The plaintiff, Jack Werner, brought this action against the Upjohn Co. and Dr. Ralph J. Carbo, an ophthalmologist, to recover damages for injuries Werner received as result of his taking the prescription drug Cleocin which was manufactured by Upjohn and prescribed by Dr. Carbo. Jurisdiction is based on diversity of citizenship and the requisite amount in controversy.

The case was tried to a jury which found that Carbo was negligent in prescribing the drug and that his negligence proximately caused or contributed to plaintiff's injury; that Upjohn was negligent in either marketing or selling Cleocin and such negligence proximately caused or contributed to plaintiff's injury; that Upjohn was negligent in failing to warn properly of the dangerous side effects of Cleocin; that Upjohn breached either an express or implied warranty in its sale of Cleocin to Carbo; and that Upjohn was not liable in strict liability for marketing an unreasonably dangerous drug. Based on these findings the jury awarded damages of $400,000. The defendants appeal.

Upjohn argues that there is insufficient evidence on which the jury could find that Upjohn was negligent; that the jury verdicts are inconsistent; that a subsequently revised warning was improperly admitted into evidence; that several of the jury instructions were improper; and that it could not be found liable even if the warning was inadequate because Dr. Carbo admitted that he did not read the warning which was given. Dr. Carbo argues that the plaintiff failed to establish the basic elements of a malpractice case; that the jury instructions were improper; and that, since the jury found Upjohn's warning inadequate and no other evidence of malpractice was present, the jury verdict against him cannot stand. Both defendants argue that it was error to refuse a jury instruction on contributory negligence and to fail to reduce the damage award to present value. The plaintiff filed a cross-appeal arguing that the trial court erred in refusing to allow the plaintiff to present evidence on punitive damages.

We hold that the district court erred in admitting the subsequent warning into evidence. We therefore reverse and remand for a new trial.

The prescription drug at the base of this action is Cleocin (generic name Clindamycin HCl) a broad-spectrum antibiotic. The Food and Drug Administration (FDA) first approved Cleocin for general use in 1970, and it soon became a popular alternative antibiotic for those persons who were allergic to penicillin. As the use of Cleocin increased Upjohn began to receive reports of side effects from Cleocin use, such as diarrhea and colitis. The appearance of these side effects was reported to the FDA and several medical studies were done both independently and by Upjohn which sought to clarify the incidence and extent of these side effects. The plaintiff argued that Upjohn knew of serious side effects in 1974 and failed to act on this information until March 1975. Upjohn argued that its 1974 warning contained all the relevant information available at that time. In any event, as a result of the reported side effects and subsequent studies, the warning information accompanying Cleocin went through several revisions.1 Throughout this period the central concern over Cleocin was its capacity to cause pseudomembranous colitis (PMC) in some patients. Prior to 1974 the incidence of PMC in Cleocin users was thought to be quite low, but in late 1973 a Dr. Tedesco published a study which found signs of PMC in 10 per cent of a test group who took the drug. However, it was not clear that the PMC found in Tedesco's patients was the same as the more commonly known and admittedly extremely serious PMC since Dr. Tedesco's patients recovered without any permanent problems. As a result of Tedesco's study and information provided by Upjohn and others a new warning was issued in the summer of 1974. This new warning was included in the package insert accompanying Cleocin and was the subject of what is called a "Dear Doctor" letter which was mailed to every physician in the United States. The adequacy of this warning at the time it was released is one of the central issues in this case. The 1974 warning stated:

Severe and persistent diarrhea, which may be accompanied by blood and mucus, and which may be associated with changes in large bowel mucosa diagnosed as "pseudomembranous colitis," has been reported in association with the administration of Cleocin HCl (clindamycin HCl hydrate).

When significant diarrhea occurs (usually more than 5 bowel movements daily), the drug should be discontinued or, if necessary, continued only with close observation of the patient (large bowel endoscopy has been recommended). Mild cases of colitis may respond to drug discontinuance alone. Moderate to severe cases should be managed promptly with fluid, electrolyte and protein supplementation as indicated. Antiperistaltic agents opiates, meperidine, and diphenoxylate with atrophine may prolong and/or worsen the condition. Systemic corticoid and corticoid retention enemas may help relieve the colitis. Other causes of colitis should also be considered.

Note: Diarrhea has been observed to begin up to several weeks following cessation of therapy with Cleocin HCl. The physician must be alert to this possibility.

The plaintiff first visited Dr. Carbo in Maryland on December 10, 1974 for treatment of a chalazion on his eyelid. Dr. Carbo prescribed Cleocin for the plaintiff's condition and stated that he advised plaintiff that he might experience some nausea, vomiting or diarrhea, and told him that if he experienced these side effects he should stop using the drug and report to him. The plaintiff testified that Carbo gave him no such warning. Carbo's warning was based on the 1973 package insert. He admitted that he did not read the 1974 Dear Doctor letter before prescribing Cleocin to the plaintiff even though it was available in his office.

The plaintiff began to feel nauseous approximately five days after he started taking the drug and stopped using it. About a week later he reported this fact to Carbo who testified that he advised the plaintiff to discontinue the drug and call him if any further problems developed. Once again, plaintiff denied receiving this warning from Carbo. Several days later while in New York City the plaintiff developed severe diarrhea and dehydration. During this time he was taking Lomotil, an antiperistaltic agent, for the diarrhea which according to the 1974 warning (but not the 1973 warning) was contraindicated for PMC. On January 11, 1975, a Dr. Axelrod sigmoidiscoped the plaintiff and discovered the presence of PMC. He then questioned the plaintiff and determined for the first time that he had taken Cleocin. The Lomotil was discontinued but the plaintiff's condition worsened. A large portion of his colon had to be removed and he underwent other operations to restore his excretory functions. The plaintiff presently is under dietary restrictions, has diarrhea, cannot engage in strenuous athletic activity, and tires easily.

As mentioned a central issue in the case against Upjohn and an important collateral issue in the case against Dr. Carbo was the adequacy of the 1974 warning, and both the plaintiff and Upjohn presented a great deal of evidence on the true dangers and incidence of side effects of Cleocin and whether the warning adequately conveyed the facts. Upjohn argued that the incidence of side effects was quite low and that if properly and timely treated the plaintiff would not have suffered permanent injury. Upjohn also relied heavily on the argument that everything which happened to the plaintiff was warned of in the 1974 warning. The plaintiff argued that, based on Upjohn's knowledge at the time the 1974 warning was released, the warning was inadequate and made Cleocin unreasonably dangerous and Upjohn negligent. Plaintiff also argued that even if the warning were adequate it was negated by Upjohn's advertising and the activities of Upjohn's detail men.

No evidence introduced by the plaintiff was more important on the adequacy of the 1974 warning, than the evidence he introduced, over objection, of a warning published in March 1975 which expanded on the 1974 warning. The 1975 warning stated:

Clindamycin can cause severe colitis which may end fatally. Therefore, it should be reserved for serious infections where less toxic antimicrobial agents are inappropriate, as described in the INDICATIONS section. It should not be used in patients with nonbacterial infections, such as most upper respiratory tract infections. The colitis is usually characterized by severe, persistent diarrhea and severe abdominal cramps and may be associated with the passage of blood and mucus. Endoscopic examination may reveal pseudomembranous colitis.

When significant diarrhea occurs, the drug should be discontinued or, if necessary, continued only with close observation of the patient. Large bowel endoscopy has been recommended.

Antiperistaltic agents such as opiates and diphinoxylate with atrophine (Lomotil) may prolong and/or worsen the condition.

Diarrhea, colitis, and pseudomembranous colitis have been observed to begin up to several weeks following cessation of therapy with clindamycin.

Upjohn moved to exclude all reference to the 1975 warning. The court denied this motion and allowed the plaintiff to introduce the warning into evidence. When the warning was offered into evidence, over objection, the district judge instructed the jury that the warning was not to be used as evidence of negligence or culpable conduct. However, the court did not inform the jury at that time for what purpose it was to be considered. Later, after all evidence was in, the judge instructed the jury that they should only consider the 1975 warning on the issue of feasibility, obviously referring to Federal Rule of Evidence 407.

Federal Rule of Evidence 407, which enacts the common law rule excluding subsequent remedial measures to prove negligence, does, however, permit evidence of subsequent remedial measures to be used to prove the feasibility of such measures, but only if feasibility is controverted by the defendant. But, despite the judge's limiting instruction on this point it is clear that the 1975 warning was used by the plaintiff to prove negligence. The warning was mentioned several times during the course of the trial when expert witnesses were asked whether the information in the 1975 warning should have been included in the 1974 warning if that information had been available to Upjohn at the time the 1974 warning was given. This use of the warning has no conceivable connection to feasibility. The plaintiff's expert witnesses were asked time after time whether Upjohn "should have" included the information in the 1975 warning in the 1974 warning. The obvious inference is that if Upjohn should have used the subsequent warning earlier, then it was negligent for failing to do what it "should have" done. This use of the 1975 warning clearly was impermissible and constitutes reversible error. Columbia & Puget Sound R. Co. v. Hawthorne, 144 U.S. 202, 12 S. Ct. 591, 36 L. Ed. 405 (1892).

Plaintiff argues that the defendant waived any objection along this line because it failed in its objection to state specifically the grounds for objection as required by Federal Rule of Evidence 103(a)(1). Rule 103(a)(1), however, requires specific objection only where the specific ground would not be clear from the context. In the case at bar, Upjohn, for example, had already filed a pre-trial motion with supporting memoranda asking that all references to the 1975 warning be suppressed. From our examination of the record, we have no doubt that the ground for objection was clear to everyone.

Plaintiff also made improper use of the 1975 warning in closing argument to the jury. Plaintiff's counsel stated:

If we were to boil this down to the least common denominator, if you were to be asked tomorrow, after, hopefully, the case is decided and done with, what was the biggest single thing that Upjohn did wrong that impells (sic) you to bring in a verdict for the plaintiff against Upjohn, would you not have to say Upjohn knew of the fatalities and didn't tell the doctors until much later, after this whole episode was over and done? That they knew it should not be used for other than serious, life-threatening infections? That they knew there were other less toxic antibiotics which were appropriate, and they told people that later? That they knew that they had, according to their people, cut off promotion of the drug in the middle of '74, but didn't tell the doctors that? Don't you think, honestly, and you have to answer this to yourselves, don't you think any one of those four things was a violation of a duty which any ethical drug manufacturer owes the public, to keep the public informed, or keep the doctors informed so the public can be informed?

Instead of that, they held it under their belt; they continued the advertising for at least four more months, in many instances. They held off the change until March of '75; (the date the 1975 warning was released) nine months after they knew this, and they cashed in as much as they could.

But, in this delay, from July 1 to March of '75, before they really cut and told the doctors what they should have told them long ago, in that nine month period, . . .

The clear import of this argument is that Upjohn breached its duty of due care by failing to adopt the 1975 warning in 1974. Thus, plaintiff used the substance of the subsequent warning to prove antecedent negligence and thereby violated Rule 407. The defendant failed to object to the closing argument quoted above, but such an objection is not required in this circuit. Leathers v. General Motors Corp., 546 F.2d 1083 (4th Cir. 1976). In Leathers during closing argument plaintiff's counsel had invoked the "Golden Rule" argument which in essence, asked the jury to decide the case as they would want it decided if they were the plaintiff. Despite lack of objection to the argument we reversed and remanded for a new trial stating that the defendant was not required to make a contemporaneous objection because to do so might only have emphasized the impermissible point and have antagonized the jury. The error in closing argument in the case at bar was at least as damaging as the argument in Leathers. Here, plaintiff was permitted to use evidence of subsequent remedial measures to argue negligence by Upjohn. Such argument is strictly forbidden by the policy-based Rule 407. If the Golden Rule argument is sufficiently out of bounds to excuse the contemporaneous objection requirement then this clear violation of Rule 407 is no less so.

Plaintiff also argues that since an instruction was given limiting the use of the 1975 warning to the issue of feasibility that any error caused by plaintiff's departure from this rule is harmless because it was cured by the limiting instruction. This argument is without merit. In the abstract, of course, evidence, may be admissible for one purpose and inadmissible for another. The usual solution for this problem is to admit the evidence to prove the permissible inference, but limit its use by jury instructions. This does not mean, however, that the party offering evidence for a limited purpose is free to use it for a forbidden purpose, over objection, during the course of the trial and insulate reversal by pointing to a limiting instruction given at the close of the case.

Plaintiff's position is further weakened by the fact that the feasibility of giving the 1975 warning in 1974 simply was not in issue in the case. As stated above, Rule 407 does allow evidence of subsequent remedial measures to be used to prove the feasibility of such measures, but only if feasibility is controverted by the defendant. The Advisory Committee Notes to Rule 407 state in terms that "unless a genuine issue" of feasibility is present the exception should not apply. The plaintiff contends that the issue of feasibility was controverted sufficiently to invoke the exception. His first argument is that no concession was made by defendant on this issue. However, it is clear from the face of the rule that an affirmative concession is not required. Rather, feasibility is not in issue unless controverted by the defendant.

In an effort to show a controverted issue of feasibility the plaintiff points to two statements Upjohn made to the FDA during negotiations on what was to be the exact content of the 1974 warning in which Upjohn purportedly argued that a stronger warning would be "misleading" or would have a "potential for much harm." When these statements are examined in context it is apparent that plaintiff's argument is without merit. The first statement was made in reply to the FDA's suggestion that the phrase "and more rarely in PMC" be added to the 1974 warning. Upjohn's reply suggested a warning which stated: "Antibiotic associated 'PMC' can occur." In rejecting use of the word "rarely" Upjohn noted that Dr. Tedesco's study indicated that incidence was higher than "rarely" and that in general Upjohn preferred to avoid terms such as "rarely" or "infrequently", "since there is a distinct possibility of misleading." Thus, it appears that Upjohn's proposed exclusion of the word "rarely" strengthened, rather than weakened, the FDA's proposed warning, and certainly had nothing to do with controverting the feasibility of a stronger warning.

The second example cited by plaintiff arose out of an FDA request that the 1974 package insert include a warning that Cleocin not be used for minor infections. Such a warning ultimately was adopted in substance in the 1975 warning. In objecting to the FDA's proposal Upjohn stated that "there is no precise definition of minor or trivial infections and the physician is adequately guided by the limitation of the approved INDICATIONS. Such a statement has a potential for much harm and little or no benefit." The above quoted statement does not controvert the feasibility of writing a warning that Cleocin not be used for minor infections. It merely questions the necessity of such a warning. This distinction points up the basic position of Upjohn in this case that the 1974 warning was adequate. Upjohn does not argue that it could not have written a stronger warning, it argues that the 1974 warning was adequate given the knowledge it had at that time. This defense simply does not raise an issue of feasibility.

Thus, plaintiff's efforts to expand the feasibility exception must fail. Since feasibility was not in issue in the case the evidence should have been excluded. See Advisory Committee Notes to Rule 407. Rule 407 is designed to protect the important policy of encouraging defendants to repair and improve their products and premises without the fear that such actions will be used later against them in a lawsuit. Several exceptions to the rule have developed, but it is clear that they must be narrowly construed if the central policy behind the rule is to be effectuated. If feasibility were to be found at issue in the case at bar, it is difficult to imagine a situation where it would not at least arguably be in issue. As Professor McCormick has said:

(T)he extrinsic policy of encouraging remedial safety measures is the predominant reason for holding evidence of these measures to be privileged. It is apparent that the free admission of such evidence for purposes other than as admissions of negligence is likely to defeat this paramount policy. It is submitted that before admitting the evidence for any of these other purposes, the court should be satisfied that the issue on which it is offered is of substantial importance and is actually, and not merely formally in dispute, that the plaintiff cannot establish the fact to be inferred conveniently by other proof, and consequently that the need for the evidence outweighs the danger of its misuse.

McCormick on Evidence, § 275, at 668-69 (2d ed. 1972). In sum, feasibility is not in issue in this case, and even if it were, it is clear that the use made by the plaintiff of the 1975 warning was not so limited.

Alternately, plaintiff argues that even if the 1975 warning was not admissible to show feasibility it was admissible on other grounds. It is arguable that this question is not directly before us on appeal because the 1975 warning was in fact used to prove negligence, and thus the fact that the evidence might have been admissible on some other ground not raised at trial should not prevent reversal and a new trial. Especially because this case must be tried again, and we do not believe that the 1975 warning was admissible, we examine plaintiff's other arguments for admitting the evidence.

Plaintiff argues, and we agree, that the exceptions listed in Rule 407 ownership, control or feasibility of precautionary measures (if controverted), and impeachment, are illustrative and not exhaustive. See Advisory Committee Notes to Rule 407. However, we note once again that Rule 407 promotes an important policy of encouraging subsequent remedial measures. If this policy is to be effectuated we should not be too quick to read new exceptions into the rule because by so doing there is a danger of subverting the policy underlying the rule.

Plaintiff's principal argument along this line is that the 1975 warning is admissible because the suit includes a strict liability claim. The courts have split on this proposition. We suppose, without making a numerical comparison, that the majority of courts would admit the evidence for one reason or another. Compare e. g., Smith v. E. R. Squibb & Sons, Inc., 405 Mich. 79, 273 N.W.2d 476 (1979); Price v. Buckingham Mfg. Co., 110 N.J.Super. 462, 266 A.2d 140 (1970); with Robbins v. Farmers Union Grain Terminal Ass'n., 552 F.2d 788 (8th Cir. 1977); Ault v. Int'l Harvester, 13 Cal. 3d 113, 117 Cal. Rptr. 812, 528 P.2d 1148 (1974).

In support of his argument to admit the evidence plaintiff first argues that Rule 407 in terms only bars the evidence to prove negligence or culpable conduct and that the rule therefore does not allow for exclusion on the strict liability issue. We disagree. The mere fact that Rule 407 by its terms only excludes evidence of subsequent precautionary measures to prove negligence or culpable conduct does not necessarily mean that the evidence should be admissible to prove strict liability.2 It is clear that in enacting the Federal Rules of Evidence Congress did not intend to wipe out the years of common law development in the field of evidence, indeed the contrary is true. The new rules contain many gaps and omissions and in order to answer these unresolved questions courts certainly should rely on common law precedent. See Redden & Saltzburg, Federal Rules of Evidence Manual 411-413 (1975). This is true with respect to Rule 407 which merely enacts the common law rule. Rule 407 bars evidence of subsequent precautionary measures to prove negligence or culpable conduct with exceptions for control, ownership or feasibility (if controverted) and impeachment. The rule simply does not speak in terms to the question of whether the evidence should come in to prove strict liability. To resolve this question we must examine the policy behind Rule 407 and the common law basis for the rule, and then determine if admitting the evidence as evidence of strict liability is more akin to the use of it to prove negligence, or if it is closer to one of the recognized exceptions to the rule. That is to say, would the policy behind the common law rule be served or subverted if evidence of subsequent precautionary measures be admitted to prove strict liability.

We first note that Rule 407 excludes evidence of subsequent precautionary measures to prove negligence or culpable conduct. Culpable conduct normally involves something more than simple negligence and implies conduct which is "blamable; censurable; involving the breach of a legal duty or the commission of a fault. . . . (I)t implies that the act or conduct spoken of is reprehensible or wrong, but not that it involves malice or a guilty purpose." Black's Law Dictionary (4th ed. 1968). Thus, Congress has determined that such evidence should be excluded not only in cases involving negligence but where the defendant is charged with culpable conduct. Strict liability on the other hand involves conduct which is technically less blameworthy than simple negligence, since the plaintiff need not prove a breach of duty by the defendant other than placing the product on the market. From a policy standpoint it follows that if the rule expressly excludes evidence of subsequent repairs to prove culpable conduct that the same should be true for strict liability. Stated another way, if the common law and Congress were willing to exclude the evidence on the issue of culpable conduct, the result should be no different on policy grounds as long as strict liability is not distinguishable on some other ground.

Plaintiff argues that a fundamental distinction exists between negligence and strict liability since in a negligence action it is the reasonableness of the defendant's conduct which is in issue while in a strict liability case the issue is whether the product is unreasonably dangerous. Thus, the argument goes, in a negligence action the focus is on the defendant while in strict liability the focus is on the product. We concede the obvious distinction between negligence and strict liability, but we do not believe that this distinction should produce a different result. The rationale behind Rule 407 is that people in general would be less likely to take subsequent remedial measures if their repairs or improvements would be used against them in a lawsuit arising out of a prior accident. By excluding this evidence defendants are encouraged to make such improvements. It is difficult to understand why this policy should apply any differently where the complaint is based on strict liability as well as negligence. From a defendant's point of view it is the fact that the evidence may be used against him which will inhibit subsequent repairs or improvement. It makes no difference to the defendant on what theory the evidence is admitted; his inclination to make subsequent improvements will be similarly repressed. The reasoning behind this asserted distinction we believe to be hypertechnical, for the suit is against the manufacturer, not against the product.

Of course, exceptions must arise where the defendant attempts to make offensive use of the exclusion of this evidence. Thus, if the defendant denies ownership or control, contends that no such repair or improvement was possible, or makes statements conflicting with the fact of repair, then the plaintiff should be allowed to make use of subsequent remedial measures. As previously noted, the list of exceptions in Rule 407 is illustrative, not exhaustive, but each of the listed exceptions deals with situations where the defendant might gain a direct benefit over and above the fact of exclusion and it seems to us that new exceptions to the rule should follow this rationale if the policy behind the rule is to be protected.

The defendant argues that we should adopt the rule of some courts, that the rule against admitting evidence of subsequent remedial measures not apply in product liability cases. See Ault v. Int'l Harvester, 13 Cal. 3d 113, 117 Cal. Rptr. 812, 528 P.2d 1148 (1974). We are not persuaded. The court assumes that the product is defective, and thus overlooks the situation where the product is not defective but could be made better. The manufacturer who undertakes precautionary measures in this setting will face the risk of liability for an injury caused by the earlier nondefective version of the product based on evidence of his subsequent act which made the product safer but in no way supports an inference that the initial version of the product was defective. Furthermore, if the rationale of Ault were applied in federal courts it might effectively override Rule 407 since its reasoning that evidence of subsequent precautionary measures is admissible in product liability cases might apply with equal force to a negligence cause of action. However, Congress, in enacting Rule 407, has chosen to exclude such evidence as proof of negligence and the Ault court offers no basis for distinguishing between the two theories. Smyth v. Upjohn Company, 529 F.2d 803 (2nd Cir. 1975) rejects the holding of Ault as do we.3

Our conclusion is supported by the close similarity between negligence and strict liability. The elements of both are the same with the exception that in negligence plaintiff must show a breach of a duty of due care by defendant while in strict liability plaintiff must show the product was unreasonably dangerous. The distinction between the two lessens considerably in failure to warn cases since it is clear that strict liability adds little in warning cases. Under a negligence theory the issue is whether the defendant exercised due care in formulating and updating the warning, while under a strict liability theory the issue is whether the lack of a proper warning made the product unreasonably dangerous. Though phrased differently the issue under either theory is essentially the same: was the warning adequate? Chambers v. G. D. Searle & Co., 441 F. Supp. 377, 381 (D.Md.1975), aff'd 567 F.2d 269 (4th Cir. 1977); Smith v. E. R. Squibb & Sons, Inc., 405 Mich. 79, 273 N.W.2d 476, 480 (1979); Prosser, Torts § 99, esp. p. 659 n. 73 (4th ed. 1971).

Any remaining distinction in theories disappears when a failure to warn case involves an unavoidably dangerous drug which the product in this case admittedly was. The Restatement of Torts (2d) § 402A, comment k makes it clear that a drug manufacturer is not to be held strictly liable for injuries caused by an unavoidably dangerous new drug if the warning is adequate.4 The standard for liability under strict liability and negligence is essentially the same. Chambers, supra at 381.

Since the issue under either theory is the same, a decision which admits evidence of subsequent remedial measures on a strict liability theory in a failure to warn case involving an unavoidably dangerous drug would promote substance over form and subvert the policy behind excluding evidence of subsequent remedial measures. We decline the invitation to follow the other courts which may have so held, and prefer the reasoning of Chambers, exactly to the point here, and Smith, very nearly so.

The plaintiff also argues that the 1975 warning is admissible to prove breach of an express warranty, to prove causation, or to show that plaintiff was entitled to punitive damages. Once again, even were we to decide any of these points in plaintiff's favor, a new trial would be required because the evidence was used to prove negligence. In any event, we hold, for the same reasons stated in our refusal to allow the 1975 warning to come in to prove strict liability, that the policy behind Rule 407 does not allow for a new exception to prove causation, breach of express warranty or entitlement to punitive damages on the facts of this case.

Plaintiff's final argument on the admissibility of the 1975 warning is that it was required by the FDA, and that since defendant Upjohn had no choice in the matter, the policy behind Rule 407 would not be served by excluding the evidence. While this argument may have a surface plausibility, we do not believe the evidence should be admitted for this reason. First, when a third party has required the change the relevance of the evidence is one more step removed from the central issue in the case, the adequacy of the warning at the time the product was marketed. When a third party makes a subsequent change the inference is removed not only in time but in who made the change. This is especially true in the case at bar since the FDA itself approved the 1974 warning at the time it was first used. Thus, it would appear that any separate relevance of the FDA's actions with regard to the 1975 warning is overridden by the fact that the FDA approved the 1974 warning as well.

Furthermore, plaintiff's argument overlooks the dual responsibility for preparing warnings for prescription drugs. The FDA can require that a warning be changed, but this does not tell the whole story. In addition to its broad regulatory sanctions it is clear that it also relies on voluntary compliance and compromise in determining the content of warnings and advertising for prescription drugs. See Dixon, Drug Product Liability § 5.05. This cooperative aspect of prescription drug regulation is apparent in the correspondence between Upjohn and FDA in the case at bar in arriving at warnings and their changes. Furthermore, FDA regulations allow drug companies to change the labels and warnings accompanying their products without prior FDA approval where new side effects are discovered by the company. 21 C.F.R. § 314.8(d)-(e). Subsection (d) states in pertinent part:

(d) Changes of the following kinds proposed in supplemental new-drug applications should be placed into effect at the earliest possible time:

(1) The addition to package labeling, promotional labeling, and prescription drug advertising of additional warning, contraindication, side-effect, and precaution information.

(e) It will be the policy of the Food and Drug Administration to take no action against a drug or applicant solely because changes of the kinds described in paragraph (d) of this section are placed in effect by the applicant prior to his receipt of a written notice of approval of the supplemental new-drug application: Provided, That all the following conditions are met:

(1) The supplemental new-drug application providing a full explanation of the basis for the changes has been submitted, plainly marked on the mailing cover and on the supplement "Special new-drug application supplement Changes being effected."

(2) The applicant specifically informs the Food and Drug Administration of the date on which such changes are being effected, and submits to the Administration 12 printed copies of any revised labeling to be placed in use, identified with the new-drug application number.

(3) All promotional labeling and all drug advertising are promptly revised consistent with the changes made in the labeling on or within the drug package.

The quoted regulations demonstrate that plaintiff's argument is without merit. If subsequent warnings are admitted to prove antecedent negligence simply because FDA required or might have required the change, then drug companies may be discouraged from taking early action on their own and from participating fully in voluntary compliance procedures. Thus, we think that the FDA's regulatory power should not be read as in conflict with the protective policy of Rule 407. The FDA's regulations and policies encourage early unilateral action by the drug companies to improve their warnings, and Rule 407 promotes the same goal. We therefore hold that FDA regulations in the area of drug labeling do not require a new exception to Rule 407. Indeed, to find such an exception would subvert the policy goals of both Rule 407 and FDA Regulations 314.8(d)-(e).

One further point deserves specific mention for it serves as an alternate ground for a new trial for Upjohn. It is claimed that the verdict for Upjohn on the issue of strict liability is inconsistent with the verdict against Upjohn on the issue of negligence. We agree. The effect of the jury verdict on negligence was to find that Upjohn failed to use due care to give an adequate warning of the propensities of the drug marketed, and, in the same breath, the verdict on strict liability found that the drug marketed with such an inadequate warning was not unreasonably dangerous. The verdicts in the context of this failure to warn case involving prescription drugs are obviously inconsistent and cannot stand. The same may be said of the verdict against Upjohn so far as it may be upon an implied warranty, for the only evidence that the drug was not reasonably fit concerned the warning.

There are other claims of error, many of which concern the sufficiency of evidence, including a cross appeal claiming error in refusing to submit the question of punitive damages to the jury. We express no opinion on any of these, since the case must be tried again and the questions now before us may not come up in a new trial or be presented in a different form upon somewhat different facts. We are not obligated in each case to pass upon claims concerning sufficiency of evidence, Stonehocker v. General Motors, 587 F.2d 151 (4th Cir. 1978), and we think this is a case in which we should not do so.

We also vacate the judgment against Dr. Carbo and remand for a new trial. It is true that most of our discussion has concerned the adequacy of the warning given by Upjohn, but much, or even most, of the trial was devoted to that subject. There is no doubt the whole trial centered about that question. Dr. Carbo's duty to his patient was so bound up in the warnings given by Upjohn, that we do not believe a judgment against him should be allowed to stand when the record contains error going to the most fundamental question in the case: the adequacy of the warning. The probability of prejudice is too great. Under 28 U.S.C. § 2106, we think the judgment against Dr. Carbo must also be vacated and remanded for a new trial. United States v. Caudle, et al, 606 F.2d 451 (4th Cir. 1979).

The judgment appealed from must be vacated and the case remanded for a new trial.

VACATED and REMANDED.