Full opinion text
MEMORANDUM AND ORDER SARIS, District Judge. Table of Contents I. INTRODUCTION..........................................................121 II. BACKGROUND...........................................................122 A. A Brief History of Neurontin Litigation.................................122 B. The Legal Backdrop and the Current Motion............................123 C. A Scientific Primer....................................................125 1. Evidence of Causation in Medicine & the Courtroom....................125 a. Epidemiological Studies and Statistics............................125 b. Non-epidemiological Evidence...................................127 2. Neurotransmitters..................................................127 3. How Gabapentin Works: The Scientific Debate.........................128 III. DISCUSSION.............................................................130 A. The Court’s Gatekeeping Role..........................................130 B. Epidemiological Evidence of an Association between Neurontin and Suicide.............................................................132 1. Hiking Bradford Hill................................................132 2. The FDA Study....................................................133 a. Dr. Gibbons’ Critique...........................................137 b. Does the FDA Study Withstand Dr. Gibbons’Critique? .............139 c. Drug Specific Evidence .........................................140 3. The Collins and McFarland Study....................................144 C. Plaintiffs’ Theory of Biological Plausibility.............................145 1. Does Gabapentin Increase the Amount of GABA in the Brain? ...........145 2. Does Gabapentin Lead to a Decrease in Monoamines?...................145 a. In vitro studies ................................................145 b. In vivo studies.................................................147 c. Kumho Wrestling..............................................149 3. Does a Decrease in Monoamines Lead to Depression and Suicidality?.....150 a. Serotonin and Mood............................................151 b. Serotonin and Suicide..........................................151 D. Additional Evidence: Adverse Event Data...............................153 1. Adverse Events in Pre- and Post-Approval Clinical Trials...............153 2. Dechallenge/Reehallenge Events .....................................154 3. Periodic Safety Update Reports......................................156 4. Peer-Reviewed Literature...........................................156 5. Trimble’s Research for Defendants...................................156 E. Expert-Specific Challenges............................................157 F. Conclusion...........................................................158 IV. ORDER ..................................................................159 I. INTRODUCTION In these products liability cases, Plaintiffs allege that they, or their decedents, suffered suicide-related injuries when their doctors prescribed the drug Neurontin, manufactured by defendants Pfizer and Warner-Lambert Company (“Defendants”). Specifically, they allege that Neurontin caused behavioral disturbances, depression, and ultimately suicidal actions (including completed suicide) in over one hundred individuals. Defendants have moved to exclude Plaintiffs’ expert testimony on the issue of general causation— that is, the question of whether Neurontin has the capacity to cause the alleged suicide-related events (suicide attempt, gesture, ideation, and/or completed suicide). (Docket No. 1157.) After a three-day hearing, conducted jointly with Justice Marcy S. Friedman of the New York State Supreme Court, and review of the voluminous briefing and submissions, the motion is DENIED. II. BACKGROUND A. A Brief History of Neurontin Litigation Defendants manufacture and distribute the prescription drug Neurontin, or gabapentin (generic). In December 1993, the Food and Drug Administration (“FDA”) approved Neurontin for use as an “adjunctive therapy” (i.e., second-line treatment) in the treatment of partial seizures in adults with epilepsy. See In re Neurontin Mktg. and Sale Practices Litig., 244 F.R.D. 89, 92 (D.Mass.2007). In May 2002, the FDA approved Neurontin for the management of post-herpetic neuralgia (pain resulting from nerve damage caused by shingles or herpes zoster) in adults. Id. In the late 1980s and early 1990s, Parke-Davis, a division of Warner-Lambert, filed patent applications for Neurontin as a treatment for depression, neurodegenerative disease, mania, and bipolar disorder; Parke-Davis, however, never sought FDA-approval for any of these indications. Id. Without FDA approval of Neurontin for indications beyond epilepsy and post-herpetic neuralgia, the law prohibited Defendants from marketing or promoting Neurontin for other (“off-label”) uses. Id. As this Court detailed extensively in a prior opinion, Defendants face allegations that they nevertheless engaged in an extensive off-label marketing scheme, promoting Neurontin for a variety of off-label uses, including psychiatric disorders such as bipolar, mood, and anxiety disorders. See id. at 92-103 (detailing the alleged off-label marketing scheme). In July 2004, Defendants pled guilty to two criminal counts related to the off-label marketing and misbranding of Neurontin. Civil actions were filed in federal courts across the country, all of which have been consolidated into the multi-district litigation presently before this Court. The Neurontin multi-district litigation has two distinct parts: (1) “Sales and Marketing” actions brought by consumer purchasers and third party payors for damages stemming from Defendants’ alleged fraudulent off-label marketing scheme and (2) “Products Liability” actions alleging injuries resulting from the prescription and subsequent use of Neurontin. The current motion relates to the products liability actions alleging suicide-related injuries caused by Neurontin. B. The Legal Backdrop and the Current Motion In order to prevail in a pharmaceutical personal injury case, a plaintiff must establish two types of causation: general and specific. In re Bextra and Celebrex Mktg. Sales Practices and Prod. Liab. Litig., 524 F.Supp.2d 1166, 1171-72 (N.D.Cal.2007) (consumers alleging cardiovascular injury in a products liability suit against drug manufacturer); In re Rezulin Prods. Liab. Litig., 369 F.Supp.2d 398, 401-02 (S.D.N.Y.2005) (diabetes patients alleging liver injuries in products liability actions against drug manufacturer). As explained in the Federal Judicial Center’s Reference Manual on Scientific Evidence, “General causation is established by demonstrating, often through a review of scientific and medical literature, that exposure to a substance can cause a particular disease.... Specific, or individual, causation, however, is established by demonstrating that a given exposure is the cause of an individual’s disease____” Mary Sue Henifin et al., Reference Guide on Medical Testimony, in Reference Manual on Scientific Evidence 439, 444 (Fed. Judicial Ctr.2d ed.2000) (hereinafter “Reference Guide on Medical Testimony ”). Only general causation— whether Neurontin is capable of causing suicide-related events — is at issue in this motion. Plaintiffs offer the testimony of three experts — Dr. Michael Trimble, Dr. Stefan Kruszewski, and Dr. Cheryl Blume— to establish general causation. They opine that, by altering the brain chemistry of its users, Neurontin has the biological capacity to cause mood and behavioral changes that predictably result in suicidality. As the Plaintiffs’ key witness on the issue of biological plausibility, Dr. Trimble puts forth a three-step explanation for how Neurontin can cause .such changes: First, gabapentin (Neurontin) increases the amount of GABA (gamma-aminobutyric acid), a neurotransmitter, in the brain. Second, this increase of GABA leads to a decrease of other neurotransmitters in the brain, like serotonin and norepinephrine. And third, the decrease of serotonin and norepinephrine can prompt behavioral disturbances, depression, and suicidal behavior. Thus, according to Plaintiffs, taking Neurontin can increase a patient’s risk of suicidality. All three of Plaintiffs’ experts present a variety of scientific evidence, including human and animal studies, case reports, and scientific literature, as support for the position that Neurontin can increase the risk of suicidality in patients. Defendants challenge the admissibility of Plaintiffs’ expert testimony on general causation, asserting that it is insufficiently reliable under Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), and Federal Rules of Evidence 702 and 703. Defendants contend that Plaintiffs lack scientific evidence demonstrating a statistically significant association between Neurontin and suicide-related events. According to Defendants, this purported omission is a fatal flaw in Plaintiffs’ case. Defendants also attack Plaintiffs’ experts’ three-step theory of biological plausibility, contending that, as a whole, it lacks widespread acceptance in the scientific community and, more specifically, that particular pieces of the theory are contradicted by current scientific literature. Defendants offered three experts of their own — Dr. Charles Taylor, Dr. Robert Gibbons, and Dr. Anthony Rothschild — to support their contentions, each of whom submitted a report and testified at the hearing. C. A Scientific Primer 1. Evidence of Causation in Medicine & the Courtroom Epidemiology is the field of public health and medicine that studies the incidence, distribution, and etiology of disease in human populations with the purpose of better understanding disease causation. Michael D. Green et al., Reference Guide on Epidemiology, in Reference Manual on Scientific Evidence 333, 335 (Fed. Judicial Ctr.2d ed.2000) (hereinafter “Reference Guide on Epidemiology”). “Epidemiology focuses on the question of general causation (i.e., is the agent capable of causing disease?).” Id. at 336. a. Epidemiological Studies and Statistics Epidemiological studies are research studies designed to identify associations between a drug and a disease. These studies generally “identiffy] agents that are associated with an increased risk of disease in groups of individuals, quantify] the amount of excess disease that is associated with an agent, and provide[ ] a profile of the type of individual who is likely to contract a disease after being exposed to an agent.” Id. at 335-336. “An association is not equivalent to causation,” and so epidemiological studies, on their own, “cannot objectively prove causation.” Id. at 336, 374. Instead, an identified association must be evaluated by researchers to determine whether the association is causal. Id. at 374 (“[Cjausation is a judgment for epidemiologists and others interpreting the epidemiological data.”). Nevertheless, epidemiological studies are often offered as evidence supporting a theory of general causation in the courtroom. See In re Viagra Prods. Liab. Litig., 572 F.Supp.2d 1071, 1078 (D.Minn.2008). There are many types of epidemiological studies. The “gold standard” for determining the relationship between a drug and a health outcome is a randomized, double-blind, placebo-controlled clinical trial. Reference Guide on Epidemiology, supra, at 338. In such a trial, subjects are assigned randomly to one of two groups: one receives the drug and the other does not, often receiving a placebo instead. Id. at 338. The study is also “double-blind,” meaning that neither the participants nor those conducting the study knows which group is receiving the actual drug and which group is receiving the placebo. Id. These studies are often used to evaluate new drugs or medical treatments, but because ethical constraints preclude exposing human subjects to agents believed to cause adverse effects, these experimental studies cannot be undertaken to investigate a suspicion that a drug increases the risk of suicide. Id. at 338-39; Giles v. Wyeth, 500 F.Supp.2d 1048, 1058 (S.D.Ill.2007) (“Suicide presents researchers seeking to study it with both ethical and practical difficulties.”). Thus, most epidemiological studies are “observational,” not experimental like the studies described above. Observational studies compare subjects already exposed to the drug to those not exposed. Reference Guide on Epidemiology, supra, at 339. Oftentimes, epidemiological studies lack the statistical power needed for definitive conclusions, either because they are small or the suspected adverse effect is particularly rare. Id. at 380; see, e.g., Giles, 500 F.Supp.2d at 1058 (noting that, “[a]s a rare event, studying [suicide] for purposes of causation requires a huge number of participants”). The technique of meta-analysis, where study results are pooled “to arrive at a single figure to represent the totality of the studies reviewed,” was developed to address such situations. Reference Guide on Epidemiology, supra, at 380. Meta-analysis “systematizes] the time-honored approach of reviewing the literature” and provides a “standardized framework with quantitative methods for estimating risk.” Id. Meta-analysis is “most appropriately]” used to pool randomized experimental trials, but “if carefully performed it may also be helpful for observational studies.” Id. at 361 n. 76. One statistical approach for expressing an association in epidemiologic research, including meta-analyses, is an odds ratio. Id. at 350. An odds ratio “expresses in quantitative terms the association between exposure to an agent and a disease.” Id. An odds ratio uses a null value of 1.0. An odds ratio greater than 1.0 means that there is a positive association, while an odds ratio lower than 1.0 indicates a negative association. Another method for investigating whether a drug is associated with a particular event is risk difference, or attributable risk. This measure “represents the amount of disease among exposed individuals that can be attributed to the exposure.” Id. at 351. Unlike odds ratios, risk differences have a null value of zero. Thus, a risk difference greater than zero suggests a positive association, and a risk difference less than zero suggests a negative association. The risk difference method is often used in meta-analyses where many of the individual studies (which are all being pooled together in one, larger analysis) do not contain any individuals who developed the investigated side effect. Whereas such studies would have to be excluded from an odds ratio calculation, they can be included in a risk difference calculation. Regardless of which method — odds ratio or risk difference — is used, a finding of an association is often assessed statistically to determine whether the result likely represents a true association or simply random error. Reference Guide on Epidemiology, supra, at 354. A study found to have “results that are unlikely to be the result of random error” is labelled “statistically significant.” Id. Statistical significance, however, does not indicate the strength of an association found in a study. Id. at 359. “A study may be statistically significant but may find only a very weak association; conversely, a study with small sample sizes may find a high relative risk but still not be statistically significant.” Id. To reach a “more refined assessment of appropriate inferences about the association found in an epidemiologic study,” researchers rely on another statistical technique known as a “confidence interval.” Id. at 360 (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 confidence interval provides an indication of the precision of the risk figure found in the study. Id. at 389. b. Non-Epidemiological Evidence Medical researchers also rely on a variety of non-epidemiological evidence when assessing causation. For example, researchers often examine and analyze case reports, descriptions of a particular patient’s clinical history and symptoms. As explained in the Reference Manual on Scientific Evidence: “Case reports lack controls and thus do not provide as much information as controlled epidemiological studies do.... Causal attribution based on case studies must be regarded with caution. However, such studies may be carefully considered in light of other information available.” Reference Guide on Medical Testimony, supra, at 475. Scientists also look beyond studies on living humans, turning to animals and cell and tissue cultures. Animal studies have the advantage of being able to be conducted as true experiments, with exposure controlled and measured. However, extrapolation from animal studies to humans entails some risks, as physiological differences and dosage differences can complicate comparisons. In re Rezulin Prods. Liab. Litig., 369 F.Supp.2d at 406-07 (citing Reference Guide on Epidemiology, supra, at 345-46). Experiments on cell and tissue cultures (either human or animal) are often called in vitro studies, as a means of distinguishing them from in vivo studies, meaning on live humans and animals. Id. While in vitro studies are often used, extrapolation from laboratory conditions to live patients can also be problematic. Id. 2. Neurotransmitters Plaintiffs’ theory of how Neurontin causes increased risk of suicide (referred to as a theory of biological plausibility) centers on the impact that Neurontin has on neurotransmitters in the brain. Accordingly, a basic understanding of neurotransmitters is needed to assess Plaintiffs’ theory and Defendants’ challenges to it. A neurotransmitter is a natural signaling chemical contained in the body. Communication between cells is essential to the effective functioning of any complex multicellular organism, and the major mode of intercellular communication is the transmission of chemical substances, specifically neurotransmitters. Principles of Pharmacology: The Pathophysiologic Basis of Drug Therapy 59 (David E. Golan, et al. eds., 2005). A nerve cell, or neuron, is connected to another nerve cell, by a specialized junction called a synapse. In a scenario where Nerve Cell A is going to transmit a signal to Nerve Cell B, Nerve Cell A is called the presynaptic cell, and Nerve Cell B is called the postsynaptic cell. Principles of Neural Science 22-23 (Eric R. Kandel et al. eds., 4th ed.2000). The transmission process begins when an electrical signal, known as an action potential, travels through the presynaptic cell (Nerve Cell A) down to its tip, known as the presynaptic terminal. Id. at 21-23. At the presynaptic terminal are collections of synaptic vesicles, each holding thousands of specific neurotransmitters. Id. at 182. Also at the presynaptic terminal are ion channels, designed for rapid information processing; these channels open and close like gates in response to particular electric or chemical signals. Id. at 105-7,182-3. When the signal (action potential) reaches the presynaptic terminal of Nerve Cell A, it causes the channels to open, leading to an influx of calcium ions. This, in turn, triggers the opening of the synaptic vesicles and the release of the packaged neurotransmitters into the area between the two nerve cells, known as the synaptic cleft. Id. at 183, fig. 10-7. Some of these released neurotransmitters then bind to receptor molecules on the post-synaptic cell (Nerve Cell B), causing that cell’s ion channels to open or close, and ultimately creating either an excitatory or inhibitory reaction. Id. at 183-84. Under Plaintiffs’ theory, Neurontin affects several different neurotransmitters in the human brain. First, Plaintiffs contend that Neurontin increases the amount of the neurotransmitter known as GABA in the brain. GABA is the primary inhibitory neurotransmitter in the brain and spinal cord. Golan, supra, at 147; Kandel, supra, at 214. Second, Plaintiffs contend that the increase in GABA leads to a decrease in several monoamine neurotransmitters, namely serotonin, norepinephrine, and dopamine. Serotonin is associated with depression and aggression, while norepinephrine is related to movement and depression. (Trimble Rep. 8); see Principles of Neural Science, supra, at 283. Third, an established theory, known as the monoamine theory of depression, states that decreased levels of serotonin and/or norepinephrine neurotransmission causes depression. {See Trimble Rep. 8) (“Early observations were that drugs which depleted the brain’s reserves of monoamines (serotonin, dopamine, and norepinephrine in particular) led to depression.”); cf. Golan, supra, at 184. Plaintiffs’ experts also present scientific literature that links low levels of serotonin in the brain to suicidal behavior. 3. How Gabapentin Works: The Scientific Debate Plaintiffs’ and Defendants’ experts concur that the scientific mechanism by which gabapentin works is not fully understood. (See also Label for Neurontin (Supp. No. 041), April 23, 2009, at 2) (stating that the mechanism by which gabapentin exerts its pain-relief and anticonvulsant actions is “unknown”). Developed as an antiepileptic compound, “[gjabapentin was originally conceived to be similar in chemical structure (and therefore in function) to ... GABA.” (Taylor Rep. 5.) As the primary inhibitory neurotransmitter in the brain, GABA has been viewed as a key to designing a therapeutic strategy for epilepsy, a condition precipitated by a lack of inhibition (or an increase in excitation) in some areas of the brain. Thus, many antiepileptic drugs were designed to counteract the over-excitation in an epileptic’s brain by increasing the amount of GABA in the brain. (See Hr’g Tr. 52-3, June 19, 2008 (Trimble.)) Such drugs are often referred to as “GABAergic.” Over time, though, researchers began to question whether gabapentin acted in the same way as other GABAergic agents and set out to investigate exactly how the drug worked. Some studies, published in the peer-reviewed literature, have indicated that, rather than acting on GABA-related receptors, gabapentin binds to a subunit of the calcium ion channel located at the edge of a presynaptie nerve cell. (Taylor Rep. 5-6.) Specifically, it is theorized that gabapentin acts at a specific part of the calcium channel known as the alpha-2delta subunit. (Id.) Debate over the chemical and pharmacological properties of gabapentin has significant repercussions for this litigation. In advancing their theory of general causation, Plaintiffs characterize gabapentin as a GABAergic drug and rely on studies of other GABAergic drugs which have been shown to lead to negative effects on mood and behavior. (See, e.g., Trimble Rep. 18) (discussing studies that have “shown that ... AEDs [antiepileptic drugs] which increase brain GABA lead to negative effects on mood and behavior”). In fact, when hired by Warner Lambert to investigate the relationship between gabapentin and behavioral disturbances in the mid-1990s, Dr. Trimble (now one of Plaintiffs’ experts) advised the company that one of the strongest associations with anticonvulsant drugs generally was to depression. (Trimble Rep. 29; see Michael Trimble, Psychosis with Gabapentin (Neurontin), May 20, 1995) (Pis.’ Ex. 17.) And, more recently, the FDA has documented a statistically significant association between GABAergic antiepileptic drugs and an increased risk of suicide. Defendants, however, contend that the belief that gabapentin is GABAergic (which they admit has been published in peer-reviewed literature) has been “mostly discounted by subsequent research.” (Taylor Rep. 6; see id. at 11-16) (listing and discussing studies which suggest that gabapentin has little, if any, GABAergic qualities). Defendants’ expert Dr. Taylor insists that “gabapentin is both chemically and pharmacologically distinct” from the GABAergic drugs referenced by Plaintiffs’ experts. (Taylor Rep. 4.) In sum, the parties — and most significantly their two very qualified experts, Dr. Trimble and Dr. Taylor — fundamentally disagree as to whether gabapentin is properly deemed GABAergic and, significantly, whether Plaintiffs’ reliance on comparisons to and extrapolations from studies examining other GABAergic drugs renders their theory of causation unreliable under Daubert, III. DISCUSSION A. The Court’s Gatekeeping Role The admission of expert evidence is governed by Federal Rule of Evidence 702, which codified the Supreme Court’s holding in Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), and its progeny. See United States v. Diaz, 300 F.3d 66, 73 (1st Cir.2002); see also Fed.R.Evid. 702 advisory committee’s note. Rule 702 states: If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise, if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (3) the witness has applied the principles and methods reliably to the facts of the case. Fed.R.Evid. 702. The trial court must determine whether the expert’s testimony “both rests on a reliable foundation and is relevant to the task at hand” and whether the expert is qualified. Daubert, 509 U.S. at 597, 113 S.Ct. 2786; Diaz, 300 F.3d at 73 (“[A] proposed expert witness must be sufficiently qualified to assist the trier of fact, and ... his or her expert testimony must be relevant to the task at hand and rest on a reliable basis.... ”). An expert’s methodology is the “central focus of a Daubert inquiry,” but a court “may evaluate the data offered to support an expert’s bottom-line opinions to determine if that data provides adequate support to mark the expert’s testimony as reliable.” RuizTrocke v. Pepsi Cola of P.R. Bottling Co., 161 F.3d 77, 81 (1st Cir.1998); see Bonner v. ISP Techs., Inc., 259 F.3d 924, 929 (8th Cir.2001) (deeming it clear that “it is the expert witnesses’ methodology, rather than their conclusions, that is the primary concern of Rule 702” and stating that a court cannot exclude testimony asserting a “novel” conclusion if the methodology and its application are reliable). Because “the admissibility of all expert testimony is governed by the principles of Rule 104(a),” the proponents of the expert testimony must establish these matters by a preponderance of the evidence. Fed.R.Evid. 702 advisory committee’s note (citing Bourjaily v. United States, 483 U.S. 171, 107 S.Ct. 2775, 97 L.Ed.2d 144 (1987)). “The proponent need not prove to the judge that the expert’s testimony is correct, but she must prove by a preponderance of the evidence that the testimony is reliable.” Moore v. Ashland Chem., Inc., 151 F.3d 269, 276 (5th Cir.1998). Daubert itself listed four factors which should guide judges in this determination: (1) whether the theory or technique can be and has been tested; (2) whether the technique has been subject to peer review and publication; (3) the technique’s known or potential rate of error; (4) the level of the theory’s or technique’s acceptance within the relevant discipline. United States v. Mooney, 315 F.3d 54, 62 (1st Cir.2002) (citing Daubert, 509 U.S. at 593-94, 113 S.Ct. 2786). “These factors, however, are not definitive or exhaustive, and the trial judge enjoys broad latitude to use other factors to evaluate reliability.” Mooney, 315 F.3d at 62 (citing Kumho Tire, 526 U.S. 137, 153, 119 S.Ct. 1167, 143 L.Ed.2d 238 (1999)); see United States v. Vargas, 471 F.3d 255, 261 (1st Cir.2006) (“The trial court enjoys broad latitude in executing its gate-keeping function; there is no particular procedure it is required to follow.”); Hollander v. Sandoz Pharm. Corp., 289 F.3d 1193, 1206 (10th Cir.2002) (noting that “different courts relying on essentially the same science may reach different results” when evaluating evidence under Daubert). In Kumho Tire, the Supreme Court was careful to emphasize that the trial judge must exercise her gatekeeping role with respect to all expert evidence, but that how she might exercise that role would necessarily vary depending on the type of testimony at issue. See Kumho Tire, 526 U.S. at 150, 119 S.Ct. 1167; United States v. Frazier, 387 F.3d 1244, 1262 (11th Cir.2004) (“Exactly how reliability is evaluated may vary from case to case, but what remains constant is the requirement that the trial judge evaluate the reliability of the testimony before allowing its admission at trial.”); Amorgianos v. Amtrak, 303 F.3d 256, 266 (2d Cir.2002) (recognizing that “the Daubert inquiry is fluid and will necessarily vary from case to case”). Under Kumho Tire, the critical inquiry is whether the expert “employs in the courtroom the same level of intellectual rigor that characterizes the practice of an expert in the relevant field.” 526 U.S. at 152, 119 S.Ct. 1167; Rider v. Sandoz Pharm. Corp., 295 F.3d 1194, 1197 (11th Cir.2002). When, for example, “the factual basis of an expert’s testimony is called into question, the district court must determine whether the testimony has ‘a reliable basis’ in light of the knowledge and experience of the relevant discipline.” Crowe v. Marchand, 506 F.3d 13, 17 (1st Cir.2007) (quoting Kumho Tire, 526 U.S. at 148, 119 S.Ct. 1167). The Court’s vigilant exercise of this gatekeeper role is critical because of the latitude given to expert witnesses to express their opinions on matters about which they have no firsthand knowledge, and because an expert’s testimony may be given substantial weight by the jury due to the expert’s background and approach. See Daubert, 509 U.S. at 595, 113 S.Ct. 2786; Kumho Tire, 526 U.S. at 148, 119 S.Ct. 1167 (noting that experts enjoy “testimonial latitude unavailable to other witnesses”); United States v. Hines, 55 F.Supp.2d 62, 64 (D.Mass.1999) (“[A] certain patina attaches to an expert’s testimony unlike any other witness; this is ‘science,’ a professional’s'judgment, the jury may think, and give more credence to the testimony than it may deserve.”). The Court must, however, keep in mind the Supreme Court’s admonition that, “[vigorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means of attacking shaky but admissible evidence.” Daubert, 509 U.S. at 596, 113 S.Ct. 2786. If an expert’s testimony is within “the range where experts might reasonably differ,” the jury, not the trial court, should be the one to “decide among the conflicting views of different experts.” Kumho Tire, 526 U.S. at 153, 119 S.Ct. 1167. “Only if the expert’s opinion is so fundamentally unsupported that it can offer no assistance to the jury must such testimony be excluded.” In re Viagra Prods. Liab. Litig., 572 F.Supp.2d at 1078 (quoting Bonner, 259 F.3d at 929-30). As the First Circuit has stated: Daubert does not require that a party who proffers expert testimony carry the burden of proving to the judge that the expert’s assessment of the situation is correct. As long as an expert’s scientific testimony rests upon “good grounds, based on what is known,” Daubert, 509 U.S. at 590, 113 S.Ct. 2786 (internal quotation marks omitted), it should be tested by the adversary process — competing expert testimony and active cross-examination — rather than excluded from jurors’ scrutiny for fear that they will not grasp its complexities or satisfactorily weigh its inadequacies, see id. at 596, 113 S.Ct. 2786. In short, Daubert neither requires nor empowers trial courts to determine which of several competing scientific theories has the best provenance. It demands only that the proponent of the evidence show that the expert’s conclusion has been arrived at in a scientifically sound and methodologically reliable fashion. Ruiz-Troche, 161 F.3d at 85. It is with these principles in mind that the Court assesses Defendants’ motion to exclude. B. Epidemiological Evidence of an Association between Neurontin and Suicide 1. Hiking Bradford Hill Defendants’ first attack on Plaintiffs’ evidence of general causation is that Plaintiffs fail to cite to any epidemiological study demonstrating a statistically significant association between Neurontin and suicide-related events. Without solid evidence of such an association, Defendants contend that Plaintiffs’ experts’ theory of how Neurontin might induce suicide-related events in patients is irrelevant and inadmissible. Referring to the Bradford Hill criteria (discussed further below), Defendants contend that failing to first identify an association between the drug and the negative effect before beginning a causation analysis violates the generally accepted methodology for establishing causation. Epidemiologic studies, while considered to be “powerful evidence of causation,” are not required to prove causation in a pharmaceutical personal injury case. See, e.g., Rider, 295 F.3d at 1198-99 (citing Eighth, Tenth, and Eleventh Circuit decisions holding that epidemiology is not required to prove causation in a toxic tort case); In re Meridia Prods. Liab. Litig., 328 F.Supp.2d 791, 800-01 (N.D.Ohio 2004) (surveying the pre- and post-Daubert landscape and concluding that “no court has held that epidemiological evidence is necessary to establish general causation when other methods of proof are available”), aff'd, 447 F.3d 861 (6th Cir.2006). While an epidemiological study is not per se required, establishing general causation without some “confirmatory” evidence of an association between the drug and the negative effect can be an uphill battle. See Lynch v. Merrell-Nat’l Labs., 830 F.2d 1190, 1194 (1st Cir.1987) (holding, in a pre-Daubert opinion, that without “confirmatory epidemiological data,” animal studies and extrapolations from studies of analogous drugs cannot establish causation in human beings); see e.g., Rider, 295 F.3d at 1202 (noting that plaintiffs can prove causation via non-epidemiological evidence but holding the evidence in that case to be insufficient). Defendants are correct in their assertion that an association is the starting point for the Bradford Hill criteria, one accepted approach to establishing causation. See Dunn v. Sandoz Pharms. Corp., 275 F.Supp.2d 672, 678 (M.D.N.C.2003) (concluding, in a Daubert inquiry, that an epidemiological study demonstrating an association is a prerequisite for proper application of Bradford Hill criteria). Developed by Sir Bradford Hill in the 1960s, the criteria are nine factors which researchers often consider when judging whether an observed association is truly causal. The Bradford Hill criteria are: (1) strength of the association; (2) consistency; (3) specificity of the association; (4) temporality; (5) dose-response curve; (6) biological plausibility; (7) coherence (with other knowledge); (8) experiment; and (9) analogy. A. Bradford Hill, The Environment and Disease: Association or Causation?, 58 Proc. Royal Soc’y Med. 295 (1965) (Defs.’ Ex. 16); see Gannon v. United States, 571 F.Supp.2d 615, 626 (E.D.Pa.2007) (listing and discussing the nine Bradford Hill factors) Amorgianos v. Nat'l R.R. Passenger Corp., 137 F.Supp.2d 147, 167-68 (E.D.N.Y.2001) (same). These factors are viewed as guidelines, and it is acknowledged that each factor need not be fulfilled in order for a researcher to proclaim causation. See A. Bradford Hill, supra, at 11 (“None of my nine viewpoints can bring indisputable evidence for or against the cause-and-effect hypothesis and none can be required as a sine qua non.”); Reference Guide on Epidemiology, supra, at 375 (referring to the criteria as “guidelines” and noting that the “drawing of causal inferences is informed by scientific expertise” as opposed to a strict scientific methodology). Several courts have recognized the Bradford Hill criteria as a generally accepted “tool for determining whether an epidemiological study establishes causation.” Dunn, 275 F.Supp.2d at 678-79 (citing “the small number” of reported federal and state cases discussing the criteria). Other courts have found the Bradford Hill criteria neither “necessary [n]or helpful.” In re Phenylpropanolamine (PPA) Prods. Liab. Litig., 289 F.Supp.2d 1230, 1243 n. 13 (W.D.Wash.2003). And in the context of a general causation challenge, failure to satisfy the Bradford Hill criteria does not doom admission under Daubert. See, e.g., In re Viagra Prods. Liab. Litig., 572 F.Supp.2d at 1081 (“The Court agrees that the Bradford Hill criteria are helpful for determining reliability but rejects Pfizer’s suggestion that any failure to satisfy those criteria provides independent grounds for granting its Daubert Motion.”); Dunn, 275 F.Supp.2d at 680 (rejecting an expert’s Bradford Hill-based testimony because he lacked evidence of an association but separately considering “whether [the plaintiff] can establish general causation independent of the Bradford Hill criteria”). Although courts have not embraced the Bradford Hill criteria as a litmus test of general causation, both parties repeatedly refer to the criteria, seemingly agreeing that it is a useful launching point and guide. Accordingly, this Court will begin its inquiry by evaluating Plaintiffs’ evidence of an association between Neurontin and suicide-related events, the starting point for an investigation under the criteria. 2. The FDA Study Plaintiffs trumpet a recent study conducted by the FDA that contains epidemiological data supporting their experts’ theories of general causation. In early 2005, the FDA initiated an inquiry into whether the use of antiepileptic drugs (“AEDs”) led to an elevated risk of suicidality, defined as suicidal behavior or ideation. After collecting data from manufacturers, the FDA conducted a meta-analysis, analyzing reports of suicidality from 199 placebo-controlled clinical studies covering eleven different AEDS, including Neuron-tin (gabapentin). (Statistical Review and Evaluation: AntiEpileptic Drugs and Suicidality, May 23, 2008, at 5) (Docket No. 1332, Ex. A) (hereinafter “Statistical Review”). The analysis included 27,863 patients treated with an AED and 16,029 patients in placebo groups. (Statistical Review at 5.) On January 31, 2008, the FDA issued an Alert entitled “Information for Healthcare Professionals-Suicidality and Antiepileptic Drugs” (“FDA Alert”), stating: “[P]atients receiving antiepileptic drugs had approximately twice the risk of suicidal behavior or ideation ... compared to patients receiving placebo.” (FDA Alert 1) (Pis.’ Ex. 31.) The Alert reported that the increased risk of suicidal behavior or ideation was “statistically significant,” and noted that “[flour of the patients who were taking one of the antiepileptic drugs committed suicide, whereas none of the patients in the placebo group did.” (Id. at 2.) The Alert also reported that patients treated for epilepsy, psychiatric disorders, and other conditions “were all at increased risk for suicidality when compared to placebo,” stating that there “did not appear to be a specific demographic subgroup of patients to which the increased risk could be attributed” and that the “results were generally consistent among the eleven drugs.” (Id. at 1.) The Alert advised that all patients treated with AEDs should be monitored closely for depression and suicidality and other unusual changes in behavior, explaining that “[s]ymptoms such as anxiety, agitation, hostility, mania and hypomania may be precursors to emerging suicidality.” (Id. at 2.) The Alert cautioned: This information reflects FDA’s current analysis of available data concerning these drugs. Posting this information does not mean that FDA has concluded there is a causal relationship between the drug products and the emerging safety issue. Nor does it mean that FDA is advising health care professionals to discontinue prescribing these products. FDA intends to update this document when additional information or analyses become available. (Id. at 1) (emphasis added); (see FDA Amicus Br. 5 (stating that the Alert “does not constitute a conclusion by FDA that the drugs subject to the Alert actually cause the adverse event,” but that, “[o]n the other hand, the disclaimer should not be read to suggest that FDA has concluded that the drugs are not causally linked to the adverse events at issue”)). In May 2008, the FDA released a “Statistical Review and Evaluation” describing the methodology and analysis it used in evaluating the collected data. The Statistical Review’s Executive Summary states: “In conclusion, antiepileptic drugs are associated with increased risk of suicidality relative to placebo in randomized placebo-controlled trials. The effect appears consistent among the group of 11 drugs.” (Statistical Review 6.) The review also revealed that the eleven drugs had been divided into three subgroups chosen by the medical officers from the FDA’s Division of Neurology. Gabapentin, along with four other drugs, was placed within the GABAergic and GABAmimetic drug group. {Id. at 13.) When tested by drug group, the GABAergic/GABAmimetic group demonstrated a statistically significant association with increased risk of suicidal behavior or ideation. {Id. at 32, fig. 7.) Finally, the statistical review revealed that, in addition to the primary analysis, the FDA also conducted three sensitivity analyses to examine the robustness of its primary analytical method. On July 10, 2008, the FDA held a Joint Meeting of the Peripheral and Central Nervous System Drugs Advisory Committee (“PCNS”) and the Psychopharmacologic Drag Advisory Committee (“PDAC”). The meeting was open to the public and offered “interested persons” the opportunity to “present data, information, or views, orally or in writing.” Meeting Notice, 73 Fed.Reg. 32588 (June 9, 2008). The Committee heard presentations from representatives of two drug manufacturers of AEDs contained within the FDA’s study, including Defendant Pfizer, and additional statements from an FDA Safety Reviewer. Discussing the significance and methodology of the FDA’s meta-analysis, Dr. Russell Katz, the FDA Director of the Division of Neurology Products, stated: “We’re unequivocally comfortable with ... saying that this establishes causality.” (Transcript of the Joint Meeting of the Peripheral and Central Nervous System Drugs Advisory Committee and the Psychopharmacologic Drag Advisory Committee, July 10, 2008, at 90) (Docket No. 1365, Ex. G) (hereinafter “FDA Hr’g Tr.”). Emphasizing that the agency had applied its regular methodology for determining causality, Dr. Katz then repeated his position that the FDA is “quite comfortable with saying there is causality.” (FDA Hr’g Tr. 90.) At the close of the meeting, the Committee members voted on four questions. First, the Committee, with twenty in favor and one abstention, voted to affirm the FDA’s overall finding of an increase in suicidality for the eleven AEDs analyzed. Second, the Committee, with eighteen in favor and three against, affirmed the FDA’s conclusion that the finding of increased suicidality should apply to each of the eleven drags in the analyses. Third, the Committee, with fifteen in favor, five against, and one abstention, affirmed the FDA’s conclusion that the finding should apply to all currently approved chronically administered AEDs, including drugs beyond the eleven included in the analysis. Finally, the Committee rejected, by a vote of fourteen to four with three abstentions, a proposal to place a “black box” warning (the most serious available) on the labels for all AEDs. However, the Committee did approve, by a vote of seventeen to four, a proposed labeling change for AED medication guides. On December 16, 2008, the FDA announced that it had completed its analysis and, based on the outcome of its review, is requiring all manufacturers of antiepileptie/anticonvulsant drugs to include a warning in their labeling and to inform patients of the risks of suicidal thoughts and actions. (See Updated FDA Alert, Dee. 16, 2008 (Docket No. 1600, Ex. A.)) The FDA stated that the general consistency of results among drugs with “varying mechanisms of action and across a range of indications suggests that the risk applies to all antiepileptic drugs used for any indication,” id,., but did not articulate a theory as to how such drugs increased the risk of suicidal thoughts and actions in patients. (Press Release, FDA News, FDA Requires Warnings about Risk of Suicidal Thoughts and Behavior for Antiepileptic Medications, Dec. 16, 2008) (stating that the biological reasons for the increased risk are “unknown”) (Docket No. 1600, Ex. D.) Neurontin’s revised label now contains an extensive seven paragraph warning, detailing the results of the meta-analysis and stating, inter alia: Suicidal Behavior and Ideation Antiepileptic drugs, including Neurontin, increase the risk of suicidal thoughts or behavior in patients taking these drugs for any indication. Patients treated with any AED for any indication should be monitored for the emergence or worsening of depression, suicidal thoughts or behavior, and/or any unusual changes in mood or behavior---- The risk of suicidal thoughts or behavior was generally consistent among drugs in the data analyzed. Neurontin Label, April 23, 2009, at 10. The label also contains a section discussing information that should be provided to patients: 1 Patients, their caregivers, and families should be counseled that AEDs, including Neurontin, may increase the risk of suicidal thoughts and behavior and should be advised of the need to be alert for the emergence or worsening of symptoms of depression, any unusual changes in mood or behavior, or the emergence of suicidal thoughts, behavior, or thoughts about self-harm.... Patients should be advised that Neurontin may cause dizziness, somnolence and other signs of CNS depression. Neurontin Label, April 23, 2009, at 13. It is widely recognized that, when evaluating pharmaceutical drugs, the FDA often uses a different standard than a court does to evaluate evidence of causation in a products liability action. Entrusted with the responsibility of protecting the public from dangerous drugs, the FDA regularly relies on a risk-utility analysis, balancing the possible harm against the beneficial uses of a drug. Understandably, the agency may choose to “err on the side of caution,” Rider, 295 F.3d at 1201, and take regulatory action such as revising a product label or removing a drug from the marketplace “upon a lesser showing of harm to the public than the preponderance-of-the-evidence or more-like-than-not standard used to assess tort liability.” McClain v. Metabolife Int’l, Inc., 401 F.3d 1233, 1250 (11th Cir.2005) (quoting Glastetter v. Novartis Pharm. Corp., 252 F.3d 986, 991 (8th Cir.2001)). In fact, FDA regulations provide that the agency can issue an Alert or warning label even before causation is established, (Hr’g Tr. 128-9, June 19, 2008 (Blume)), and the agency has, in a recent guidance document, stated that it has “begun taking a more comprehensive approach to making information on potential drug risks available to the public earlier.” (FDA Amicus Br. 2) (quoting Guidance: Drug Safety Information-FDA’s Communication to the Public (March 2007)). This earlier disclosure allows “healthcare professionals and patients [to] ... consider the information when making decisions about medical treatment” even when there may be “uncertainties in the data.” Id. at 3. As such, the decision by the FDA to require warnings on a drug label, without more, does not suffice to establish causation. Plaintiffs argue persuasively that even if this Court does not consider the FDA study as definitive proof of general causation, the study nevertheless qualifies as powerful epidemiological evidence establishing an association between Neurontin and suicidality. Whether the FDA meta-analysis is properly termed an epidemiological study is an area of dispute between the parties. Defendants’ expert Dr. Gibbons acknowledged that some would call it an epidemiological study, but disagreed with that characterization. (Hr’g Tr. 299-300, June 20, 2008.) He takes the position that, because the meta-analysis combined the results of many small clinical trials, it does not bear the hallmark characteristic of an epidemiological study, which, he says, typically involves very large populations. (Id.) On the other hand, Plaintiffs’ experts Blume and Trimble — who are not epidemiologists or biostatisticians but work regularly with such data — both testified that the study qualified as an epidemiological study. (Hr’g Tr. 57, June 19, 2008 (Trimble); Hr’g Tr. 170,174, June 20, 2008 (Blume.)) Moreover, the Reference Guide on Epidemiology suggests that a meta-analysis can itself be deemed an epidemiological study. See Reference Guide on Epidemiology, supra, at 380. For reasons discussed below, the Court finds that the FDA study is an epidemiological study that may be considered on the question of whether the Plaintiffs have produced evidence of an association between Neurontin and suicidality (the starting point for a causal inquiry under Bradford Hill). a. Dr. Gibbons’ Critique Defendants strenuously argue that the FDA study contains methodological flaws, which produce a misleading picture of the data. In their view, the data actually suggest that only two of the eleven drugs— neither of which is Neurontin — are associated with increased suicidality. Thus, Defendants maintain that any claim that the FDA Alert demonstrates an association between Neurontin and suicidality is erroneous. Defendants’ expert Dr. Robert Gibbons critiques the FDA study, stating that it fell prey to what he describes as a serious, but not uncommon, problem with meta-analyses: that a subset of the studies can “drive the overall results, making it appear as if there’s an overall effect that’s consistent.” (Hr’g Tr. 380-81, July 23, 2008.) Here, Dr. Gibbons points the finger at two of the eleven drugs: “My postmortem on the FDA Alert is that the entire analysis was driven by lamotrigine and topiramate. If those two drugs were not a part of FDA’s analysis, there would be no FDA Alert on anticonvulsants.” The issue raised by Dr. Gibbons is often referred to as “trial heterogeneity.” (Hr’g Tr. 320-21, June 20, 2008.) Gibbons highlights the fact that sixty-one percent of all of the suicidality events (thoughts and behaviors) observed in the entire meta-analysis came from the lamotrigine and topiramate studies, even though these two drugs account for only thirty-eight percent of the total data. (Supp. Expert Rep. of Robert D. Gibbons, July 11, 2008 at 2) (Docket No. 1363, Ex. A.) (hereinafter “Gibbons July Rep.”). Moreover, he points out that these two drugs were the only ones in the analysis which demonstrated a statistically significant increased risk in both the primary and secondary sensitivity analyses. (Gibbons July Rep. 2; Statistical Review at 24, fig. 2; 26, fig. 4.) While several other drugs, including Neurontin, demonstrated a positive association with suicidality events, these associations were not statistically significant. As further evidence for his position, Dr. Gibbons offered his own analysis of the data, in which he separated lamotrigine and topiramate from the other nine AEDs in the meta-analysis. While lamotrigine and topiramate together demonstrated a statistically significant association, the other nine AEDs in combination did not. (Gibbons July Rep. ¶ 3.) The “revelation” that these two drugs are driving the entire study, according to Dr. Gibbons, undercuts two FDA conclusions upon which Plaintiffs rely: (1) that the results were consistent among the eleven drugs, and (2) the FDA’s finding of a statistically significant association within the GABAergie or GABAmimetic drug subgroup, a group which included both topiramate and Neurontin. With respect to the GABAergic/GABA mimetic subgroup, Gibbons again insists that topiramate was the driving force behind the statistically significant finding; without topiramate, the remaining four drugs in the GABAergie subgroup together have an incidence rate of suicide events “virtually identical” to patients treated with placebo and thus producing no signal, much less one with statistical significance. (Gibbons ¶ 4; Hr’g Tr. 320, June 20, 2008 (Gibbons.)) Gibbons also points out that topiramate, which was found to increase the risk of suicidality in patients approximately two-and-a-half times, was the only one of the five drugs which, when analyzed independently, produced a statistically significant association. Grouping these five drugs together was therefore, according to Gibbons, empirically inappropriate: [Tjhese five drugs are showing very different results. Topiramate is going one way, and all of the other drugs are going the other way.... [F]rom an empirical basis, based solely on the numbers ... it’s inappropriate to apply a statistical method that assumes that all five of these drugs have a common risk.... You cannot come up with a reliable conclusion about the pooled risk when you have that level of heterogeneity. (Hr’g Tr. 321-22, June 20, 2008.) Having laid out his argument as to why the FDA’s overall finding of an increased risk of suicidality and its particular finding of an increased risk within the GABAergie subgroup are flawed, Dr. Gibbons next turns his eye to the gabapentin-specific data. In the FDA’s primary analysis, gabapentin yielded an odds ratio of 1.57, indicating a positive association between the drug and suicidal behavior or ideation. However, Gibbons emphasizes that the perceived association had an extremely wide confidence interval (.12 to 47.66) and was not statistically significant; thus the possibility that the association occurred by chance cannot be ruled out. (FDA Statistical Review 24, fig. 2.) Dr. Gibbons also attacks the FDA study for employing an odds ratio methodology, in which the FDA excluded all studies where no suicidality events occurred (“zero event” studies) from its primary meta-analysis. See discussion supra, at Part ILC.l.a. Gibbons maintains that because only three of the forty-nine studies on Neurontin submitted to the FDA by Defendants included qualifying incidents of suicidality (two incidents of suicidal ideation in Neurontin-treated patients and one in a patient receiving a placebo), this approach translated to the exclusion of the vast majority of the gabapentin data from the initial FDA analysis, arguably skewing the gabapentin-specific analysis. The FDA, however, did not ignore this limitation. As part of its sensitivity analysis, the FDA applied a risk difference analysis (an alternative method which does not require the exclusion of zero event studies), see discussion supra, Part II.C.1.a, to all of the data submitted for each of the eleven drugs. Here, the gabapentin-specific data yielded a 0.28 risk difference, indicating a positive association. Gibbons emphasizes that it is a small association that was deemed not statistically significant. Gibbons characterized this finding as evidence indicating that “for gabapentin, [there was] no increased risk of suicidality observed.” (Gibbons July Rep. ¶ 5.) In sum, Dr. Gibbons concludes that, in light of what he deems serious methodological and analytical flaws, the FDA’s statement that “the results were generally consistent among all the different drug products” is not supported by the data. (Gibbons July Rep. ¶ 6.) Gibbons contends that the agency selected “really bad methods” for examining “whether or not the effects they were seeing applied to all of the drugs.” (Hr’g Tr. 401-02, July 23, 2008.) For Gibbons, the bottom line is that, in his view, neither the FDA study nor any evidence put forth by Plaintiffs demonstrates that gabapentin itself is associated with suicidality or supports a conclusion of a causal link between gabapentin and suicidality. (Gibbons May Rep. 16.) b. Does the FDA Study Withstand Dr. Gibbons’Critique? Dr. Gibbons has presented a powerful critique of the FDA’s statistical analysis and its conclusion that the increased risk of suicidality detected in its meta-analysis is “consistent among the group of 11 drugs.” (FDA Statistical Review 6.) However, on balance, Dr. Gibbons’ criticism of the FDA’s statistical methods and conclusions — particularly its conclusion of consistency in effect among the eleven drugs— affects the weight that should be given to the study, not its admissibility. First, the underlying data for the FDA’s meta-analysis were placebo-controlled, clinical studies, the “gold standard” for epidemiological evidence. And, as the Reference Manual on Scientific Evidence explains, pooling such evidence is the “most appropriate” application of the meta-analysis method. Reference Guide on Epidemiology, supra, at 380. Moreover, the FDA has continued to stand by its statistical methods and conclusions even after considering the very concerns raised by Dr. Gibbons. As detailed in the Statistical Review, the FDA conducted several sensitivity analyses “to examine the robustness” of the results generated by the study’s primary analysis. (FDA Statistical Review 43.) In fact, two of the sensitivity analyses addressed the key concerns raised by Gibbons. One sensitivity analysis specifically examined the question of trial heterogeneity, the primary critique raised by Dr. Gibbons. The FDA applied two separate tests to evaluate whether all eleven drugs were properly grouped together in the primary analysis or whether the drugs’ “treatment effects” were too heterogenous — or different from each other — to justify such pooling. While the first test was inconclusive, the second model produced a result that convinced the FDA that “trial heterogeneity was not a major concern” and that, therefore, it was appropriate to group all eleven drugs together in the meta-analysis. (FDA Statistical Review 27.) At the Advisory Committee meeting three statisticians expressed general approval of the FDA’s technical approach to the study’s analyses. Then, the entire group was asked whether they agreed with the Agency’s overall finding of an increase in suicidality for the eleven AEDs analyzed. The Committee voted yes, with a tally of twenty to zero (with one abstention). The second question was whether “the Committee agree[s] with the Agency’s conclusion that [the] finding of increased suicidality should apply to all, or each, of the drugs included in the analysis, despite the observation that the odds ratio for two of the drugs was below one.” (FDA Hr’g Tr. 94.) Again, the Committee signaled its agreement with the FDA, with eighteen members voting “yes” and three voting “no.” Perhaps most persuasive is that these votes each came after the Committee members had heard Dr. Christopher Wohlberg, on behalf of Pfizer, argue that it was improper to pool the eleven drugs together and that the effects were not consistent across all eleven drugs. (Id. at 37.) In fact, Dr. Wohlberg articulated the very point Dr. Gibbons emphasized in his testimony: that topiramate and lamotrigine were driving the results of the study. (Id. at 38.) Yet, even after this testimony, this blue ribbon Committee voted to affirm the FDA’s methods and conclusion of consistency among the eleven drugs. In addition, Plaintiffs’ expert Dr. Sander Greenland, a statistician and epidemiologist, rebuts many of Dr. Gibbons’ arguments. Greenland concludes that “[t]he FDA Analysis was in all respects conducted properly and followed the current best standards of practice.” (Decl. of Sander Greenland, July 22, 2008, at 2.) Dr. Greenland also declared that he “concur[s] with the ... vote of the ... A