Full opinion text
MEMORANDUM AND ORDER VRATIL, District Judge. Mark and Cheryl Miller claim that their 13 year old son, Matthew, committed suicide because he took Zoloft, a prescription drug which Pfizer Inc. manufactured for the treatment of depression. See Amended Complaint (Doc. # 92) filed December 16, 1999. In that regard, plaintiffs seek to hold Pfizer liable for failure to warn and marketing misrepresentations under common law negligence theory and the Kansas Products Liability Act, K.S.A. § 60-3304. See Pretrial Order (Doc. # 171) filed March 6, 2000. This matter comes before the Court on various motions regarding plaintiffs’ primary expert witness, Dr. David Healy: Defendant Pfizer Inc’s Refiled Motions To Exclude The Testimony Of Dr. David Healy And For Sanctions Against Dr. David Healy For Violations Of The Court’s Protective Order (Doc. # 534) filed November 1, 2001; Plaintiffs’ Post-Hearing Memorandum, And Request For Independent Expert Verification (Doc. #545) filed November 27, 2001; Defendant Pfizer Inc.’s Opposition To Plaintiffs’ Post-Hearing Memorandum And Request For Independent Expert Verification; Request For Sanctions (Doc. # 547) filed November 29, 2001; and Defendant Pfizer Inc’s Motion In Limine No. 9 To Exclude The Use At Trial Of The Hindmarch Study, Dr. Healy’s “Healthy Volunteer Study,” And Dr. Healy’s Statistical Analysis Of The Pfizer Meta-Analysis (Doc. # 304B) filed April 28, 2000. After carefully considering the parties’ arguments and briefs in support, the Court is ready to rule. Background Sertraline hydrochloride is a selective serotonin reuptake inhibitor (“SSRI”) drug which is generically known as sertraline. Pfizer markets it as Zoloft. Dr. David Healy, a Welsh psychiatrist and academic neuropsychopharmacologist, proposes to testify that Zoloft can cause depressed patients to commit suicide and that it did so in the case of Matthew Miller. In his declaration of May 9, 2000, Dr. Healy states that his opinion is based upon “research that went into ... aforementioned publications, review of depositions and testimony in the Fentress v. Eli Lilly case, depositions and trial testimony in the For-syth v. Eli Lilly case, exhibits in both cases, depositions in the Miller v. Pfizer case, thousands of pages of Pfizer documents, the “healthy volunteer” study I conducted, as well as published and unpublished clinical trial material of which I was in possession.” In an article titled Zoloft And Suicide: Causal Mechanisms, which is attached to his declaration of August 13, 1999, Dr. Healy cites 54 depositions, exhibits and articles on which he has also relied. Because Dr. Healy has explained the basis for his opinions in this manner, he has opened an almost limitless arena of material for the Court to review in evaluating the admissibility of his testimony. Of course this is precisely the situation which the mandatory disclosure requirements of Fed.R.Civ.P. 26(a)(2) seek to avoid. See Burton v. R.J. Reynolds Tobacco Co., 203 F.R.D. 636, 639 (D.Kan.2001). Nonetheless the Court has carefully reviewed every article in the record which Dr. Healy has cited in his various declarations. Aside from Dr. Healy’s own articles, none of them directly support his claim that Zoloft causes akathisia, which in turn causes suicide in depressed patients. For the purposes of the pending motions, the Court focuses on the articles and research that Dr. Healy has authored or on which he appears to place particular emphasis: his own “Healthy Volunteer Study,” his statistical analysis of pre-existing Pfizer data (the so-called Pfizer “Meta-Analysis”), his application of Koch’s Postulates, and his evaluation of three studies — one by Dr. Ian Hindmarch (the so-called “Hind-march Study”) and two challenge-dechallenge-rechallenge studies by other authorities. Dr. Healy is an accomplished researcher in the area of neuropsychopharmacology, and his credentials are not in dispute. Indeed, he is an eminent psychiatrist and neuropsychopharmacologist who has made important contributions to the history of psychiatry and many significant clinical contributions. Notwithstanding his credentials, Pfizer asks the Court to exclude Dr. Healy’s opinions, along with certain studies and calculations on which he relies — the Hindmarch Study, the Healthy Volunteer Study, and the Pfizer Meta-Analysis. Pfizer does not seek to exclude the two challenge-dechallenge-rechallenge studies, although it does discount them as “case reports” and “anecdotal evidence.” To assist the Court in evaluating Dr. Healy’s opinions under Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993), the Court appointed two independent experts: John Concato, M.D., M.S., M.P.H. and John M. Davis, M.D. Dr. Concato, a professor at the Yale University School of Medicine, is a general internist and clinical epidemiologist who is a preeminent authority in his field. Dr. Davis is a professor in the department of psychiatry at the University of Illinois at Chicago; he has practiced in the field of neuropsychopharmacology since 1965, spending seven years with the National Institute of Mental Health. He was one of the first scientists to study affective disorders and the theory that reserpine can lead to suicide. Dr. Davis was a member of the FDA’s Psychopharmaco-logical Drug Advisory Committee (“PDAC”) which examined the initial drug application for Zoloft, and concluded that it was safe and effective for the treatment of depression. On September 5, 2001, after reviewing materials which the parties jointly submitted, the experts submitted their Report Of Independent Experts (Doc. # 502) (“Independent Expert Report”). On November 19 and 20, 2001, the Court held a Daubert hearing at which the parties had an opportunity to question the experts and present other relevant evidence. Legal Standard Rule 702, Fed.R.Evid., provides that a witness who is qualified by knowledge, skill, experience, training or education may testify in the form of opinion or otherwise as to scientific, technical or other specialized knowledge if such testimony will assist the trier of fact to understand the evidence or to determine a fact in issue, “if (1) the testimony is based upon sufficient facts or data, (2) the testimony is the product of reliable principles and methods, and (8) the witness has applied the prinei-ples and methods reliably to the facts of the case.” Analysis I. Defendant’s Motions To Exclude Dr. Healy’s Testimony (Doc. # 534) And Use At Trial Of The Hindmarch Study, The Healthy Volunteer Study And The Pfizer Meta-Analysis (Doc. # 304b) Dr. Healy proposes to testify about (1) general causation (that Zoloft causes suicide); (2) specific causation (that Zoloft caused Matthew Miller to commit suicide); (3) suicidology and the incidence of teen suicide; (4) the adequacy of Pfizer’s warnings about Zoloft; and (5) the regulation of pharmaceuticals by the FDA. A. General Causation Dr. Healy’s view as to general causation is a distinctly minority view. The American College of Neuropsychopharmacology, the FDA’s PDAC, and the Medicines Control Agency in his own United Kingdom have all reached contrary conclusions. For this reason and others, Pfizer argues that Dr. Healy’s opinions and methodologies are unreliable and inadmissible under Daubert and Federal Rules of Evidence 106, 403, 702 and 703. In concluding that Zoloft causes suicide in some patients, Dr. Healy relies heavily upon case reports and his own studies and calculations, and disavows the need for randomized controlled trials (“RCTs”) and epidemiological studies. His opinion rests most heavily on his own Healthy Volunteer Study, his Meta-Analysis of Pfizer data (which allegedly yields a relative risk calculation of 2.19), his application of Koch’s Postulates, and his evaluation of three studies — the Hindmarch Study and two challenge-dechallenge-rechallenge studies. Because these studies are critical to an understanding of his testimony, they are briefly summarized as follows: Healthy Volunteer Study: In 1999, Dr. Healy conducted a double-blind study of 21 healthy volunteer' health care workers from the North Wales Department of Psychological Medicine in Bangor, Wales, of which Dr. Healy is the director. Dr. Healy and his research assistant recruited the subjects from administrative, medical and nursing staff who ranged in age from 28 to 52 years. According to Dr. Healy, none were on concurrent drug treatment, all were free of medical conditions and none had a history of psychiatric illness. Under the study, each volunteer took either sertraline (Zoloft) or reboxetine for 15 days. After this period, the volunteers were drug-free for two weeks. For the next 15 days, each volunteer switched and used the other drug. None of the subjects had suicidal ideation while taking re-boxetine. During the second regimen, two volunteers who were taking Zoloft developed “clear suicidal ideation” (one of which reached “extremely serious proportions”), along with elements of akathisia and emotional blunting. Akathisia is “[a] syndrome characterized by an inability to remain in a sitting position, with motor restlessness and a feeling of muscular quivering.” Other subjects had these features but did not become suicidal. The study did not use a placebo control. Because of the nature of the study, it involved extensive interaction between research personnel and participants during the study and afterwards in focus groups; furthermore, the volunteers were “more or less aware” of Dr. Healy’s views and the fact that he expected them to react more poorly on Zoloft than on reboxetine. Meta-Analysis of Pfizer Data: On April 15, 2000 — after Dr. Healy’s deposition and four iterations of expert reports in this case — plaintiffs filed their second supplemental responses to Pfizer’s first requests for admissions. At that time they first disclosed that Dr. Healy had completed a “meta-analysis” of Pfizer data and that his review indicated a “relative risk of suicidal acts on Zoloft, compared to placebo and corrected for exposure times of 2.19 (i.e. 2.31 vs. 1.0555).” According to plaintiffs, Dr. Healy’s analysis was as follows: All Zoloft treated MDD cases-23 suicidal acts in 6,549 subjects = 2.31% per year; All TCA treated-2 suicidal acts in 922 subjects = 1.37% per year; All placebo treated-3 suicidal acts in 1,042 subjects = 1.055% per year. The Court cannot locate in the record a more precise explanation of Dr. Healy’s calculation. It may derive from premark-eting clinical trials for Zoloft which set forth data and incidence rates for suicides and suicide attempts in Zoloft-treated, placebo-treated and active-control-treated depressed patients. It may derive from reports which Pfizer provided the FDA in connection with its application to market Zoloft for the treatment of pediatric patients with obsessive compulsive disorder. Neither the Pfizer experts nor the independent experts have been able to replicate the calculation. Koch’s Postulates: Koch’s Postulates are a series of factors which are designed to assist in determining general causation and whether a drug may cause a certain reaction. The factors are strength of association, consistency with other research, alternative explanations, temporal relationship, biological plausibility, specificity of association and dose-response relationship. Reference Manual On Scientific Evidence (“Reference Manual”) (1st ed.1994). In his analysis of Koch’s Postulates, Dr. Healy did not attempt to directly assess whether Zoloft causes suicidal ideation or behavior. Instead he focused on whether SSRI drugs have a favorable therapeutic effect on depressive disorders (Dr. Healy concluded that the strength of this association is “in fact ... rather weak”) and the link between SSRI drugs and akathisia (Dr. Healy concluded that this association was much stronger, if not conclusive). He therefore concluded that “there is greater specificity in the relationship between SSRI and akathisia than there is between the SSRI and the response of a depressive disorder.” With regard to the association between SSRI drugs and akathisia, based on his application of Koch’s Postulates, Dr. Healy concluded that (1) as to strength of association, akathisia is a clear-cut event which is strongly associated with some insult to the central nervous system such as an infective process or drug intake and, in the case of Zoloft, akathisia may occur in at least a mild form in over 50 per cent of the patients who are taking the drug; (2) as to temporal relationship, akathisia has a fairly predictable temporal onset in relation to the intake of Zoloft, i.e. 24 to 48 hours following the institution of treatment or an escalation of dose (a phenomenon which builds to a peak over one to two weeks as the drug dosage rises in the body); (3) as to biological plausibility, there appears to be a “clear biological mechanism” by which SSRI drugs produce akathisia; (4) as to specificity of association, not all antidepressants appear capable of producing akathisia and “there is ... greater specificity in the relationship between an SSRI and akathisia than there is between the SSRI and the response of a depressive disorder;” and (5) as to dose-response relationship, there appears to be a clear dose-response relationship between SSRI intake and akathisic phenomena. Dr. Healy did not address two postulates: consistency with other research and alternative explanations. On the other hand, he added four factors: salience of association, consistency of association, support by experiment and support by analogy. As to salience of association (which he explained as “the difficulty in otherwise explaining the effect”), Dr. Healy stated that the association between SSRI drugs and akathisia may be “very salient,” in that the patient may be visibly distressed and clearly deteriorated “with no ... means of explaining what is happening other than by invoking an action of a drug.” As to consistency of association, Dr. Healy stated that akathisia has a physiological input which appears to occur across all personality types, and aka-thisia in response to a particular agent strongly predicts renewed akathisia on further exposure to the triggering agent. As to support by experiment, Dr. Healy stated that in his own tests, healthy volunteers who took droperidol all became akathisic and up to five became suicidal, with two becoming acutely suicidal within hours of treatment. Although Dr. Healy does not explain what droperidol is, or how it relates to this litigation, he posits that this experiment constitutes “strong evidence” in favor of the proposed mechanism of action whereby an SSRI may produce sui-cidality. Finally, as to support by analogy, Dr. Healy states that according to case reports, patients committed suicide on Prozac (another SSRI drug) and reserpine (a drug which is active on the serotonin system and according to Dr. Healy “probably” causes akathisia by similar mechanisms to Zoloft) even though they were “not depressed at all” and had “no nervous condition on which the suicide could be blamed.” According to Dr. Healy, these case studies provide “strong evidence” of a causal relationship between SSRI intake and suicidality. Hindmarch Study: In 1988, Dr. Ian Hindmarch conducted a double-blind placebo-controlled crossover study to assess the effects of sertraline (Zoloft), alone and with diazepam (Valium), on the psychomotor performance of 12 healthy volunteers at the University of Leeds, U.K. On April 14, 2000, after Dr. Healy’s deposition and various expert reports, plaintiffs served supplemental discovery responses. In plaintiffs’ second supplemental responses to Pfizer’s first requests for admissions, they stated that according to Dr. Healy, the Hindmarch study demonstrated a relative risk greater than two for “emotional blunting” and probable aggressive, assaul-tive or suicidal behavior in persons taking Zoloft. These claims are not explained in the record, which reveals that (1) the Hindmarch study dose began at 150 mg (three times more than the recommended 50 mg starting dose for Zoloft and three times more than the dose allegedly taken by Matthew Miller); (2) none of the volunteers had any diagnosed psychiatric condition; and (3) it appears that none of the study volunteers reported assaultive behavior, depression, akathisia, or homicidal or suicidal behavior, ideation or attempts. Challenge-Dechallenge-Rechallenge Articles: In concluding that Zoloft causes suicide, Dr. Healy relied upon two challenge-dechallenge-rechallenge studies. The first was a 1991 report by Anthony J. Rothschild, M.D. and Carol A. Locke, M.D., which discussed three patients (all of whom had histories of suicidal ideation) who made serious suicide attempts during treatment with fluoxetine (Prozac). During re-treatment with fluoxetine, all three developed severe akathisia — stating that the akathisia made them feel suicidal and, in the context of their depressive episodes, had precipitated their prior suicide attempts. The akathisia and suicidal ideation abated after fluoxetine was discontinued or propranolol (an anti-akathisia medication) was added. The authors noted that they had observed the emergence of suicidal ideation secondary to akathisia in only three of 1,500 patients (0.2%) — perhaps in part because they had lowered the fluoxetine dose when side effects developed and promptly recognized and treated the akathisia. They noted the hypothesis that suicidal ideation and suicide occur secondary to the emotional distress of akathisia, but also cautioned that “[i]t remains unclear whether there exists a common pharmacologic basis for akathisia and suicidal ideation or acts.” Because it is “conceivable that the development of aka-thisia in patients with a past history of suicidal or homicidal ideation is particularly problematic,” the authors added that “clinicians should remain alert to the development of akathisia in patients taking fluoxetine and to the fact that patients are often unable to distinguish akathisia from the ongoing symptoms of their psychiatric illness,” i.e. patients need to be reassured that the symptoms are side effects of medication and are treatable. The second challenge-dechallenge-rechallenge study involved six case reports in an article by Dr. Robert King, et al. In these reports, self-injurious ideation or behavior appeared or intensified during fluoxetine treatment of six patients, age 10 to 17 years, who were among 42 young patients receiving fluoxetine for obsessive compulsive disorder (“OCD”). Before they received treatment, four of the six patients had major risk factors for self-destructive behavior, including depression or prior suicidal ideation or self injury. The article examined three possible explanations for the emergence of suicidal ideation and self-destructive behavior in the six subjects: (1) that it was coincidental, in that while all six subjects had received fluoxetine, they all had complex longstanding psychiatric difficulties; (2) that it was a consequence of the treatment or was related to medication-induced agitation, disorganization or mood changes; and (3) that it was due to a specific effect of fluoxetine on the regulation of aggression directed either outward or toward the self. The authors concluded that while the adverse side effects appeared to be more than coincidental as to four of the six patients, “[distinguishing among these three alternatives [was] not possible on the basis of the uncontrolled clinical material presented here.” Accordingly, they suggested a need for more research. Pfizer first argues that Dr. Healy’s testimony on general causation will not satisfy the pertinent indicia of scientific reliability or assist the trier of fact in this case. Under Dau-bert, a district court must determine at the outset, pursuant to Rule 104(a), whether expert testimony is reliable and will assist the trier of fact to understand or determine a fact in issue. See Daubert, 509 U.S. at 592, 113 S.Ct. 2786. In determining whether a particular scientific theory or technique is rehable, the district court may consider several nondispositive factors: (1) whether the proffered technique can and has been tested; (2) whether the technique or theory has been subject to peer review; (3) the known or potential rate of error; and (4) the general acceptance of a technique in the relevant scientific community. Id. at 593-94, 113 S.Ct. 2786. The proponent of expert testimony must show “a grounding in the methods and procedures of science which must be based on actual knowledge and not subjective belief or unaccepted speculation.” Mitchell v. Gencorp. Inc., 166 F.3d 778, 780 (10th Cir.1999). An expert’s qualifications are relevant to the reliability inquiry. See United States v. Taylor, 154 F.3d 675, 683 (7th Cir.1998); In re Indep. Serv. Org. Antitrust Litig., 85 F.Supp.2d 1130, 1163 (D.Kan.2000). 1. Methodology The first Daubert factor asks whether a theory is susceptible to testing and whether the theory has been subjected to such testing. See Mitchell, 165 F.3d at 780 (citing Daubert, 509 U.S. at 593, 113 S.Ct. 2786). Pfizer appears to concede that Dr. Healy’s theory can be tested, but argues that it has not been tested because Dr. Healy has not conducted and does not rely on any RCT or large-scale epidemiological study which demonstrates a causal relationship between Zoloft and suicide in depressed patients. Absent such studies, Pfizer faults Dr. Healy for relying on tests which involve SSRI drugs other than Zoloft and drugs which are not even SSRI drugs like Zoloft, or which involve other patient populations or other adverse side effects or events. Plaintiffs admit that Dr. Healy does not rely on the type of RCTs or large-scale epidemiological studies which Pfizer insists are requisite to an expert opinion as to causation, but they argue that his hypothesis has been otherwise tested in two challenge-dechallenge-rechallenge studies, his own Healthy Volunteer Study and Meta-Analysis of Pfizer data, as well as by Koch’s Postulates. Plaintiffs are correct that RCTs and large-scale epidemiological studies are not essential to the formation of an expert opinion as to causation. See Bonner v. ISP Tech., Inc., 259 F.3d 924, 929 (8th Cir.2001) (expert not required to rely on epidemiological studies; testimony allowed when district court found expert rebanee on temporal connection, animal studies, studies of chemicals with similar structures, study of drug at issue and the manner of acting on nerve pathways, and plaintiffs medical records showed rebable methodology under Daubert); Brasher v. Sandoz Pharm. Corp., 160 F.Supp.2d 1291, 1296-98 (N.D.Ala.2001) (expert testimony abowed despite lack of large-scale epidemiological study when experts relied on animal studies, case reports, Adverse Drug Reaction reports to FDA, generally accepted notion in the medical community and ruled out ab other possible causes; large-scale study nearly impossible because adverse reaction was rare event and study would unethically expose patients to risk of harm); Ruff v. Ensign-Bickford Indus., Inc., 168 F.Supp.2d 1271, (D.Utah 2001) (lack of epidemiological data no bar to expert testimony that relied on numerous animals, discredited the one study relied upon by defendants and ruled out alternative causes). At least as to Koch’s Postulates, the independent experts do not disagree that Dr. Healy has identified a method of testing the potential cause-effect relationship of sertaline and suicide. For purposes of this order the Court assumes that challenge-dechallenge-rechallenge studies, double-blind studies like Dr. Healy’s Healthy Volunteer Study, and meta-analysis of data from randomized controlled trials or large-scale epidemiological studies are all legitimate tests which may be applied to determine whether Zoloft induces suicidal ideation or behavior. The Court therefore finds that Dr. Healy’s hypothesis is amenable to testing and that he has at least identified tests which have some bearing on the issue of general causation. While Pfizer argues that his methodology is fatally flawed, that argument is addressed by Daubert’s fourth factor, which the Court discusses below. 2. Peer Review And Publication The second Daubert factor asks whether the expert’s theory or technique has been submitted to peer review and publication. Pfizer argues that Dr. Healy has not published any peer-reviewed article which states that Zoloft causes some people to commit suicide, and that his Meta-Analysis (including his relative risk calculation of 2.19) has not been subject to peer review. According to the independent experts, Dr. Healy is an accomplished investigator who has published several articles which are directly relevant to the issue of a possible association between sertraline and suicide, including Suicide In The Course Of The Treatment of Depression, 13 J. of Psycho-pharmacology 4-99 (1999), and Emergence Of Antidepressant Induced Suicidality, 6 Primary Care Psychiatry 23-28 (2000) (concerning Dr. Healy’s Healthy Volunteer Study). The experts characterize the first article as a “review” or “thought-piece” on possible mechanisms that would explain why patients commit suicide while taking antidepressants and the second article as “a randomized, controlled trial of two agents, given to 20 patients (total), that also included a component of qualitative analysis regarding changes in emotional state.” Although neither article proposed new or modified methods to determine causality and the editors and reviewers did not endorse Dr. Healy’s methodology, both pieces were peer-reviewed and deemed suitable for publication. Both articles theorize that SSRI drugs in general, and sertraline in particular, causes suicide in some patients. Pfizer is therefore incorrect in its argument that Dr. Healy has not published any peer-reviewed article which claims that Zoloft causes some people to commit suicide. On the other hand, Dr. Healy has not subjected to peer review his calculation, based on. his Meta-Analysis of Pfizer data, that Zoloft carries a relative risk of suicide of 2.19. Thus, while the theory behind Dr. Healy’s work has been subject to limited peer review and publication, his relative risk calculation has not. 3. Rate Of Error And Proper Controls Pfizer argues that Dr. Healy’s Healthy Volunteer Study cannot properly form a basis for his opinions on causation because (1) it was not a placebo-controlled test; (2) it did not result in a single suicide attempt; (3) its methodology has not been explained; (4) the sample size is too small to yield statistically significant analysis; (5) the cross-over design could lead to erroneous results; and (6) the volunteers were medical employees in the hospital department that Dr. Healy directed. To some degree, many of these problems result from the fact that the Healthy Volunteer Study was not designed to research suicide but to test the effects of sertraline and reboxetine on the well-being of healthy subjects. According to the independent experts, the lack of a placebo control in this context is not so much scientifically improper as it is problematic in evaluating the “big picture” question of a possible association between sertraline and suicide. In other words, “[w]hen results from the study are extrapolated to the issue of sertraline possibly causing suicide, ... a question can be raised what potential information is unavailable due to the absence of a placebo group ... [f]or example, ... would patients receiving placebo have reported any fluctuations in their emotional state when asked to keep a diary, etc.” The experts noted that although this scenario might seem unlikely, the absence of placebo controls “may affect the interpretation of the study findings.” The independent experts also criticized the design of Dr. Healy’s Healthy Volunteer Study in two important respects. First, its methodology involved intensive interaction between participants and research personnel, as well as a focus group at the end of the study — strategies which (along with a corresponding emphasis on patient narratives) represented qualitative rather than quantitative research. Second, because of the small sample size and the qualitative nature of the research, statistical analysis — while it theoretically could be done — would not produce a statistically significant result or provide a sufficient basis for calculating a statistically significant relative risk. During the Daubert hearing, Dr. Healy admitted that his Healthy Volunteer Study was not designed to research treatment-emergent suicidality and that any conclusions regarding that phenomena were a by-product of the study results — not the design. He also acknowledged that his study was not well designed for relative risk calculations. Dr. Healy also agreed with the independent experts that subliminal communication between the researchers and the volunteers may have influenced the results. In particular, before the study Dr. Healy gave the subjects detailed information regarding the side effects of each drug. Subjects who had certain side effects might therefore deduce which drug they had received. Dr. Healy also admitted that three subjects were on medication despite the fact that the research protocol was that the subjects would not be on medication, other than oral contraceptives. In addition, even though the subjects allegedly had no histories of psychiatric illness, one of them had a history of depression. At the hearing, Pfizer took aim on the two subjects who allegedly had adverse effects on sertraline. Initial mental health examinations were not done on either volunteer. Dr. Healy claimed that mental health screenings were supposed to be completed later, but did not address the fact that the researchers did not have a mental health base line for either subject. Furthermore, Dr. Healy could not rule out a carry-over effect from the reboxetine which the volunteers had taken before they started on Zoloft. Dr. Healy opined that any reboxetine effect was probably small since reboxetine on its own did not cause agitation in any volunteer, but he did not address what effect reboxetine with sertraline may have had. Of the two volunteers who developed adverse reactions on sertraline, one lost her grandmother during the reboxetine period and during that time was unable to sleep or eat and felt annoyed, unhappy, miserable and angry. The other felt terrible (like she had a flu bug) during the reboxetine period and she could not eat or sleep. Her bad experience on sertraline was preceded by a migraine headache, which . Dr. Healy opines may have been drug-induced. Dr. Healy did not defend the design of his Healthy Volunteer Study in a manner which adequately addresses the criticisms of Pfizer or the independent experts. The Court must therefore conclude that his Healthy Volunteer Study does not include controls sufficient to prove that Zoloft causes akathisia-induced suicide in depressed patients. 4. General Acceptance Of Methodology Pfizer argues that the medical community does not agree that Zoloft causes suicide, and that Dr. Healy cannot testify otherwise. The “general acceptance” factor bears on the technique used, however, rather than the conclusion reached. As noted above, Dr. Healy has reached his conclusion by conducting his own so-called Healthy Volunteer Study, using the results of two challenge-dechaflenge-rechallenge articles, conducting a Meta-Analysis of Pfizer’s own clinical trial data and applying Koch’s Postulates to his hypothesis that Zoloft causes suicide. The Court examines whether Dr. Healy’s technique in each of these areas constitutes generally accepted methodology. a. Healthy Volunteer Study Pfizer argues that Dr. Healy’s Healthy Volunteer Study did not adequately test his theory that Zoloft causes suicide because he did not initially provide his methodology or use a placebo control. While Dr. Healy has now provided his methodology, the independent experts concluded that “[t]he healthy volunteer study is not a sufficient basis for calculating any statistically significant relative risk, due to issues involving study design and sample size.” They noted that because “the phenomenon of patients reporting symptoms while taking placebo is more the rule than the exception,” the study raised a question “what potential information is unavailable due to the absence of a placebo group”— e.g., would patients receiving placebo have reported any fluctuations in their emotional state when asked to keep a diary, etc. More importantly, Dr. Concato and Dr. Davis criticized Dr. Healy’s methodology because (1) it involved extensive interaction between research personnel and participants; and (2) in the circumstances, statistical calculations represented a statistical “exercise” rather than a statistically significant result. Furthermore, Dr. Healy and his research assistant personally recruited the subjects from the department of which Dr. Healy is the director and each volunteer was “more or less aware” of Dr. Healy’s expected outcome. From these criticisms, especially in light of the small sample size, the Court must readily conclude that Dr. Healy’s Healthy Volunteer Study does not represent a generally accepted methodology for testing the hypothesis that Zoloft causes suicide. b. Challenge-Dechallenge-Rechallenge Reports In concluding that Zoloft causes suicide, Dr. Healy relied upon two challenge-de-challenge-rechallenge studies which are basically case reports regarding nine individuals (of approximately 1,542) who developed suicidal ideation or behavior during treatment with fluoxetine (Prozac). According to the independent experts, heavy reliance on case reports is not an accepted methodology for determining the strength of association which Dr. Healy advocates between Zoloft and suicide. A major problem is that “suicides would be expected among patients taking sertra-line, and the occurrence of such events does not establish causality.” It is therefore generally accepted metholodogy to “control” or “adjust” for factors (such as propensity to commit suicide) that may influence the strength of the cause-effect association. In this context, the independent experts noted that [djetermining the strength of an association requires at least two groups of subjects, one exposed to the agent of interest (sertraline), the other not so exposed, so that rates of the outcome event (suicide) can be determined and compared.... [T]he lack of a comparison group would ... make a series of case reports unsuitable for determining the quantitative strength of an association, regardless of other attributes. Also, as Dr. Davis noted at the Daubert hearing, a rechallenge study can consist of merely anecdotal evidence with many possible explanations — including coincidence. Dr. Davis explained: [A]needotal studies are important because the brilliant clinician might discover something that everybody else missed and it might be proven ten years later.... So I’m not putting down anecdotal information, but there’s a lot of room for coincidence, particularly in patients who were recurrently suicidal. So could be suicidal one week, okay the next week, suicidal another week. And you could get things occurring by coincidence. So it’s hard to take that as definitive proof. Id. Dr. Healy’s citation of the Rothschild study is particularly problematic because after it was published, the authors issued a letter to the editor which refuted the analysis that Dr. Healy now attempts to draw from their work. They explained that their article did not stand for the proposition that fluoxetine causes suicide, but rather that the development of akathisia (as a side effect of fluoxetine) was one of several factors that led depressed patients to attempt suicide. Use of a small number of challenge-dechallenge studies to test a hypothesis is inadequate and not generally accepted methodology. See Hollander v. Sandoz Pharm. Corp., 95 F.Supp.2d 1280, 1284 n. 10 (W.D.Okla.2000) (three challenge-de-challenge reports insufficient to be of consequence). The Court finds that cases reports — even those gathered and reported as challenge-dechallenge-rechallenge studies — do not represent a generally accepted methodology for evaluating whether Zoloft causes suicidal behavior in otherwise depressed patients. c. Meta-Analysis On April 15, 2000, plaintiffs claimed that through calculations which involved Pfizer’s own data, Dr. Healy had confirmed the hypothesis that Zoloft causes suicide. As noted above, according to plaintiffs, his review indicated a “relative risk of suicidal acts on Zoloft, compared to placebo and corrected for exposure times of 2.19 (i.e. 2.31 vs. 1.0555).” Pfizer questions those calculations and has submitted an affidavit which suggests that they are impossible. Indeed, the independent experts have reported that they cannot replicate Dr. Healy’s relative risk calculation of 2.19. Armed with this information in advance of the Daubert hearing, Dr. Healy made some effort to explain and defend his calculation. When discussing the calculation with Dr. Concato, he explained that he had calculated the number based on patients and not patient years. Even with this information, Dr. Concato could not verify the validity of Dr. Healy’s calculation. Dr. Concato and Dr. Davis expressed concern about Dr. Healy’s calculation and a desire for Dr. Healy to produce more information to validate the number. As the Court explained, however, Dr. Healy’s opportunity to declare the bases for his opinion and produce evidence to support it has come and gone. Because the experts cannot replicate Dr. Healy’s calculation based on the information that he timely provided and Dr. Healy does not defend it based on that information, the Court has no alternative but to conclude that it does represent generally accepted methodology. d. Koch’s Postulates As noted above, Koch’s Postulates are a series of factors which are designed to assist in determining general causation and whether a drug may cause a certain reaction. Reference Manual (1st ed.1994). Because Dr. Healy’s conclusion as to general causation is so heavily dependent on his application of Koch’s Postulates, the Court asked its independent experts to evaluate this methodology in particular detail. After completing their review, Dr. Concato and Dr. Davis determined that while some elements of his approach were customary and accepted methods of applying Koch’s Postulates, other elements were inconsistent with Koch’s Postulates and their application. Overall, because Dr. Healy had focused on the association between SSRI drugs and akathisia — rather than the direct association between SSRI drugs and suicide — the independent experts concluded that Dr. Healy’s application of Koch’s Postulates was not generally accepted by the scientific community. As discussed below, they also cited specific deficiencies of Dr. Healy’s methodology in applying the postulates. Strength Of Association The first postulate concerns the strength of the association between the exposure and effect; it is measured by relative risk. The higher the relative risk, the greater the likelihood that the relationship is causal. See Reference Manual (2nd ed.2000) at 376. “[Relative risk is one of the cornerstones for causal inferences.” Id. at 376. As noted above, Dr. Healy concluded with respect to this factor that akathisia is a clear-cut event which is strongly associated with some insult to the central nervous system such as drug intake and, in the case of Zoloft, akathisia may occur in at least a mild form in over 50 per cent of the patients who are taking it. Pfizer challenges Dr. Healy’s application of this postulate on two grounds, claiming that Dr. Healy has no scientific evidence that Zoloft patients are at increased risk of akathisia, and that unless Zoloft creates a relative risk of greater than 2.0, Dr. Healy’s testimony is inadmissible. Preliminarily, the Court rejects Pfizer’s argument that unless Zoloft is shown to create a relative risk greater than 2.0, Dr. Healy’s testimony is inadmissible. According to Dr. Concato and Dr. Davis, the magnitude of relative risk is a continuum and a threshold of 2.0 is an arbitrary cutoff. They explained that even strong associations may turn out to be false and that the rigor of the study design and the quality of the data collection are more important than the precise magnitude of the risk. Furthermore, statistical significance — as expressed by confidence intervals, P values or both — is important in determining the stability of a relative risk calculation. The Court recognizes that other courts have adopted a 2.0 relative risk calculation as the benchmark for admissibility of expert testimony. See Allison v. McGhan Med. Corp., 184 F.3d 1300, 1315 n. 16 (11th Cir.1999) (relative risk greater than .2.0 allows inference that agent caused illness; risk of 1.24 insufficient); Pozefsky v. Baxter Healthcare Corp., No. 92-314LEKRWS, 2001 WL 967608 (N.D.N.Y. Aug. 16, 2001) (breast implant plaintiff must show relative risk greater than 2.0). On this record, especially in light of the testimony of its independent experts, this Court declines to do so. See Magistrini v. One Hour Martinizing Dry Cleaning, 180 F.Supp.2d 584, 606 (D.N.J.2002) (“[A] relative risk of 2.0 is not so much a password to a finding of causation as one piece of evidence, among others for the court to consider in determining whether an expert has employed a sound methodology in reaching his or her conclusion.”); Pick v. Am. Med. Sys., Inc., 958 F.Supp. 1151, 1160 (E.D.La.1997) (“A relative risk above 1.0 is statistically significant, even if not sufficient, by itself, to establish causation by a preponderance of the evidence.”). Pfizer also challenges Dr. Healy’s application of this postulate on the ground that he has no scientific evidence that Zoloft patients are at increased risk of akathisia. The Court need not address this issue in light of its finding — based on the testimony of its independent experts — that generally accepted methodology in this case required Dr. Healy to consistently test the strength of association between SSRI drugs and suicide (the outcome of interest) — rather than the association between SSRI drugs and akathisia (which is purported to be part of the chain of events that lead to suicide, rather than an independent outcome) or, in other cases, the association between SSRI drugs and improvement in symptoms of depression. Furthermore, as the independent experts explained, determining the strength of association “requires at least two groups of subjects, one exposed to the agent of interest (sertraline), the other not exposed, so that rates of the outcome event (suicide) can be determined and compared,” In concluding that akathisia may occur in at least a mild form in over 50 per cent of the patients who take Zoloft, Dr. Healy reasons as follows: (1) clinical trial data suggest that up to 25 per cent of patients who take Prozac may have “very marked akathisia;” (2) “as a matter of certainty,” the method of data collection in clinical trials “significantly under-estimates the extent of the problem;” (3) in the case of SSRI-induced sexual dysfunction, such data collection methods reported an incidence rate of five per cent when the “true incidence is of the order of 50%;” (4) general placebo controlled trials suggest that nervousness and anxiety occur in 10 to 30 per cent of the population taking SSRI drugs; (5) tests by Eli Lilly suggest that such data collection methods “may underestimate the burden of side-effects by a six-fold factor;” and (6) if this applies to “akathisia spectrum effects” in the case of Zoloft, akathisia “may occur in at least a mild form in over 50% of patients taking the drug. This reasoning says nothing about the extent to which nervousness, anxiety and akathisia appear in the population at large — let alone the population of untreated patients with depression or depressed patients who are treated with non-SSRI drugs.” Even if Dr. Healy is not legally required to show a relative risk of at least 2.0, he must rely on studies which yield results that are statistically significant regarding the causal relationship at issue in this case. See Magistrini, 180 F.Supp.2d at 605 (expert’s failure to adequately address relative risks weighed in court’s conclusion that methodology was unreliable); Siharath v. Sandoz Pharm. Corp., 131 F.Supp.2d 1347, 1358 (N.D.Ga.2001) (expert’s failure to derive statistically significant relative risk greater than 2.0 rendered testimony inadmissible). The independent experts were not able to derive statistically significant results from the materials which Dr. Healy provided. For these reasons, the Court must conclude that Dr. Healy did not apply the first of Koch’s Postulates in a manner which satisfies Daubert. Consistency With Other Research Koch’s Postulates ask whether the proposed association is consistent with other research. Reference Manual (2nd ed.2000) at 379. Dr. Healy does not discuss the fact that other research is contrary to his conclusion. According to the independent experts, his failure to do so constitutes a misapplication of Koch’s Postulates — which are “commonly applied in the context of other research, including in this case the consideration that depression itself may have led to suicide.” Dr. Healy’s misapplication of Koch’s Postulates is highly material, given the large body of research from RCTs with placebo controls which refute or fail to support his opinions. Dr. Healy has not directly responded to this fact, other than to attack the drug companies who funded and analyzed the data and disavow the need for RCTs and large scale epidemiological studies. This response in itself is apparently inconsistent with generally accepted methodology, which credits the role of RCTs and large-scale epidemiological studies as “very pertinent” in evaluating the relative risks of potential associations between sertraline and suicide. The Court therefore concludes that in failing to discuss the consistency of his hypothesis with other research, Dr. Healy has not used generally accepted methodology. Alternative Explanations Koch’s Postulates provide that to avoid reaching an erroneous conclusion, a scientist should rule out alternative explanations. See Reference Manual (2nd ed.2000) at 374. Dr. Healy omitted this factor from his application of Koch’s Postulates. Pfizer contends that Dr. Healy has committed a major error in failing to show that Zoloft (as opposed to other factors such as depression itself) causes suicide in depressed patients. The independent experts stated that “[wjhether or not the concept is identified as a separate postulate, the need to rule out alternative explanations is a fundamental tenet of scientific reasoning,” and that Dr. Healy had not given adequate consideration to the possibility of other explanations. Failing to consider alternative explanations is not a generally accepted methodology of applying this postulate. Temporal Relationship According to Koch’s Postulates, a temporal relationship is necessary to find a causal relationship. Reference Manual (2nd ed.2000) at 376. “If the exposure occurs after the disease develops, it cannot cause the disease.” Id. As noted above, Dr. Healy found that akathisia has a fairly predictable temporal onset in relation to the intake of Zoloft, i.e. 24 to 48 hours following the institution of treatment or an escalation of dose (a phenomenon which builds to a peak over one to two weeks as the drug dosage rises in the body). Pfizer attacks Dr. Healy’s application of this postulate because he claims that SSRIs do not begin treating depression for two to three weeks, but that they cause suicide after one or two weeks of therapy. Dr. Davis testified, however, that it is possible for sertraline to produce negative side effects before it begins to work in a beneficial way. Pfizer also contends that Dr. Healy also does not show a temporal relationship between Zoloft ingestion and Zoloft-induced suicide as opposed to depression-induced suicide. Pfizer concedes that Matthew Miller ingested Zoloft before he committed suicide, but insists that in order to satisfy this postulate, Dr. Healy must definitively show that Zoloft (as opposed to depression) caused the suicide. Pfizer’s position begs the question, however, since a definitive answer to this question would render the other Postulates superfluous. The possibility that something other than Zoloft caused Matthew’s suicide is addressed by the postulate concerning alternative explanations, and also by arguments about specific causation. Dr. Healy’s proffered testimony satisfies the generally accepted methodology for this postulate. Biological Plausibility Under Koch’s Postulates, the factor of biological plausibility can be difficult and “depends upon existing knowledge about the mechanisms by which the disease develops.” Reference Manual (2nd ed.2000) at 378. “The saliency of this factor varies depending on the extent of scientific knowledge about the cellular and subcellu-lar mechanisms through which the disease process works.” Id. As to this factor, Dr. Healy stated that there appears to be a “clear biological mechanism” by which SSRI drugs produce akathisia. He does not identify this “clear biological mechanism,” other than to state that it can be “blocked or ameliorated” through other drugs. Pfizer notes that in applying this factor, Dr. Healy connects Zoloft and aka-thisia rather than Zoloft and suicide, and that his expert reports reveal no discussion that directly connects akathisia (or Zoloft) to suicide. Dr. Concato and Dr. Davis noted that Dr. Healy’s method of determining biological plausibility — that Zoloft leads to aka-thisia and emotional blunting which in turn causes suicide — may be a suitable explanation if there is a relationship between ser-traline and suicide. Dr. Healy provided no explanation regarding the biological plausibility of his theory that Sertraline causes either akathisia or suicide. His application of this postulate does not constitute generally accepted methodology. Specificity Of Association Under Koch’s Postulates, specificity of association will be found if exposure to the agent is only associated with a single disease. See Reference Manual (2nd ed.2000) at 379. Lack of specificity of association does not necessarily undermine the case for causation when there is a biological explanation for its absence. Id. This factor can be helpful, however, in strengthening the case for causation. Id. In addressing this factor, Dr. Healy determined that not all antidepressants appear capable of producing akathisia and that “there is ... greater specificity in the relationship between an SSRI and akathi-sia than there is between the SSRI and the response of a depressive disorder.” In other words, Dr. Healy admits that not all antidepressants cause akathisia. From the record, it appears that even specific classes of antidepressants (such as SSRI drugs) do not all produce akathisia. In fact, SSRI drugs which may appear to produce akathisia in some patients at certain doses at certain times do not have that effect on other patients. In applying this postulate, Dr. Healy does not pretend to address any association between akathisia and suicide or Zoloft and suicide. The Court must therefore conclude that he has not used generally accepted methodology in addressing this postulate. Dose Response Relationship This postulate addresses that idea that the more intense the exposure to a particular agent, the greater the risk of disease. See Reference Manual (2nd ed.2000) at 377. Some agents do not exhibit a dose-response relationship when there is a threshold phenomenon (the exposure must exceed a certain dose to produce the disease). See id. Therefore this factor may be helpful, but it is not essential in proving causation. See id. As to dose-response relationship, Dr. Healy stated that there appears to be a clear dose-response relationship between SSRI intake and akathi-sic phenomena. Again, however, he does not address a dose-response relationship between SSRI drugs and suicide. In this regard he has not used a generally accepted methodology in his analysis of this postulate. Additional Factors In applying Koch’s Postulates, Dr. Healy added four factors which Mr. Koch himself did not include: salience of association, consistency of association, support by experiment and support by analogy. In its initial analysis of Dr. Healy’s work, the Court was unsure whether these factors have been or currently are part of Koch’s Postulates. Dr. Concato and Dr. Davis found that Dr. Healy’s additional postulates were unnecessary, however, and merely repeated information which was addressed in already-existing postulates. Thus, utilizing these postulates does not constitute generally accepted methodology. As noted above, as to salience of association (which he explained as “the difficulty in otherwise explaining the effect”), Dr. Healy stated that the association between SSRI drugs and akathisia may be “very salient,” in that the patient may be visibly distressed and clearly deteriorated “with no ... means of explaining what is happening other than by invoking an action of a drug.” The independent experts noted that in this context, “salience of the association” appears to invoke a criterion of “obvious” cause and effect (e.g., an anaesthetic and falling asleep). Koch’s Postulates were developed and applied in the context of chronic diseases, however, for which the association of cause and effect was not obvious. As such, this factor per se is not a part of Koch’s Postulates. Rather, in a more general sense, the existing postulates (such as temporal association) may be said to address the overall salience of the. association. As to support by experiment, Dr. Healy cited his Healthy Volunteer Study, stating that in his tests, healthy volunteers who took droperidol became akathisic and up to five became suicidal with two becoming acutely suicidal within hours of treatment. Dr. Healy posits that this experiment constitutes “strong evidence” in favor of the proposed mechanism of action whereby an SSRI drug may produce suicidality. The independent experts explained that “support by experiment” is not a category of Koch’s Postulates because experiments generate information for other (existing) postulates but do not constitute postulates unto themselves. Therefore Dr. Healy’s approach would tend to provide an indirect and possibly vague endorsement of a proposed cause and effect association. As to support by analogy, Dr. Healy states that according to case reports, patients on Reserpine (a drug active on the serotonin system which “probably causes akathisia by similar mechanisms to Zoloft”) and Prozac (another SSRI drug) committed suicide even though they were “not depressed at all” and had “no nervous condition on which the suicide could be blamed.” According to Dr. Healy, these case studies by analogy provide “strong evidence” of a causal relationship between SSRI intake and suicidality. According to the independent experts, “support by analogy” appears to be a variation on the postulates that address biological plausibility and coherence with existing information, and it does not appear to be necessary. Finally, Dr. Healy has added a factor called “consistency of association.” In applying it, he states that akathisia has a physiological input which appears to occur across all personality types, and akathisia in response to a particular agent strongly predicts renewed akathisia on further exposure to the triggering agent. Given his failure to address the direct linkage between Zoloft and suicide and the overwhelming deficiencies in his application of Koch’s Postulates, the addition of this extraneous factor is perhaps best characterized as harmless error. In any event, the Court concludes that it is subsumed in established factors and that it does not constitute any part of a generally accepted methodology for applying Koch’s Postulates. In summary, Dr. Healy has misapplied, failed to satisfy, or failed to address six of the seven postulates (strength of association, consistency with other research, alternative explanations, biological plausibility, specificity of association and dose-response relationship). To compound this problem, he has invented other factors and variously applied or misapplied them in ways which make it impossible to discern what his conclusions would be if they rested only on generally accepted methodology. Dr. Healy adequately addressed only one of the postulates (temporal relationship). While “[t]here is no formula or algorithm that can be used to assess whether a causal inference is appropriate based on the guidelines,” see Reference Manual (2nd ed.2000) at 375, the Court concludes by overwhelming evidence that Dr. Healy’s application of the postulates is not generally accepted methodology. In their report the independent experts issued the follow conclusion: The theory that sertraline increases the risk of suicide rests largely on case reports of suicide among patients taking sertraline, a posited connection between akathisia (as a side effect of sertraline use) and suicide, and the report of suicidal ideation in two healthy volunteers given sertraline. These “lines of evidence” only partially satisfy Koch’s postulates. In reviewing the Daubert factors, the Court finds that the flaws in Dr. Healy’s methodology — particularly his Healthy Volunteer Study, his Meta-Analysis of Pfizer data and his application of Koch’s Postulates- — are glaring, overwhelming and unexplained, and that his testimony on general causation (which is premised on those methodologies) must therefore be excluded from evidence. Consideration of such evidence would only serve to confuse and mislead the jury and waste the time and resources of the parties and the Court. Plaintiffs concede that Dr. Healy has not cited the Hindmareh Study as a basis for his expert opinion and that they will not cite that study as part of their case in chief on the issue of general causation. That leaves Dr. Healy with a handful of case reports to substantiate his distinctly minority position that to a reasonable degree of medical certainty, Zoloft causes otherwise depressed patients to commit suicide. The Court has already determined that heavy reliance on case reports is not an accepted methodology for determining the strength of association between Zoloft and suicide. The Court therefore has no alternative but to sustain Pfizer’s motion to prohibit Dr. Healy’s testimony on the subject of general causation and to prohibit plaintiffs from using at trial the Hind-march Study, the Healthy Volunteer Study and the Pfizer Meta-Analysis. B. Specific Causation Dr. Healy proposes to testify as to specific causation: that in his opinion, based on a review of notes by Matthew, reports by Matthew’s father, and the deposition of Matthew’s psychiatrist, Zoloft more likely than not caused Matthew to commit suicide and no other factor in his background or psychosocial situation could explain his suicide. Pfizer argues that Dr. Healy’s testimony is not adequate to show specific causation and that at the time of his deposition on March 27, 2000, Dr. Healy had not even bothered to review much of the factual evidence regarding Matthew’s symptoms, behavior and social and family circumstances. Indeed, Dr. Healy’s contention ignores undisputed evidence that Matthew Miller engaged in suicidal thoughts and behavior before he first took Zoloft. Plaintiffs took Matthew to see a psychiatrist because he had been struggling with mood, anger and school-related issues. In a diagnostic consultation, Matthew told the psychiatrist th