees regarding FDA requirements and how to make a
determination regarding the clinical impact of FDA actions on the public. Id. at 56, ¶14.
Since leaving the FDA in 1995, Dr. Parisian has served as a consultant on
issues related to the FDA and is the owner and founder of MD Assist, Inc. (“MD
Assist”), a firm specializing in FDA regulation of products. Id. at 1, ¶1. She is also
the author of FDA Inside and Out, a textbook on the history, rules, and regulations
of the FDA. Id. at 2, ¶3. At MD Assist, she designs and markets new medical
products; presents marketing applications to the FDA; designs, conducts, and
reviews clinical studies, creates and evaluates marketing applications; drafts
product labeling; investigates potential adverse events; and instructs about the
14

requirements of the FDA. She has provided litigation support since 1997. Id. at 7,
¶20.
Dr. Parisian claims that she reached her opinions in this case using the
methodology that she was first trained to use at the FDA for clinical review and
health risk assessment, as well as her own education, training, and experience. Id.
at 8, ¶21. Jones’s response brief states that her primary role is threefold: she will
explain the responsibilities of a pharmaceutical drug manufacturer in the context
of the FDA’s regulatory scheme both before and after a drug is placed on the
market; she will provide opinions on the failure of Novartis to act as a reasonable
pharmaceutical manufacturer by failing to comply with FDA regulations; and she
will explain the meaning of specialized terminology. Parisian Response, (Doc.
166-39) at 5.
2.

Dr. Parisian’s Opinions

In her report, Dr. Parisian offers the following opinions:
(1)There is a trend in the United States for increasing numbers of
non-healing atypical femur fractures in women. This trend is a
national public health safety issue, (doc. 125-22) at 9-10;
(2) Novartis failed to adequately design, study, test, research, or
investigate whether Reclast could safely be used by postmenopausal patients, id. at 10-33;
(3) Novartis failed to effectively use pharmacovigilance
15

surveillance tools to identify a signal with zoledronic
acid/Reclast, id. at 33-41;
(4) Novartis is required to submit truthful and accurate
information to the FDA. Novartis failed to adequately and timely
update the FDA with relevant safety information regarding the
risk of femur fracture, id. at 41-50;
(5) It is not the responsibility of the FDA to ensure the adequacy
of a drug label; that responsibility lies with the drug sponsor, id.
at 50-51;
(6) Novartis did not adequately warn patients of the risks of
Reclast in its patient brochure, id. at 52;
(7) As the Reclast sponsor, Novartis was required to provide
physicians with adequate warnings about its drugs. Novartis did
not adequately warn prescribing physicians regarding the risk of
femur fracture associated with Reclast, id. at 52-58;
(8) Novartis, through its sales force, including Kelly Pyron,
helped delay providing health care providers with adequate and
truthful information about the long-term risks of Reclast.
Novartis’ drug advertising and communications with health care
providers were not fair and balanced, id. at 58-66;
(9) Novartis’ conduct regarding acknowledging a risk of femur
fractures is similar to its conduct regarding bisphosphonaterelated osteonecrosis of the jaw, id. at 66-71;
(10) Ernesteen Jones fit the profile for Novartis’ marketing
strategy, id. at 71-74;
(11) The report on Ms. Jones’s adverse events is indicative of
Novartis’ failure to conduct meaningful pharmacovigilance, id. at
74-75;

16

(12) Novartis’ actions regarding Reclast do not support that
patient safety was its highest priority. Its actions were misleading
and directly impacted the quality of life of patients, including
Ernesteen Jones, id. at 75.
Dr. Parisians’s supplemental report offers the following opinions:
(1) Novartis submitted a 2008 report to the FDA that was
inaccurate and misleading and downplayed the risk of AFF with
Reclast, (doc. 125-23) at 6-8;
(2) Novartis in 2015 submitted inaccurate and misleading
information to the FDA in its Errata Document regarding AFF, id.
at 8-25;
(3) Opinions on the Expert Report of Dr. David Feigal, id. at 2643.
3.

Novartis’ Motion To Exclude Dr. Parisian’s Testimony Is
Granted in Part and Otherwise Is Denied

Novartis first argues that all of Dr. Parisian’s testimony should be fully
excluded because she lacks adequate expertise, she does not employ a reliable
methodology, and her opinions would not assist the jury.
The court finds that Dr. Parisian is qualified, based on her experience at the
FDA as a Medical Officer, to offer testimony about regulatory requirements for
the testing, marketing, and development of prescription drugs. Even though her
time at the FDA was primarily spent on medical devices, Dr. Parisian has
extensive experience with the regulation of drugs. Her report further demonstrates

17

that she has particularized knowledge of the FDA standards that apply to drug
manufacturers.
She has also followed an appropriate methodology. Experts are permitted to
draw conclusions from a set of observations that are based on their extensive and
specialized experience. Kumho Tire, 526 U.S. at 156, 119 S. Ct. at 1178. Dr.
Parisian is a regulatory expert who has drawn conclusions based on her review of
regulatory filings and company documents in light of FDA regulations and her
experience at the FDA. She claims she used the methodology she was trained to
use at the FDA “for clinical review and health risk assessment, as well as her
scientific and medical education, professional training, and experience.” Parisian
Report, (Doc. 125-22) at 7-8, ¶¶ 20-21. However, the court’s general acceptance
of her methodology does not mean that all aspects of her testimony will be
admitted.
Further, to the extent that her opinions are admissible, her assessment would
be helpful to the jury to understand the complex, technical topic of FDA
regulations. See In re Mirena IUD Products Liability Litigation, 169 F. Supp. 3d
396, 474 (E.D.N.Y. 2016) (“Expert testimony from a regulatory expert on
complicated schemes like the FDA’s statutory framework, as well as opinions on
the adequacy of a drug’s label and the reasonableness of a pharmaceutical
18

company’s conduct, are useful in assisting the trier of fact.”).4
Novartis relies on Hogan v. Novartis Pharms. Corp., No. 06-CV-260, 2011
WL 1533467 (E.D.N.Y. Apr. 24, 2011), which is one of the few cases where the
court fully excluded all testimony by Dr. Parisian.5 However, Hogan differs from
the current litigation because, in that case, the plaintiff brought common law
claims that were unrelated to FDA regulations, so testimony from a regulatory
expert was not needed. Id. at *2. Here, Jones argues that FDA regulations impact
her claims, so Dr. Parisian’s testimony and her expertise would be helpful to a
jury. Parisian Response, (Doc. 166-39) at 15.
Novartis also relies on In re Prempro Prod. Liab. Litig., 554 F. Supp. 2d
871 (E.D. Ark. 2008), aff’d in part, rev’d in part, 586 F.3d 547 (8th Cir. 2009), to

4

The court in In re Mirena also found that the methodology used by Dr. Parisian and
other regulatory experts does not have to strictly conform to the four-factor model found in
Daubert because those factors apply to scientific expert opinions, not non-scientific regulatory
reports. 169 F. Supp. 3d at 480.
5

Dr. Parisian has served as a plaintiff’s regulatory expert in almost 100 cases against
Novartis. Of those cases, only a handful have excluded her testimony completely, and Novartis
cites repeatedly to those cases in its briefing. See In re Trasylol Prods. Liab. Litig., 709 F. Supp.
2d 1323, 1345 (S.D. Fla. 2010); Kaufman v. Pfizer Pharmaceuticals, No. 02-CV-22692, 2011
WL 7659333 (S.D. Fla. Aug. 4, 2011) (finding Dr. Parisian failed to provide reliable bases for
her testing opinion); Lopez v. I-Flow Inc., 2001 WL 1897548 (D. Ariz. Jan. 26, 2011) (excluding
Dr. Parisian’s testimony because she gave narrative testimony and formed improper legal
conclusions). However, the overwhelming number of courts who have analyzed Dr. Parisians
qualifications and methodology have allowed her to testify, at least in part. In any event, each
court addressing a Daubert challenge to an expert, such as this court, must decide the challenge
based on its application of relevant law to the facts before it.
19

demonstrate that Dr. Parisian’s testimony should be excluded. The district court in
In re Prempro initially allowed her to testify in both the compensatory and
punitive damages phases of the trial. However, the district court struck her
punitive damages testimony post-trial because Dr. Parisian had mainly read from
documents and provided unhelpful narrative testimony. 586 F.3d at 570. The
Eighth Circuit held that the district court’s error in admitting Dr. Parisian’s
testimony in the punitive damages phase was prejudicial error that warranted a
new trial on punitive damages. Id. at 573. While In re Prempro guides the court to
scrutinize Dr. Parisian’s testimony closely for narrative testimony or other
impermissible statements, it does not suggest that Dr. Parisian’s testimony should
be excluded in full.
To support her claim that Dr. Parisian’s testimony is credible, Jones cites to
several cases where Dr. Parisian’s testimony was permitted in part and excluded in
part. See, e.g., Kruska v. Novartis Pharms. Corp., 28 F. Supp. 3d 920, 934 (D.
Minn. 2014) (denying motion to fully exclude Dr. Parisian’s testimony); Lemons v.
Novartis Pharms. Corp., 849 F. Supp. 2d 608, 613 (W.D.N.C. 2012) (“Exclusion
of [Dr. Parisian’s] testimony in the entirety would be inappropriate.”).
As explained below, Dr. Parisian is qualified and will be permitted to testify
about the FDA regulatory process. However, her testimony will be limited and
20

excluded in part. Jones agrees that Dr. Parisian will not testify as to Novartis’
intent or state of mind, so any testimony on these issues is not admissible. Parisian
Report, (Doc. 125-22) at 9, ¶ 23. Additionally, Dr. Parisian is not qualified and
will not be permitted to testify regarding causation or “causal association”;
whether and when Novartis was put on “notice”; whether any advertising or
marketing changes might have impacted the opinion of a physician; studies
conducted on other BP drugs; pharmaceutical industry standards; correlations
between AFF and osteonecrosis of the jaw; and other improper legal conclusions.
a.

Causation and “Causal Association”

Novartis argues that Dr. Parisian lacks the expertise to testify on either
causation or “causal association.” In her deposition, Dr. Parisian agreed that she is
not a causation expert and does not intend to offer a medical opinion about
causation. Parisian Deposition, (Doc. 125-13) at 13(46:9-47:2) (“For Mrs. Jones –
in terms of one patient, I’m not the causation person. It has to be the surgeon
who’s taking care of her.”). As a regulatory expert rather than a medical expert,
Dr. Parisian does not have the expertise or qualifications to testify as to general or
specific causation.
Jones argues, however, that Dr. Parisian should be able to testify relating to
“causal association,” as the term is used in FDA regulations. See 21 C.F.R.
21

201.57(c)(6)(i) (providing that a drug’s “labeling must be revised to include a
warning about a clinically significant hazard as soon as there is reasonable
evidence of a causal association with a drug; a causal relationship need not have
been definitively established.”). In her Response brief, Jones attempts to
distinguish Dr. Parisian’s use of causal association terminology from medical
causation. Parisian Response, (Doc. 166-39) at 27-31; see Parisian Deposition,
(Doc. 125-13) at 78-79(309:13-310:1)(“[I]n terms of the way the FDA is talking
about causal association, it’s not that high of a bar. It’s not causation.”)). Novartis
counters that Dr. Parisian is engaging in “verbal semantics” and is attempting to
create a back-door method of testifying to medical causation without violating her
stipulation. Parisian Reply Brief, (Doc. 175) at 7.
Dr. Parisian’s attempts to distinguish her opinion as a “causal association”
opinion rather than a causation opinion fall short of the mark. First, Dr. Parisian
has failed to sufficiently define causal association, as found in 21 C.F.R.
201.57(c)(6)(i), or distinguish it from medical causation.6 She states in her
deposition that in order to make a label change pursuant to FDA regulations, you

6

See Dopson-Troutt v. Novartis Pharms. Corp., No. 06-CV-1708, 2013 WL 1344755 at
*3 (M.D. Fla. April 2, 2013) (finding Dr. Parisian failed to meaningfully distinguish causal
association from causation, so her regulatory causation discussion would not assist the jury’s
decision-making process); Rowland v. Novartis Pharms. Corp., 9 F. Supp. 3d 553, 562 (W.D. Pa.
2014) (excluding Dr. Parisian from offering “causation testimony of any kind”).
22

must have “reasonable evidence of a causal association.” Parisian Deposition,
(Doc. 125-13) at 78(309:22-310:1). Dr. Parisian will be permitted to tell the jury
what FDA regulations require in terms of label changes, including what action is
required if a drug manufacturer has “reasonable evidence” of a causal association
between a drug and an injury. However, she will not be permitted to testify
whether Novartis had reasonable evidence or whether Novartis should have
changed its label for Reclast.
Despite her assertions to the contrary, Dr. Parisian has implicitly provided
her own causation opinion in both her report and her deposition. See Parisian
Supplemental Report, (Doc. 125-23) at 38-39, ¶ 93 (“[bisphosphonates] may not
be the only cause of AFF, but it is one drug class that has been linked to AFF.”);
Id. at 38, ¶92 (“it is my opinion that . . . there is ‘reasonable evidence of a causal
association’ sufficient to trigger Novartis’ duty to update its Reclast label to
provide adequate risk information.”) (emphasis in original).
In her Supplemental Expert Report, Dr. Parisian states that she has based
her causation association opinion on “the evidence of [bisphosphonates] and
zoledronic acid”; the “occurrence of femur fractures in Study 2301"; the “lack of
adequate pharmacovigilance practices to work up femoral fractures”; and the
“literature.” Parisian Supplemental Report, (Doc. 125-23) at 38, ¶ 92. However,
23

Dr. Parisian cites to no specific medical literature establishing a causal association,
and she neither specifies why Study 2301 demonstrated a causal association nor
explains why Novartis should be responsible for the lack of adequate
pharmacovigilance practices. Id. Dr. Parisian is not an endocrinologist and does
not otherwise have the expertise to review medical literature on femoral fractures
and opine on whether a correlation between a disease and an alleged injury
amounts to a causal association.
As another district court recently determined, Dr. Parisian is not qualified to
opine “either that the risk was clinically significant or that there was reasonable
evidence of causal association”; she could “testify as to what Defendants should
have done in terms of investigation based on the post-marketing information
available to them, but may not opine that that information amounted to evidence of
causal association sufficient to warrant a label change.” In re Mirena, 169 F. Supp.
3d at 476 n.76.
Jones cites to Hill v. Novartis Pharms. Corp., No. 06-CV-939, 2012 WL
5451809 (E.D. Cal. Nov. 7, 2012) to support her argument that Dr. Parisian is
qualified to opine as to whether a pharmaceutical company should be responsible
for failing to warn of a causal association. However, the court in Hill merely
denied the defendant’s motion to exclude Dr. Parisian’s testimony on causal
24

association as a “causation opinion in disguise” without providing any explanation
to help guide this court’s analysis. Therefore, this court will not rely on Hill as
persuasive authority.
Dr. Parisian’s expertise lies in FDA regulations and in corporate compliance
with FDA regulations. Her expertise does not lie in reviewing medical literature or
determining whether causal associations exist between products and clinically
significant hazards. Therefore, she has the expertise to explain the technical
terminology of 21 C.F.R. 201.57(c)(6)(i) to a jury, but she does not have the
expertise to assess whether a causal association existed that would impact
Novartis’ alleged failure to warn. Therefore, none of Dr. Parisian’s testimony as to
the existence of either causation or causal association will be permitted in front of
a trier of fact in this case.
b.

Labeling

Novartis also argues that Dr. Parisian does not have sufficient experience to
discuss labeling and product warnings. However, this argument overlooks Dr.
Parisian’s extensive experience with reviewing proposed labeling and providing
comments for patient brochures. For example, Dr. Parisian’s report states that
while she was at the FDA’s Office for Device Evaluation, she trained medical
officers on labeling review, among other types of assessments. Parisian Report,
25

(Doc. 125-22) at 4, ¶9. Further, during her time at the FDA, she helped design an
FDA epidemiologic study that was created to capture risk to the public and
“trigger the appropriate label changes for both the involved drugs and devices in
order to protect the public.” Id. at 4, ¶8.
Several other courts that have scrutinized Dr. Parisian’s qualifications and
only allowed her testimony in part have allowed Dr. Parisian to testify about
communications between the FDA and Novartis regarding Novartis’ labeling. See
Lemons v. Novartis Pharms. Corp., 849 F. Supp. 2d 608, 614 (W.D.N.C. 2012)
(“Dr. Parisian’s experience and expertise with the FDA and its regulatory scheme
does render her fit to offer testimony on the issue of Novartis’ interactions with the
FDA on the subject of labeling.”); Rowland v. Novartis Pharms. Corp., 9 F. Supp.
3d 553, 561 (W.D. Pa. 2014) (allowing Dr. Parisian to testify about the FDA”s
requirements for labeling and the adequacy of the product warnings and labels due
to her FDA experience and expertise). The court in Rowland, however, found Dr.
Parisian’s testimony regarding alternative actions physicians might have taken had
they received different warnings required an “impermissible degree of speculation
from Dr. Parisian, as she is not an oncologist.” Id. at 562; see also Stambolian v.
Novartis Pharms. Corp., No. 12-CV-4378, 2013 LEXIS 173016 at *29 (C.D. Cal.
Dec. 6, 2013) (“Dr. Parisian is not qualified to offer testimony as to what decisions
26

prescribing doctors would have made had they been given different labeling
information . . . Dr. Parisian is not an oncologist. Therefore, such testimony would
be outside the scope of her expert knowledge.”).
Novartis cites to this court’s decision in Thomas v. Evenflo Co., No. 2:02CV-2001, 2005 WL 6133409, at *14 (N.D. Ala. August 11, 2005), aff’d 205 F.
App’x 768 (11th Cir. 2006), where the plaintiff was not permitted to testify on
labeling because he did not follow accepted methodology in determining warnings
were defective and did not provide alternative warning language. However, unlike
the expert in Thomas, this court finds that Dr. Parisian has the expertise to discuss
the adequacy of warnings and labels based on her contributions to the FDA study
that discussed label changes and the associated risks to the public.7 Dr. Parisian
substantiates her methodology by stating that she employed the same approach
while at the FDA, and her opinion on Novartis’ label cites frequently to applicable
FDA regulations. Parisian Report, (Doc. 125-22) at 50-51.
This court finds persuasive the distinction other courts have drawn between
allowing Dr. Parisian to discuss FDA label compliance and adequacy of labeling
7

Plaintiff cites to In re Ethicon, MDL No. 2327, 2016 LEXIS 119460 at *18 (S.D.W.
Va. Sept. 2, 2016), to support her claim that Dr. Parisian is qualified to testify to the accuracy of
product warnings and labels. However, the court in In re Ethicon construed plaintiff’s counsel’s
position as a concession that Dr. Parisian would not testify about the adequacy of the labels, so it
did not reach the question of whether Dr. Parisian was qualified to testify on label adequacy,
rather than accuracy.
27

versus allowing her to speculate about the potential impact of a label change on
prescribing physicians. Therefore, Dr. Parisian will be permitted to testify to
communications between Novartis and the FDA regarding Reclast label changes
and the adequacy of the Reclast label, but she will not be permitted to testify to the
potential impact a Reclast label change would have had on treating physicians if
the labeling change had occurred.
c.

“Notice” to Novartis, including Study 2202

Novartis seeks to prevent Dr. Parisian from testifying that certain events
should have put Novartis on notice that Reclast could cause harm. First, for the
purposes of this litigation, Dr. Parisian will not be permitted to testify to any
studies, reports, or trials that were published after September 2011, when Jones
alleges that she first developed symptoms of thigh pain, as they could not have put
Novartis - or Jones’s treating physicians - on notice.
Second, Dr. Parisian was involved in clinical trials as a medical officer in
the FDA’s Office of Device Evaluation. See Parisian Report, (Doc. 125-22) at 4,
¶9 (“I participated in the review of proposed clinical trials, the review of premarketing applications . . . and training new medical officers” in clinical trial
evaluation). Therefore, she is qualified to testify about the operations of clinical
trials and the FDA review process for proposed clinical trials.
28

However, Dr. Parisian will not be permitted to testify to “notice” as a
method of circumventing the court’s ruling preventing her from discussing
causation, and Jones’s assertion that Dr. Parisian should be permitted to testify
about all notice to Novartis is overbroad. See Parisian Response, (Doc. 166-39) at
32. Dr. Parisian is a regulatory expert, yet in her report she opines that certain drug
potencies, a study on match factory workers in the 19th century, and causal
associations between other BP drugs and femur fractures should all have put
Novartis on “notice.” See Parisian Report, (Doc. 125-22) at 10-13, ¶¶ 30-36. Dr.
Parisian is not an expert on bisphosphonate medications, a toxicologist, or a
pharmacologist, and she will not be permitted to opine that Novartis should have
conducted safety evaluations more quickly or should have noticed a link between
bisphosphonates and femur fractures earlier. See Kruska v. Novartis Pharms.
Corp., 28 F. Supp. 3d 920, 934 (D. Minn. 2014)(finding that Dr. Parisian is not an
expert on bisphosphonate medications and may not offer testimony on that issue).
In particular, Novartis objects to Dr. Parisian’s reliance on Study 2202 to
demonstrate that Novartis was put on notice. Study 2202 was a clinical trial where
children with a severe genetic disorder called osteogenesis imperfecta (“OI”) were
injected with zoledronic acid either once or twice a year. Parisian Report, (Doc.
125-22) at 19, ¶58. While Study 2202 was conducted on pediatric patients with
29

OI, Reclast is a drug designed for postmenopausal women with osteoporosis. See
Parisian Reply Brief, (Doc. 175) at 14. As the Eleventh Circuit has stated, “even
minor deviations in chemical structure can radically change a particular
substance’s properties and propensities.” Rider v. Sandoz Pharm. Corp., 295 F.3d
1194, 1201 (11th Cir. 2002) (internal quotation and citation omitted).
Though Dr. Parisian is qualified to testify generally to the review process
for Study 2202, as well as the results, she does not have the expertise to opine that
Study 2202 put Novartis on “notice” or to compare the chemical structures of the
drug used in Study 2202 with Reclast.
d.

Advertising and Marketing

Novartis argues that Dr. Parisian is not qualified to discuss a pharmaceutical
company’s marketing or advertising to physicians. It also argues that her opinion
is not reliable because she has formed improper conclusions about the extent to
which Reclast advertising and marketing was communicated to one of Jones’s
prescribing physicians. (Doc. 125-15) at 22. Jones counters that her experience at
the FDA fully qualifies her to render her opinion on advertising and marketing.
(Doc. 166-39) at 33.
Dr. Parisian is qualified to give her opinion on whether Novartis’ marketing
and advertising complied with FDA regulations. While at the FDA, Dr. Parisian
30

helped draft FDA guidance documents that outlined “the requirements for
obtaining FDA’s marketing approval and the FDA Safety Alerts directed to
healthcare providers and their patients.” Parisian Report, (Doc. 125-22) at 5, ¶ 13.
Further, while serving as an instructor in the FDA’s Staff College for training
FDA reviewers, she was responsible for reviewing marketing applications and
“teach[ing] medical officers the process for evaluation and review required by
FDCA directed to product marketing.” Id. at 4-5, ¶10. Therefore, Dr. Parisian has
the expertise to opine on whether Novartis’ marketing complied with FDA
regulations.
Dr. Parisian’s report cites to 21 C.F.R. § 202.1, which sets out the
requirements for prescription drug advertisements, as well as 21 C.F.R. §§ 201,
202, 314, and 352. See id. at 61, ¶200; 66, ¶213. Her report also cites to 21 C.F.R.
§ 314.80, which outlines the requirements of post-marketing reporting of adverse
drug experiences. Id. at 61, ¶200 (“It is also required as part of pharmacovigilance
(21 C.F.R. 314.80) to have investigated and obtained accurate post-market safety
information about the risks of Reclast.”). In some areas of her report, she properly
ties her analysis of Novartis’ marketing and sales practices to pertinent FDA
regulations. See id. (concluding that a directive to the Novartis sales team was “not
permitted by the FD&C Act or implementing regulations”).
31

However, other areas of her report impermissibly extend beyond compliance
with FDA advertising and marketing and into the “intent” or “state of mind” of
Novartis. See id. at 58, ¶190 (“Novartis chose not to accurately update its sales
force or prescribers about the growing risk of association.”). Her report also
speculates about whether Jones’s prescribing physician, Dr. Traylor, was actually
provided with certain sales and marketing information before prescribing Reclast.
Id. at 58-59, ¶191. Allowing Dr. Parisian to testify as to the actions prescribing
doctors would have taken with adequate labeling amounts to impermissible
speculation by her. In these speculative statements, Dr. Parisian does not outline a
sufficiently sound methodology for assessing the adequacy of advertising or
marketing and does not apply the language of any FDA regulation to an allegedly
improper marketing mechanism.
Novartis’ briefing cites to other courts which have found that Dr. Parisian
does not possess the qualifications to assess whether Novartis’ communication of
risks to health care providers, through drug sponsors or members of the sales team,
was appropriate. See Lemons v. Novartis Pharms. Corp., 849 F. Supp. 2d 608, 615
(W.D.N.C. 2012) (“Dr. Parisian does not possess the requisite experience or
expertise, as an employee or insider of a pharmaceutical drug sponsor, to opine on
the conduct of Novartis.”); Kruska v. Novartis Pharms. Corp., 28 F. Supp. 3d 920,
32

935 (D. Minn. 2014) (“Dr. Parisian may testify as to whether Novartis’s marketing
complied with FDA regulations, as that is within her area of expertise, but she will
not be permitted to expand the discussion beyond those as she has in the past.”);
see also Deutsch v. Novartis Pharms. Corp., 768 F. Supp. 2d 420, 468 (E.D.N.Y.
2011) (“Dr. Parisian is not qualified to opine on the ethical standards in the
pharmaceutical industry, not is she qualified to testify as to any obligations
Novartis may have had to the medical community in addition to the FDA
requirements.”) (emphasis added).
Jones cites to Earp v. Novartis Pharms. Corp., No. 5:11-CV-680, 2013 WL
4854488 (E.D.N.C. Sept. 11, 2013) for the proposition that Dr. Parisian should be
allowed to provide all of the advertising and marketing opinions in her report and
deposition. However, the court in Earp merely summarily states that her expertise
at the FDA qualified her to testify to the approval, advertising, and marketing of
drugs without any further specificity or helpful parameters. Id. at *4. Jones also
cites to Stambolian, 2013 LEXIS 173016 at *29, where the court found that Dr.
Parisian was qualified to speak to the FDA’s requirements for labeling and
marketing. However, the court in Stambolian also excluded Dr. Parisian’s
testimony on the actions physicians might have taken with adequate labeling or
marketing and other matters outside the scope of her knowledge. Id. at *28.
33

Therefore, like with the court’s “labeling” analysis, the court finds that Dr.
Parisian is qualified to opine whether Novartis complied with FDA advertising
and marketing regulations. However, Dr. Parisian will not be permitted to extend
her testimony beyond that scope.
e.

Novartis’ 2008 and 2015 Reports to the FDA

Novartis argues that Dr. Parisian should not be permitted to testify
regarding the adequacy of Novartis’ 2008 and 2015 reports to the FDA because
her opinions do not properly cite to FDA regulations and her opinion would be
both unreliable and unhelpful to a jury. Parisian Brief, (Doc. 122-15) at 24-25.
Jones counters that Dr. Parisian has the expertise to testify that Novartis’
communications in its 2008 and 2015 reports to the FDA were misleading and
inaccurate. Parisian Response, (Doc. 166-39) at 33-36. Novartis also argues that
any omission it may have made in its reports was a product of “inadvertent human
error,” so Dr. Parisian’s opinions on the adequacy of Novartis’ reporting are
irrelevant. Parisian Reply, (Doc. 175) at 13.
Generally speaking, an assessment of the reasonableness of a
pharmaceutical company’s conduct in conforming with FDA regulations and in
communicating with the FDA would be helpful to a jury. Further, it is well within
Dr. Parisian’s area of expertise to discuss a company’s compliance with FDA
34

regulations.
Jones cites to In re Mirena, 169 F. Supp. 3d at 481, where the court allowed
Dr. Parisian to testify as to the reasonableness of another pharmaceutical
company’s conduct so long as Dr. Parisian did not attempt to offer narrative
testimony or merely regurgitate statements from company documents with little
analysis. Jones also cites to Deutsch v. Novartis Pharms. Corp., 768 F. Supp. 2d
420 (E.D.N.Y. 2011), where the court found that Dr. Parisian had the expertise to
discuss the “reasonableness of Novartis’ conduct in its interactions with the FDA.”
Id. at 468 (emphasis added).
Novartis argues that Dr. Parisian should not be allowed to opine on
communications between the FDA and Novartis regarding the 2008 and 2015
reports because she does not cite to specific FDA standards that were purportedly
violated. To support its claim, Novartis cites to In re Trasylol Prods. Liab. Litig.,
where the court barred Dr. Parisian’s testimony in full because she did not back up
her opinions that the company behaved inappropriately with a citation to any FDA
standard that was violated. 709 F. Supp. 2d 1323, 1350 (S.D. Fla. 2010). Novartis
also cites to this court’s opinion in Thomas, 2015 WL 6133409 at *15, where this
court excluded the expert’s opinion as nothing more than a “bare assertion”
without any citations to the record to support his opinions.
35

However, Novartis does not dispute that Dr. Parisian is qualified to testify
to the communications between the FDA and a pharmaceutical company,
particularly regarding requests made by the FDA. Further, Dr. Parisian’s
Supplemental Report, which criticizes Novartis’ 2008 and 2015 reports as
inaccurate and misleading, specifically references FDA regulations and applies
those regulations to Novartis’ conduct. See Supplemental Report, (Doc. 125-23) at
¶¶ 65, 75, 83, 94. She also discusses in detail FDA terminology and processes,
including the process of conducting a safety labeling change notification. Id. at 35,
¶83.
Therefore, Dr. Parisian has demonstrated that she applied appropriate
methodology in formulating her opinions on Novartis’ 2008 and 2015 reports to
the FDA. She will be permitted to testify regarding the adequacy of these
communications, and Novartis will be able to critique her methodology on crossexam.
f.

Intent and State of Mind

Jones agrees that Dr. Parisian will not testify as to Novartis’ intent or state
of mind. Parisian Brief, (Doc. 125-15) at 22; see also Parisian Report, (Doc. 12522) at 9, ¶ 23 (“There are no unsupported opinions intended to be offered
regarding the ‘state of mind’ or ‘intent’ of Novartis.”). Accordingly, any testimony
36

on these issues will be excluded.
However, Jones argues that Dr. Parisian should still be able to express
opinions on what Novartis was “aware” of or “knew” based on the information
that was in its possession. She cites to Block v. Woo Young Med. Co., 937 F. Supp.
2d 1028, 1045 (D. Minn. 2013), where the court excluded testimony on corporate
motive or intent but allowed Dr. Parisian to testify to “the state of the medical
literature, the state of FDA approval, and other information about which [the
company] should have been aware.” Jones also cites to In re Mirena, 169 F. Supp.
3d at 479, where the court allowed Dr. Parisian to testify to what the company
“knew,” in the sense of what information was in its possession. The court also
allowed Dr. Parisian to testify to what the FDA “would have done in a typical
situation when presented with a set of facts” and to what the FDA or the
company’s intent was, only to the extent that it was clear from public documents.
Id. at 479-80.
Novartis, in comparison, relies on Jenkins v. Novartis Pharms. Corp, 2012
WL 6213494 at *6 (E.D. Tenn. Dec. 13, 2012), where the court allowed Dr.
Parisian to testify about when certain materials were submitted to the FDA but
excluded testimony on what Novartis “knew” because knowledge is an issue
properly reserved for the jury. See id. (“Dr. Parisian has no specialized knowledge
37

or scientific/medical expertise that provides her with a superior ability to judge
Novartis’s knowledge, and there is no basis for finding that the jury needs her
assistance in evaluating Novartis’s knowledge.”); see also In re Trasylol, 709 F.
Supp. 2d at 1338 (excluding Dr. Parisian from addressing Novartis’ intent, motive,
or bad faith); Bartoli v. Novartis Pharms. Corp., No. 3:13-CV-0724, 2014 WL
1515870 at *5 (M.D. Pa. April 7, 2014) (same); Lopez v. I-Flow, Inc., 2011 WL
1897548 at *11 (E.D.N.C. Jan. 9, 2012) (excluding Dr. Parisian from testifying to
the knowledge, state of mind, or motivations of either the pharmaceutical company
or the FDA itself).
Regardless of Jones’s reassurances, any testimony on Novartis’ state of
mind or intent is excluded as a question for the jury rather than for Dr. Parisian.
Dr. Parisian may, however, testify regarding what information was or should have
been, pursuant to specific FDA regulations, in Novartis’ possession.
g.

References to Osteonecrosis of the Jaw (“ONJ”)

Novartis argues that Dr. Parisian should be excluded from opining on
bisphosphonate-related osteonecrosis of the jaw (“BRONJ”) or osteonecrosis of
the jaw generally (“ONJ”), as she does in both her report and supplemental report.
Jones counters that because Zometa, another Novartis drug that was associated
with BRONJ, and Reclast, the subject of this litigation, are “chemically identical,”
38

it is proper for Dr. Parisian to discuss the entire class of drugs. Parisian Response,
(Doc. 166-39) at 38. However, arguments of counsel are not evidence. Moreover,
even if the two drugs are “chemically identical,” even Jones admits the dosages
are different, as are the alleged injuries. Id.
As the Eleventh Circuit has explained, even “minor deviations in chemical
structure can radically change a particular substance’s properties and
propensities.” Rider v. Sandoz Pharm. Corp., 295 F.3d 1194, 1201 (11th Cir.
2002) (internal quotation and citation omitted) (excluding an expert witness from
testifying that, because other drugs in the same class cause a certain effect, a drug
from that class must cause the same effect as well). Jones has not cited any case
law to support her claim that a regulatory expert should be able to opine on a
different type of injury, associated with a different drug, than the injury that is the
subject of the current litigation. Further, Jones has not established why any
discussion of chemical analogies would help, rather than confuse, a jury.
Therefore, Dr. Parisian is prohibited from testifying regarding ONJ, BRONJ, and
Novartis’ drug Zometa.
4.

Conclusion as to the Admissibility of Dr. Parisian’s
Testimony

Based on the foregoing, Novartis’ Motion To Strike Dr. Parisian’s

39

testimony (doc. 108) is due to be DENIED to the extent that it seeks to exclude
Dr. Parisian’s testimony in its entirety. However, Novartis’ Motion is due to be
GRANTED IN PART and DENIED IN PART as to the particular areas of Dr.
Parisian’s testimony that it seeks to exclude.
Dr. Parisian may testify on the following issues: (1) Novartis’ interactions
with the FDA on the subject of labeling; (2) compliance and/or noncompliance
with FDA regulations on advertising and marketing; (3) Novartis’ conduct in
communicating with the FDA, including the 2008 and 2015 reports; and (4) what
information Novartis should have had, based on specific FDA regulations.
Dr. Parisian may not testify on the following issues: (1) causation, including
“causal association”; (2) how a change in labeling might have impacted the
decision of a prescribing physician; (3) whether or not Novartis was put on
“notice,” including by Study 2202; (4) how a change in advertising and marketing
might have impacted the decision of a prescribing physician; (5) Novartis’ state of
mind, intent, or motive, including what Novartis “knew”; and (6) comparisons
between ONJ, BRONJ, or the drug Zometa.
B.

Dr. William B. Hinshaw

Jones also offers the opinions of Dr. Hinshaw into evidence. Dr. Hinshaw
has offered both an expert report (doc. 125-26, the “Hinshaw Report”) and a
40

supplemental expert report (doc. 125-25, the “Hinshaw Supplemental Report”).
Dr. Hinshaw was deposed on June 1, 2016, and the deposition transcript was filed
into the record (doc. 125-21, the “Hinshaw Deposition”).
On August 15, 2016, Novartis filed a Motion To Strike Dr. Hinshaw’s
expert testimony (Doc. 112) and a brief in support of its Motion (doc. 125-20, the
“Hinshaw Brief”). On September 19, 2016, Jones filed a Response opposing
Novartis’ Motion To Strike. (Doc. 166-41, the “Hinshaw Response”). On October
14, 2016, Novartis filed a reply brief in support of its Motion To Strike. (Doc. 180,
the “Hinshaw Reply”).
1.

Dr. Hinshaw’s Qualifications

Dr. Hinshaw is Jones’s designated scientific expert and offers opinions on
both general and specific causation. Hinshaw Response, (Doc. 166-41) at 4. Dr.
Hinshaw is a bioorganic chemist and a practicing gynecologist, specializing in
menopause and clinical management of bone fragility. Hinshaw Report, (Doc.
125-26) at 2. He received a B.S. degree in Chemistry in 1963 from Duke
University and a Ph.D. in Biomimetic Organic Chemistry in 1967 from Stanford
University. Id. He received a medical degree from Albany Medical College in
1978. Id. Subsequently, he completed a four-year residency in Obstetrics and
Gynecology at Albany Medical Center Hospital, serving as the chief senior
41

resident. Id.
Since 1983, Dr. Hinshaw has worked as a board-certified gynecologist.
Hinshaw Curriculum Vitae, (Doc. 125-26) at 21-23. Approximately ninety percent
of his current gynecological practice is devoted to “hormonal aberrations of
menopause which are associated with problems of bone fragility.” Hinshaw
Response, (Doc. 166-41) at 7. Dr. Hinshaw served as an assistant investigator
during a clinical trial involving a non-Reclast BP drug around 1997. Hinshaw
Deposition, (Doc. 125-21) at 7 (18:14-19:23).
In 2012, Dr. Hinshaw authored a peer-reviewed article titled “Using
Medical Chemistry to Solve an Old Medical Mystery” that discussed unusual
femur fractures suffered by match factory workers in the nineteenth century.
Hinshaw CV, (Doc. 125-26) at 24. In total, Dr. Hinshaw has published twenty-one
peer-reviewed articles, five of which pertained to phosphorous chemistry, bone
fragility, and fractures. See id. In 2012, he co-authored an article titled “Atypical
Femur Fractures: 81 Individual Personal Histories,” that was published in the
Journal of Clinical Endocrinology and Metabolism. Id. In fall 2016, Dr. Hinshaw
co-authored an article, titled An Evaluative History of Bisphosphonate Drugs:
Dual Physiologic Effects of Pyrophosphate as Inspiration for a Novel
Pharmaceutical Class, that was accepted by the Journal of Osteoporosis for
42

publication. See (Doc. 169).
Jones claims that Dr. Hinshaw has reviewed scientific literature on the
chemistry and kinetics of BPs on bone and has studied early scientific literature of
the effect of pyrophosphates on bone. Hinshaw Response, (Doc. 166-41) at 9.
Jones states that Dr. Hinshaw has primarily been offered “to explain to the jury the
science of bone self-repair and Reclast’s deleterious effect on the bone’s natural
abilities to grow, repair, model, and re model.” Id. at 4.
2.

Dr. Hinshaw’s Opinions

In Dr. Hinshaw’s Expert Report,8 he offers the following opinions:
(1) The history of bisphosphonates for the prevention and
treatment of osteoporosis and osteopenia is one of conflicting
values, perspectives, and interests that has contributed to the
present confusion on the meaning of those terms, (doc. 125-26)
at 4-5;
(2) At the time of its distribution to Dr. Traylor, Reclast was
defective. Reclast was not reasonably safe as intended to be used.
Reclast was defective in design. The design was unreasonably
dangerous, id. at 5;
(3) Novartis breached its promise or warranty to Ernesteen Jones
that Reclast was suitable or fit to treat post-menopausal
osteoporosis in the general population, a population to which Ms.
Jones belonged, by seeking indications granted from data
produced in trials employing very restricted and non-

8

Dr. Hinshaw offers the same 13 opinions in his supplemental expert report as well. See
Hinshaw Supplemental Report, (Doc. 125-25).
43

representative populations, id. at 5-6;
(4) Novartis and its predecessor companies, including Sandoz
Pharmaceuticals and its consultants, had ample notice of the
propensity of the bisphosphonates such as zoledronic acid to
occasion fractures of the femur and general embrittlement of bone
to permit avoidance of its introduction or early warning of this
propensity, id. at 6-8;
(5) Novartis breached its duty to the clinical population
candidates for the treatment by Reclast by making effectiveness
representations based upon unrepresentative and manipulated
clinical trials, id. at 9;
(6) There was a safer and practical alternative design that Novartis
could have used at the time the Reclast was manufactured. The
safer alternative design would have reduced or eliminated the
harm to Ernesteen Jones, id. at 9;
(7) Novartis breached its duty to act as a reasonably prudent
pharmaceuticals manufacturer in its design, research and testing
of Reclast. Novartis failed to conduct appropriate clinical studies
to determine the safety of long-term use of bisphosphonate drugs.
Furthermore, Novartis failed to recognize, acknowledge or
evaluate the increased hazard ratio of atypical femoral fractures
in the pivotal clinical trials, although the existence of the risk was
well-known at or near the beginning of that trial, id. at 10-11;
(8) Novartis breached its duty to act as a reasonably prudent
pharmaceuticals manufacturer in its advertising, marketing,
distribution, promotion and sale of Reclast, id. at 11-12;
(9) Novartis breached its duty to act as a reasonably prudent
pharmaceuticals manufacturer in its labeling of Reclast, id. at 12;
(10) Novartis did not give Ernesteen Jones an adequate warning
about the danger of femur fracture associated with the use of
44

Reclast, id. at 12;
(11) Novartis knew or should have known that Reclast could
create danger when used as intended in its customary manner, id.
at 13;
(12) The defective design of Reclast causes atypical femur
fractures, id. at 13-17;
(13) The defective design of Reclast caused the injuries to
Ernesteen Jones, id. at 17-19.
3.

Dr. Hinshaw Is Qualified To Testify in This Case

Novartis mischaracterizes Dr. Hinshaw’s statements on causation as
“eccentric”and describes him as a “some-time chemist.” Hinshaw Brief, (Doc.
125-20) at 4. Dr. Hinshaw has multiple degrees in chemistry and has done
extensive research on biometric pathways, gynecology, post-menopausal
medicine, and bone fragility. He has also published a number of peer-reviewed
articles, several of which pertain to bone fragility and fractures. Given that Dr.
Hinshaw does have extensive training, experience, and expertise in these fields, he
will not be excluded from testifying based on his qualifications. However, as
discussed below, Dr. Hinshaw’s testimony and opinions will be excluded because
the methodology he relied upon in forming his opinions has not been shown to be
reliable.
4.

Dr. Hinshaw’s Methodology Is Not Reliable
45

Dr. Hinshaw’s opinion has been offered by Jones to establish both general
and specific causation. General causation refers to the “general issue of whether a
substance has the potential to cause the plaintiff’s injury.” Chapman, 766 F.3d at
1307 (citing Guinn, 602 F.3d at 1248 n.1). Specific causation refers to “the issue
of whether the plaintiff has demonstrated that the substance actually caused injury
in her particular case.” Guinn, 602 F.3d at 1248 n. 1.
a.

General Causation Opinion
i.

Bradford Hill Methodology

Dr. Hinshaw primarily relies on the Bradford Hill methodology to reach his
conclusion that Reclast generally causes atypical femoral fractures. Hinshaw
Response, (Doc. 166-41) at 6. Novartis disputes his use of Bradford Hill as both
inapplicable and unreliable.
Sir Bradford Hill was a world-renowned epidemiologist who articulated a
nine-factor set of guidelines in his seminal methodological article on causality
inferences. Milward v. Acuity Speciality Products Group, Inc., 639 F.3d 11, 17
(1st Cir. 2011) (citing Arthur Bradford Hill, The Environment and Disease:
Association or Causation?, 58 Proc. Royal Soc’y Med. 295 (1965)). The Bradford
Hill criteria are nine factors “widely used in the scientific community to assess
general causation.” In re Stand ‘N Seal Products Liab. Litig., 623 F. Supp. 2d
46

1355, 1372 (N.D. Ga. 2005)(citing Gannon v. United States, 292 F. App’x 170,
173 (3d Cir. 2008).
Sir Bradford Hill’s article explains that “one should not conclude that an
observed association between a disease and a feature of the environment (e.g., a
chemical) is causal without first considering a variety of ‘viewpoints’ on the
issue.” Milward, 639 F.3d at 17. The nine viewpoints look to:
(1) the strength of the association;
(2) the consistency of the association;
(3) the specificity of the association;
(4) the temporal relationship of the association;
(5) whether there is a dose-response relationship;
(6) whether causation is biologically plausible;
(7) the coherence of the association;
(8) the presence of experimental evidence; and
(9) evidence by analogy.”9
In re Stand ‘N Seal, 623 F. Supp. 2d at 1372.
While the Eleventh Circuit has not yet directly commented on the Bradford
Hill criteria, the reliability of the methodology is strengthened by the number of
other circuit courts and district courts within this Circuit who have approved of an

9

The Reference Manual on Scientific Evidence lists slightly different Bradford Hill
factors than Sir Bradford Hill’s original factors. However, the factors are “largely the same” In re
Lipitor (Atorvastatin Calcium) Mktg., Sales Practices & Prod. Liab. Litig., 174 F. Supp. 3d 911,
916 n.3 (D.S.C. 2016). Cf. Michael D. Green, D. Michael Freedman & Leon Gordis, “Reference
Guide on Epidemiology,” REFERENCE MANUAL ON SCIENTIFIC EVIDENCE, p. 600 (3d ed. 2011)
(“2011 Reference Guide on Epidemiology”) with Sir Austin Bradford Hill, The Environment and
Disease: Association or Causation?, 58 Proc. Royal Soc’y Med. 295, 295-300 (1965).
47

expert’s use of the criteria. Furthermore, the Third Restatement of Torts states that
if an association is found between a substance and a disease, “epidemiologists use
a number of factors (commonly known as the ‘Hill guidelines’) for evaluating
whether that association is causal or spurious.” Restatement, §28 cmt. c(3). As the
Bradford Hill methodology has received wide acceptance and approval in the
scientific community, Novartis’ claims that the methodology is totally unreliable
and unacceptable misrepresent the position of the scientific and legal communities
and are not well taken by this court.
In his expert report and supplemental report, Dr. Hinshaw applied all nine
Bradford Hill factors to reach his conclusion that Reclast causes AFF. Hinshaw
Report, (Doc. 125-26). This was wise, as another circuit recently excluded an
expert as unreliable whose testimony relied on the Bradford Hill methodology but
only discussed three of the nine factors in a limited and sparse manner. See In re
Nexium Esomeprazole, 2016 WL 6298741, at *2 (9th Cir. October 28, 2016)
(unpublished).
However, Dr. Hinshaw’s reliance on the Bradford Hill methodology suffers
from a fatal flaw: he cannot point to a study that establishes a causal association
between Novartis’ drug Reclast and AFFs. Both parties cite to the 2011 Reference

48

Guide on Epidemiology,10 which is published by the Federal Judiciary Center, and
the Restatement to support their respective Bradford Hill claims. These resources
explain that the Bradford Hill factors cannot be applied without first establishing a
causal association. The 2011 Reference Guide on Epidemiology states,
In assessing causation, researchers first look for alternative explanations
for the association, such as bias or confounding factors . . . [o]nce this
process is completed, researchers consider how guidelines for inferring
causation from an association apply to the available evidence. We
emphasize that these [Bradford Hill] guidelines are employed only after
a study finds an association to determine whether that association
reflects a true causal relationship.
2011 Reference Guide on Epidemiology at 598-99 (emphasis added). A
supplementary note further explains that “in a number of cases, experts attempted
to use these guidelines to support the existence of causation in the absence of any
epidemiologic studies finding an association . . . [t]here may be some logic to that
effort, but it does not reflect accepted epidemiologic methodology.” Id. at 599 n.
141.
The Restatement similarly describes the use of the Bradford Hill
methodology as a two-step process. The first step requires that reliable medical
studies establish an association between a substance and a disease. Restatement, §
28 cmt. c(3). Then, “if an association is found, epidemiologists use a number of
10

2011 Reference Guide on Epidemiology at 597-606.
49

factors (commonly known as the ‘Hill guidelines’) for evaluating whether that
association is causal or spurious.” Id. (emphasis added); see also id. “[h]owever,
even when epidemiology finds an association, the observational (rather than
experimental) nature of these studies requires an examination [using the Bradford
Hill factors] of whether the association is truly causal or spurious.” The Bradford
Hill factors help determine whether an association is causal, not whether there is
an association at all.
Novartis cites to several courts who have required that an expert who
intends to rely on Bradford Hill methodology must first cite to an epidemiologic
study finding an association. These courts allege that any other type of study
establishing causal association would not be sufficient to apply the Bradford Hill
factors. See, e.g., Dunn v. Sandoz Pharms. Corp., 275 F. Supp. 2d 672, 678-79
(M.D.N.C. 2003) (“[T]he Bradford Hill criteria is a method for determining
whether the results of an epidemiological study can be said to demonstrate
causation and not a method for testing an unproven hypothesis.”) In Dunn, the
expert’s testimony was excluded as unreliable because he developed a hypothesis
and attempted to use the Bradford Hill criteria to prove that assertion, rather than
using the factors to evaluate whether an association already shown by a reliable
study demonstrates causation. Id. at 680; see also In re Lipitor, 174 F. Supp. 3d at
50

925 (“Courts exclude expert testimony that attempts to start at step two, applying
the Bradford Hill criteria without adequate evidence of an association.”).
However, the Eleventh Circuit and a substantial number of other Circuit
courts have rejected the epidemiologic study threshold for sufficient proof of
general causation. See Rider v. Sandoz Pharms. Corp., 295 F.3d 1194, 1198 (11th
Cir. 2002)(“It is well-settled that while epidemiological studies may be powerful
evidence of causation, the lack thereof is not fatal to a plaintiff’s case.”); see also
In re Lipitor, 174 F. Supp. 3d at 916 (“While a causation opinion need not be
based on epidemiological studies . . . it is well established that the Bradford Hill
method used by epidemiologists does require that an association be established
through studies with statistically significant results.”) (internal citation omitted)
(emphasis in original).
In accordance with the Eleventh Circuit’s holding in Rider, Dr. Hinshaw
was not required to rely on an epidemiological study showing a causal association
between the drug and the injury alleged. However, he did need to establish that the
causal association existed based on existing medical literature.
ii.

Hinshaw Relies on a Class-Wide Association
Between BPs and AFFs Rather Than an
Association Between Reclast and AFFs

First, Dr. Hinshaw admits in his deposition that he is not able to point to
51

“any study in the peer-reviewed literature that defines a statistically-significant
AFF association for Reclast specifically.” Hinshaw Deposition, (Doc. 125-21) at
47(181:11-15). The reliability of his Bradford Hill methodology is called into
question by this concession that there is no statistically significant causal
association between Reclast and AFFs.
To overcome this hurdle, Dr. Hinshaw bases his general causation opinion
on a causal association found between the entire class of BP drugs, of which
Reclast is one type, and femoral fractures. His report claims that the association
requirement is met because bisphosphonate use, not Reclast use, is “associated
with a detectable increase in the incidence of non-hip femoral fractures. The
ASBMR Task Force has recognized this association and called it ‘robust.’”
Hinshaw Report, (Doc. 125-26) at 13, ¶ 12.
The 2013 American Society for Bone and Mineral Research (“ASBMR”)11
Task Force Report (the “2013 ASBMR Report”) included the following update on
the class-wide association: “[a]lthough the Task Force still holds the opinion that a
causal relationship between BPs and AFFs has not been established, evidence for

11

The ASBMR is a “professional, scientific, and medical society established to bring
together clinical and experimental scientists who are involved in the study of bone and mineral
metabolism.” See ABOUT ASBMR, http://www.asbmr.org/About/Overview.aspx (last visited Jan.
24, 2017).
52

an association has continued to accumulate in the 2 years since the first report was
published and is quite robust.” (Doc. 117-1) at 11 (emphasis added). Dr. Hinshaw
agreed at deposition that the Task Force’s viewpoint was a “fair characterization.”
Hinshaw Deposition, (Doc. 125-21) at 38(143:19-21)).
This report, however, did not look or pertain specifically to any causal
association between Reclast (only one drug in the class of BPs) and AFFs, and
neither Dr. Hinshaw’s report and deposition nor Ms. Jones’s briefing cites to any
authority stating that the association between a specific drug and an alleged injury
may be extrapolated from a class-wide association. In fact, a number of courts
have criticized the use of extrapolation from a class-wide study to an individual
drug, as explained below.
iii.

Dr. Hinshaw Improperly Extrapolates From
Studies on the Class of BP Drugs To Form his
Causal Association Opinion

Dr. Hinshaw has not substantiated his claim that a causal association
between Reclast and AFFs may be extrapolated from a class-wide association
between BPs and femoral fractures. In Rider v. Sandoz Pharms. Corp., 295 F.3d
1194 (11th Cir. 2002), the Eleventh Circuit affirmed the decision of a district court
that “drew a careful distinction between clinical process, in which conclusions
must be extrapolated from incomplete data, and the scientific method, in which
53

conclusions must be drawn from an accepted process, and concluded that the
plaintiffs’ experts were relying on the former.” Id. at 1196.
One of the reasons the Eleventh Circuit upheld the district court’s decision
was based on the plaintiffs’ failure to explain why their presumption that a specific
drug had the same effects as other drugs in that class was valid:
[P]laintiffs suggest that because bromocriptine is an ergot alkaloid, it
causes vasoconstriction. Although some other ergot alkaloids do cause
vasoconstriction, plaintiffs offered insufficient evidence for the district
court to find that bromocriptine does so as well. This is not a case where
the Court finds the evidence offered to be unreliable. In this case the
record contains no evidence at all of this hypothesis. Instead, it contains
principally speculation and conjecture. Because the ergot alkaloid class
of drugs has a wide range of effects, it is not obvious that bromocriptine
should have the same effects as other drugs in that class.
Id. at 1201 (emphasis added). The court concluded that “scientific evidence must
‘fit’ the plaintiff’s theory of causation,” and in this case, the court would have
been required to “make several scientifically unsupported ‘leaps of faith’ in the
causal chain” in order to admit the plaintiff’s evidence. Id. at 1202 (citing Joiner,
522 U.S. at 152, 118 S. Ct. at 522).
Some extrapolation is permitted and sometimes even necessary. See Joiner,
522 U.S. at 146, 118 S. Ct. At 519 (“[C]onclusions and methodology are not
entirely distinct from one another. Trained experts commonly extrapolate from
existing data.”). However, the Supreme Court in Joiner also made clear that
54

“nothing in either Daubert or the Federal Rules of Evidence requires a district
court to admit opinion evidence that is connected to existing data only by the ipse
dixit of the expert.” Id.
The Tenth Circuit in Hollander v. Sandoz Pharms. Corp., 289 F.3d 1193
(10th Cir. 2002), for example, held that the district court did not abuse its
discretion in concluding that the methodology of an expert who relied on the
Bradford Hill criteria was unreliable because there was “simply too great an
analytical gap” between studies comparing the class of drugs and the disease and
the experts’ conclusion that the specific drug in question caused the alleged injury.
Id. at 1208 (citing Joiner, 522 U.S. at 146, 118 S. Ct. at 512). The court elaborated
that “the studies in question [did] not directly address the relationship between
[the specific drug] and [the alleged injury]” and critiqued the plaintiff for
presenting “no expert analysis as to how one might extrapolate” from the drug’s
effect on a group with one syndrome to another group who took the drug for a
different purpose. Id.
Novartis cites to several examples in Dr. Hinshaw’s deposition where he
distinguished certain BP drugs from others in the class. Dr. Hinshaw admitted that
one could not reliably extrapolate results from studies of oral BPs administered
either weekly or monthly to non-oral BPs infused once a year, like Reclast:
55

Q.

[referring to one of the BP studies] (By Mr. Klein) But they did
find a statistical difference – but they did find a safety signal for
some of the other bisphosphonates; is that correct?

A.

That’s correct. They were comparing a drug which is given on a
regular periodic basis to a drug that was given on a once-a-year
basis. And this is a completely invalid comparison from the
beginning because the presence of a drug in the environment in
which the healing of the bone is occurring is paramount to the
rate at which it heals.

Q.

So you believe – so, just to clarify, Alendronate and the other oral
bisphosphonates are given either on a weekly or monthly –

A.

Periodically, in one way or another.

Q.

And Zoledronic acid is a once-a-year annual infusion; correct?

A.

Yes.

Q.

If I understand what you just said, its not valid to extrapolate
results from studies looking at oral bisphosphonates (inaudible);
is that correct?

A.

It’s not valid because of that fact and because of the fact that
under the best of circumstances it’s problematic whether the
parameter being examined has anything to do with the problem
that we’re interested in, not healing.

Hinshaw Deposition, (Doc. 125-21) at 47(178:3-179:14) (emphasis added); see
also id. at 73(282:22-283:16):
Q:

Doctor, do you agree different bisphosphonates can have different
impacts on the material properties of bone?

A.

Probably so. It’s – one problem is that, in measuring impact of
56

bisphosphonates on the material properties of bone, the oral
bisphosphonates are usually given orally and intravenous
bisphosphonates are usually given intravenously and it really
produces a set of circumstances that are only assumed to be
parallel. They’re not likely to be parallel. So – but I do agree
there’s considerable differences in the bisphosphonates. There’s
less difference if they’re all compared intravenously, which can
be done and has been done in some circumstances.
(Emphasis added). Novartis also cites to a study, which Dr. Hinshaw discussed in
his deposition, that found a statistically significant loss in bone fatigue life in
orally-administered BPs but not in zoledronic BPs like Reclast. Id. at 74(286:1321). Dr. Hinshaw also agreed that different BPs have different “affinities” for bone
and that zoledronic acid BPs have the “highest affinity for bone amongst the drugs
in the United States.” Id. at 74(289:5-16).
Dr. Hinshaw concedes that, at least on some level, a comparison between
BP drugs administered once a year and oral BPs administered more frequently is
invalid. As stated above, Dr. Hinshaw also admitted in his deposition that there is
no study in peer-reviewed literature that found a statistically-significant
association between Reclast and AFFs. Id. at 47(181:11-15).
In her response, Jones relies on the decision of another district court in this
Circuit to support her claim that “it is not a requirement in the Eleventh Circuit to
prove the exact mechanism of injury.” Hinshaw Response, (Doc. 166-41) at 15

57

(citing McBride v. Houston Cty. Health Care Auth., 2015 WL 3648995 (M.D. Ala.
2015)). Though that may be true, an association must first be established between
the drug and the alleged injury in order for an expert to rely on the Bradford Hill
methodology to form a general causation opinion. Because Dr. Hinshaw has not
first established that there is an association between Reclast and AFFs, his general
causation opinion will be excluded.12
iv.

Weight of the Evidence Methodology

Jones also briefly argues that Dr. Hinshaw properly used a “weight of the
evidence” methodology in addition to the Bradford Hill criteria. Hinshaw
Response, (Doc. 166-41) at 25. Novartis counters in its reply brief that Dr.
Hinshaw did not disclose his use of a weight of the evidence methodology in
either his initial expert report or his supplemental report and also did not apply the
methodology reliably. Hinshaw Reply, (Doc. 180) at 13-14.
Generally speaking, weight of the evidence methodology can be considered

12

Novartis also argues that Dr. Hinshaw has conceded that even the class-wide
epidemiology studies are conflicting. Hinshaw Brief, (Doc. 125-20) at 11, n. 9; see Hinshaw
Deposition, (Doc. 125-21) at 55(212:8-13) (agreeing “that the epidemiology that exists in the
literature is conflicting”). However, Dr. Hinshaw goes on to clarify that Novartis’ attempt to
group all the epidemiological studies on BPs together and call them conflicting is like
“comparing apples and oranges” because “we are comparing reports that use totally different
techniques for evaluating the problem”). Id. at 55(212:18-19 and 211:20-25). Therefore, Dr.
Hinshaw was actually referring to Novartis’ grouping of certain studies when he used the term
“conflicting,”’ not necessarily the studies themselves.
58

reliable. See Milward, 639 F.3d at 17 (“This ‘weight of the evidence’ approach to
making causal determinations involves a mode of logical reasoning often
described as ‘inference to the best explanation,’ in which the conclusion is not
guaranteed by the premises.”). The methodology involves
six general steps, some of which may be implicit. The scientist must (1)
identify an association between an exposure and a disease, (2) consider
a range of plausible explanations for the association, (3) rank the rival
explanations according to their plausibility, (4) seek additional evidence
to separate the more plausible from the less plausible explanations, (5)
consider all of the relevant available evidence, and (6) integrate the
evidence using professional judgment to come to a conclusion about the
best explanation.
Id. at 17-18.
Jones claims that Dr. Hinshaw applied the methodology when he
“considered the communication between the FDA and NPC regarding AFFs,
documentation regarding clinical trials, postmarket reports and voluminous
scientific studies.” Hinshaw Response, (Doc. 166-41) at 25. However, Dr.
Hinshaw has not described the process he used or the steps he took in applying
this methodology, including whether he ranked plausible rival explanations, and
Jones’s briefing provides no clarity either. Dr. Hinshaw’s testimony on his
purported weight of the evidence methodology is therefore unreliable and would
not assist a trier of fact. As both Dr. Hinshaw’s “weight of the evidence” and

59

Bradford Hill methods were applied unreliably, his general causation opinion is
due to be excluded.
b.

Specific Causation Opinion

Dr. Hinshaw also opines in his Expert Report (doc. 125-26) and
Supplemental Report (doc. 125-25) that Reclast caused the injuries to Jones. See
Doc. 125-26 at 17-19. Like his general causation opinion, his specific causation
opinion must meet the Eleventh Circuit’s standard for admissibility. For the
following reasons, the court finds that Dr. Hinshaw applied an unreliable
methodology in forming his specific causation opinion, so his specific causation
opinion is also due to be excluded.
i.

Dr. Hinshaw Has not Conceded That He Is
Unqualified To Offer Specific Causation
Testimony

Novartis first claims that Dr. Hinshaw admitted in his deposition that he is
unqualified to opine on specific causation. Dr. Hinshaw was previously offered as
an expert witness in Barnes v. Merck Sharp & Dohme Corporation, CV-2011900393.00, in the Circuit Court of Montgomery County, Alabama. (Doc. 125-26)
at 3. A portion of his deposition from Barnes was read to Dr. Hinshaw by defense
counsel during his deposition. Novartis claims that (1) Dr. Hinshaw admitted in
his previous testimony in the Barnes case that he was not qualified to offer a
60

specific causation opinion; and (2) Dr. Hinshaw, in his deposition in this case, reaffirmed his testimony in Barnes. (Doc. 125-20) at 16-17.
However, Dr. Hinshaw’s Barnes deposition testimony is not the slam-dunk
that Novartis considers it to be. Dr. Hinshaw was asked, “[h]ow would you go
about determining whether a particular medicine a patient was taking may have
contributed to the development of a fracture?” Hinshaw Deposition, (Doc. 125-21)
at 90(350:12-16). In response, he stated that “I don’t think, as an individual
practitioner, that I could determine that a particular medicine that a particular
patient was taking was associated with the cause of the fractures . . . I don’t feel
that I’m qualified to do that.” Id. at 90(350:16-23) (emphasis added). Though Dr.
Hinshaw agreed that he “still feel[s] that way” in this case, id. at 90(350:24-25), he
appears to be referring to his opinion as a treating physician or prescribing
practitioner, not as an epidemiologist. Novartis has misconstrued Dr. Hinshaw’s
agreement with his statement in his prior deposition testimony as a full concession
that he is never qualified to offer specific causation testimony.
Moreover, because the court has not been provided with Dr. Hinshaw’s full
deposition in the Barnes case, it will not base its decision whether to exclude Dr.
Hinshaw on a discussion of an incomplete portion of that prior deposition.
ii.

Dr. Hinshaw Improperly Used Bradford Hill, a
61

General Causation Methodology, To Form his
Specific Causation Analysis
Dr. Hinshaw has stated that he relied on the Bradford Hill criteria to form
his specific causation opinion as well as his general causation opinion. While
Bradford Hill may be used to reliably establish general causation, Jones has not
demonstrated that Bradford Hill is an accepted criteria for determining specific
causation.
Numerous courts have referred to the Bradford Hill criteria as a useful tool
to analyze general, rather than specific, causation. See, e.g., Gannon, 292 F. App’x
at 173 (“Specifically, the Court found that [the expert’s] testimony failed to satisfy
the ‘Bradford Hill’ criteria, which are nine factors widely used in the scientific
community to assess general causation.”) (emphasis added); In re Stand ‘N Seal,
623 F. Supp. 2d at 1372 (same); see also 2011 Reference Guide on Epidemiology
at 598-600 (listing application of Bradford Hill criteria only as a method of
determining general causation).
Jones does not cite to any authority approving of the use of the Bradford
Hill criteria alone to support a specific causation opinion. Instead, she argues that
because Dr. Hinshaw utilized both the Bradford Hill methodology and a
differential diagnosis, his specific causation opinion is reliable. In his deposition,

62

Dr. Hinshaw stated:
Q.

In your report, you say that you arrive at your specific
causation opinion based on Bradford-Hill; is that correct?

A.

Yes.

Q.

Can you cite me any paper – any peer-reviewed literature in
which Bradford-Hill was used to arrive at specific causation in
a particular patient?

A.

A bisphosphonate patient or any patient?

Q.

Any patient.

A.

Any patient.

Q.

Specific Causation.

A.

No, I can’t.

Q.

Do any of the Bradford-Hill criteria involve ruling out other
potential causes in a particular patient?

A.

No.

Q.

Did you perform a differential diagnosis on Ms. Jones?

A.

To a limited extent. To the extent of which I was offered the
medical record.

Hinshaw deposition, (Doc. 125-21) at 91(355:14-356:14).
Jones cites to Estate of Tobin v. SmithKline Beacham Pharms., No. 00-CV25, 2001 LEXIS 26397 (D. Wyo. May 8, 2001) to support her claim that Dr.

63

Hinshaw’s specific causation opinion is reliable because it was based both on
Bradford Hill criteria and on a differential diagnosis. However, the court in Estate
of Tobin provided little analysis as to why it allowed an expert’s specific causation
opinion, other than stating that while the expert relied to some extent “on
differential diagnosis in reaching his opinions regarding general and specific
causation,” he also “relie[d] upon the Bradford Hill Factors . . . in reaching his
opinion.” Id. at *46. It is not clear whether the Bradford Hill criteria were used to
form both the expert’s specific and general causation opinions in this case, and
Jones did not direct the court to any other authority where the combination of
Bradford Hill and differential diagnosis methodologies was sufficient to allow in
an expert’s specific causation testimony.
Moreover, Dr. Hinshaw’s alleged reliance on a differential diagnosis is
itself questionable because he concedes that his differential diagnosis analysis was
“limited,” and he did not discuss the methodology in his expert report.
iii.

Differential Diagnosis Methodology Has
Generally Been Accepted by the Eleventh Circuit

Differential diagnosis is “a medical process of elimination whereby the
possible causes of a condition are considered and ruled out one-by-one, leaving

64

only one cause remaining.” Hendrix, 609 F.3d at 1195.13 The differential diagnosis
method, when applied under circumstances that “ensure reliability,” can “provide a
valid basis for medical causation opinions.” Id. To ensure reliability, the court
must consider the reasonableness of applying the differential diagnosis approach
“to the facts of this case and the validity of the experts’ particular method of
analyzing the data and drawing conclusions therefrom.” Id. On several occasions,
the Eleventh Circuit has referred to differential diagnosis as a “scientifically
accepted methodology.” Chapman, 766 F.3d at 1307 (citing Guinn, 602 F.3d at
1248 n.1).
As the Eleventh Circuit has explained,
Differential diagnosis includes three steps: (1) the patient's condition is
diagnosed, (2) all potential causes of the ailment are considered, and (3)
differential etiology14 is determined by systematically eliminating the
possible causes. McClain, 401 F.3d at 1252. A reliable differential
analysis “need not rule out all possible alternative causes,” but “it must
at least consider other factors that could have been the sole cause of the
plaintiff's injury.” Guinn, 602 F.3d at 1253. Differential diagnosis,

13

The Restatement similarly states that differential diagnosis involves identifying a cause
“by eliminating the possibility that other known and alternative causes were responsible for the
outcome.” Restatement §28 cmt. c(3).
14

As the Eleventh Circuit has explained, “[t]echnically, differential diagnosis refers to a
method of determining which of two diseases a patient suffers from, whereas differential etiology
is a term used to describe the process by which the cause of an injury is determined. Following
the trend among federal courts, however, we will use the term differential diagnosis to refer to
both concepts.” Guinn, 602 F.3d at 1253 n.6. As both parties in this case use the term
“differential diagnosis” to describe the latter process, this court will do the same.
65

“however, will not usually overcome the fundamental failure of laying
a scientific groundwork for the general toxicity of the drug and that it
can cause the harm a plaintiff suffered.” McClain, 401 F.3d at 1252.
Chapman, 766 F.3d at 1308-09 (concluding that the expert did not properly follow
the necessary steps for a differential diagnosis because she relied on temporal
proximity and failed to fully explore other potential causes of the plaintiff’s
alleged injury). A differential diagnosis is flawed if it presumes the existence of
general causation or if it is improperly based solely on a temporal relationship
between the drug and the plaintiff’s injuries. Kilpatrick v. Breg, 613 F.3d 1329,
1334 (11th Cir. 2010).
iv.

Dr. Hinshaw’s Testimony at Deposition That He
Relied on a Differential Diagnosis Violates Rule
26

Novartis critiques Dr. Hinshaw’s use of the differential diagnosis
methodology on both a substantive and procedural level. Novartis claims that
because Dr. Hinshaw did not discuss his reliance on a differential diagnosis in
either his expert report or the supplemental report, he cannot discuss the
methodology for the first time during his deposition. (Doc. 125-20) at 18. Novartis
believes that Dr. Hinshaw’s specific causation testimony should be excluded on
that basis alone, in reliance on Hamlett v. Carroll Fulmer Logistics Corporation,
176 F. Supp. 3d 1360 (S.D. Ga. 2016). In Hamlett, the magistrate judge excluded
66

the expert’s newly-presented opinions, which he supplied for the first time at
deposition, because “Rule 26(a)(2) deters procrastination and sandbagging,” and
the expert “did both.” 176 F. Supp. 3d at 1365.
Rule 26 requires that the disclosure of the identity of expert witnesses shall
be accompanied by a written report that contains a “complete statement of all
opinions the witness will express and the basis and reasons for them,” as well as
“the facts or data considered by the witness” in forming his or her opinion. FED. R.
CIV. P. 26(a)(2)(B)(i) (emphasis added).15 Expert reports must include both “how”
and “why” the expert reached a certain result, not just conclusory opinions.
Hamlett, 176 F. Supp. 3d at 1365. The court in Hamlett concluded that it was
“unacceptable to make a party wait, and thus be surprised, at a deposition . . . the
defendants were prejudiced because they were denied, prior to the deposition, the
full opportunity to digest [the expert’s] information and formulate their deposition
questions based on the same.” Id.
Jones counters16 that, although he does not explicitly describe it as a

15

The Advisory Committee Notes further clarify that Rule 26(a)(2) “imposes an
additional duty to disclose information regarding expert testimony sufficiently in advance of trial
that opposing parties have a reasonable opportunity to prepare for effective cross-examination.”
FED. R. CIV. P. 26(a)(2), Advisory Committee Notes, 1993 Amendment, 146 F.R.D. 633.
16

In a footnote, Jones also argues that because Dr. Hinshaw was cross-examined by
defense counsel at deposition, Novartis will “not be surprised at trial.” (Doc. 166-41) at 27 n. 10.
She cites to one of this court’s previous opinions, Diasource, Inc. v. Gamble, 2005 WL 6000494
67

differential diagnosis, Dr. Hinshaw “reference[d] other conditions he ruled out as
being the sole cause of her fractures” on page 13 of his Supplemental Report.
(Doc. 166-41) at 27 n.10. However, page 13 of Dr. Hinshaw’s Supplemental
Report merely discusses one condition associated with fragility fractures,
hyperparathyroidism (“HPP”), and rules out HPP based on a lack of causal
association in the medical literature. Hinshaw Supplemental Report, (Doc. 125-25)
at 14. Despite directing the court’s attention to this specific page, Jones fails to
demonstrate that Dr. Hinshaw sufficiently considered, and then eliminated, other
risk factors associated with femoral fractures in preparing his expert report.17 In
fact, all of Jones’s citations in her Response brief, where she alleges that Dr.
Hinshaw utilized a differential diagnosis, derive from his deposition testimony,

(N.D. Ala. August 18, 2005), for this proposition. However, the plaintiff in Diasource, unlike in
the current case, submitted a supplemental affidavit that “at least minimally complied with Rule
26(a)(2)(B),” so this court concluded that the defendants had enough information from the
affidavit to select their own expert, depose plaintiff’s expert, and avoid surprise at trial. Id. at *2.
Jones does not cite to any authority for the proposition that an expert who first reveals the “basis
and reasons” for his opinion for the first time at deposition, and has not included those bases and
reasons in his expert report or any supplemental filing, has not violated Rule 26's disclosure
requirements.
17

These other factors, as Novartis points out, include age and steroid use. Dr. Hinshaw
does discuss how Ms. Jones’s age and cortico-steriod use are “associated with an increasing risk
of fracture,” in his deposition, but he does not identify, and subsequently eliminate, these factors
in any differential diagnosis analysis in his reports. Hinshaw Deposition, (Doc. 125-21) at
91(356:16-25). In fact, Dr. Hinshaw admitted during deposition that at least some studies show
an increased or significant association between glucocorticoid steroids and AFFs, which
undermines his opinion that Reclast causes AFFs. Id. at 96(376:20-377:25).
68

not from his expert report or supplemental report. See Hinshaw Response Brief,
(Doc. 166-41) at 27-28.
Therefore, Dr. Hinshaw violated Rule 26, and did not put Novartis on
notice, by failing to sufficiently provide the methods and bases that he alleges he
used in opining on specific causation in his expert reports.
v.

The Lack of Reliable General Causation Opinion
Makes a Specific Causation Opinion Using a
Differential Diagnosis Methodology Unreliable

Finally, Dr. Hinshaw’s opinion is unreliable because general causation has
not been established between Reclast and AFFs. The Eleventh Circuit has
explained that if no expert has been offered who can provide an admissible general
causation opinion, then an expert may not rely on a differential diagnosis to prove
specific causation. See McClain, 401 F.3d at 1253 (“A valid differential diagnosis,
however, only satisfies a Daubert analysis if the expert can show the general
toxicity of the drug by reliable methods.”). In McClain, the medical articles and
analogies the experts relied on were not sufficient to establish general causation,
and “[i]n the absence of such a foundation for a differential diagnosis analysis, a
differential diagnosis generally may not serve as a reliable basis for an expert
opinion on causation in a toxic tort case.” Id.; see also id. (“[A]n expert does not
establish the reliability of his techniques or the validity of his conclusions simply
69

by claiming that he performed a differential diagnosis on a patient.”)
The court has already ruled that Dr. Hinshaw’s general causation opinion is
based on unreliable methodology. Therefore, based on the lack of reliable general
causation opinion, Dr. Hinshaw’s differential diagnosis is unreliable, and his
specific causation opinion will be excluded.
c.

Dr. Hinshaw Is not Qualified To Opine on Novartis’
Compliance with FDA Regulations

Dr. Hinshaw will not be permitted to testify as to whether Novartis’ clinical
trials and warnings were adequate and in compliance with FDA regulations.
Though Jones claims that Dr. Hinshaw “certainly intends to defer to Dr. Parisian
as to specific FDA regulations,” the brief also argues that his opinion is relevant to
whether Novartis had “reasonable evidence of a causal association with a drug”
such that a warning regarding femur fractures was warranted. Hinshaw Response,
(Doc. 166-41) at 29 (citing 21 C.F.R. § 201.57(c)(6)(i)). His eighth and ninth
opinions also conclude that Novartis did not act as a “reasonably prudent
pharmaceutical manufacturer” in its advertising, marketing, promotion,
distribution, sale, and labeling of Reclast. (Doc. 125-26) at 11-12.
Dr. Hinshaw has experience reviewing clinical trial documents and,
according to his deposition, was an assistant investigator in a trial involving

70

another bisphosphonate drug. Hinshaw Response, (Doc. 166-41) at 32 (citing
Hinshaw Deposition, (Doc. 125-21) at 7(19:5-23)). However, Dr. Hinshaw’s
qualifications do not extend to FDA labeling and marketing regulations. He is
qualified to discuss the scientific components of studies that were the subject of
communications between Novartis and the FDA, but he may not opine on the
adequacy of those communications themselves or on whether Novartis acted as a
reasonably prudent manufacturer should.
Dr. Hinshaw will also not be permitted to testify to whether Novartis
complied with FDA regulations, including whether Novartis failed to report
fractures to the FDA in July 2008. See Hinshaw Response, (Doc. 166-41) at 32.
Fortunately for Jones, Dr. Parisian does have the expertise to discuss the adequacy
of Novartis’ communications with and disclosures to the FDA, and she will be
permitted to do so, with certain limitations.
5.

Conclusion as to the Admissibility of Dr. Hinshaw’s
Testimony

Although Dr. Hinshaw is qualified to testify in this case, other than to
Novartis’ compliance with FDA regulations, he did not employ a reliable
methodology in forming either his general or specific causation opinions.
Therefore, Dr. Hinshaw’s testimony is due to be EXCLUDED in full, and the

71

Motion to strike his testimony (doc. 112) is due to be GRANTED.
C.

Dr. Wayne A. Taylor

Jones also offers the opinions of Dr. Taylor into evidence. Dr. Taylor has
offered an expert report, dated August 31, 2015, which includes his Curriculum
Vitae. (Doc. 125-17, the “Taylor Report”). Dr. Taylor also offered a supplemental
expert report, dated May 2, 2016. (Doc. 125-18, the “Taylor Supplemental
Report.”). Dr. Taylor was deposed by Novartis on June 3, 2016, and the deposition
transcript was filed into the record. (Doc. 125-16, the “Taylor Deposition”).
On August 15, 2016, Novartis filed a Motion To Strike Dr. Taylor’s expert
testimony (Doc. 116). A day later, Novartis filed a brief in support of its Motion
(doc. 125-19, the “Taylor Brief”). On September 19, 2016, Jones filed a Response
opposing Novartis’ Motion To Strike. (Doc. 166-40, the “Taylor Response”). On
October 14, 2016, Novartis filed a reply brief in support of its Motion To Strike.
(Doc. 177, the “Taylor Reply”).
In his report, Dr. Taylor offers the following opinions: (1) Novartis’ clinical
trials H2301 and L2310 do not demonstrate the safety of Reclast; (2) Novartis’
clinical trial H2301 offers “evidence of causality” between Reclast and AFFs; and
(3) Novartis provided incomplete and biased information to the FDA.
1.

Dr. Taylor’s Qualifications
72

According to his curriculum vitae, Dr. Taylor received a bachelor of science
in mathematics from Purdue University in 1976. (Doc. 117-3) at 3, the “Taylor
CV.” He received an M.S. in Statistics from Purdue University in 1978 and a
Ph.D. in Statistics from Purdue University in 1983. Id. at 3. Dr. Taylor worked as a
statistician at Baxter Healthcare Corporation for 22 years and at one point was
responsible for their Six Sigma program.18 Id. at 2. In 2000, Dr. Taylor retired
from his position as the Director of Quality Technologies at Baxter and founded
Taylor Enterprises, Inc. (“Taylor Enterprises”). Id.
At Taylor Enterprises, Dr. Taylor develops software packages and provides
consulting and training on statistics. Id. He is a leading expert on acceptance
sampling in the pharmaceutical, medical device, and diagnostics industries, and
his articles on selecting statistically valid sampling plans are used by the FDA in
their new inspector training. Id. He has taught two courses on acceptance sampling
to over 5,000 students and 50 companies, including the Center for Devices and
Radiological Health. Id.
According to his deposition testimony, Dr. Taylor has designed statistical
method sections for clinical trials of pharmaceutical products and medical devices.

18

Dr. Taylor testified at deposition that he was involved in health hazard review boards
that dealt with the safety of drugs manufactured by Baxter. Taylor Deposition, (Doc. 125-16) at 8
(25:21-22).
73

Taylor Deposition, (Doc. 125-16) at 5(13:14-14:17). He also claims to have
expertise in contributing information on variables that could affect the outcome of
a study to clinical trial design experts. Id. at 42(161:14-23).
2.

Dr. Taylor Is Qualified To Offer Some, but not All, of his
Opinions
a.

Dr. Taylor Was not Designated as a Causation
Expert

Novartis first briefly argues that Dr. Taylor should not be permitted to offer
a causation opinion because he was not designated as a general causation expert in
this case. Taylor Brief, (Doc. 125-19) at 7 n.5. Jones’s Response brief provides
little clarity as to the purpose for which Dr. Taylor’s opinion is offered. His
opinion certainly does not appear to be offered for specific causation, as Dr.
Taylor’s expert report does not even reference Ms. Jones. Novartis alleges that Dr.
Taylor was not designated as a general causation expert in Jones’s Rule 26(a)(2)
disclosures. Taylor Brief, (Doc. 125-19) at 7 n.5; see Jones’s Expert Disclosures,
(Doc. 119-1). Yet in his expert report and during his deposition, Dr. Taylor offers
causality opinions. Taylor Expert Report, (Doc. 125-17) at 3(claiming Study
H2301 “offers evidence of causality”).19

19

During his deposition, Dr. Taylor was asked, “Are you offering a causation opinion in
this case?” Taylor Deposition, (Doc. 125-16) at 37(143:11-12). Dr. Taylor responded, “[r]elative
to the analysis I did in this report, yes.” Id.
74

Jones does not respond to Novartis’ expert designation argument or provide
any evidence that Dr. Taylor was properly designated as a causation expert.
Instead, she counters that Dr. Taylor is not offering a medical causation opinion
but rather is offering a causation opinion based on statistical methods. Taylor
Response, (Doc. 166-40) at 12.
While Rule 26(a)(2) does require that retained experts provide a written
report with a complete statement of all experts’ opinions and the facts and data to
support those opinions, the Rule does not require that an expert specifically be
designated as a causation expert. See FED. R. CIV. P. 26(a)(2)(B). Novartis has
failed to point to any binding authority stating that a retained expert must be
specifically referred to as a “causation” expert before that expert can offer
causation opinions, and this court has not found any. Novartis does not claim that
it was prejudiced in any way by such a technical omission. Accordingly, Dr.
Taylor’s causation opinions will not be excluded on this basis alone.
b.

Dr. Taylor Is not Qualified To Offer a Causation
Opinion

Dr. Taylor lacks the expertise to offer a general causation opinion,
particularly an opinion based on his improper re-analysis of medical data. He has
expertise in designing statistical method sections for clinical trials and in

75

contributing statistical information on variables that could affect the outcome of a
study to clinical trial design experts. (Doc. 125-16) at 5(13:14-14:17); 42(161:1423). However, Dr. Taylor has never served as the principal designer of the
protocol of a clinical trial. Id. at 5(15:23-16:2). He does not have a medical degree
and does not have any medical expertise in bisphosphonate use or atypical femur
fractures. Id. at 3(7:13-20); 31(118:14-20). Dr. Taylor himself admitted that his
review of the medical literature was limited to ten journal articles. Id. at
60(235:14-16).
Given that Dr. Taylor does have extensive training and expertise in statistics
in the pharmaceutical and medical device industries, he will not be excluded from
opining on statistical methods or other matters within his expertise. However, as
discussed below, he is not qualified to offer his general causation opinion.
Additionally, his general causation opinion will be excluded because the
methodology Dr. Taylor relied upon in forming it is unreliable and will not assist
the trier of fact.
3.

Dr. Taylor’s Methodology Is Unreliable
a.

The Background Risk Methodology Has Been
Accepted by the Eleventh Circuit

Jones’s briefing states that Dr. Taylor relied on a background risk of the

76

disease methodology in forming his causation opinions. Taylor Response Brief,
(Doc. 166-40) at 4. Novartis counters that, while the background risk methodology
is widely accepted, Dr. Taylor did not satisfy the criteria of the methodology.
A reliable methodology must take into account the background risk of a
specific disease, which is “the risk that everyone faces of suffering the same
malady that a plaintiff claims without having exposure to the same toxin.”
McClain, 401 F.3d at 1243. As the Eleventh Circuit has explained,
The background risk is not the risk posed by the chemical or drug at
issue in the case. It is the risk a plaintiff and other members of the
general public have of suffering the disease or injury that plaintiff
alleges without exposure to the drug or chemical in question. The
background risks include all those causes of a disease, whether known
or unknown, excluding the drug or chemical in question.
Id. (emphasis in original). A failure to assess or identify the background risk of a
disease is a “serious methodological deficiency” because, “[w]ithout a baseline,
any incidence may be coincidence.” Chapman, 766 F.3d at 1307-08 (internal
citations omitted).
Ignoring available evidence about background risks may be fatal to an
expert’s general causation opinion. Kilpatrick, 613 F.3d at 1342 (finding that the
expert impermissibly “ignored such background risks. While recognizing the
existence of idiopathic (or unknown) causes of [the disease], he dismissed them by

77

merely stating that the risk of idiopathic [disease] is essentially zero. The failure to
take into account the potential for idiopathically occurring [disease] placed the
reliability of [the expert’s] conclusions into further doubt.”).
Jones correctly points out that in Chapman, McClain, and Kilpatrick, unlike
with Dr. Taylor’s opinion, the experts altogether failed to assess the background
risk of the relevant disease in question. Taylor Response, (Doc. 166-40) at 16.
However, an unreliable application of a background risk methodology leads to the
same result as a failure to consider the background risk at all: the expert’s opinion
will be excluded.
Novartis claims that Dr. Taylor’s application of the methodology was
flawed for five reasons: (1) Dr. Taylor ignored other estimates of background rates
of AFFs; (2) Dr. Taylor improperly re-categorized fractures described in a paper
by Dr. Black; (3) Dr. Taylor improperly compared the lower bound of a Reclast
rate confidence interval to a point estimate in another paper; (4) Dr. Taylor
improperly used a one-tailed, rather than two-tailed, test to calculate the
confidence interval; and (5) Dr. Taylor improperly included a spiral fracture in the
Reclast-treated portion of his study.
b.

Dr. Taylor Ignored Other Background Rates

Dr. Taylor based his opinion on a historical, or background, rate of .6 per
78

10,000 patient years, as set out in a paper by Dr. Feldstein. Taylor Response, (Doc.
166-40) at 9 (referencing A. Feldstein et al., Incidence and Demography of Femur
Fractures With and Without Atypical Fractures, 27 J. Bone Miner Res. 997 (2012)
(the “Feldstein Paper”)). Jones claims that Dr. Taylor relied on this historical rate
because the Feldstein Paper “used a large database with hundreds of thousands of
patients (millions of patient years)” and “discusses other papers with different
background rates.” Taylor Response, (Doc. 166-40) at 9-10. Jones claims that Dr.
Taylor also considered background rates of AFFs of .17 per 10,000 patient years20
and .32 per 10,000 patient years,21 both of which were smaller studies. Id. at 10.
However, Dr. Taylor’s reliance on the results found in the Feldstein Paper
conflicts with the findings of the 2010 and 2013 ASBMR Task Force Reports. The
2010 ASBMR Task Force Report22 determined that “the incidence of atypical

20

This rate is derived from B. Edwards et al., Bisphosphonates and Nonhealing Femoral
Fractures: Analysis of the FDA Adverse Event Reporting System (FAERS) and International
Safety Efforts, J. Bone Joint Surg. Am. 2013 (the “Edwards article,” Doc. 139-9).
21

This rate is derived from R. Meier et al., Increasing Occurrence of Atypical Femoral
Fractures Associated with Bisphosphonates Use, 172 ARCHIVES OF INTERNAL MED. E1, E6
(2012)(the “Meier article,” Doc. 139-10).
22

Shane E, et al., Atypical Subtrochanteric and Diaphyseal Femoral Fractures: Report of
a Task Force of the American Society for Bone and Mineral Research, J. Bone Miner Res. 2010
(“the 2010 ASBMR Report”). Reports that linked long-term use of BPS with atypical femur
fractures led the ASBMR to appoint a task force of 28 multidisciplinary experts to address key
questions related to the problem. Id. at 2. The task force reviewed relevant articles from January
1990 to April 2010 in forming their analysis. Id. at 3.
79

femoral fractures has come to medical attention principally in the setting of BP use
and that the incidence in the general population not exposed to BPs is unknown.”
The 2010 ASBMR Report, (Doc. 117-9) at 4 (emphasis added). The more recent
2013 ASBMR Report noted that every study of AFFs included patients who had
not been treated with bisphosphonates, “suggesting that the ‘background rate’ of
AFF in osteoporosis patients is not zero.”23 (Doc. 117-1) at 15.
Further, Jones admits that the more recent ASBMR Report found that “the
historical rate was still unknown,” yet she fails to explain why Dr. Taylor’s use of
the Feldstein Paper background rate is reliable when other reports have found that
the background rate is still unknown or unclear. Taylor Response, (Doc. 166-40)
at 16. Jones conclusorily states that it is “reasonable and reliable” for Dr. Taylor to
use the Feldstein Paper rate, but she fails to provide any support for this assertion.
Id.
Dr. Taylor further conceded in his deposition that he did not come across
any papers published prior to the Feldstein Paper that identified a background rate
for AFFs, partly because he “did not extensively look for other articles” and

23

Shane E, et al., Atypical Subtrochanteric and Diaphyseal Femoral Fractures: Second
Report of a Task Force of the American Society for Bone and Mineral Research, J. Bone Miner
Res. 2013 (“the 2013 ASBMR Report”)(Doc. 117-1) at 15. The 2013 ASBMR Report reviewed
studies on AFFs that were published since 2010. Id.
80

apparently relied only on the articles and rates cited in the Feldstein Paper. Taylor
Deposition, (Doc. 125-16) at 21(77:25-78:16). He also admitted that he did not
conduct a search for any relevant articles published after the Feldstein Paper and
provided no explanation as to why he chose not to do so. Id. at 21(78:23-25).
If any step in Dr. Taylor’s methodology is flawed, it is a fatal defect under
Daubert, as “any step that renders the analysis unreliable under the Daubert
factors renders the expert's testimony inadmissible.” McClain, 401 F.3d at 1245
(emphasis in original, internal citations omitted). Dr. Taylor’s failure to
substantiate his use of the Feldstein Paper rate, or explain why he did not review
other articles to determine that his use of that rate was reliable, draws the
admissibility of his background risk methodology into doubt.
c.

Dr. Taylor Conducted an Improper Re-Analysis
of Data in Dr. Black’s 2010 Analysis

Novartis also takes issue with several other aspects of Dr. Taylor’s
methodology. In particular, Novartis critiques Dr. Taylor’s re-analysis of data
originally published in an article by Dr. Dennis Black in the New England Journal
of Medicine in 2010. Taylor Brief, (Doc. 125-19) at 15-17. Dr. Black co-authored
a report that re-analyzed the results of three large, randomized bisphosphonate
trials, including Novartis’ trial H2301, for evidence of atypia in hip and femur

81

fractures. Id. at 11 (citing Black DM, et al., Bisphosphonates and Fractures of the
Subtrochanteric or Diaphyseal Femur, N. Engl. J. Med. 362: 1761-71, 1768
(2010)) (the “Black 2010 Analysis”, Doc. 117-12). Dr. Taylor’s background risk
methodology relies on the data and findings of Dr. Black’s analysis, so in order for
Dr. Taylor’s opinions to be considered reliable, his re-analysis of Dr. Black’s data
and findings must also be reliable.
Jones does not dispute Novartis’ description of Dr. Black as a worldrenowned statistician and epidemiologist with over twenty years of experience
investigating osteoporosis and treatment therapies. Taylor Brief, (Doc. 125-19) at
11. Jones further does not dispute Novartis’ summary of Dr. Black’s analysis.
Novartis states that Dr. Black’s report concluded:
There was no statistically significant difference between the occurrence
of subtrochanteric and diaphyseal femur fractures in the Reclast-treated
population and the placebo-treated population. Dr. Black was not able
to fully assess other indicia of atypia, such as non-comminution, and
therefore expressly did not determine whether any of these fractures
were, in fact, AFFs.
Id. at 11-12 (emphasis added).
Novartis argues that Dr. Taylor improperly took the six femur fractures that
Dr. Black identified in his analysis and re-classified three of those fractures as
AFFs, despite lacking the medical background or expertise to do so. Taylor Brief,

82

(Doc. 125-19) at 15-17; Taylor Reply, (Doc. 177) at 11-12.
Dr. Black’s analysis states that “in all three studies, atypical fractures could
not be fully assessed, since radiographs were not generally available.” The Black
2010 Analysis, (Doc. 117-12) at 9. Dr. Black’s analysis further explains,
In the three studies we reviewed, even among women with up to 10
years of bisphosphonate exposure, the risk of fracture of the
subtrochanteric or diaphyseal femur ranged from one to six cases per
10,000 patient-years. If we had been able to assess atypical
characteristics (e.g., cortical thickness and fracture morphology) and
limit our consideration to atypical fractures, we probably would have
excluded some additional events; this would have further reduced the
risk.
Id. (emphasis added). Dr. Black’s analysis makes clear that he was not able to, and
did not, analyze the prevalence of AFFs.
Dr. Taylor admitted at deposition that he is not a medical doctor and does
not have any medical expertise that would allow him to evaluate whether or not a
fracture is atypical. Taylor Deposition, (Doc. 125-16) at 31(118:14-20). He further
admitted that he took Dr. Black’s assessment, added Dr. Black’s statements on
whether there was a presence of trauma associated with those fractures, and
concluded that three of the fractures were in fact atypical femoral fractures. Id. at
31-33(120:13-125:19). Upon further questioning, Dr. Taylor admitted,
Q: Okay, its possible, is it not, that all of those fractures are
subtrochanteric or diaphyseal and still don’t qualify as atypical femur
83

fractures, isn’t it?
A: That I don’t know.
Q: You don’t know, but you looked at the table on trauma that’s in
[Dr. Black’s] paper and you made the decision that they did, those
three did qualify as atypical femoral fractures, but the two in the
placebo did not, isn’t that correct?
A: That is correct, based on the presence or absence of trauma.
Id. at 125:8-19. Dr. Taylor’s deposition testimony reveals that he only compared
each of the fractures in Dr. Black’s analysis with the presence of trauma associated
with each incident. However, trauma is only one of the five major features of an
atypical femoral fracture that are required to satisfy the case definition of an
AFF.24 Neither Dr. Taylor’s expert report nor his deposition testimony discusses
whether he found evidence of the other four necessary major features in forming
his conclusion that three of the femur fractures identified by Dr. Black were AFFs.
Novartis argues that, as a statistician, Dr. Taylor is not qualified to make a

24

The 2010 ASBMR Report states that evidence of all five major features of an atypical
femoral fracture is required to satisfy the AFF case definition. (Doc. 117-9) at 4. The five major
features required to find an AFF are (1) location “anywhere along the femur from just distal to
the lesser trochanter to just proximal to the supracondylar flare”; (2) association “with no trauma
or minimal trauma, as in a fall from a standing height or less; (3) “transverse or short oblique
configuration”; (4) Non-comminution; and (5) “Complete fractures extend through both cortices
and may be associated with a medial spike; incomplete fractures involve only the lateral cortex.”
Id. The 2013 ASBMR Report slightly altered these mandatory major features in the AFF case
definition by relaxing the “non-comminution” factor to include “minimal comminution” and
adding a sixth major requirement that there be a periosteal stress reaction at the fracture site.
(Doc. 117-1) at 2.
84

determination that associations between fractures and trauma means that some
fractures do not meet the case definition for AFFs, while other fractures do.
Novartis further argues that, even if Dr. Taylor were qualified, his methodology is
flawed because he only relied on the presence or absence of trauma, rather than all
five major features of an AFF, in forming his conclusions that certain fractures in
Dr. Black’s analysis were AFFs.
Jones does not respond to this line of criticism. Instead, she merely
distinguishes Dr. Black’s analysis from Dr. Taylor’s by stating that he relied on a
different methodology and had a different object of his analysis. Taylor Response,
(Doc. 166-40) at 17-18. Her Response also claims that there are approximately
fifty published studies showing a statistically significant AFF association with
bisphosphonate use, but she does not specifically cite or reference the findings of
any of these studies. Id. at 18. She merely concludes that Dr. Taylor’s re-analysis
is sound and any critique of his methodology should be reserved for crossexamination. Id.
Dr. Taylor’s re-analysis of Dr. Black’s analysis meets the same fate as that
of the expert in Perry v. United States, 755 F.2d 888 (11th Cir. 1985). In Perry,
the expert re-diagnosed a number of cases in the data collection that other doctors
had diagnosed differently in order to present his evidence of a statistical
85

relationship between a disease and a vaccine. Id. at 891. The Eleventh Circuit
affirmed the district court’s exclusion of the expert’s testimony because the expert
reached a conclusion as to the connection between the disease and the vaccine
“before commencing his research . . . a scientist who has a formed opinion as to
the answer he is going to find before he even begins his research may be less
objective than he needs to be in order to produce reliable scientific results.” Id. at
892.
In Jack v. Wellcome, Inc., 239 F. Supp. 2d 1308 (N.D. Ga. 2002), the court
similarly excluded an expert’s testimony as unreliable because it was not
supported
by sufficient facts or data. Instead, his conclusion is more properly
classified as a theory. This theory he derived from a reclassification and
analysis of data from prior studies. His methodology, however, is far too
speculative for the court to allow. [The expert] has reclassified data
based on his belief that certain events should properly be classified
under different categories than those the original researchers chose. He
has done so largely without review of any facts specific to any specific
patient or event. Rather, he has essentially hypothesized a possible or
probable scenario whereby the reclassification is justified. This post hoc
approach is certainly not reliable absent a particularized reason for the
individual reclassification . . . he did not rely on the raw data from the
previous studies.
Id. at 1316 (emphasis added).
Jones does not explain why Dr. Taylor is qualified to determine that certain

86

of the fractures in Dr. Black’s analysis were AFFs, when Dr. Black’s analysis
itself expressly denied making such a finding. Jones also does not explain why
finding an absence of trauma in relation to a fracture is sufficient to determine that
a fracture meets all five factors in the case definition of an AFF. Additionally, like
in Jack, Dr. Taylor has not claimed he relied on the raw data from Dr. Black’s
analysis; instead, he performed a post hoc analysis based on the evidence as
presented by Dr. Black. Dr. Taylor concluded that Dr. Black’s data showed AFFs
and worked backwards from that step without sufficiently demonstrating that these
fractures met the appropriate case definition, so his re-analysis is unreliable.
Because his re-analysis formed the basis of his background risk methodology, see
Taylor Response, (Doc. 166-40) at 17, his methodology is independently
unreliable for this reason.
d.

Dr. Taylor’s Inclusion of a Spiral Fracture From
Dr. Black’s Analysis Was Improper

Novartis also alleges that Dr. Taylor improperly failed to exclude a spiral
fracture from his re-analysis of the Black 2010 Report’s findings. Novartis claims
that one of the three fractures that Dr. Taylor labeled as an AFF was actually a
spiral fracture, which is expressly excluded from the AFF case definition. Taylor
Brief, (Doc. 125-19) at 21. Novartis argues that because these spiral fractures are

87

excluded from the AFF definition, the spiral fracture identified by Dr. Black
should not have been considered an AFF by Dr. Taylor in his re-analysis. Taylor
Brief, (Doc. 125-19) at 12-13, 21-22.
The 2010 ASBMR Report explicitly excludes “intertrochanteric fractures
with spiral subtrochanteric extension” from the case definition of AFFs. See 2010
ASBMR Report, (Doc. 117-9) at 4. Dr. Taylor’s report re-summarizes the results
from the H2301 study that were analyzed and reported in the Black 2010 Analysis.
Taylor Report, (Doc. 125-17) at 4. Dr. Taylor states that the Black 2010 Analysis
reported six cases of fractures that were the result of 5 events, as one patient broke
both hips simultaneously. Id. After determining which of these five events was
“associated with no or minimal trauma, as in a fall from a standing height or
less,”25 Dr. Taylor concluded that the three fractures in the treatment group
receiving Zoledronic acid were Atypical Femoral Fractures. Id. at 3-4.
However, in his deposition, Dr. Taylor admitted that he did not know
whether a spiral fracture was disqualified from being labeled an AFF and
conceded that the presence of only two AFFs in the treatment group that received
Zoledronic acid, rather than three, would alter his analysis:

25

Association with no trauma or minimal trauma is only one of five major features of an
AFF. Evidence of all five major features is required to satisfy the AFF case definition. (Doc. 1179) at 4.
88

Q: [Reading from Dr. Black’s 2010 analysis of the three fractures
identified in the treated group] In three women, receiving Zoledronic
acid, one had simultaneous fractures of both femoral shaft[s]. One had
a transverse fracture but no beaking and no focal or generalized cortical
thickening on radiography, and one had a spiral fracture on radiography,
although the morphologic features were not described as atypical in case
reports.
Do you see that?
A: Uh-huh.
Q: Do you know whether a spiral fracture disqualifies a fracture from
being considered atypical under the minor – major feature requiring
transverse or short oblique configuration?
A: No, I don’t.
Q: So you don’t know whether that’s actually an – the fact that it’s spiral
disqualifies it as an atypical femur fracture or not, do you?
A: No, I don’t.
Q: We’d have to rely on a doctor to figure that out, wouldn’t we?
Ms. Taylor: We object to that.
Mr. Johnston: Q: A medical doctor.
Ms. Taylor: We object.
The Witness: Yes.
Mr. Johnston: Q: Your position is that three AFFs in the Reclast arm of
2301 compared to the single point rate in the Feldstein article results in
a statistically significant differen[ce] in AFFs and Reclast treated
patients, right?
89

A: Yes.
Q: What if there were only two patients in the AFF side of the equation,
two out of 11,580?
A: That would alter my analysis.
....
A: Certainly if there were only two fractures that qualified as atypical
femur fractures in the Reclast treated population, the lower bound of
your number would be less than 7.1 per 100,000 patient years, wouldn’t
it?
A. Yes, it would lower the lower bound.
Q: And it may in fact lower the lower bound below Feldstein’s rate of
5.9 events per 100,000 patient years, couldn’t it?
A. If it lowers it, it may.
Taylor Deposition, (Doc. 125-16) at 34-35(131:8-133:16) (citing Black 2010
Analysis, (Doc. 117-12) at 6-7).
Dr. Taylor’s deposition testimony reveals that he failed to determine
whether the fractures identified in the 2010 Black Analysis met all five of the
major features associated with the case definition of an AFF. His testimony also
reveals his unfamiliarity with the need to differentiate spiral fractures from AFFs.
Dr. Taylor’s report shows that he erroneously relied on the spiral fracture in the
2010 Black Analysis in forming his opinion, and his deposition testimony

90

demonstrates that, had the spiral fracture been excluded, his analysis would
change. Dr. Taylor’s inclusion of a spiral fracture in his analysis, in contravention
of the AFF case definition in the 2010 ASBMR Report, renders his methodology
unreliable for this independent reason.
e.

Dr. Taylor’s Comparison of the Lower Bound of
the Confidence Interval to the Feldstein Point
Estimate Was Improper

Novartis also believes that Dr. Taylor improperly compared the lower bound
of the confidence interval to the point estimate in the Feldstein Paper. Taylor
Brief, (Doc. 125-19) at 17-19. Jones counters that Dr. Taylor’s comparison of the
lower bound of the confidence interval was “reasonable and reliable.” Taylor
Response, (Doc. 166-40) at 19.
Both parties agree that a point estimate and a confidence interval are both
statistical estimates. A point estimate is “an estimate of the value of a quantity
expressed as a single number.” David H. Kaye & David A. Freedman, “Reference
Guide on Statistics,” REFERENCE MANUAL ON SCIENTIFIC EVIDENCE 284, 292 (3d
ed. 2011) (“2011 Reference Guide on Statistics”). A confidence interval, in
comparison, is an
estimate, expressed as a range, for a parameter. For estimates such as
averages or rates computed from large samples, a 95% confidence
interval is the range from about two standard errors below to two
91

standard errors above the estimate. Intervals obtained this way cover the
true value about 95% of the time, and 95% is the confidence level or the
confidence coefficient.
Id. at 284-85; see also Siharath v. Sandoz Pharms. Corp., 131 F. Supp. 2d 1347,
1357 (N.D. Ga. 2001), aff’d sub nom. Rider v. Sandoz Pharms. Corp., 295 F.3d
1194 (11th Cir. 2002) (citing to an older version of the Reference Guide) (defining
a confidence interval as a “range of values calculated from the results of a study,
within which the true value is likely to fall; the width of the interval reflects
random error.”). Novartis also describes the confidence interval as the “full range
of values that the true background rate can take.” Taylor Brief, (Doc. 125-19) at
17.26,27
In his analysis, Dr. Taylor opined that the lack of overlap between the

26

As the 2011 Reference Guide on Statistics further explains, “a confidence interval for
the population average is centered at the sample average; the desired confidence level is obtained
by adding and subtracting a suitable multiple of the standard error. Statisticians who say that the
population average falls within 1 standard error of the average will be correct about 68% of the
time. Those who say ‘within 2 standard errors’ will be correct about 95% of the time, and those
who say ‘within 3 standard errors’ will be correct about 99.7% of the time, and so forth.” Id. at
244.
27

Additionally, the 2011 Reference Guide on Statistics explains, “[t]he main point is that
an estimate based on a sample will differ from the exact population value, because of random
error. The standard error gives the likely size of the random error. If the standard error is small,
random error probably has little effect. If the standard error is large, the estimate may be
seriously wrong. Confidence intervals are a technical refinement . . . .” Id. at 246. The Reference
Guide also explains that courts have often misinterpreted the confidence level as the chance that
repeated estimates will fall into the confidence interval. Id. at 247. This interpretation is
incorrect; rather, the confidence level “indicates the percentage of the time that intervals from
repeated samples would cover the true value.” Id.
92

Feldstein Paper point estimate (0.60) and his calculated rate of the lower
confidence interval (0.71) meant that there was a statistically significant difference
in the number of AFFs in Reclast-treated patients compared to the background
rate. Taylor Brief, (Doc. 125-19) at 17; Taylor Response, (Doc. 166-40) at 13. The
upper bound of the Feldstein Paper’s confidence interval was 0.74, however, and
the Feldstein Paper’s confidence interval was 95% (0.46-0.74). (Doc. 125-19) at
17. Dr. Taylor testified as follows at deposition regarding point estimates and
confidence intervals:
Q: Why do statisticians use confidence intervals instead of point
estimates? Why do you do that?
A: They often report them both in combination, but its [sic] basically
with the point estimate you have no idea how precise it is, and so the
confidence interval helps indicate how precise the estimate is.
Taylor Deposition, (Doc. 125-16) at 22(84:4-9). Dr. Taylor further testified,
Q: And so, given that, with the overlapping confidence intervals, you
can’t say that the rate of AFFs in the Reclast treated patients is
statistically significant different than the background rate as
represented by Feldstein if you take into account the confidence
interval in Feldstein, can you?
A: Using this approach of the confidence interval in Feldstein and the
estimate that I have that falls within the interval so I can’t – using this
analysis declare that it’s statistically different, no.
Q: So you can find that they’re statistically different using this
difference analysis, its [sic] not the analysis you use, but this different
93

analysis, you would not conclude there’s a statistically significant
difference, correct?
A: No. I agree.
Id. at 26 (97:21-25 to 98:1-11) (emphasis added). Novartis argues that Dr.
Taylor’s methodology in his report goes against his own testimony at deposition
by unreliably omitting the full confidence interval and relying only on the lower
bound of the confidence interval. Taylor Brief, (Doc. 125-19) at 19. Novartis
believes that Dr. Taylor should have taken the full confidence interval into
consideration and that his failure to do so is in contravention of standard practice
for statisticians. Id. at 18; Taylor Reply, (Doc. 177) at 13.
Aside from providing the court with definitions for point estimates and
confidence intervals, Jones provides the court with no support for her proposition
that Dr. Taylor’s comparison was reasonable and reliable. See id at 19. She does
not explain why Dr. Taylor’s analysis was both reasonable and reliable and cites to
no authority, other than Dr. Taylor’s own deposition, to support this claim. Jones’s
summary conclusion that Dr. Taylor has offered reliable evidence is not sufficient
to demonstrate his reliability to the court. Yet again, Jones asserts that if Novartis
disagrees with the reliability of Dr. Taylor’s testimony, its arguments go only to
the weight of the evidence and should be properly argued during cross-

94

examination after Dr. Taylor has been permitted to testify.
Throughout her Response, Jones continually relies on In re Chantix
(Varenicline) Prods. Liab. Litig., 889 F. Supp. 2d 1272, 1282 (N.D. Ala. 2012)
(Johnson, J.), where the district court found that the defendant’s arguments to
exclude plaintiff’s expert went to the weight of his testimony rather than its
admissibility. However, in In re Chantix, the court also found that “nothing
inherent in the defendant’s objections to [the expert’s] methodology addresse[d]
the reliability of his findings.” Id. Here, Novartis’ objections go to the heart of the
reliability of Dr. Taylor’s testimony. Therefore, Jones’s numerous citations to In re
Chantix are not persuasive. If the expert’s testimony is unreliable, it may not be
presented to a jury at all.
Jones has failed to adequately demonstrate why Dr. Taylor’s use of the
lower bound of the confidence interval rather than the full confidence interval was
appropriate. Therefore, Dr. Taylor’s opinion is independently unreliable for this
reason.
f.

Jones Has Failed To Demonstrate That Dr.
Taylor’s Use of a One-Tailed Rather Than TwoTailed Test To Calculate the Confidence Interval
Was Reliable

Novartis also claims that Dr. Taylor’s use of a one-tailed, rather than two-

95

tailed, test to calculate the lower bound of the confidence interval deviated from
both standard statistical practice and FDA guidance. Taylor Brief, (Doc. 125-19)
at 19-21. Jones counters that his use of a one-tailed test was appropriate, given that
Dr. Taylor’s objective was to compare the historical rate of AFF to the rate of AFF
in the Reclast-treated group. Taylor Response, (Doc. 166-40) at 20-22.
A one-sided, or one-tailed, hypothesis “[e]xcludes the possibility that a
parameter could be, for example, less than the value asserted in the null
hypothesis. A one-sided hypothesis leads to a one-sided (or one-tailed) test.” 2011
Reference Guide on Statistics at 291. A two-sided, or two-tailed, hypothesis is an
“alternative hypothesis asserting that the values of a parameter are different form –
either greater than or less than – the value asserted in the null hypothesis. A twosided alternative hypothesis suggests a two-sided (or two-tailed) test.” Id. at 300.
The terms “one-tailed” and “two-tailed” refer to the “tails or ends of the bell-shape
curve, which represents in graph form a random normal distribution.” Palmer v.
Shultz, 815 F.2d 84, 93 (D.C. Cir. 1987) (internal quotations and citations
omitted)(discussing whether reliance on a one-tailed or two-tailed test is
appropriate in the context of a Title VII discrimination case).
Essentially, the two tests are two different ways of measuring probability.
Palmer, 815 F.2d at 94; see also Barrow v. Bristol-Myers Squibb, No. 96-CV-689,
96

1998 WL 812318, at *23 n.217 (M.D. Fl. Oct. 29, 1998) (“Whether the study was
one-tailed or two-tailed is important because it determines the way relative risk
will be spread. In a two-tailed study, the risk is spread between two hypotheses,
while in a one-tailed study, all risk is allocated to one hypothesis.”).
A two-tailed test is more demanding than a one-tailed test because it
requires greater departures from the expected numbers for a given level
of statistical significance; this implies that a two-tailed test has a larger
Type II error rate and less power than a one-tailed test with the same
significance level. Since power and statistical significance are both
desirable attributes, two-tailed tests should not be used where one-tailed
tests are appropriate, especially if power is an issue.
Michael O. Finkelstein & Bruce Levin, “Testing Statistical Hypotheses,” Statistics
for Lawyers (2d ed. 2001) at 120-22 (Doc. 117-17) at 6. On the other hand,
Many scientific researchers recommend two-tailed tests even if there are
good reasons for assuming that the result will lie in one direction. The
researcher who uses a one-tailed test is in a sense prejudging the results
by ignoring the possibility that the experimental observation will not
coincide with his prior views. The conservative investigator includes
that possibility in reporting the rate of possible error. This routine
calculation of significance levels, especially when there are many to
report, is most often done with two-tailed tests. Large randomized
clinical trials are always tested with two-tails.
Id. at 5. Secondary sources cited by both parties indicate to the court that,
depending on the specific circumstances or the objective of the test, either a onetailed or a two-tailed test may be appropriate.

97

Novartis argues that Dr. Taylor is not able to point to medical literature
where only a one-tailed confidence interval was used. In his deposition, Dr. Taylor
was asked,
Q: And have you ever seen a medical paper that is published in the
literature that only addressed a one-tailed confidence interval?
A: I don’t recall that I have, but the objective here is different than what
generally is [sic] literature reporting estimates of which a two-sided is
appropriate for, whereas we’re trying to make a determination as to
whether the clinical trial data has evidence in it. That’s a different
objective.
Taylor Deposition, (Doc. 125-16) at 28(107:14-22). He further explained,
A: So in statistics, the decision to do a one sided or two sided interval
is dependent upon the objective of the analysis. If it’s a report where
you’re just estimating rates and reporting information and odds ratio, a
two sided approach is best for just an estimate, but when you’re trying
to compare numbers, specifically a number in one standard whether its
higher than it or lower than it, it makes sense to use the one sided that’s
appropriate for the comparison you are making.
Id. at 28 (108:8-16). Jones argues that a one-tailed test was appropriate in this case
because “the test is only interested in the one-tail (i.e. the lower tail). If the
historical rate is lower than the confidence interval, then it will necessarily be
lower than the higher confidence interval.” Taylor Response, (Doc. 116-40) at 21.
Novartis believes that the assumptions underlying Dr. Taylor’s methodology not
only contradict Dr. Taylor’s own deposition testimony but also show that Dr.

98

Taylor impermissibly allowed bias into his analysis. Taylor Brief, (Doc. 125-19) at
15.
As the proponent of Dr. Taylor’s testimony, Jones bears the burden under
Rule 702 of establishing the qualifications, reliability, and helpfulness of
testimony. Frazier, 387 F.3d at 1260. Jones’s briefing does not help clarify why
Dr. Taylor’s use of a one-tailed test was reliable, and Dr. Taylor may confuse the
jury by providing testimony on the subject. Therefore, Dr. Taylor’s testimony is
independently unreliable because Jones is not able to establish that his use of a
one-tailed, rather than two-tailed, test was appropriate.
g.

Dr. Taylor’s Opinion on Novartis’ Clinical Trials
(H2301 and L2310) Should Be Excluded

In his expert report, Dr. Taylor also opines that Novartis’ two clinical trials,
H2301 and L2310, do not demonstrate the safety of Reclast relative to AFFs
because the studies were “underpowered for definitive conclusions” and Novartis
“did not collect all the necessary information to classify many adverse events as
Atypical Femur Fractures.” Taylor Report, (Doc. 125-17) at 2. Novartis believes
that these opinions should be excluded because the trials were approved and
deemed appropriate by the FDA. Taylor Brief, (Doc. 125-19) at 17-18. Novartis
also argues that Dr. Taylor has limited experience with clinical trials and therefore

99

does not have the expertise to offer this opinion. Id. at 5-6.
As above discussed, Dr. Taylor does not have the expertise to offer opinions
on compliance with FDA regulatory standards. Additionally, Dr. Taylor does not
indicate in what way(s) either trial fails to meet the safety standards established by
the FDA. Further, Dr. Taylor admitted during his deposition that clinical trials may
never be adequately powered to find some sufficiently rare adverse events, which
can only be detected once the medicine is released into the general population
after being approved for marketing. Taylor Deposition, (Doc. 125-16) at 9(29:2330:8). Therefore, Dr. Taylor’s opinions on clinical trials H2301 and L2310 are not
reliable and will not assist the trier of fact. They will be excluded.
h.

Dr. Taylor’s Opinion That Novartis Provided the
FDA With Incomplete Data Should Be Excluded

In his Report, Dr. Taylor claims that Novartis provided incomplete and
biased data to the FDA. He opines that in over a third of the adverse event reports
given to the FDA in 2008, either no mechanism was reported or the event detail
was not provided. Taylor Report, (Doc. 125-17) at 6-7; see Taylor Deposition,
(Doc. 125-16) at 51(198:20-23) (“I believe [Novartis] should have collected, and
at least ask the question as to what were the events that led up to the fracture. It’s
just a co-variant that should be included in this type of study.”). Yet Dr. Taylor

100

admits, conversely, that he is not attempting to offer “an opinion in this case that
Novartis used an inadequate methodology to respond to the FDA in 2008.” Id. at
53(206:8-11).
As stated above, Dr. Taylor is not qualified to and will not be permitted to
opine on the adequacy of Novartis’ communications with the FDA, including in its
2008 Response to Request for Information. See (Doc. 191-5). Even assuming that
Dr. Taylor has the expertise to opine on the sufficiency of the statistical data
provided by Novartis to the FDA, his deposition testimony reveals that he did not
review the protocol of Novartis’ clinical trial before determining whether or not
the information provided to the FDA was incomplete.
For example, when asked why he thought that information on trauma was
not included in the clinical trial, Dr. Taylor resorts to speculation:
Q: If an investigator fails to report trauma, what is the clinical
sponsor supposed to do?
A: I guess it wasn’t in the study design. The study - clinical trials
study director shouldn’t let it get to that point as part of the study
design.
Q: Did you look at the efficacy end point evaluation for fractures and
what the design there was?
A: No, I did not.
....
101

Q: Let me ask you another question. Did you look at the – did you
look at the case report forms for the adverse reactions to see what
efforts Novartis had made to try to obtain the information on trauma?
A: No, I did not look at those.
Q: Those weren’t provided to you?
A: No.
Q: So it’s possible that Novartis tried to get that information and that
effort is documented, but they never could get it, isn’t that correct?
A: That’s correct.
Taylor Deposition, (Doc. 125-16) at 51-52(200:16-23 and 204:13-23) (emphasis
added).
Dr. Taylor’s testimony demonstrates that, having not been provided with all
necessary information himself, he is not qualified to form an opinion on Novartis’
alleged failure to provide the FDA with complete data. It may in fact be the case
that Novartis failed to provide the FDA in 2008 all the information in its
possession. However, Dr. Taylor is not qualified to opine on this topic since he
failed to review the underlying case reports and is not a regulatory expert. His
opinion that Novartis provided incomplete data to the FDA will also be excluded.
4.

Conclusion as to the Admissibility of Dr. Taylor’s
Testimony

Dr. Taylor lacks the expertise to offer the general causation opinion he
102

intends to offer, and he did not employ a reliable methodology in forming his
opinion. Therefore, Dr. Taylor’s testimony is due to be EXCLUDED in full, and
the Motion To Strike his testimony (doc. 116) is due to be GRANTED.
D.

Non-Retained Experts

Jones also offers the opinions of two non-retained experts, Dr. James
Worthen and Dr. Timothy Mark Ricketts, into evidence. As neither Dr. Worthen
nor Dr. Ricketts was specifically retained by Jones for the purposes of this
litigation, neither doctor was required to provide a written report pursuant to Rule
26(a)(2)(B) of the Federal Rules of Civil Procedure.28 Instead, pursuant to Rule
26(a)(2)(C), Jones was required to disclose (1) the subject matter on which Dr.
Worthen and Dr. Rickets were expected to present evidence and (2) a summary of
the facts and opinions to which they were expected to testify. FED. R. CIV. P.
26(a)(2)(C). These brief statements may be found in “Plaintiff’s Expert
Disclosures,” which has been admitted into evidence. (Doc. 119-1).
Dr. Worthen was deposed on April 16, 2015 (doc. 119-4, the “Worthen
Deposition”), and Dr. Ricketts was deposed on April 14, 2015 (doc. 119-5, the

28

Rule 26(a)(2)(B) states as follows, “Unless otherwise stipulated or ordered by the
court, this disclosure must be accompanied by a written report – prepared and signed by the
witness – if the witness is one retained or specially employed to provide expert testimony in the
case.” FED. R. CIV. P. 26(a)(2)(B).
103

“Ricketts Deposition”). On August 15, 2016, Novartis filed a Motion To Strike the
testimony of Dr. Worthen and Dr. Ricketts (doc. 118) and a brief in support of its
Motion (doc. 119, the “Non-Retained Experts Brief”). On September 19, 2016,
Jones filed a Response opposing Novartis’ Motion To Strike. (Doc. 144, the “NonRetained Experts Response”). On October 14, 2016, Novartis filed a reply brief in
support of its Motion To Strike. (Doc. 181, the “Non-Retained Experts Reply”).
For the following reasons, Novartis’ Motion To Strike the testimony of Dr.
Worthen and Dr. Ricketts is due to be GRANTED to the extent that either expert
intends to offer an opinion on causation.
1.

Dr. Worthen
A.

Dr. Worthen’s Qualifications

Dr. Worthen is a board-certified orthopaedic surgeon who has been
continuously practicing orthopaedic surgery since 2005. (Doc. 119-1) at 3,
(“Plaintiff’s Expert Disclosures”). He attended the University of Alabama
Birmingham (“UAB”) Medical School and completed his residency in orthopedic
surgery at UAB. Id. Jones claims that Dr. Worthen has treated thousands of
patients who suffered from femur fractures and has treated several patients who
have been treated for osteoporosis with bisphosphonate drugs. Id.
On October 26, 2011, Dr. Worthen was on call at St. Vincent’s Hospital
104

when Jones arrived with a femoral fracture. Worthen Deposition, (Doc. 119-4) at 6
(20:1-9). Dr. Worthen performed surgery on both of Jones’s femoral fractures.
(Doc. 119-1) at 3.29 Jones plans to offer Dr. Worthen to testify based on personal
knowledge gained from his care and treatment of Jones, including his own
examinations, diagnoses, and courses of treatment. Id. Dr. Worthen intends to
opine that (1) the complete femoral fracture of Jones’s right femur was an atypical
femur fracture; (2) Reclast “probably caused” Jones’s AFF in her right femur; (3)
the stress fracture of Jones’s left femur “was caused by the infused Reclast
following a history of bisphosphonate use.” Id. at 3-4.
b.

Dr. Worthen Is not Qualified To Offer an Expert
Opinion on Causation

Novartis argues that Dr. Worthen is not qualified to offer an expert opinion
about the cause of Jones’s femoral fractures. Novartis claims that this court’s
previous decision in Harvey v. Novartis Pharms. Corp., 895 F. Supp. 2d 1206
(N.D. Ala. 2012) provides very persuasive authority that Dr. Worthen is not
qualified and his opinions are not reliable.
In Harvey, this court excluded causation testimony by the plaintiff’s non-

29

The first surgery was performed on October 27, 2011. (Doc. 144) at 4. She
subsequently experienced a second femoral fracture, id. at 5, and Dr. Worthen operated on that
second fracture on February 10, 2012. Id.
105

retained expert because he was not qualified to testify to the cause of the
plaintiff’s injury, and his attempt to rule out alternative competing explanations
for plaintiff’s injury using a differential diagnosis was not reliable. Id. at 12121213. Even though the surgeon in Harvey was purportedly aware of the
relationship between osteonecrosis and BP use and had extensive surgical training,
he had no practical experience in identifying the causes of that type of injury and
testified that he assumed that BP use had caused the plaintiff’s injury. Id. at 1212;
see id. at 1211 (“[The surgeon] testified that he does not conduct medical or
scientific research, has never researched osteonecrosis of the jaw, has never
researched bisphosphonates, has never published or submitted a paper on either
osteonecrosis of the jaw or bisphosphonates, and has never taught on either
subject.”). As the court explained, “the issue before the court is not if Harvey had
osteonecrosis of the jaw, but what caused her to have it.”) Id. at 1212 (emphasis in
original).
Jones has understandably attempted to distinguish this court’s opinion in
Harvey from this litigation, but her attempts fall short of the mark. Like in Harvey,
Dr. Worthen has little expertise in determining the cause of AFF. In his deposition,
Dr. Worthen testified that prior to treating Jones, he had never made any
presentations or conducted any research on AFFs. Worthen Deposition, (Doc. 119106

4) at 26(99:19-100:2). Similar to the expert in Harvey, Dr. Worthen has never
published or submitted any academic publications on BPs, and he is not an
endocrinologist. Id. at 25(94:14-95:6). Dr. Worthen further testified that he has
never conducted and is not even aware of a study that evaluated the rate of AFFs
in patients who received Reclast to treat osteoporosis. Id. at 25(95:7-12).
Dr. Worthen is not a member of the ASBMR and has not read either the
2010 or the 2013 ASBMR Task Force’s reports that discuss whether there is a
correlation between BP use and AFFs. Id. at 29(110:11-111:23); see id. at
29(112:13-21) (“[Y]ou know, I obviously, didn’t spend the time doing the – doing
the research. And – and – and there are some prominent names on the list that are
part of this task force. There – you know, from a causal association, yes, we
haven’t – I guess, none of the articles that I was – have seen claim that they’ve
claimed a causal association.”).
As this court has previously concluded, “[e]xperience in a particular field is
not enough to qualify an expert; the expert must have experience with the issue
before the court.” Harvey, 895 F. Supp. 2d at 1209. Dr. Worthen’s education,
training, and experience do not qualify him to testify regarding the causes of Ms.
Jones’s femoral fractures. Additionally, as explained below, because Dr. Worthen
failed to perform a reliable differential diagnosis, his testimony is independently
107

due to be excluded because it is not sufficiently reliable.
c.

Dr. Worthen Does not Employ a Reliable Differential
Diagnosis Methodology

Dr. Worthen asserts that his causation opinion was reliably based on a
differential diagnosis methodology. Worthen Deposition, (Doc. 119-4) at 12
(42:10-13). Novartis argues that Dr. Worthen’s differential diagnosis is unreliable
because he did not rule out other potential causes of Jones’s injury; he did not
have access to her full medical history and records; and he did not sufficiently
establish the existence of general causation required in order to render his
differential diagnosis reliable.
A differential diagnosis is “accomplished by determining the possible
causes for the patient's symptoms and then eliminating each of these potential
causes until reaching one that cannot be ruled out or determining which of those
that cannot be excluded is the most likely.” Guinn, 602 F.3d at 1253 (11th Cir.
2010) (citing Westberry v. Gislaved Gummi AB, 178 F.3d 257, 262 (4th Cir.
1999)).
The Eleventh Circuit has specifically addressed the reliability of differential
diagnoses. See Guinn, 602 F.3d at 1253; see also Kilpatrick, 613 F.3d at 1342–43;
McClain, 401 F.3d at 1252–53. In Guinn, the Eleventh Circuit explained that a

108

reliable differential diagnosis has two steps. 602 F.3d at 1253. First, the expert
must identify and consider the factors which could be the “sole cause of the
plaintiff's injury.” Id. Second, the expert “must provide a reasonable explanation”
for his conclusion that the other potential causes identified in step one are not the
sole cause of the plaintiff's injury. Id. A court must examine for itself whether the
expert properly considered other potential causes for the plaintiff's injury and
properly excluded them. Id.
i.

Dr. Worthen Failed To Reliably Rule out
Jones’s Steroid Use and Other Risk
Factors as Causes of her Injuries

Dr. Worthen’s differential diagnosis fails the Eleventh Circuit’s reliability
test. First, Dr. Worthen failed to rule out other conceivable causes of Jones’s
injury, which is undoubtedly partly due to the fact that he did not have full access
to her medical records. He admitted at deposition that he did not have access to
Jones’s prior medical records, so his diagnosis relied upon Jones’s self-reported
medical history rather than a full and extensive review of her records. Worthen
Deposition, (Doc. 119-4) at 25(96:2-16). He did not know, for example, the
duration, extent, or timing of Jones’s use of bisphosphonates other than Reclast.
Id. at 26(97:7-19)).
In Haller v. AstraZeneca Pharms. LP, 598 F. Supp. 2d 1271 (M.D. Fla.
109

2009), the court excluded an expert’s specific causation testimony as
impermissibly “hurried, cursory, and incomplete” because he did not properly
consider other factors that could have caused the plaintiff’s injury. Id. at 1295-96.
As the district court explained,
[t]he Court accepts as a general proposition that a physician need not
necessarily examine a patient, interview that patient, or speak with the
patient’s treating physician(s) in order to render opinions regarding
diagnosis, prognosis, course of treatment and perhaps even causation.
In other words, the Court does not deem it necessarily fatal that an
expert medical witness has relied on medical records alone to reach a
specific causation opinion. However, [the plaintiff’s] particular
circumstances expose the failings of applying to this specific case what
might otherwise be considered a generally acceptable methodology . .
. [g]iven this background [of a pattern of weight gain and loss and a
recent release from prison], simple logic and common sense dictated that
[the expert] do more than merely “skim” [the plaintiff’s] medical records
and summaries to ascertain the cause of [the alleged injury].
Haller, 598 F. Supp. 2d 1271, 1294-95 (M.D. Fla. 2009)(emphasis added).
Like in Haller, Dr. Worthen did not sufficiently determine whether there
were pre-existing factors other than Jones’s Reclast injections that could have
caused the alleged injury, particularly Jones’s previous high-dose steroid use to
treat a different condition. When questioned at deposition about whether he
considered Jones’s steroid use, Dr. Worthen responded as follows:
Q: Did you consider any other medications other than bisphosphonates?
A: Yeah. So there’s other – there’s other medications that slow — that
110

make bones weaker. The form of steroid use that she has had before,
that’s a consideration for fractures. But, usually, with steroid use, they’re
more associated with the typical osteopenic fractures or osteoporosis
fractures, not with atypical femur fractures.
Q: So, what methodology did you use to rule out steroid use, then, as a
potential cause?
A: It being an atypical femur fracture.
Worthen Deposition, (Doc. 119-4) at 32(124:1-16). However, he also testified that
“you can’t rule [steroid use] out” as a possible cause of an atypical fracture. Id. at
33(126:14-127:2). Dr. Worthen’s deposition testimony on his exclusion of steroid
use as a potential cause is circular, self-contradictory, and would likely be
confusing to a jury.
Jones had a number of other risk factors, including age and postmenopausal
status, and Dr. Worthen’s testimony does not demonstrate that he reliably ruled
out these other factors, or even if he was aware of their existence. Dr. Traylor, the
rheumatologist who first prescribed Reclast to Jones to treat her osteoporosis,
stated that she had pre-existing “persistent osteoporotic T-scores” and continued to
exhibit “persistent osteoporosis” at the time of her third infusion of Reclast. (Doc.
119-2) at 25(95:10-23). Additionally, Dr. Traylor testified that Jones was
diagnosed with dorsal kyphosis as of January 2009, prior to her first Reclast
injection, which is a “sign” of “ compression fractures that are unrecognized.” Id.
111

at 29(112:9-20)). Regardless of whether Dr. Worthen was aware of these preexisting conditions based on the medical history he was provided by Jones, he
failed to rule out these alternative causes.
Because Dr. Worthen never sufficiently explained why other potential
causes, such as Jones’s steroid use, could not have caused her injury, his opinion
that Reclast caused Jones’s femoral fractures amounts to mere conjecture and
speculation. Under Guinn, Dr. Worthen’s failure to adequately and reasonably
explain why other factors could not have contributed to Jones’s injury renders his
differential diagnosis not sufficiently reliable. 602 F.3d at 1253. Therefore, his
specific causation opinion is excluded as unreliable.
ii.

Without a Reliable General Causation
Opinion, a Differential Diagnosis
Methodology Is Unreliable

The Eleventh Circuit has previously explained that if no expert can offer a
reliable general causation opinion, it would be unreliable for an expert to use a
differential diagnosis to prove specific causation. See McClain, 401 F.3d at 1253
(“A valid differential diagnosis, however, only satisfies a Daubert analysis if the
expert can show the general toxicity of the drug by reliable methods.”). The
Eleventh Circuit concluded that, because the medical articles and analogies the
experts relied on were not sufficient to establish general causation, “[i]n the
112

absence of such a foundation for a differential diagnosis analysis, a differential
diagnosis generally may not serve as a reliable basis for an expert opinion on
causation in a toxic tort case.” Id; see also McClain, 401 F.3d at 1253 (“[A]n
expert does not establish the reliability of his techniques or the validity of his
conclusions simply by claiming that he performed a differential diagnosis on a
patient.”).
Dr. Worthen, who has not conducted research on BPs, does not attempt to
provide a general causation opinion himself. In his deposition, Dr. Worthen also
admitted he had not read any medical literature where the author claimed to have
found a causal association between BP use and AFFs. Worthen Deposition, (Doc.
119-4) at 29(112:18-21). Jones’s assertion that Dr. Worthen’s specific causation
opinion is reliable because it is “based on the information provided by and
generally accepted in the scientific community,” (doc. 144) at 21, mischaracterizes
the scientific consensus and fails to establish the general toxicity of Reclast.
The court has already ruled that Dr. Hinshaw’s general causation opinion is
based on unreliable methodology, and the general causation opinion of Jones’s
alternative expert, Dr. Taylor, has also been excluded. Dr. Parisian has also been
excluded from opining on causation or causal association. Without the foundation
of a general causation opinion, Dr. Worthen’s testimony is both unreliable and
113

irrelevant. See Rink, 400 F.3d at 1294 (affirming the district court’s exclusion of a
treating physician as irrelevant because the doctor relied on foundational
testimony that the court determined was unreliable); Allison v. McGhan Med.
Corp., 184 F.3d 1300, 1320 (11th Cir. 1999) (affirming exclusion of expert
witness on relevance grounds because two other experts were excluded that were
crucial to the expert’s relevance). Therefore, Dr. Worthen’s differential diagnosis
is unreliable and irrelevant for this independent reason, and his specific causation
opinion will be excluded.
2.

Dr. Ricketts
a.

Dr. Ricketts’s Qualifications

Dr. Ricketts is an internist and general practitioner who is board certified in
internal medicine. Plaintiff’s Expert Disclosures, (Doc. 119-1) at 5. He attended
UAB Medical School and completed a residency and internship in internal
medicine. Id. Dr. Ricketts has provided Ms. Jones with medical care since 2004.
(Doc. 119-1) at 5. Dr. Ricketts claims that Jones presented to him with prodromal
pain in her legs, which he claims is a sign of impending femur fractures. Id. at 6.
He also previously treated Jones for polymyalgia rheumatica, which attacks
muscles in the upper shoulder or hips. Ricketts Deposition, (Doc. 119-5) at
7(23:19-25).
114

Jones plans to offer Dr. Ricketts to testify that Jones’s Reclast infusions
caused her femur fractures, that Jones’s medical conditions were of the type
described in medical literature caused by BP use, and that there was no other cause
of Jones’s fractures. (Doc. 119-1) at 5. As Dr. Ricketts is equally as unqualified to
testify as Dr. Worthen and offers opinions that are equally as unreliable, he will be
excluded from testifying as a causation expert.
B.

Dr. Ricketts Is not Qualified To Offer an Expert
Opinion on Causation

In his deposition, Dr. Ricketts stated that he has seen thousands of patients
with osteoporosis, some of whom have had osteoporosis-related fractures. (Doc.
119-5) at 4(10:17-24). However, he testified that he did not directly treat Jones’s
osteoporosis and took a “long absence” from his practice during the time when
Jones was treated for her femoral fractures. Id. at 11(37:18-24). Dr. Ricketts also
did not review or have copies of any of Jones’s femoral x-rays in her chart and
merely relied on the orthopedist’s evaluation of the fracture because he has “no
reason to doubt this guy.” Id. at 17(62:24-63:-20). Additionally, he stated that
Jones “would have been [his] first and only” patient to present with a femoral
fracture. Id. at 11(37:10-14).
Jones argues that, as her “primary care physician who has managed her

115

overall health for over ten years, Dr. Ricketts is more than qualified to opine to
any of her healthcare related questions.” Non-Retained Experts Response, (Doc.
144) at 24. While Dr. Ricketts may provide his opinion as to treatment decisions
he made, his education, training, and expertise do not qualify him to provide
expert testimony as to whether Jones’s Reclast injections actually caused her
femoral fractures. Therefore, his causation opinions are due to be excluded.
C.

Dr. Ricketts Does not Employ a Reliable Differential
Diagnosis Methodology

Additionally, his testimony is independently due to be excluded because it
relies substantially on the opinions of another physician and is not itself reliable.
In her Response, Jones claims that Dr. Ricketts consulted with Dr. Worthen but
also formed his own, independent differential diagnosis. (Doc. 144) at 24.
However, Dr. Ricketts does not actually state anywhere in his deposition that he
used a differential diagnosis methodology. Further, he admits in his deposition that
he is not in a position to reliably testify about the cause of her fracture:
Q:

Okay. So if you were called to testify at trial, you wouldn’t be
testifying about – you wouldn’t have an opinion about the – the
cause of her fracture. Is that something that the orthopedist would
be in a position to have an opinion on?
Ms. Taylor: Object to the form.

A:

Right now I would have the opinion that it did cause it. Not that
116

I diagnosed it, but I trust this surgeon. I saw the pathology
reports.30 And I would have to testify that it caused it.
Q.

Okay. So the basis for you stating that would be the orthopedist’s
opinion?

A:

Orthopedist’s opinion.

Q:

Okay.

A:

Yeah. Absolutely.

Q:

Okay. So you would – you’d be able to relay what your
understanding was if the orthopedist –

A:

Absolutely.
Ms. Taylor: Object to the form.

A:

I can tell exactly what the orthopedist told me.

Q:

So you would have had conversations with the orthopedist?

A:

Yes.

Ricketts Deposition, (Doc. 119-5) at 9(30:23-31:17). Dr. Ricketts’ testimony
demonstrates that he did not himself diagnose Jones and would not himself be able

30

Novartis also claims that the pathology reports that Dr. Ricketts refers to are nonexistent, so there is no way that Dr. Ricketts could rely on them in forming any opinion.
Novartis’ claim is based on Dr. Worthen’s testimony at his own deposition that he could not
recall any bone pathology being performed on bone fragments from Jones’s surgeries. Worthen
Deposition, (Doc. 119-4) at 33(127:15-22). Even if Dr. Ricketts was able to review pathology
reports that Dr. Worthen did not see or does not remember, Jones has not identified any exhibits
offered into evidence as the pathology report in question. Therefore, the pathology report may not
form a reliable basis for Dr. Ricketts’s differential diagnosis, to the extent that one was
conducted.
117

to form a specific causation opinion based on his own expertise or methodology.
Lastly, as above explained, a differential diagnosis will only satisfy a
Daubert inquiry if the expert can show that the general toxicity of the drug has
been reliably established. McClain, 401 F.3d at 1253. Dr. Ricketts has not reliably
established that an association exists between Reclast injections and femoral
fractures. Therefore, for the independent reason that Dr. Ricketts cannot show that
Reclast has generally been shown to cause AFFs, Dr. Ricketts’ specific causation
testimony will be excluded.
3.

Conclusion as to the Admissibility of Dr. Worthen’s and Dr.
Ricketts’s Testimony

Neither Dr. Worthen nor Dr. Ricketts employed a reliable methodology in
forming their respective opinions that Reclast caused Jones’ femoral fractures.
Therefore, the Motion to Strike the testimony of the Non-Retained Experts (doc.
120) is due to be GRANTED IN PART. The Motion is due to be GRANTED as
to both doctors to the extent that either intends to offer a causation opinion and is
otherwise DENIED.
V.

CONCLUSION
For the above reasons, Novartis’ Motion To Strike the expert testimony of

Dr. Suzanne Parisian (doc. 108) is HEREBY DENIED to the extent that it seeks

118

to exclude Dr. Parisian’s testimony in its entirety. However, Novartis’ Motion is
due to be GRANTED IN PART and DENIED IN PART as to the particular
areas of Dr. Parisian’s testimony that it seeks to exclude.
Novartis’ Motion To Strike the expert testimony of Dr. Hinshaw (doc. 112)
is HEREBY GRANTED.
Novartis’ Motion To Strike the expert testimony of Dr. Taylor (doc. 116) is
HEREBY GRANTED.
Novartis’ Motion To Strike the expert testimony of the non-retained experts,
Dr. Worthen and Dr. Ricketts (doc. 118), is HEREBY GRANTED to the extent
that either expert intends to offer an opinion on causation and is otherwise
DENIED.
DONE and ORDERED this the 26th day of January.

VIRGINIA EMERSON HOPKINS
United States District Judge

119