Lynch v. Merrell-National Laboratories, 646 F. Supp. 856 (D. Mass. 1986)
October 17, 1986
MERRELL-NATIONAL LABORATORIES DIVISION OF RICHARDSON-MERRELL, INC., Defendant.
United States District Court, D. Massachusetts.
*857 Philip J. Crowe, Jr., Lubin & Meyer, Boston, Mass., for plaintiffs.
Charles Choate, and Larry Kenna, Choate Hall & Stewart, Boston, Mass., for defendant, Frank C. Woodside, III, and Lynda E. Roesch, Dinsmore & Shohl, Cincinnati, Ohio, of counsel.
MEMORANDUM AND ORDER
MAZZONE, District Judge.
This matter is before the Court on the defendant's motion for summary judgment. The case involves a claim for damages brought by the parents on behalf of a minor plaintiff, Margo Lynch, and individually as well, against the defendant Merrell-National Laboratories ("Merrell Dow"). The plaintiffs allege that Margo Lynch sustained injury in the form of the congenital absence of her right forearm as a result of the ingestion of the defendant's prescription pharmaceutical, Bendectin, by plaintiff's mother, Margaret Lynch, during her pregnancy. In support of its motion for summary judgment, the defendant has filed an extensive record, consisting of excerpts of deposition and trial testimony of nineteen expert witnesses, numerous Bendectin epidemiological studies, and an appendix containing a list of other Bendectin cases, an index of the Bendectin multi-district litigation including juror questionnaires, instructions, special questions and an exhaustive and thoughtful ruling by the trial judge in that litigation denying the plaintiffs' motions for judgment notwithstanding the verdict and for new trial. The plaintiffs have opposed the motion and have supported their opposition with a memorandum.
The two issues are: (1) whether the plaintiffs are collaterally estopped from relitigating the issue of Bendectin's role in the causation of birth defects; and (2) whether there is any factual dispute on the issue of causation. As to the first issue, the defendants claim the plaintiffs should be bound by the result of the multidistrict trial in which the jury concluded that Bendectin did not cause human birth defects. The plaintiffs say that because they did not participate in that trial, they are not bound by that result, and further, that the result is not conclusive because other cases have produced a different result favorable to them. As to the second issue, the defendant says that even if the plaintiffs were not bound by the results of the earlier trial, the plaintiffs' proof on the issue of causation is insufficient as a matter of law to establish causation. The plaintiffs say that their evidence, consisting of a re-analysis of epidemiological studies, in vivo and in vitro animal studies, and studies of analogous chemical structures create an issue of fact for the jury. A brief review of the history of this litigation follows.
Bendectin, a product manufactured by the defendant for use in the treatment of nausea and vomiting during pregnancy, has been the subject of numerous cases in which it has been alleged that Bendectin was the cause of human birth defects. In 1982, in order to deal with the volume of Bendectin cases, the Judicial Panel on Multidistrict Litigation assigned Chief Judge Carl B. Rubin of the United States District Court for the Southern District of Ohio, to handle pretrial proceedings in the Bendectin multidistrict litigation. In re Richardson-Merrell, Inc. "Bendectin" Prod. Liab. Litig. (II), MDL No. 846, 533 F. Supp. 489 (Jud.Pan.Mult.Lit.1982). The panel found that the cases involved common questions of fact and that centralization of these actions pursuant to Section 1407 would "best serve and promote the just and efficient conduct of the litigation." The panel found that common factual questions arose from the allegations in each action that birth defects were caused by the mother's ingestion of Bendectin and that centralization was "necessary in order to prevent duplication of discovery, avoid inconsistent pretrial rulings, and conserve the resources of the parties, their counsel and the judiciary." Transfer Order of the Judicial Panel on Multidistrict Litigation No. 486, February 9, 1982.
Upon completion of multidistrict discovery, Judge Rubin consolidated for trial *858 all cases originally filed in the Northern and Southern Districts of Ohio, and adopted an "opt-in" procedure, allowing plaintiffs in cases filed in other districts to participate in the proceedings upon application of plaintiffs' counsel. Approximately 1174 plaintiffs were represented at that trial. For reasons not clear in the record, the plaintiffs in the present case, although participating in the multidistrict discovery, elected not to participate in the trial. They chose instead to have their case returned to this court for a separate proceeding.
The large number of cases involved in the consolidated trial required unusual and innovative procedures. To assist counsel in their jury selection, the standard jury questionnaire form was supplemented by an additional questionnaire, designed jointly by all counsel. The additional questionnaire was mailed to prospective jurors and returned before the beginning of trial. Based upon an examination of that second questionnaire and by agreement of counsel, some prospective jurors were excused. The jury ultimately selected was composed of five women and one man. The presentation of evidence at the consolidated trial required a total of 21 days, and nothing suggests that the trial was anything but capably and skillfully handled by Judge Rubin and trial counsel. Ten expert witnesses were presented on behalf of the plaintiffs, eight on direct and two on rebuttal, and nine expert witnesses were presented on behalf of the defendant. Plaintiffs required 41 hours 15 minutes to present their case both on direct, cross and rebuttal, while the defendant required 31 hours, 30 minutes to present its case on direct and on cross. Final argument and instruction consumed an additional day. After deliberations, the jury returned a unanimous verdict in favor of Merrell-Dow.
The jury unanimously answered in the negative the following special interrogatory:
Have the plaintiffs established by a preponderance of the evidence that ingestion of Bendectin at therapeutic doses during the period of fetal organogenesis is a proximate cause of human birth defects?
Judge Rubin denied a motion for judgment notwithstanding the verdict. In re: Richardson-Merrell, Inc. Bendectin Products, 624 F. Supp. 1212 (S.D.Ohio 1985). The plaintiffs' appeal is pending. On March 12, 1985, Judge Rubin ordered that those cases that did not elect to become part of the consolidated litigation be returned to the originating court "for such further proceedings as may be appropriate."
The case is now before this Court for further proceedings. In a nutshell, the plaintiffs in the present case seek to relitigate the issue of whether Bendectin causes human birth defects. Plaintiffs here assert that they should not be precluded from relitigating this issue because they were not named parties in the consolidated trial, nor were they in "privity" with any parties to that proceeding. They claim that they had no financial or proprietary interest in the consolidated case, nor did they supply an attorney or in any manner exercise control over that litigation. Furthermore, plaintiffs claim that the exercise of their discretionary right not to intervene in the earlier, consolidated litigation, should not preclude the assertion of their claim here and that they would be denied their constitutional right to a trial if they were precluded. I turn, then, to the first question: whether the doctrine of collateral estoppel should be invoked under these facts to preclude plaintiffs from relitigating the issue of Bendectin's alleged responsibility for human birth defects.
The doctrine of collateral estoppel, and the related doctrine of res judicata, may function to prevent the relitigation in subsequent actions of an issue or issues decided in earlier actions. Collateral estoppel, or issue preclusion, may protect litigants from the burden of relitigating an identical issue with the same party or his privy, and serve to promote judicial economy through the prevention of needlessly repetitive litigation. Parklane Hosiery Co. v. Shore, 439 U.S. 322, 326, 99 S. Ct. 645, 649, 58 L. Ed. 2d 552 (1979). The doctrine has generally been applied against persons who were either parties to a prior suit or in privity with such parties. Plaintiffs assert that collateral estoppel should only be applied against persons who were either parties to a prior suit or in strict privity with such parties.
The plaintiffs' argument is not without merit. In United States v. Mendoza, 464 U.S. 154, 158-159 n. 3, 4, 5, 104 S. Ct. 568, 571-72 n. 3, 4, 5, 78 L. Ed. 2d 379, the Supreme Court said:
Under the judicially developed doctrine of collateral estoppel, once a court has decided an issue of fact or law necessary to its judgment, that decision is conclusive in a subsequent suit based on a different cause of action involving a party to the prior litigation. Montana v. United States, 440 U.S. 147, 153 [99 S. Ct. 970, 973, 59 L. Ed. 2d 210] (1979). Collateral estoppel, like the related doctrine of res judicata,3 serves to "relieve parties of the cost and vexation of multiple lawsuits, conserve judicial resources, and, by preventing inconsistent decisions, encourage reliance on adjudication." Allen v. McCurry, 449 U.S. 90, 94 [101 S. Ct. 411, 414-415, 66 L. Ed. 2d 308] (1980). In furtherance of those policies, this Court in recent years has broadened the scope of the doctrine of collateral estoppel beyond its common-law limits. Ibid. It has done so by abandoning the requirement of mutuality of parties. Blonder-Tongue Laboratories, Inc. v. University of Illinois Foundation, 402 U.S. 313 [91 S. Ct. 1434, 28 L. Ed. 2d 788] (1971), and by conditionally approving the "offensive" use of collateral estoppel by a nonparty to a prior lawsuit. Parklane Hosiery, supra.4
In Standefer v. United States, 447 U.S. 10, 24 [100 S. Ct. 1999, 2008, 64 L. Ed. 2d 689] (1980), however, we emphasized the fact that Blonder-Tongue and Parklane Hosiery involved disputes over private rights between private litigants. We noted that "[i]n such cases, no significant harm flows from enforcing a rule that affords a litigant only one full and fair opportunity to litigate an issue, and [that] there is no sound reason for burdening the courts with repetitive litigation." 447 U.S. at 24 [100 S. Ct. at 2008]. Here, as in Montana v. United States, supra, the party against whom the estoppel is sought is the United States; but here, unlike in Montana, the party who seeks to preclude the Government from relitigating the issue was not a party to the earlier litigation."5
3 Under res judicata, a final judgment on the merits bars further claims by parties or their privies on the same cause of action. Montana v. United States, 440 U.S at 153 [99 S. Ct. at 973]; Parklane Hosiery Co. v. Shore, 439 U.S. 322, 326 n. 5 [99 S. Ct. 645, 649 n. 5, 58 L. Ed. 2d 552] (1979). The Restatement of Judgments speaks of res judicata as "claim preclusion" and of collateral estoppel as "issue preclusion." Restatement (Second) of Judgments § 27 (1982).
4 Offensive use of collateral estoppel occurs when a plaintiff seeks to foreclose a defendant from relitigating an issue the defendant has previously litigated unsuccessfully in another action against the same or a different party. Defensive use of collateral estoppel occurs when a defendant seeks to prevent a plaintiff *860 from relitigating an issue the plaintiff has previously litigated unsuccessfully in another action against the same or a different party. Parklane Hosiery, supra, at 326 n. 4 [99 S. Ct. at 649 n. 4].
5 In Montana we held that the Government was estopped from relitigating in federal court the constitutionality of Montana's gross receipts tax on contractors of public construction firms. That issue has previously been litigated in state court by an individual contractor whose litigation had been totally financed and controlled by the Federal Government. Montana v. United States, supra, [440 U.S.] at 151, 155 [99 S. Ct. at 972, 974] see n. 9, infra.
In Mendoza, the Court disapproved the application of the doctrine of collateral estoppel to the government, but noted that in a companion case, United States v. Stauffer Chemical Co., 464 U.S. 165, 104 S. Ct. 575, 78 L. Ed. 2d 388 (1983), the doctrine was applied to the government under certain circumstances when the parties to the two lawsuits were the same.
This Circuit has spoken earlier with caution on this issue. Griffin v. Burns, 570 F.2d 1065, 1072 (1st Cir. 1978) (absence of privity prevents collateral estoppel. But certain inroads have been made upon the rule of mutuality); Walsh v. International Longshoremen's Association, AFL-CIO, 630 F.2d 864, 874 (1st Cir.,1980) (if the jurisdictional questions in the two cases are the same, relitigation of the issue is barred by collateral estoppel); General Foods v. Massachusetts Department of Public Health, 648 F.2d 784, 790 (1st Cir.1981) (rejecting doctrine of virtual representation).
More recently, this Circuit has applied the doctrine of collateral estoppel under certain circumstances. Pignons S.A. de Mechanique v. Polaroid Corp., 701 F.2d 1 (1st Cir.1983). In that case, the court said:
Pignons had a fair opportunity to make these arguments and to introduce this evidence the first time. The law requires the courts to offer Pignons nothing more, for collateral estoppel implements "the principle that one opportunity to litigate an issue fully and fairly is enough." Continental Can Co. v. Marshall, 603 F.2d 590, 594 (7th Cir.1979). Once a plaintiff has had a chance to prove a fact, he cannot reopen the matter simply by stating that he wishes to introduce more or better evidence.
See also Fiumara v. Fireman's Fund Insurance Companies, 746 F.2d 87-92 (1st Cir.1984) (the significant question is whether a party has had a full and fair opportunity for judicial resolution of the same issue); O'Reilly v. Malon, 747 F.2d 820 (1st Cir. 1984) (the party is not entitled to more than one full and fair opportunity for judicial resolution of the same issue).
Other circuits have invoked the doctrine of collateral estoppel when, as in this case, a plaintiff has elected as a matter of litigation strategy to forego an opportunity to intervene in an action. In National Wildlife Fed'n v. Gorsuch, 744 F.2d 963 (3rd Cir.1983), the Third Circuit held that collateral estoppel was properly invoked against plaintiffs who, though not parties to an earlier action, bypassed an opportunity to intervene in the earlier action. Like the plaintiffs in Gorsuch, the plaintiffs here "were not outsiders unaware of litigation in progress that would ultimately affect their interests." Gorsuch, 744 F.2d at 971-72. In Bronson v. Board of Educ., 525 F.2d 344 (6th Cir.1975), the Court of Appeals for the Sixth Circuit refused to allow relitigation of racial discrimination claims by new plaintiffs following an earlier action on the same issues brought by different plaintiffs, noting that it would be inequitable to require the defendant school board to repeatedly battle the "same charge of improper conduct if it has been vindicated in an action brought by a person or group who validly and fairly represent those whose rights are alleged to have been infringed." Bronson, 525 F.2d at 349.
In Gerrard v. Larsen, 517 F.2d 1127, 1135 (8th Cir.1975), the court stated that the determination of who should be bound by a prior adjudication ought to be conducted "on a case by case basis by an examination of underlying facts and circumstances rather than by reliance solely upon the formal status of persons against whom an estoppel is asserted."
*861 Finally, there is some respected authority that where there has been adequate litigation, the virtual representation theory may play a narrow role. 18 Wright, Miller & Cooper, Federal Practice and Procedure § 4457 at 502 (1981). See also Crane v. Commissioner of Department of Agriculture, Food and Natural Resources, 602 F. Supp. 280 (D.Me.1985). Other commentary provides a checklist to determine whether collateral estoppel is appropriate:
1. Is the issue identical to the issue in the first action?
2. Was the precise issue actually litigated and decided by the factfinder in the first action?
3. Considering all the objections, motions and rulings which related to this issue, was it finally judicially determined in the first action?
4. Was the judgment in the first action actually dependent on the determination of this issue?
21 Federal Procedure Lawyers Edition § 51:212.
This somewhat extensive review was necessary, first, to illustrate the evolution of the doctrine of collateral estoppel and its application, and secondly, to decide whether the doctrine should be applied here. After examination of the underlying facts and circumstances of this case, I conclude that the plaintiffs here should be bound by the results of the consolidated trial.
The issue of causation with respect to Bendectin and birth defects was extensively litigated in the consolidated trial, and the outcome of that litigation should be accorded finality. The claims by plaintiffs in the consolidated trial were representative of all potential claims involving birth defects allegedly caused by the maternal ingestion of Bendectin, and all of the opinion witnesses designated by plaintiffs in this case testified at the consolidated trial. The plaintiffs here unconvincingly argue that their claim differs from those in the consolidated trial. They emphasize that a trial before this Court would focus on the specific incident of Margaret Lynch's ingestion of Bendectin. But the central issue in any proceeding involving Bendectin and its role in birth defects is that of causation, and that issue has already been fully litigated in the consolidated trial.
If this case were to be tried, there would be no reason to depart from the careful, thoughtful and fair procedure adopted by Judge Rubin in the earlier trial. Bifurcation would certainly be appropriate. Fed. R.Civ.P. 42(b). The jury's attention should be focused quickly and without distraction if both sides are to receive a fair resolution. The issue is complex; approximately 19 experts would be expected to testify, and the 4 cases tried to date lasted an average of 38 trial days. The jury would be selected in the same manner, employing the same questionnaire and voir dire procedures, and receiving essentially the same instructions. Margo Lynch would not be present in the courtroom, although she could observe the proceedings from the adjoining lobby by electronic means. The jury would be instructed at the outset that while the immediate issue was a medical-legal one, the case involved real people and birth deformity. Therefore, this would not be a mere "academic debate between the medical and scientific experts for each side," as the plaintiffs describe the consolidated trial. The plaintiffs attribute the defendant's verdict in the consolidated trial to the fact that "no individual cases were discussed" or that "confronted with the causation issue in a vacuum, the jury returned a verdict for the defendant." Not only does this argument belittle the jury's role, but it demonstrates the plaintiffs' hope that the presence of Margo Lynch in the courtroom would produce a different and favorable result. This is a case in which there would be no new medical or scientific evidence, no new expert opinion, no new studies, no new data or theories. The plaintiffs point out that a plaintiff's verdict was returned in Oxendine v. Merrell Dow Pharmaceuticals, 506 A.2d 1100 (D.C.C.A.1986), but that state court trial did not have the full record of the consolidated trial, nor was the case tried under the same procedures. It produced the very *862 inconsistency which the doctrine of collateral estoppel was designed to prevent.
The plaintiffs' constitutional arguments are not persuasive. A significant commonality of interest, sufficient to overcome the plaintiffs' due process objections, existed between the Lynches and the more than 1100 plaintiffs participating in the consolidated trial. The "opt-in" procedure adopted by Judge Rubin was not a commitment that every out-of-state plaintiff would receive a separate jury trial. Under the multidistrict litigation panel's orders, the case could not be transferred for trial, but only for pre-trial proceedings. The cases involving plaintiffs who chose not to participate in the consolidated trial were returned to the originating court for appropriate proceedings. As stated earlier, the record is silent as to the reasons plaintiffs elected not to participate in the consolidated trial. Were the record to reflect a worthy explanation, a lack of confidence in trial counsel, or a compelling distinction between the Lynches' claims and the others, I would have analyzed the doctrine in that light. But the plaintiffs have not provided a single sound reason for their decision. One permissible inference is that the plaintiffs elected not to participate because they felt they would not be bound by a defendants' verdict, but could take advantage of a plaintiffs' verdict by negotiating a favorable settlement or, ironically, by asserting issue preclusion or offensive collateral estoppel against the defendant. This would be the type of tactical or procedural maneuvering that should be discouraged and not rewarded. Imagine the consternation of the over 1100 plaintiffs involved in the prior litigation if the plaintiffs here were allowed to obtain an inconsistent adjudication by nothing more than a change in forum.
I am not unmindful of the harshness this ruling visits on the plaintiffs. But what it boils down to is that the plaintiffs want a second chance. Conclusiveness of adjudication is an important principle. Justice is achieved when a competent jury hears all the relevant and material evidence by competent witnesses on the issue, presented by able, competent counsel, and returns a unanimous verdict. That was achieved in the prior trial and I conclude the plaintiffs here are bound by that result.
The second assertion by the defendant in support of its motion for summary judgment is that the plaintiffs' proof on the issue of causation is insufficient to permit a reasonable trier of fact to conclude that Bendectin caused Margo Lynch's birth defect. Because of the extensive proceedings conducted in the consolidated trial before Judge Rubin, this Court is able to conduct a review of all of the evidence available to the parties with respect to the issue of Bendectin's causal role in human birth defects. All of the evidence which might provide reliable support for an opinion as to whether Bendectin causes birth defects is in possession of the parties and is before the Court. In any trial, the same evidence would be presented to the factfinder. After a thorough review of the records of prior testimony, expert depositions and affidavits, and available scientific studies, I conclude that, even if the plaintiffs were not collaterally estopped from litigating the issue, the plaintiffs have failed to show that they can provide admissible evidence on the issue of causation sufficient to permit a reasonable trier of fact to conclude that Bendectin caused Margo Lynch's birth defect.
*863 At the outset, this Court must recognize that despite recent advances in the field of teratology, the causation of most birth defects remains unknown. This has been admitted by plaintiffs' own expert, Dr. Alan K. Done, who has testified that the causation of 75 to 80 percent of birth defects is unknown (Done (Oxendine) Dep. 322; Done (Mekdeci) Tr. 1971; Done (Koller) Dep. 105-107), and that 95 percent of limb reduction defects, the type suffered by Margo Lynch, are of unknown origin. (Done (Oxendine) Dep. 327). None of the plaintiffs' witnesses claim that there is a scientific or medical consensus that Bendectin causes birth defects. Dr. Done has admitted that "obviously" there is not complete agreement that Bendectin is teratogenic. (Done (Koller) Dep. 107).
In support of its summary judgment motion, the defendant notes that more studies have been done of Bendectin than of any other drug used in pregnancy, and the defendant has collected and submitted more than 25 human epidemiological studies concerning Bendectin which have been published in the scientific literature.
Epidemiological studies rely on "statistical methods to detect abnormally high incidences of disease in a study population and to associate those incidences with unusual exposures to suspect environmental factors." Done, "A Commentary on the Use of Epidemiological Evidence in Demonstrating Cause-In-Fact." 7 Harv.Envtl.L. Rev. 429, 431 (1983). The majority of epidemiological studies presented here are of two main types cohort studies and case control studies. "Cohort" studies are those in which the incidence of birth defects among groups of persons exposed to Bendectin and groups of persons not exposed to Bendectin are compared. "Case control" studies start with a group of persons who have the birth defects in question ("cases"), match this group with another group (the "controls") not having the birth defects in question, and compare the frequency of exposure to the drug among the cases and the controls.
Several courts have recognized the important role epidemiological evidence may play in demonstrating causation of illness or disease. In the recent swine flu cases, courts consistently held that, in the absence of either epidemiological evidence or a scientific understanding of causation, a finding of causation must be rejected as speculative. See Terrell v. United States, 517 F. Supp. 374, 379 (N.D.Tex.1981) ("[e]pidemiological evidence is critical in discovering the etiology, or cause of diseases, particularly those which may have numerous causes") and Iglarsh v. United States, No. 79C2148, slip op. at 5 (N.D.Ill. Dec. 9, 1983) [Available on WESTLAW, DCTU database] ("[i]n the absence of a statistically valid epidemiological study, even the plaintiff's treating physicians, or any clinician for that matter, is unable to attribute a plaintiff's injury to the swine flu vaccination.").
Epidemiological studies on causation also played a major role in the recent "Agent Orange" litigation. In re "Agent Orange" Prod. Liab. Litig., 611 F. Supp. 1221 (D.C. N.Y.1985). In that litigation, Judge Weinstein reviewed "a number of sound epidemiological studies" on Agent Orange, finding that these were "the only useful studies having any bearing on causation", id. at 1231, and concluding that "the epidemiological studies alone demonstrate that on the basis of present knowledge, there is no question of fact. Agent Orange cannot now be shown to have caused plaintiff's numerous illnesses." Id. at 1241.
This conclusion is equally true with respect to the epidemiological studies of Bendectin and Margo Lynch's birth defect. The epidemiological studies presented by the defendant, including eleven cohort *864 studies and four case-control studies, indicate that there is no statistically significant association between Bendectin and birth defects. Also, none of the numerous published epidemiological studies of Bendectin has reported a statistically significant association between Bendectin and limb reduction defects. The cohort studies have not reported any disproportionate occurrence of limb reduction defects, and all of the case-control studies of limb reduction defects have failed to find any association with Bendectin usage.
The findings of these cohort and case-control studies have been confirmed by other epidemiological studies of population groups. A study by Dr. Steven H. Lamm indicates that, while Bendectin usage declined from more than 1 million new therapy starts in 1979 to zero in 1984, there was no change in the incidence of limb reduction defects. Thus, no connection between fluctuations in the use of Bendectin and the incidence of limb reduction defects has been demonstrated.
The several epidemiological studies of Bendectin also withstand a combined statistical analysis. Dr. Brian MacMahon computed a combined estimate of relative risk for ten of the cohort studies and found no statistically significant difference between the rate of birth defects among those exposed and those not exposed. Dr. MacMahon's results were replicated by Dr. Lamm in a similar study and, in an affidavit prepared for this case, Dr. Lamm has concluded that the "totality of the evidence of these studies does not indicate any increased risk of infants with congenital malformations among pregnancies with exposure to Bendectin."
The defendants have asked this Court to find, pursuant to Federal Rules of Evidence 401-403 and Federal Rule of Evidence 703, that the plaintiffs' proposed opinion testimony is inadmissible. Rules 401 and 402 require this Court to exclude evidence that does not conform to accepted scientific methodology and that is not of a type upon which experts could reasonably rely. Furthermore, under Rule 403, this Court must exclude relevant evidence "if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury." Rule 703 requires that an expert, considering data not admitted into evidence, base his opinion only upon those data that are reasonably relied upon by experts in that particular field. In applying Rule 703, the Court "must focus not only on the kinds of information that experts in a particular field generally rely upon, but also on whether in the particular case before the Court and on the particular subject to which the expert testimony is addressed, facts or data not admissible in evidence can reasonably be relied upon by an expert witness." Saltzburg and Redden, Federal Rules of Evidence Manual (3d ed.1982). Therefore, this Court must examine the evidence upon which the plaintiffs' opinion testimony will be based to determine whether it is of a type that is relied upon by experts in the field of teratology, and whether reliance on that evidence to explain Margo Lynch's specific birth defect is reasonable. The Court must also be mindful of whether the probative value of this evidence is sufficient to outweigh the potential danger of unfair prejudice, confusion of the issues, or misleading of a jury.
The plaintiffs seek to rely on four major types of evidence to support their opinion testimony that Bendectin was the cause of Margo Lynch's birth defect: re-analysis of the epidemiological studies presented by the defendant, in vivo and in vitro animal studies, and studies of analogous chemical structures.
As noted above, the epidemiological studies presented by the defendant indicate no statistically significant association between Bendectin and either birth defects generally or limb reduction defects in particular. The plaintiffs have presented no epidemiological evidence indicating a statistically significant association between Bendectin and any type of birth defect, and none of the plaintiffs' witnesses has conducted a study of Bendectin and birth defects.
*865 The only epidemiological evidence offered by the plaintiffs is a re-analysis by Dr. Shanna Swan, criticizing a study by Dr. Jose F. Cordero and Dr. Godfrey P. Oakley, Jr. of the Center on Disease Control, and attempting to realign the data in the original study to demonstrate a causal connection between Bendectin and birth defects. The defendant claims that Dr. Swan's study is methodologically flawed and untrustworthy in that it was conducted during litigation, was financed by plaintiffs' counsel, was prepared in substantial part by employees of plaintiffs' counsel, was developed without a written, objective protocol, and was initiated with an a priori understanding of what the likely results of the re-analysis would show. Dr. Swan has conceded that she was "potentially biased in performing this study." (Swan (MDL) Tr. 1059; Swan (Cordova) Tr. 2946-47).
Even if this Court were to find the methodology of Dr. Swan's re-analysis credible, this Court still could not accept result-oriented re-analysis of epidemiological studies and criticisms of others' methodology, such as that performed here by Dr. Swan, as reliable data upon which to base an opinion on causation. The plaintiffs cannot merely rely on criticisms of the defendant's studies to establish causation. The fallacy of such an attempt was recognized by the court in Koller v. Richardson-Merrell, Inc. which ruled in its February 25, 1983 order that the plaintiffs were "precluded from presenting alleged methodological flaws and other errors in epidemiological and animal studies in their case in chief unless plaintiffs can make a satisfactory showing that absent the flaws, the studies would affirmatively demonstrate that Bendectin is a teratogen at the appropriate level of statistical significance." The plaintiffs have made no such showing in this case.
A second basis for plaintiffs' opinion testimony are extrapolations from in vivo animal studies, in vitro animal studies, and studies done with chemical structures analogous to Bendectin. The in vivo studies relied upon by plaintiffs' experts are those in which dosages of Bendectin or doxylamine, a compound found in Bendectin, were administered to pregnant rats, rabbits, and monkeys. In vitro studies are those in which a chemical or drug were tested on something less than an intact, live pregnant animal. Several in vitro studies of Bendectin have been reported, including studies of the influence of Bendectin or doxylamine on frog nerve cells and rat brain cells. In the chemical structure studies, researchers attempt to draw conclusions concerning Bendectin through studies conducted on drugs which are chemically similar to Bendectin.
There are several limitations inherent in the use of the animal and other experimental data relied upon here by the plaintiffs, and these limitations have been conceded in prior testimony by plaintiffs' witnesses. Dr. Done, for example, has admitted that the fact that Bendectin may be teratogenic in a species of animals does not necessarily mean it will be teratogenic in humans. (Done (Oxendine) Tr. 942; Done (Koller) Dep. 78; Done (Oxendine) Tr. 684-685; Done (MDL) Tr. 464-65, 550-53, Done (Cordova) Tr. 1253-54). Dr. Michael Melnick has testified that, on the basis of animal studies, "one can only speak of birth defects in the animal in the animal one has studied," (Melnick (Cordova) Tr. 1934-36, see Melnick (MDL) Tr. 1226-27, 1374-75), and Dr. Mark Edward Thoman has stated that animal studies "have to be taken in conjunction with the fact that there are differences between animals and the human." (Thoman (Koller) Dep. 71).
It is also difficult to extrapolate findings from animal studies to human studies because of the large doses used in the animal studies. Dr. William J. Scott has testified that almost any substance can be shown to be teratogenic at some dose in some species. (Scott (MDL) Tr. 2179-80, Scott (Cordova) Tr. 5817-18). Dr. Done has recognized that if one gives too much of virtually anything to a pregnant animal, malformation may be produced "if it makes the mother sick enough." (Done (Oxendine) Tr. 844; see Done (MDL) Tr. 541-42).
*866 None of the animal studies submitted by the plaintiffs provide evidence of teratogenicity at doses comparable to the human therapeutic dose of Bendectin. These animal studies are therefore lacking in probative value and must be found inadmissible. As Judge Weinstein explained in the Agent Orange litigation:
There is no evidence that plaintiffs were exposed to the far higher concentrations involved in both the animal and industrial exposure studies. The animal studies are not helpful in the instant case because they involve different biological species. They are of so little probative value as to be inadmissible. See Fed.R. Evid. 401-403. They cannot be an acceptable predicate for an opinion under Rule 703.
In re "Agent Orange" Prod. Liab. Litig., 611 F. Supp. at 1241.
There is no evidence here that Margaret Lynch was exposed to the types of doses involved in the animal studies relied upon by the plaintiffs' experts, and studies conducted on rats, rabbits, and monkeys are not helpful and are of little probative value in this case. The animal studies of Bendectin are therefore inadmissible in this action and can neither form a basis for expert opinion nor create a genuine issue for trial.
For similar reasons, this Court must reject the plaintiffs' proffered evidence of in vitro studies and studies of analogous chemical structures as a basis for the plaintiffs' experts testimony. The in vitro studies suffer from the same deficiencies as the in vivo animal studies they are performed in other biological species and at doses far in excess of the human therapeutic dose. This Court also cannot find, pursuant to Rule 703, that such studies are "of a type reasonably relied upon by experts in the particular field." Dr. John Hassell, the author of one such in vitro study, has expressly recognized that neither his technique nor any other in vitro system has yet been validated as an accurate predictor of teratogenicity in animals or humans. (See Defendant's App. III, Tab 40). Dr. Scott (Scott (Oxendine) Tr. 1958-66, Scott (MDL) Tr. 2202, Scott (Cordova) Tr. 5828-29) and Dr. Done (Done (Cordova) Tr. 881-82, 891-96) have also expressed their view that these tests have not yet been validated as useful screening tests of teratogenicity. Based on the in vitro studies, the plaintiffs' witnesses have generated many hypotheses as to the mechanisms by which Bendectin may cause birth defects, but these hypotheses are nothing more than speculation which might confuse and mislead a jury.
I find the studies of analogous chemical structures relied upon by plaintiffs' experts to be of even lesser probative value than the in vivo and in vitro animal studies. Plaintiffs' witnesses have admitted that, despite similarily, the "chemically analogous" drugs to which Bendectin is compared are different in some respects from Bendectin and its components, and that these differences may affect its properties. (Crescitelli (MDL) Tr. 1570-75, 1644-45, 1653, Crescitelli (Cordova) Tr. 801-05; Done (MDL) Tr. 522-23). To accept the validity of these chemical structure studies, one must accept the premise that the drugs to which Bendectin is compared are, in fact, themselves teratogenic. Admission of such evidence at trial would necessitate "minitrials" as to whether the chemically analogous drugs are indeed teratogenic. As the plaintiffs' expert, Dr. Done, has stated, chemical structure analysis can only raise "suspicion" about potential teratogenicity and cannot be relied upon alone to show causation (Done (Oxendine) Tr. 1056; Done (MDL) Dep. 113). Due to the highly speculative nature of the chemical structure analogies, I must rule that such evidence does not comport with the requirements of Rule 703 that an expert's opinion be grounded on facts or data of a type reasonably relied upon by experts in the particular field.
Thus, a careful review of the material before this Court indicates that the only relevant, probative, and non-misleading evidence on the issue of Bendectin's role in the causation of birth defects are the controlled observations of human beings, documented *867 in more than 25 published epidemiological studies. The data from these studies do not indicate any statistically significant association between Bendectin and the type of birth defect suffered by Margo Lynch. This Court finds that the evidence submitted by the plaintiffs in support of their expert testimony does not comport with the requirements of the Federal Rules of Evidence. Absent admissible and competent expert testimony grounded on evidence comporting with the requirements of the Federal Rules of Evidence, the plaintiffs cannot raise a genuine issue of material fact concerning Bendectin's role in the causation of Margo Lynch's birth defect.
The plaintiffs are collaterally estopped from relitigating the issue of Bendectin's alleged role in the causation of birth defects, and the plaintiffs' proof on the issue of causation is insufficient to permit a reasonable trier of fact to conclude that Bendectin caused Margo Lynch's birth defect. The defendant's motion for summary judgment is granted and the complaint is to be dismissed.
 In three other cases, separate judgments on jury verdicts have been entered at the trial court level for Merrell-Dow in suits alleging birth defects caused by Bendectin. See Mekdeci v. Merrell-National Labs., No. 77-255 (M.D.Fla.), aff'd. 711 F.2d 1510 (11th Cir.1983); Cordova v. Philips Roxane Labs., No. 432-656 (Cal.Super. June 24, 1985); Will v. Richardson-Merrell, Inc., ___ F.Supp. ___, No. CV-484-118 (S.D.Ga. June 27, 1986). In Oxendine v. Richardson-Merrell, Inc., No. 1245-82 (D.C.Super. Sept. 1, 1983), the trial judge entered judgment notwithstanding the verdict in favor of the plaintiff, but that ruling was reversed on appeal. A petition for rehearing is pending.
 Under res judicata, it is the claim itself that is precluded. Here the Lynches' claims were never litigated, but the issue was. Hence, this memorandum speaks in terms of collateral estoppel, although the distinction between the two doctrines appears to be a blurred one at times. Kremer v. Chemical Construction Corp., 456 U.S. 461, 480-481 n. 22, 102 S. Ct. 1883, 1896-97 n. 22, 72 L. Ed. 2d 262 (1982); Nevada v. United States, 463 U.S. 110, 128-130 n. 11, 103 S. Ct. 2906, 2917-18 n. 11, 77 L. Ed. 2d 509 (1982); Gerrard v. Larsen, 517 F.2d 1127, 1134 (8th Cir.1975).
 I understand that this finding is, in some respects, contradictory to the District of Columbia Court of Appeals holding in Oxendine v. Merrell Dow Pharmaceuticals, 506 A.2d 1100 (D.C.App. 1986). The issues addressed here, however, differ from the issues addressed in the Oxendine appeal. The Oxendine court did not directly address the issue of admissibility of certain evidence and testimony presented by the plaintiff. This Court examines the admissibility of the plaintiffs' evidence and opinion testimony with respect to the Federal Rules of Evidence and examines the admissibility of evidence and testimony in light of applicable federal case law decided after argument of the Oxendine appeal.
 For the purposes of this Memorandum and Order, I have adopted a citation form for the depositions and transcripts similar to the form used by the defendants in their motion for summary judgment. The citation refers to the expert testifying, then provides the abbreviated case name in parentheses and the page number of the deposition or case transcript. The testimony referenced may be found in the defendant's Appendix II, Depositions and Trial Testimony.