Full opinion text
WISDOM, Circuit Judge: This products liability case raises significant questions concerning the scope of a drug manufacturer’s duty to warn ultimate consumers of dangers inherent in his product. In May 1970, slightly more than two weeks after she had received a dose of Wyeth Laboratories’ oral polio vaccine, eight-month-old Anita Reyes was diagnosed as having paralytic poliomyelitis. Epifanio Reyes, Anita’s father, filed suit against Wyeth Laboratories, alleg- \ ing that the live polio virus in the vac\cine had caused Anita’s polio and that Wyeth was liable for her injuries because it had failed to warn her parents pf this danger. Wyeth contends that it is not liable for Anita Reyes’s injuries under the products liability law of Texas. In addition, Wyeth raises numerous procedural and evidentiary questions. The case was tried to a jury. The jury answered special interrogatories (see Appendix A) and returned a verdict in favor of Reyes against Wyeth for the sum of $200,000. I. Twenty or thirty years ago poliomyelitis was a dread disease that especially attacked the very young. In 1952 alone, there were 57,879 reported cases of polio in the United States; 21,269 of these resulted in crippling paralysis to the victims. By 1970, when Anita Reyes contracted polio, the number of those stricken by polio had diminished dramatically; she was one of just 33 individuals to be afflicted during that year. Credit for this precipitous decline must go primarily to the medical researchers who discovered the viral nature of the disease, and were able to isolate and reproduce the virus in an inactivated or an attenuated form. See Appendix B. But credit for this remarkable achievement must also be given to such laboratories as Wyeth, which processed the polio vaccine, and to massive federal-state public health programs for the administration of the vaccine. On May 8, 1970, Anita Reyes was fed two drops of Sabin oral polio vaccine by eye-dropper at the Hidalgo County Department of Health clinic in Mission, Texas. The vaccine was administered to Anita by a registered nurse; there were no doctors present. Mrs. Reyes testified that she was not warned of any possible danger involved in Anita’s taking the vaccine. Mrs. Reyes has a seventh grade education, but her primary language is Spanish. She signed a form releasing the State of Texas from “all liability in connection with immunization”. The form contained no warning of any sort, and it is apparent from, her testimony that she either did not read the form or lacked the linguistic ability to understand its significance. About fourteen days after the vaccine was administered, Anita Reyes became ill. On May 23, 1970, she was admitted to the McAllen (Texas) General Hospital, where her disease was diagnosed as Type I paralytic poliomyelitis. See Appendix B, footnote 59. As a result of the polio, at the time of trial Anita was completely paralyzed from the waist down, her left arm had become atrophied, and she was unable to control her bladder or bowel movements. The vaccine given Anita Reyes in the Mission clinic on May 8, 1970 was part of a “lot”, No. 15509, prepared by Wyeth. Lot No. 15509 was trivalent oral polio vaccine that Wyeth had titered (mixed) from Types I, II, and III monovalent vaccine provided by Pfizer, Ltd. In response to an order placed by the Texas State Department of Health on December 23, 1969, Wyeth shipped 3500 vials of Lot No. 15509 vaccine to the State Health Department which in turn transferred 400 vials to the Hidalgo County Health Department. The jury found that vaccine from one of these vials was given to Anita Reyes. Included with every vial, each of which contained ten doses of vaccine, was a “package circular” provided by Wyeth which was intended to warn doctors, hospitals, or other purchasers of potential dangers in ingesting the vaccine. Mrs. Lenore Wiley, the public health nurse who administered the vaccine to Anita Reyes, testified that she had read the directions on this package insert, but that it was not the practice of the nurses at the Mission Health Clinic to pass on the warnings to the vaccinees or to their guardians. She testified that she gave Mrs. Reyes no warning before she administered the vaccine to Anita. On October 7, 1970, Epifanio Reyes, individually and as next friend of his minor daughter, brought this action on theories of strict products liability, breach of warranty, and negligence. In his complaint he alleged that his daughter had contracted polio from the live virus in Wyeth’s vaccine, and that Wyeth’s failure to warn him or his wife that this might occur rendered it liable for Anita’s injuries. Wyeth’s main line of defense was that Anita’s polio was not vaccine-induced at all. Wyeth produced experts who characterized virus isolated from a specimen of stool taken from Anita on the day after she was admitted to the hospital as “probably wild”. This probability was increased, according to epidemological testimony adduced on Wyeth’s behalf, by the fact! that there was a polio “epidemic” in Hi-' dalgo County at the time Anita fell ill,* In effect, the jury rejected these theories in rendering a verdict for Reyes. Wyeth asserts more than a score of separate grounds for reversal, each accompanied by extensive argument and citation. We have considered all of these contentions, but in this opinion we shall address only those challenges which we regard as raising substantial questions. II. We turn first to the substantive issues raised by this appeal. Our inquiry is bounded by the jury’s finding that Wyeth’s vaccine was the producing cause of Anita Reyes’s polio, and by those principles of products liability law we conclude would be applied by the courts of Texas. Erie R. R. Co. v. Thompkins, 1938, 304 U.S. 64, 58 S.Ct. 817, 82 L.Ed. 1188. The jury’s finding that the Wyeth vaccine was responsible for Anita Reyes’s polio is less a conclusion than a starting point. Texas courts may hold manufacturers of products which harm consumers liable for the injuries, but, of course, manufacturers are not insurers. Texas courts recognize both tort and warranty theories of products liability. This case was tried, briefed, and argued on appeal entirely on the tort theory of strict liability. Thus the differences between the two approaches, and whatever contractual trappings of warranty have not been destroyed with the crumbling of the citadel of privity, need not detain us. Rather, we turn to the theory of strict products liability as embodied in Section 402A of the Restatement (Second) of Torts (1965), and approved by the Supreme Court of Texas in McKisson v. Sales Affiliates, Tex.Sup.Ct. 1967, 416 S.W.2d 787. Basically, Section 402A subjects to liability the seller or manufacturer of a product sold “in a defective condition unreasonably dangerous” to an ultimate user or consumer whose person or property is physically harmed by the product. Moreover, one who places defective goods in the stream of commerce will be liable “to the user or consumer even though he has exercised all possible care in the preparation and sale of the product”. Restatement (Second) of Torts, Section 402A, comment a. Yet imposition of liability is by no means automatic; the elements tacit or explicit in Section 402A’s mandate must be demonstrated to the trial court’s satisfaction before the burden of the consumer’s loss will be imposed on the seller of the product: The plaintiff is faced with an arduous burden of proof. He must prove that: 1) the product in question was defective; 2) the defect existed at the time the products left the hands of the defendant; 3) that because of the defect the product was unreasonably dangerous to the user or consumer (plaintiff) ; 4) that the consumer was injured or suffered damages; 5) and that the defect (if proved) was the proximate cause of the injuries suffered. Gravis v. Parke-Davis & Co., Tex.Civ.App.1973, 502 S.W.2d 863, 868. Although we do not embrace Gravis as the all-inclusive embodiment of Texas products liability law, the framework it provides suggest a useful analysis for the issue of Wyeth’s liability here. All five elements, however, which constitute the Gravis requirements need not be discussed here. There can be no question that Anita Reyes was injured, so the fourth element is not in issue. Nor is the second, since the defect alleged, failure to warn by the manufacturer, is by definition the manufacturer’s dereliction. Moreover, to find that the plaintiff proved the first element is to conclude that he proved the third, for properly understood, “defective condition” has no meaning independent of “unreasonably dangerous”; the two terms are essentially synonymous. Thus if a product is unreasonably dangerous as marketed, the manufacturer may be held liable for injuries proximately caused by what he has produced, whether or not it was manufactured exactly as intended, that is without a production “defect”. We do not understand this approach to dispense with the principle that to prompt liability a product must reach the consuming public in a “defective condition”. Rather, by rephrasing the defectiveness requirement in terms of “unreasonable danger”, it becomes clear that the circumstances of marketing themselves can amount to a defect; the defect can be extrinsic to the product. AH that we need determine here, then, is first, whether the vaccine was unreasonably dangerous, and second, whether the showing of proximate causation was sufficient under Texas law. a) Unreasonable Danger and the Duty to Warn We begin the inquiry by asking whether the vaccine was unreasonably dangerous, that is, in a defective condition when Anita Reyes received it. It is clear, of course that the vaccine was not itself defective. Wyeth Vaccine Lot No. 15509 was exactly what its makers and the Texas public health authorities intended it to be: trivalent live-virus Sabin oral polio vaccine. The live virus which the jury concluded caused Anita’s poliomyelitis was not inadvertently included in the mixture. Indeed, it is the presence of the living but attenuated Type I, II, and III viruses which makes the Sabin vaccine so effective [See Appendix B]. Although the living virus in the vaccine does not make the vaccine defective, it does make it what the Restatement calls an “unavoidably unsafe product”,! one which cannot be made “safe” no matter how carefully it is manufactured. Such products are not necessarily “unreasonably dangerous”, for as this Court has long recognized in wrestling with product liability questions, many goods possess both utility and danger. See, e. g., Ross v. Up-Right, Inc., 5 Cir. 1968, 402 F.2d 943, 946; Helene Curtis Industries, Inc. v. Pruitt, 5 Cir. 1967, 385 F.2d 841, 850, cert. denied, 1968, 391 U.S. 913, 88 S.Ct. 1806, 20 L.Ed.2d 652. Rather, in evaluating the possible liability of a manufacturer for injuries caused by his inevitably hazardous products, a two-step analysis is required to determine first, whether the product is so unsafe that marketing it at all is “unreasonably dangerous per se”, and, if not, whether the product has been introduced into the stream of commerce without sufficient safeguards and is thereby “unreasonably dangerous as marketed”. In either case, the applicable standard, as formulated in the Restatement, is as follows: In terms of the user’s interests, a product is “unreasonably dangerous” only when it is “dangerous to an extent beyond that contemplated by the ordinary consumer”; or, to phrase it in terms of the seller’s responsibility, “so dangerous that a reasonable man would not sell the product if he knew the risk involved”. See Borel v. Fibreboard Paper Products Corp., 5 Cir. 1973, 493 F.2d 1076, 1088; Helene Curtis Industries, Inc. v. Pruitt, 385 F.2d at 850; Wade, Strict Tort Liability of Manufacturers, 19 Sw.LJ. 5, 15 (1965). In determining whether placing a commodity on the market is “unreasonably dangerous per se”, the reasonable man standard of the Restatement becomes the fulcrum for a balancing process in which the utility of the product properly used is weighed against whatever dangers of harm inhere in its introduction into commerce. Obviously, use of an unavoidably unsafe product always presents at least a minimal danger of harm, but only if the potential harmful effects of the product— both qualitative and quantitative — outweigh the legitimate public interest in its availability will it be declared unreasonably dangerous per se and the person placing it on the market held liable. Applying this standard here, the scales must tip in favor of availability. The evil to be prevented — poliomyelitis and its accompanying paralysis — is great. Although the danger that vaccinees may contract polio is qualitatively devastating, it is statistically miniscule. On balance then, marketing the vaccine is justified despite the danger. Since Sabin oral polio vaccine is not “unreasonably dangerous per se”, we move to the second step of our analysis to determine whether it is “unreasonably dangerous as marketed”, for to conclude that the maker of an unavoidably unsafe product did not act unreasonably in placing it on the market is not to relieve him of the responsibility to market it in such a way as to prevent unreasonable danger. In the case of a product .such as Sabin oral polio vaccine, this translates into a duty to provide proper warnings in selling the product. As comment k to Section 402A instructs, an unavoidably unsafe product is neither defective nor unreasonably dangerous if such a product is “properly prepared, and is accompanied by proper directions and warning”. Consequently, the Restatement requires a seller who has reason to believe that danger may result from a particular use of his product to provide adequate warning of the danger in order that the product’s potential for harm may be reduced. Failure to give such a warning when it is required will itself present a “defect” in the product and will, without more, cause the product to be “unreasonably dangerous as marketed”. Issue is joined then, on the question whether Wyeth was under a duty to warn the parents of Anita Reyes that there was a possibility, however remote, that she might contract polio from the drug designed to immunize her. If such a duty existed, the vaccine was “defective” and unreasonably dangerous as marketed, for such cautions as Wyeth advanced did not, and were not intended to, reach the Reyes family. See Davis v. Wyeth Laboratories, Inc., 9 Cir. 1968, 399 F.2d 121, 129. Wyeth does not deny that its vaccine is “unavoidably unsafe”, or contend that it was unaware of the danger. Rather, the appellant contends that if it had a duty to warn at all, that duty was discharged by the warning contained on the package insert which accompanied the vials of vaccine sold to the Texas State Department of Health. This is so, Wyeth asserts, because the Sabin trivalent oral polio vaccine in issue here is a “prescription drug”, and those who prepare such drugs are not required to warn the ultimate consumer. If the warning to the dispensing physician or authorities (here the Texas and Hidalgo County Public Health Departments) was adequate, Wyeth is not liable for any harm' caused by the vaccine. Resolution of these contentions is crucial; Wyeth concedes in its brief that “[s]ince it is undisputed that Wyeth did not warn Reyes, but only the Texas State Department of Health, a finding that the vaccine was not a prescription drug establishes as a matter of law the defectiveness of the vaccine for purposes of a prima facie case in strict products liability.” We cannot quarrel with the general proposition that where prescription drugs are concerned, the manufacturer’s duty to warn is limited to an obligation to advise the prescribing physician of any potential dangers that may result from the drug’s use. This special standard for prescription drugs is an understandable exception to the Restatement’s general rule that one who markets goods must warn foreseeable ultimate users of dangers inherent in his products. See Restatement (Second) of 'Torts, Section 388 (1965). Prescription drugs are likely to be complex medicines, esoteric in formula and varied in effect. As a medical expert, the prescribing physician can take into account the propensities of the drug, as well as the susceptibilities of his patient. His is the task of weighing the benefits of any medication against its potential dangers. The choice he makes is an informed one, an individualized medical judgment bottomed on a knowledge of both patient and palliative. Pharmaceutical companies then, who must warn ultimate purchasers of dangers inherent in patent drugs sold over the counter, in selling prescription drugs are required to warn only the prescribing physician, who acts as a “learned intermediary” between manufacturer and consumer. Although there is no question that Sabin oral vaccine is licensed for sale only as a prescription drug, the district court, in its charge to the jury, noted that the vaccine was not administered as a prescription drug at the Mission Clinic. The court charged: “if you [the jury] find that a warning should have been given, the warning had to be given to Anita and her parents, not to Mrs. Wiley, that Public Health nurse, somebody else . . . The ultimate consumer is the one that had to be warned.” The district court apparently based this instruction on the leading federal case in the area, Davis v. Wyeth Laboratories, 9 Cir. 1968, 399 F.2d 121. In Davis, the plaintiff had allegedly contracted polio from Wyeth oral vaccine distributed at a public clinic. The Ninth Circuit held that where no individualized medical judgment intervenes between the manufacturer of a prescription drug and the ultimate consumer, “it is the responsibility of the manufacturer to see that warnings reach the consumer, either by giving warning itself or by obligating the purchaser to give warning.” 399 F.2d at 131. Where there is no physician to make an “individualized balancing ... of the risks”, the Court reasoned, the very justification for the prescription drug exception evaporates. Thus, as in the case of patent drugs sold over the counter without prescription, the manufacturer of a prescription drug who knows or has reason to know that it will not be dispensed as such a drug must provide the consumer with adequate information so that he can balance the risks and benefits of a given medication himself. Moreover, just as the manufacturer cannot make this choice for its ultimate consumers, it cannot allow its immediate purchaser to choose for them. In sum, then, the manufacturer is required to warn the ultimate consumer, or to see that he is warned. 399 F.2d at 131. Wyeth does not resist the Ninth Circuit’s holding in Davis, but asserts that the instant case can be distinguished on four grounds. First, the appellant argues, Davis received his vaccine during a mass immunization program, whereas Anita Reyes ingested her vaccine at her parents’ request. Second, Wyeth stresses the fact that Davis received his vaccine from a pharmacist, but Reyes’ was administered by a public health nurse. Third, Wyeth’s active participation in the mass immunization program involved in the Davis case is contrasted to its relatively passive role here. Finally, Wyeth urges that unlike the situation in Davis, here it had no knowledge that the vaccine would not be administered as a prescription drug. None of these asserted grounds for distinguishing Davis justifies a different result here. The first two arguments are admittedly distinctions between Davis and the instant controversy, but they have no bearing on the rationale of the Davis opinion. Whether vaccine was received during a mass immunization or an on-going program, whether it was administered by nurse or pharmacist, it was, in both these cases, dispensed without the sort of individualized medical balancing of the risks to the vaccinee that is contemplated by the prescription drug exception. The third and fourth asserted bases for distinguishing Davis from this case are essentially the same: Wyeth took no active part in the vaccination process here, and did not kpow that its vaccine would be dispensed without procedures appropriate for distribution of prescription drugs. Were we to conclude that Wyeth neither knew nor had reason to know that its vaccine would be dispensed without prescription drug safeguards, we might be required to hold that the rationale in Davis is inapplicable here. But Wyeth had ample reason to foresee the way in which its vaccine would be distributed. A drug manufacturer is held to the skill of an expert in his field, and is presumed to possess an expert’s knowledge of the arts, materials, and processes of the pharmaceutical business. Included in such expertise must be a familiarity with practices and knowledge common in the drug industry as to distribution and administration of pharmaceutical products. Neal Nathanson, Professor of Epidemiology at the Johns Hopkins University School of Public Health and a witness for Wyeth testified that it was common knowledge in the drug industry that “a great majority” of vaccinees receive their Sabin vaccine in mass administrations or county clinics manned at least in part by volunteers. Moreover, Dr. Nathanson agreed that it was well known that such clinics were stocked primarily by sale of vaccine to state health departments. These clinics, as Wyeth must be presumed to know, dispense Sabin vaccine to all comers in an “assembly line” fashion; there is often neither time nor personnel to make an “individualized medical judgment” of the vaccinee’s needs or susceptibilities. See Davis v. Wyeth, supra, 399 F.2d at 131; 56 Geo.L.J. 1016, 1017 (1968); 5 San Diego L.Rev. 422, 428 (1968). Viewed in this light, the present controversy, however it differs from Davis factually, invites application of the Davis principles, and the conclusion that Wyeth was under a duty to warn Anita Reyes’s parents of the danger inherent in its vaccine. Wyeth knew or had reason to know that the vaccine would not be administered as a prescription drug, and therefore was required to warn foreseeable users, or see that the Texas Department of Health warned them. See Davis v. Wyeth, supra, 399 F.2d at 131; cf. Helene Curtis Industries v. Pruitt, supra, 385 F.2d at 861. Wyeth’s failure to warn was a breach of its duty and made the vaccine “defective” — hence “unreasonably dangerous” — as marketed. Before moving to a consideration of causation, we pause to analyze two possible arguments, one raised by Wyeth and one not, which appear to add further conceptual problems in the “duty to warn” area. The first may be disposed of quickly. Wyeth contends that whatever its duty to warn, Mrs. Reyes assumed the risk that her daughter might .contract polio, or effectively released Wyeth, by signing a “release” purporting to excuse the State of Texas from liability. Wyeth does not explain how signing a form releasing the State could excuse Wyeth from performing a duty to warn imposed on it by law, other than to cite the “unity of release” rule, apparently under the assumption that the State of Texas was its agent. Since we decline to adopt the district court’s conclusion that Texas was the agent of Wyeth in administering the vaccine, we need not consider Wyeth’s contention that the release signed by Mrs. Reyes also applied to it. Moreover, although we agree that assumption of risk is a proper defense to a product liability action in Texas (see Borel v. Fibreboard Paper Products Corp.) one cannot assume a risk of which he is unaware. Shamrock Fuel & Oil Sales Co. v. Tunks, Tex.Sup.1967, 416 S.W.2d 779, 786 and n. 8. Since Mrs. Reyes was not cognizant of the danger inherent in the vaccine, she could not have assumed the risk of its use. Somewhat more troubling is a line of Texas cases involving unusual allergic reactions to potentially dangerous products sold without a warning. These decisions may be read for the proposition that only if the product, properly prepared, will harm a substantial number of people will the manufacturer be held liable for his failure to warn. See C. A. Hoover & Son v. O. M. Franklin Serum Co., Tex.Sup.1969, 444 S.W.2d 596, 598; Alberto-Culver Co., v. Morgan, Tex.Civ.App.1969, 444 S.W.2d 770, 776-777, writ ref. n. r. e.; Cudmore v. Richardson-Merrell, Inc., Tex.Civ.App. 1966, 398 S.W.2d 640, 644, writ ref. n. r. e., cert. denied, 1967, 385 U.S. 1003, 87 S.Ct. 705, 17 L.Ed.2d 542. Essentially these are foreseeability cases; the element they would require to invoke strict liability doctrines is foreseeability. In these cases, the manufacturer not only marketed the unavoidably unsafe product without adequate warning and with the knowledge that it was inherently dangerous, but the danger to any individual consumer was sufficiently significant and knowable that (1) the manufacturer was cognizant of a need to warn so as to prevent injury and (2) the consumer might or could have altered his conduct in such a way as to lessen or avoid the danger. This ancipital coin of foreseeability is the currency of these Texas cases. What they seek to avoid is the imposition of liability for “abreactions”, hypersensitive allergic reactions so unique that the class of persons exposed to the risk is miniscule. See Cudmore v. Richardson-Merrell, Inc., 398 S.W.2d at 644. In such cases, warnings would be meaningless, since the manufacturer can convey nothing meaningful to the allergic consumer unless the latter knows he has the allergy. Without such knowledge, the consumer is apt to assume that he is not a member of the minute susceptible class and to disregard the warning. See Alberto-Culver Co. v. Morgan, 444 S.W.2d at 777. Thus with the utility of warning limited and the susceptible class minute, the manufacturer is not required to foresee that anyone will suffer an allergic reaction to his product, and is not held liable. To the instant analysis, these principles present no dilemma, for the effect of the Sabin oral vaccine upon those who contract polio from it cannot fairly be styled an “abreaction”. Although the danger of vaccine-induced polio to any one individual is small, the risk appears to be distributed evenly among that substantial segment of the population that is not naturally immune to polio. This is so because the behavior of polio virus in those who contract the disease seems to suggest more a reversion to virulence by the virus than a sensitivity in the vaccinee. Thus, if an individual is a member of the significant susceptible class, and his case should be one of those rare instances in which the vaccine strain reverts to virulence, he may contract polio. This is not a case like Cud-more, then, where the “appellant belongs to a class of people not appreciable in number who are allergic to the drug,” and in which a warning would have been futile. 398 S.W.2d at 644. Rather, proper warnings here would be addressed to a sizeable group and would be designed to alert each of its members that although the incidence of vaccine-induced polio is minute, he does belong to a substantial class, an unpredictible few of which could contract polio from the vaccine. In light of such factors, the Texas “abreaction” eases cannot dilute Wyeth’s duty to warn here. This view coincides with the objective of Section 402A to protect the consumer by his standards of reasonableness in determining the manufacturer’s duty. b) Proximate Cause Having concluded that Wyeth had a duty to warn Anita Reyes, and that, dispensed without a warning, the vaccine was unreasonably dangerous, we now turn to Wyeth’s objections to the trial court’s not instructing the jury and not submitting an interrogatory on the issue of “proximate cause”. As Wyeth correctly contends, there are two causation issues in most product liability cases. The first is that of “producing cause”: Was the defendant’s product the cause-in-fact of the plaintiff’s injuries? More specific is the question posed under the rubric of proximate causation: Did the plaintiff’s injuries result from the alleged “defect” in the defendant’s product? Unless the jury is given an opportunity to pass upon both issues, Wyeth urges, the plaintiff has not made out a prima facie case of product liability. If Wyeth is correct, reversal is required, for the trial judge made it clear that he did not believe proximate cause to be an issue in this controversy, and neither instructed the jury nor submitted an interrogatory on the question. We cannot concur in Wyeth’s premise, however. The submission of interrogatories accompanying a general verdict is within the discretion of the trial court. Moreover, although an instruction on proximate cause is appropriate in most products liability cases, a jury finding on the question was not required here. As a general proposition, both Texas Courts and this Court have required that the plaintiff in a product liability suit prove not only that his injuries were caused by the defendant’s product, but that they resulted from an alleged defect in the product. Leading commentators have concurred. Yet the rule in Texas that proximate cause must be specifically proved in products liability cases is not immutable. In some circumstances, proximate cause may be inferred as a matter of law or as a matter of fact. Procter & Gamble Mfg. Co. v. Langley, Tex.Civ.App.1967, 422 S.W.2d 773, writ dismissed, want of jurisdiction. In other cases, an instruction on producing cause which omits a consideration of proximate cause may be sufficient. C. A. Hoover and Son v. O. M. Franklin Serum Co., Tex.Sup.1969, 444 S.W.2d 596, 598. We believe, in light of the holding in Hoover and crucial dicta in Technical Chemical Co. v. Jacobs, Tex.Sup.1972, 480 S.W.2d 602, that to have omitted a specific proximate cause instruction here was not reversible error. Hoover was an action against a serum manufacturer by the owners of 25 Hereford calves which had become desperately ill after having been innoculated with the defendant’s antibiotic. The trial resulted in a verdict and judgment for the plaintiffs, which was reversed and remanded by the court of civil appeals because the trial court had submitted a special issue on “producing cause” rather than “proximate cause”, “and had thus failed to require the element of foreseeability”. The Supreme Court of Texas reversed the court of appeals, and held that where, as here, the injury’s foreseeability was not in issue, the “producing cause” instruction was sufficient. Much of the Supreme Court’s brief opinion is spent in distinguishing the issue presented in Hoover from that considered in Cudmore v. Richardson-Merrell, Inc. Cudmore required that the jury consider whether the defendant’s product was the “proximate cause” of the plaintiff’s injuries rather than merely whether it was the “producing cause”, since the latter inquiry “omits entirely the element of foreseeability. . . .” The Hoover court distinguished Cud-more on the ground that it involved an “abreaction” to a “pure drug” and thus foreseeability was a crucial issue. In such a case, a “proximate cause” instruction is required. In Hbover, however, all twenty-five innoculated cattle fell ill or died. The jury found that the cause of this widespread illness was not an unforeseeable abreaction. Thus, concluded the Texas Supreme Court, foreseeability was not in issue, and no proximate cause instruction was required. The Hoover rationale is applicable to the instant controversy. Although this is, in a sense, a “pure drug” case like Cudmore, as we have pointed out above, it involves no “abreaction”. As we have also concluded earlier, foreseeability— both as to the possibility of injuries and the manner of dispensing the drug — is not really in issue. In Hoover, “foreseeability” was found by the jury; here it was established as a matter of law. Hoover was not, of course, a “failure to warn” case. The defect involved was intrinsic to the serum which felled the plaintiff’s cattle, a fact which may differentiate Hoover from extrinsic defect eases where the product and defect are separable. See Keeton, Products Liability — Problems Pertaining to Proof of Negligence, 19 Sw.L.J. 26, 33 (1965). Yet language in Technical Chemical Co. v. Jacobs, the recent Texas Supreme Court examination of failure to warn, convinces us that even in this area, a producing cause instruction will suffice in the circumstances of this case. Technical Chemical involved a suit against the manufacturer of a can of freon which exploded in the plaintiff’s hand as he attempted to put its contents into the air conditioning unit of his automobile. It was apparent from testimony adduced at trial that Jacobs, the plaintiff, inadvertently and mistakenly attached a hose from the “high” pressure side of his air conditioning compressor to the can of freon, rather than joining a hose from the “low” pressure side, which was the proper procedure. The can exploded, a possible result which the directions on the can’s label did not mention. Jacobs admitted, however, that he had not read the directions on the can’s label. In light of this fact, the jury refused to find that Technical Chemical’s failure to warn was a cause of Jacob’s injuries. In affirming the trial verdict, the Supreme Court of Texas held that it was “incumbent upon the plaintiff to secure a jury finding that the faulty labeling was a cause of the injury”. This he had not done. 480 S.W.2d at 605. Moreover, a plaintiff who had not read such directions as were provided could hardly expect the court to conclude that the manufacturer’s failure to provide further warnings caused his injury as a matter of law. Although much of the Technical Chemical opinion speaks in absolute terms of requiring proof that the defect caused the plaintiff’s injuries, it also recognizes how “speculative” testimony in this type of case is likely to be and suggests an alternative approach, a legal “presumption that an adequate warning would have beeir^ead”. 480 S.W.2d at 606. See Restatement (Second) of Torts, § 402A, comments. As the court explained: n Such a presumption works in favor of the manufacturer when\an adequate warning is present. Where there is no warning, as in this case, however, the presumption that the user would have read an adequate warning works in favor of the plaintiff user. In other words, the presumption is that Jacobs would have read an adequate warning. The presumption, may, however, be rebutted if the manufacturer comes forward with contrary evidence that the presumed fact did not exist. 480 S.W.2d at 606. Read together, Hoover and Technical Chemical suggest a test for cases such as the one now before the Court: Where a consumer, whose injury the manufacturer should have reasonably foreseen, is injured by a product sold without a required warning, a rebuttable presumption will arise that the consumer would have read any warning provided by the manufacturer, and acted so as to minimize the risks. In the absence of evidence rebutting the presumption, a jury finding that the defendant’s product was the producing cause of the plaintiff’s injury would be sufficient to hold him liable. Such a test makes sense in this case. The jury found that the defendant’s polio vaccine caused Anita Reyes’s polio. Testimony by her mother as to what she would have done, had proper warnings been provided, would have been both speculative and self-serving. Thus we turn to the Technical Chemical presumption that a warning, had it been given, would have been heeded. Buttressing the presumption that Mrs. Reyes might have taken preventive steps is the testimony of Reyes’ expert, Dr. Ramiro Casson, that some pediatricians in Hidalgo County, at least by the time of trial, had begun administering killed-virus vaccine to infants in order to build up their level of antibodies before feeding them the live-virus drug. Tending to rebut the presumption that Mrs. Reyes would have behaved differently had she been warned was the fact that she twice returned to the Mission Clinic for further doses of vaccine, even after Anita contracted polio. Yet it is patent from her testimony that Mrs. Reyes had not, even then, been informed of the danger of the polio vaccine, and did not in fact understand what medication Anita was to receive. The legal presumption suggested by the Technical Chemical opinion thus operates here to provide the final element necessary to hold Wyeth Laboratories liable for Anita Reyes’ poliomyelitis. Aware of its unavoidable dangers and cognizant that it foreseeably would not be dispensed as a prescription drug, Wyeth nonetheless failed to warn Mrs. Reyes that its vaccine could cause polio in some few of the millions receiving the medication. Administered without a warning, the vaccine was “defective”, hence unreasonably dangerous. According to the test we have distilled above, we must assume in the absence of evidence to the contrary that Anita’s parents would have acted on the warning, had it been given. Perhaps this would have prevented her polio. It unquestionably would have avoided Wyeth’s liability. III. Having concluded that Wyeth was properly held liable under substantive law, we turn to procedural matters, specifically to Wyeth’s objections to the admission or exclusion of various testimony and exhibits, and to its assertion that the evidence was insufficient to support the verdict for the plaintiffs. a) The Impeachment of Dr. Fox Among the medical experts called to testify in Wyeth’s behalf was Dr. John D. Fox, Associate Dean and Professor of Epidemiology at the University of Washington School of Public Health. On cross-examination, after Dr. Fox had testified that Type III polio vaccine was less stable than either Type I or Type II, counsel for the plaintiffs attempted to impeach his testimony by reference to an uncertified transcript of an earlier trial in which Dr. Fox had been called as a witness in Wyeth’s behalf. There Dr. Fox had testified that Type I polio vaccine was “unstable”, with reference to its “antigenic character”, and Reyes’s counsel attempted to show that this earlier testimony contradicted Dr. Fox's testimony as to Type I’s stability here. Upon further inquiry, however, Dr. Fox was apparently successful in explaining that his previous statement as to Type I’s instability dealt only with its “antigenic instability,” that is, its tendency to become an “unvaccine-like” strain, whereas his comments on the instability of the Type III virus related to its propensity to cause polio in monkeys or man. On appeal, Wyeth contends that this line of questioning was improper and requires reversal, both because the testimony of an expert cannot be impeached through the use of prior inconsistent statements and because no proper foundation was laid. We concur in Wyeth’s conclusion that the law of Texas governs the conduct of this type of impeachment. See Wright, Law of Federal Courts, § 94 at 412-15. We cannot, however, read the cases cited by the appellant as prohibiting cross-examination of expert witnesses through the use of prior inconsistent testimony. Texas does not bar impeachment of a lay witness through the use of testimony given by him at a former trial. It is difficult to understand why this impeachment technique ought not be available when the witness is an expert rather than a layman. Indeed, it might be argued that prior inconsistency is a particularly appropriate weapon for attacking expert testimony, since demonstration of the inconsistency is designed not to show that the expert has erred, but that he is capable of error. See Cirilo v. Cook Paint and Varnish Co., Tex. Civ.App.1972, 476 S.W.2d 742, 748, writ ref. n. r. e. Not surprisingly, then, Texas courts endorse far-ranging cross-examination generally, and have permitted particularly wide latitude in the interrogation of an adversary’s expert. See Davidson v. County of Harris, Tex.Civ. App.1970, 454 S.W.2d 830, 832, writ ref. n. r. e. Thus in Hutson v. State, Tex.Cr.App.1956, 164 Tex.Cr.R. 24, 296 S.W.2d 245, 247-248, the prosecution was permitted to impeach the testimony of a county sheriff as to the defendant’s intoxication, by introducing a tape recording (made shortly after the defendant’s arrest) in which the sheriff voiced an opinion contrary to that offered at trial. Also permitted in cross-examination of real estate experts has been reference to earlier inconsistent appraisals of land, the value of which is in issue. A similar standard should apply to expert medical testimony. Wyeth concedes that a cross-examiner may properly inquire whether a medical expert has previously testified on behalf of the party for whom he offers testimony. But, Wyeth urges, any further foray into prior testimonial assertions is forbidden by Texas law. We do not so read the Texas cases. To be sure, no inquiry into the correctness of opinions expressed at earlier trials is proper; but an inconsistency, if there is one, may be demonstrated. As we noted earlier, the purpose of such impeachment is not to demonstrate that the expert has erred, but that he is capable of error. An expert in medicine is not exempt from the use of prior testimony to prove his fallibility. Wyeth also urges that even if permitting the impeachment was proper, counsel for Reyes failed to lay the proper foundation for the impeachment and that therefore the court erred in allowing the cross-examination to proceed. Texas courts and this Circuit have held that to impeach a witness by showing a prior inconsistent statement, “a foundation must first be laid by asking the witness whether he made the statement, giving him an opportunity to admit or deny the statement, to correct his testimony, and to explain the apparent contradiction”. Yet it is Reyes’s attempt to lay precisely such a foundation to which Wyeth takes exception here. Dr. Fox’s attention was drawn to his testimony in Davis v. Wyeth Laboratories, Inc., 9 Cir. 1968, 399 F.2d 121, and he was given an opportunity to examine the uncertified transcript relied upon by Reyes, out of the sight of the jury and during a luncheon recess. When court reconvened, under questioning by counsel for Reyes, Dr. Fox testified that the transcript, although “somewhat inaccurate” did contain the substance of his prior testimony. He then explained at length why his former statements were in no way inconsistent with his testimony on direct examination here. Wyeth continually objected to this attempt to lay a foundation on the ground that no foundation had been laid. Overruling such objections cannot constitute reversible error. b) The Exclusion of Dr. Nakano’s Opinion Dr. James Nakano, a virologist and the Chief of the Poliovirus Strain Characterization Laboratory in Atlanta, Georgia, was allowed to testify that, based on his analysis of Anita’s stool specimens, he had concluded that the etiological agent causing her paralysis was probably a “wild strain”. The trial court, however, refused to allow Dr. Nakano, who is not a physician, to advance his opinion as to “the probable cause of Anita Reyes’s paralysis”. Asserting that this ruling constitutes reversible error, Wyeth points out that another of its experts who is also not a practicing physician was permitted to render an opinion as to the cause of Anita's condition, as were several other experts who based their conclusions on Dr. Nakano’s laboratory findings. Aside from indicating that any error in the exclusion of Dr. Nakano’s testimony was probably harmless, such arguments have little force. As we concluded in another Texas diversity action: Generally, the admission of . tendered expert testimony is a matter within the sound discretion of the trial court, and the action of the trial court in admitting or excluding such evidence will be sustained unless clearly and manifestly erroneous. Scott v. Fancher, 5 Cir. 1966, 369 F.2d 842, 844. It is at least arguable that the effect of a disease virus on a particular human being whom he had not examined was outside of the scope of Dr. Nakano’s expertise. In any event, in light of all of the testimony on the issue, the exclusion of his testimony was not so “clearly and manifestly erroneous” as to constitute reversible error. c) The Admission of Exhibit 13 Wyeth vigorously excepts to the admission into evidence of Plaintiff’s Exhibit Number 13, a document purporting to detail the suspected and confirmed cases of polio which occurred in Hidalgo County during 1970, on the ground that counsel for Reyes failed to lay the proper foundation for admission of the document under the “business records” exception to the hearsay rule. Exhibit 13 was first referred to during the cross-examination of Mrs. Ruth McDonald, Director of Nursing for the Hidalgo County Health Department. On direct examination, Mrs. McDonald testified that the County Health Department kept statistics and records on communicable diseases and immunization of the population in Hidalgo County and there her “employer” was Dr. Copenhaver, director of the Public Health Unit in Hidalgo and Cameron Counties. On cross-examination she was shown the challenged exhibit, and testified that she had seen the report previously, “when it was made up”. Mrs. McDonald also identified certain writing on the exhibit as Dr. Copenhaver’s handwriting. This tended to link Wyeth with the vaccine given Anita Reyes. The Exhibit was marked “Plaintiff’s 13” for identification, but was not offered in evidence. The following day, Oscar Garza, another defense witness, who is a Health Programs Specialist for the Texas State Department of Health, was confronted with Plaintiff’s Exhibit 13 on cross-examination. Exhibit 13 appeared to bear a strong similarity to a list prepared by Mr. Garza, although Mr. Garza’s list evidently did not show all of the same entries, and exhibited none of the hand-» writing previously identified as Dr. Copenhaver’s. Mr. Garza demonstrated a general familiarity with the information contained in Exhibit 13, but under interrogation by counsel for Wyeth, he testified that the list was not a record prepared or retained in the regular course of business of his office. Exhibit 13 was then offered in evidence by Reyes’s counsel, and Wyeth objected on the ground that a proper foundation had not been laid. After a colloquy which revealed that the challenged exhibit had been appended to Dr. Copenhaver’s deposition in accordance with a request that he submit all relevant records, the trial court decided that the record had been sufficiently authenticated to justify admission. The considerable Texas authority Wyeth marshals to support its position is of little aid, for its assumption that Texas law controls our review of the trial court’s decision is erroneous. While state standards may be found to apply to evidentiary issues in federal courts, such questions are governed by federal standards in general, and by Fed.R.Civ.P. 43(a) in particular. Bailey v. Kawasaki-Kisen, K. K., 5 Cir. 1972, 455 F.2d 392, 397. Rule 43(a) provides: (a) Form and Admissibility. In all trials the testimony of witnesses shall be taken orally in open court, unless otherwise provided by these rules. All evidence shall be admitted which is admissible under the statutes of the United States, or under the rules of evidence heretofore applied in the courts of the United States on the hearing of suits in equity, or under the rules of evidence applied in the courts of general jurisdiction of the state in which the United States court is held. In any case, the statute or rule which favors the reception of the evidence governs and the evidence shall be presented according to the most convenient method prescribed in any of the statutes or rules to which reference is herein made. The competency of a witness to testify shall be determined in like manner. We are less sanguine than the appellants in assuming that Texas courts would not admit this exhibit as a business record. We are aware also that it is at least arguably admissible under the Federal Business Records Act. Yet neither a protracted statutory analysis nor the precarious Erie ruminations characteristic of such close cases is necessary to approve the admission of Exhibit 13 here. Rather, Rule 43(a), itself, perceived through the liberal prism of our earlier decisions, furnishes the basis for Exhibit 13’s admission. Plainly enough, the Rule is a standard of admission not exclusion; all doubts are to be resolved in favor of admissibility, and even evidence not perceived as admissible when viewed through the receptive trifocals of Rule 43(a), is not necessarily to be excluded. This expansive standard was conceived in Monarch Insurance Co. v. Spach, 5 Cir. 1960, 281 F.2d 401, 411, and born in Dallas County v. Commercial Union Assurance Co., 5 Cir. 1961, 286 F.2d 388, 395, decisions which point the way to our approval of the admission of Exhibit 13 here. In Monarch Insurance, the Court faced the difficulty of interpreting Rule 43 (a) ’s sanction of evidence admissible “under the rules of evidence heretofore applied in the courts of the United States on the hearing of suits in equity”. On its face, this formula might be read as endorsing admission of only that evidence admissible under specific pre-1938 equity precedents. But the notion that in order to admit such evidence the court must pore over ancient volumes in search of pre-Federal Rules dogma was firmly interred, so far as this Court is concerned, by Monarch Insurance. Judge, now Chief Judge Brown, writing for the Court, pointed out that the district court sits as a court of equity, and the trial judge is not pinioned to pre-1938 holdings: “his chief censor is the conscience of a Chancellor.” Dallas County v. Commercial Union Assurance Co., 286 F.2d at 395. Moreover, not only did Monarch Insurance greatly expand the scope of evidence admissible within the strictures of Rule 43(a), it also signalled the advent of a still more liberal standard by suggesting that the Rule “does not purport to prohibit the admission of other relevant material probative evidence which, in the considered exercise of judicial wisdom, is trustworthy”. 281 F.2d at 411. The other shoe was dropped a year later in Dallas County v. Commercial Union Assurance Co., when we flatly held that a federal court is not required to cram proffered evidence into one of Rule 43(a)’s pigeonholes in order to find it admissible. Nor must the court find that the exhibit or testimony is within a “readily identifiable and happily tagged” exception to the hearsay rule to admit it. 286 F.2d at 398. Rather, in the words of Dallas County, if the exhibit is “necessary and trustworthy, relevant and material”, the trial judge is free to exercise his discretion and admit it, so long as he keeps the hearing within reasonable bounds. 286 F.2d at 394, 398; Price v. United States, 5 Cir. 1964, 335 F.2d 671, 677. From our review of the transcript it is plain that reception of Exhibit 13 posed no threat to the proper conduct of the trial. Neither party contests the exhibit’s materiality or relevance ; indeed the damage which the appellants allege resulted from its admission establishes those characteristics beyond any doubt. Since Exhibit 13 was “necessary and trustworthy”, it was properly admitted under the liberal Dallas County standard. Supporting this conclusion are several post -Dallas »County decisions by this Court. In Butler v. Southern Pacific Co., 5 Cir. 1970, 431 F.2d 77, 80, cert. denied, 1970, 401 U.S. 975, 91 S.Ct. 1196, 28 L.Ed.2d 325, a proffered report was held to be necessary because information in it went “to the heart of the ease” and it was trustworthy because it was in the “direct financial interest” of the maker of the report to prepare it accurately. In the other instance, Sabatino v. Curtiss National Bank, 5 Cir. 1969, 415 F.2d 632, 636, cert. denied, 1970, 396 U.S. 1057, 90 S.Ct. 750, 759, 24 L.Ed.2d 752, trustworthiness was revealed both by the financial interest of the keeper of the record in its accuracy, and the fact that it “was kept according to a regular procedure of entries made contemporaneously with the events it recorded”. There was no question of necessity; the sponsoring witness was dead. So analyzed, the report at issue here was properly admitted. The report of suspected and confirmed polio cases, inscribed with the word “Wyeth,” was crucial in linking Wyeth to the vaccine which allegedly caused Anita Reyes’s polio. Although it does not appear from the record why Dr. Copenhaver did not testify at trial, we do not believe that Reyes was required to demonstrate his unavailability in order to meet the necessity test. Cf. Butler v. Southern Pacific Co., 431 F.2d at 80. There were also sufficient indicia of trustworthiness to justify admission. Nurse McDonald testified that she had witnessed the report’s preparation, and that it was laden with Dr. Copenhaver’s handwriting. The motivation for accuracy in a record prepared by a public health official concerning outbreaks of serious disease within an area committed to his charge is obvious. A brand of expertise sufficient to insure accuracy may be assumed; there can be no suggestion that in analyzing poliomyelitis cases Hidalgo County Public Health officials were amateurish or careless. The notation as to the origin of the lot number of the vaccine ingested by Anita Reyes shares this reasonable presumption of accuracy. Exhibit 13, necessary and trustworthy, met this Court’s strictures for admission. d) The Sufficiency of the Evidence Finally, we confront the appellant’s contentions that the trial court erred in denying its motions for a directed verdict, for judgment notwithstanding the verdict, and for a new trial. One argument urged by Wyeth in support of this conclusion — -that there was no evidence that Wyeth had a duty to warn Anita Reyes — we discussed in dealing with the substantive law. Thus here we need consider only the contentions that the evidence was insufficient to support the verdict and judgment, that the verdict was the result of passion and prejudice, and that it was against the clear weight of the evidence. The yardstick against which such motions must be measured is a federal one: On motions for directed verdict and for judgment notwithstanding the verdict the Court should consider all of the evidence — not just that evidence that supports the non-mover’s case— but in the light and with all reasonable inferences most favorable to the party opposed to the motion. If the facts and inferences point so strongly and overwhelmingly in favor of one party that the Court believes that reasonable men could not arrive at a contrary verdict, granting of the motions is proper. On the other hand, if there is substantial evidence opposed to the motions, that is, evidence of such quality and weight that reasonable and fair-minded men in the exercise of impartial judgment might reach different conclusions, the motions should be denied, and the case submitted to the jury. Boeing Co. v. Shipman, 5 Cir. 1969, 411 F.2d 365, 374. As an appellate court, our sole function is to ascertain whether there is a rational basis in the record for the jury’s verdict; we are forbidden to usurp the function of the jury by weighing the conflicting evidence and inferences and then reaching our own conclusion. There was conflicting testimony as to the cause of Anita Reyes’s polio. In answer to a special interrogatory, the jury found, despite the testimony of Wyeth’s expert witnesses to the contrary, that Anita’s polio was vaccine-induced. Essentially Wyeth complains that the jury credited the testimony of the physician who treated Anita Reyes rather than evidence offered by its parade of experts. Presented with conflicting credible evidence, such a demination was within the jury’s prerogative; expert witnesses appear to assist in the court’s decision-making process, not to control it. Houston Natural Gas Corp. v. Pearce, Tex.Civ.App.1958, 311 S.W.2d 899, 909, writ ref. n. r. e.; see Steakley, Expert Medical Testimony in Texas, 1 St. Mary’s L.J. 161-162 (1969). The rule is the same in the Fifth Circuit. See Alman Brothers Farm & Seed Mill, Inc. v. Diamond Laboratories, Inc., 5 Cir. 1971, 437 F.2d 1295, 1298. The evidence for Wyeth was somewhat persuasive; there is no doubt about it. The evidence for Reyes was substantial and at least as persuasive, looking at the record coldly. Consequently, the trial court was guilty of no error in denying Wyeth’s motions for a directed verdict and judgment n. o. v. Denial of a motion for a new trial will be reversed only upon showing of clear abuse of discretion by the trial court, Davis v. Yellow Cab Co., 5 Cir. 1955, 220 F.2d 790, 791. In light of the conflicting evidence, the denial of the motion reveals no abuse here. IV. Finally, we come to a series of contentions under the general heading of challenges to comments, instructions, and charges by the trial judge. Some of these have already been considered. Remaining are Wyeth’s objections to specific instructions by the trial court, particularly: (1) its definition of the term “epidemic,” (2) its failure to define “medically cognizable risk,” and (3) its failure to instruct the jury as to the issue of “relative risk”. Federal courts apply their own rules in the manner and method of charging the jury. Where state rights are involved, state law must, of course, be looked to for the substance of the instructions, but the form of the charges and the language employed to convey the substance are procedural, hence federal, concerns. See 9 C. Wright & A. Miller, Federal Practice and Procedure, § 2255 at 651 (1971). Where as here, the issues are submitted upon a general verdict to be accompanied by answers to specific interrogatories, the Federal Rules provide simply that, “[t]he court shall give such explanation or instruction as may be necessary to enable the jury to make answers to the interrogations and to render a general verdict. . ” Not surprisingly, then, while a litigant is entitled to have the trial judge advise the jury of his theories and claims (to the extent they are supported by the evidence adduced at trial), the actual form of the instructions is within the trial court’s discretion. Counsel cannot, therefore, require that an instruction be rendered in the categorical language that he fancies would be most beneficial to his cause. To requests for use of specific verbal formulae, definitions, or synonyms a rule of reason should apply. Unadorned language comprehensible to the average juror, not legal jargon, should characterize the instructions. Words familiar to persons of average intelligence do not require definition. In reviewing the linguistic choices made by the trial court, the charge must be read as a whole. An ambiguous phrase or inaccurate clause does not necessarily destroy an entire instruction. Delancey v. Motichek Towing Service, Inc., 5 Cir. 1970, 427 F.2d 897, 901. Rather, the test is whether the charge, viewed as a whole, was sufficiently clear and accurate to be understood by jurors of ordinary intelligence, and to be applied by them to the evidence so as to reach a proper conclusion. Wyeth’s objections to the instructions here do not present serious difficulties. The appellant first contends that the issue whether there was a polio epidemic in the Rio Grande Valley at the time Anita Reyes contracted the disease should never have been presented to the jury, and, alternatively, that the trial judge’s definition of the term “epidemic” was erroneous. It is manifest that both arguments center on the definitional dispute. Wyeth produced eight expert medical and epidemiological witnesses, each of whom testified that in his opinion a polio epidemic existed in Hidalgo County, Texas in May 1970. To arrive at this conclusion, Wyeth’s experts employed an “ep