Sterling Drug, Inc. and its division, Winthrop Laboratories, appeal from a judgment in favor of plaintiff Leota A. DeLuryea in the amount of $80,000. Mrs. DeLuryea brought suit against defendants (Sterling) for damages suffered as a result of drug dependence and tissue damage caused by the drug Talwin manufactured by Sterling. The case was submitted to the jury on theories of strict liability, negligence, and breach of express warranty. Sterling claims that the trial court erred in refusing to direct a verdict against DeLuryea for failure to establish causation; in refusing to admit the deposition of Dr. Ivie, a deceased treating physician; in admitting evidence of a change in warnings concerning Talwin’s side effects which Sterling published after DeLuryea stopped taking Talwin; in admitting certain statements allegedly made by Sterling’s detail men; and in submitting the issue of breach of express warranty. DeLuryea cross-appeals claiming error in the trial court’s refusal to submit the issue of punitive damages. Because we find the exclusion of Dr. Ivie’s testimony to be in error, we reverse.
DeLuryea’s ponytail was caught in the shaft of some machinery in the shoe factory in which she was working on January 3, 1959, and her scalp, forehead, upper eyelids and the area around her ears were torn off completely. She had surgery on several occasions and the use of painkilling medication was required. In the summer of *2241968, DeLuryea’s treating physician, Dr. McCarthy DeMere, became concerned that she had become dependent upon her principal painkilling medication, Demerol. DeLuryea was hospitalized on August 28, 1968 and Dr. Charles Cooke was called in for consultation. Dr. Clarke prescribed Talwin in lieu of Demerol. After the initial prescription by Dr. Clarke, DeLuryea obtained prescriptions for Talwin from Drs. Clopton and Futrell, two general practitioners in northeast Arkansas, and self-administered the drug in injectible form until June 3, 1974.
In the summer of 1970, Dr. DeMere became suspicious that DeLuryea was developing a dependence on Talwin and asked Dr. Mark Ivie, a psychiatrist, to see her.
In September of 1970, Dr. DeMere excised a small lesion on DeLuryea’s thigh that was thought to have been caused by an injection, but this was not directly traced to Talwin. On January 8,1971, DeLuryea had a sinus tract on her buttock that Dr. DeMere determined was caused by Talwin. Dr. DeMere advised DeLuryea that she was sensitive to Talwin and she should not take it anymore. He did not learn until 1974 that she had continued taking Talwin long after this conversation.
At trial, DeLuryea described how she had learned to administer injections of medication when her father was ill in 1949 and knew that shots were to be rotated intramuscularly. Dr. Clopton gave her further instruction as to the injection of Talwin in 1968 and the locations to use in her arms, hips, and thighs and to rotate the site of injections. Both oral Demerol and oral Talwin nauseated her. DeLuryea testified that she was certain that no one had told her not to take Talwin before 1974.
DeLuryea developed severe ulcerations on her left thigh with necrotic tissue caused by injections of Talwin over a period of time. Later she developed tissue necrosis of her other thigh and her hips, shoulders, and upper arms. Sterling stipulated that DeLuryea’s tissue problems were caused by her use of Talwin. Insofar as the evidence bears upon the issues raised on appeal, it will be discussed with reference to each such issue.
I.
Sterling claims that DeLuryea failed to make a submissible case because no doctor who prescribed Talwin for DeLuryea testified that the warnings were inadequate or that he would have acted differently if different warnings had been given. Thus, Sterling argues there was a complete absence of proof of causal connection between the warnings and the damage suffered by DeLuryea.
None of the doctors prescribing Talwin testified. One was deceased at the time of trial.
DeLuryea produced testimony from Dr. McCarthy DeMere that the warnings given prior to May 1974 regarding tissue damage at injection sites were inadequate to warn the medical profession of the danger involved. He stated that the manufacturer knew in 1967, prior to marketing, that Talwin would cause tissue damage after prolonged use. An adequate warning could have been given at that time. Dr. Richard McShane testified that Sterling’s premarketing animal studies revealed tissue damage similar to that found in humans. This, however, was not publicized until May 1974. He agreed with Dr. DeMere that the warnings prior to May 1974 were inadequate to warn of damage at injection sites.
Information concerning not only Talwin but other prescription drugs is contained in a package insert accompanying each drug. The package insert is updated as further information on the drug becomes available. The Physicians’ Desk Reference, which in the case of Talwin contained the same information as the package insert, is furnished to all physicians without charge by its publisher, Medical Economics.
Under these circumstances, with evidence of inadequate warnings, and that the warnings were placed in the hands of all physicians, which would include the three prescribing physicians, we conclude that DeLuryea’s evidence established a submissible is*225sue on negligence and proximate cause. We must, of course, consider the evidence in the light most favorable to DeLuryea. De Witt v. Brown, 669 F.2d 516, 512 (8th Cir.1982); Garoogian v. Medlock, 592 F.2d 997, 999 n. 3 (8th Cir.1979).
This court in Sterling Drug, Inc. v. Cornish, 370 F.2d 82 (8th Cir.1966), discussed the importance of warning prescribing physicians as follows:
[T]he purchaser’s doctor is a learned intermediary between the purchaser and the manufacturer. If the doctor is properly warned of the possibility of a side effect in some patients, and is advised of the symptoms normally accompanying the side effect, there is an excellent chance that injury to the patient can be avoided. This is particularly true if the injury takes place slowly, as is the case with the injury in question here.
370 F.2d at 85.
The defendant drug manufacturer in Cornish alleged that plaintiff’s doctors negligently failed to keep up with medical literature and that that negligence was an intervening proximate cause of plaintiff’s injuries. This court answered that argument by stating:
There is no question of intervening proximate cause [from the doctors failing to keep up with medical literature] in this case. The sole issue was whether appellant negligently failed to make reasonable efforts to warn appellee’s doctors. If appellant did so fail, it is liable regardless of anything the doctors may or may not have done....
370 F.2d at 85. See also, Lindsay v. Ortho Pharmaceutical Corp., 481 F.Supp. 314 (E.D.N.Y.) 1979, rev’d on other grounds, 637 F.2d 87 (2d Cir.1980); Bine v. Sterling Drug, Inc., 422 S.W.2d 623 (Mo.1970).
Under Cornish we conclude that failure to produce testimony of the prescribing doctors was not fatal to DeLuryea’s case on the issue of proximate cause. In Schenebeck v. Sterling Drug, Inc., 423 F.2d 919, 923 (8th Cir.1970), we held that one of several permissible inferences was that the doctors would have heeded the warning and altered the course of treatment had an adequate warning been given. Here there was evidence that the warnings were inadequate, that DeLuryea continued using the medication, and that the usage produced the tissue damage, all subjects of the court’s reasoning in Schenebeck. As in Schenebeck:
[W]e think the plaintiff presented sufficient evidence for the jury to find that Sterling’s failure to timely warn constituted an omission on its part which operated through a natural sequence of events to proximately cause or contribute to plaintiff’s harm. See Parke-Davis and Co. v. Stromsodt, 411 F.2d 1390 (8th Cir.1969).
423 F.2d at 923.
Sterling relies upon Douglas v. Bussabarger, 73 Wash.2d 476, 438 P.2d 829 (Wash. 1968) (en banc); Oppenheimer v. Sterling Drug, Inc., 7 Ohio App.2d 103, 219 N.E.2d 54 (1964); Chambers v. G.D. Searle & Co., 441 F.Supp. 377, (D.Md.1975), aff’d, 567 F.2d 269 (4th Cir.1977) (per curiam); and Brochu v. Ortho Pharmaceutical Corp., 642 F.2d 652 (1st Cir.1981).
The prescribing physician in Douglas specifically stated that he did not read the allegedly inadequate warnings, rather he relied on his own medical knowledge. The court concluded that any negligence in failing to warn was not the proximate cause of plaintiff’s disability. In Oppenheimer the prescribing physician did not specifically recall reading the precautions in question. The doctor stated that he relied on his own experience in prescribing the drug. The court held that the doctor’s failure to rely on the warnings was sufficient to defeat plaintiff’s claim. In Chambers the prescribing physician testified that he had no knowledge of plaintiff’s predisposition toward stroke and therefore would not have altered his prescription had there been an adequate warning about use of the drug in question by those predisposed to stroke. The court concluded that plaintiff had not furnished sufficient proof of causation.
Sterling argues that plaintiff was obligated to put on evidence similar to that in *226Brochu in which the court upheld a jury-verdict for plaintiff. We do not read Brochu as creating a pattern of evidence that is necessary for plaintiff to make a submissible case.
The specific testimony of the doctors in each of the cases relied upon by Sterling distinguishes those cases from this case in which the prescribing physicians did not testify. Chambers, 441 F.Supp. at 385; McEwen v. Ortho Pharmaceutical Corp., 270 Or. 375, 528 P.2d 522, 526 (Or.1974). The issue of proximate cause was properly submitted to the jury. While we have discussed the question in light of negligence cases, proximate cause in the strict liability submission in this case involves similar considerations.
II.
Sterling claims that the trial court erred in excluding the deposition of Dr. Mark Ivie, a psychiatrist who had treated DeLuryea in 1970 and 1971 and who was dead at the time of trial.
Dr. Ivie’s deposition was taken in DeLuryea’s workers’ compensation claim arising out of her injury in 1959. In the deposition, Dr. Ivie described DeLuryea's abuse of various painkillers from minor analgesics to narcotics. He stated that he took her off Talwin and asked her to leave Talwin alone because she was abusing it tremendously. He was not, however, optimistic that she would be able to stay away from Talwin.
The trial court granted DeLuryea’s pretrial motion in limine with respect to Dr. Ivie’s deposition and adhered to its ruling when Sterling offered the deposition at trial.
The question is whether the deposition was admissible under Rule 804(b)(1) which provides as follows:
(b) Hearsay exceptions. The following are not excluded by the hearsay rule if the declarant is unavailable as a witness:
(1) Former testimony. Testimony given as a witness at another hearing of the same or a different proceeding, or in a deposition taken in compliance with law in the course of the same or another proceeding, if the party against whom the testimony is now offered, or, in a civil action or proceeding, a predecessor in interest, had an opportunity and similar motive to develop the testimony by direct, cross, or redirect examination.
Dr. Ivie’s death satisfies the requirement that the declarant be unavailable. Fed.R. Evid. 804(a)(4).
DeLuryea argues that the deposition is not admissible because she did not have an opportunity and similar motive at the workers’ compensation hearing to develop the testimony by direct, cross or redirect examination. We do not agree.
The deposition of Dr. Ivie was taken in the workers’ compensation claim arising out of the original injuries suffered in 1959. The record demonstrates that an issue before the commission was whether certain of DeLuryea’s injuries were “the result of [her] own misconduct” and thus not job related. Respondents in the workers’ compensation action based their allegation of misconduct on the fact that Dr. Ivie had taken DeLuryea off Talwin because she was abusing it and that he had “asked her to leave Talwin totally alone.”
DeLuryea’s counsel in the worker’s compensation case asked only one question in cross-examining Dr. Ivie, whether DeLuryea’s problem with painkillers was related to the industrial accident, and Dr. Ivie responded affirmatively. DeLuryea argues that cross-examination was so limited because the sole issue before the commission was whether her condition arose out of an injury which was job related. Therefore, she argues, the cross-examination was not meaningful in the context of the subsequent products liability action. This argument ignores the fact that Dr. Ivie’s testimony related to an issue relevant to both actions, whether DeLuryea abused drugs. DeLuryea had a similar motive in the two actions in disproving the allegations of misconduct. Under these circumstances, we conclude that the purpose for which Dr. Ivie’s testimony was offered at the workers’ compensation hearing was such that DeLu*227ryea had a similar motive for testing the credibility of the testimony on cross-examination.
DeLuryea points out that she was represented by different counsel at the workers’ compensation hearing. Representation by the same counsel at both proceedings, however, is not required. United States v. Amaya, 533 F.2d 188, 191-92 (5th Cir.1976), cert. denied, 429 U.S. 1101, 97 S.Ct. 1125, 51 L.Ed.2d 551 (1977).
Furthermore, DeLuryea’s counsel’s decision to limit cross-examination in the workers’ compensation hearing does not bar use of the former testimony even though DeLuryea might later have desired fuller cross-examination. Opportunity and motivation to cross-examine are the important factors, not the actual extent of cross-examination. No unfairness is apparent in requiring DeLuryea to accept her prior cross-examination in that she was the party against whom the testimony was offered in both proceedings. Fed.R.Evid. 804(b)(1) advisory committee note.
This court has observed that:
As a general rule, depositions taken in a prior action are admissible in a subsequent action if there is substantial identity of issues and parties in the two actions.
Rule v. International Ass’n of Bridge, Structural and Ornamental Ironworkers, Local No. 396, 568 F.2d 558, 568 (8th Cir.1977). As we have shown, there is substantial identity of issues and parties in the two proceedings in question.
The district court also cited Fed.R. Evid. 4031 when it excluded Dr. Ivie’s deposition apparently on the ground that the deposition would inject the matter of workers’ compensation benefits into the case. The deposition, however, could have been identified simply as a deposition taken under oath during the lifetime of Dr. Ivie, and there would have been no occasion to mention workers’ compensation in any way.
The deposition contained direct impeachment of DeLuryea’s testimony that no one had told her to discontinue the use of Talwin prior to 1974, and specifically that Dr. Ivie had not so instructed her. DeLuryea’s abuse of Talwin was a significant issue in this case, and from a study of the entire record it is apparent that Sterling was substantially prejudiced by exclusion of this testimony. The district court abused its discretion in excluding the deposition.
The district court, however, properly excluded certain letters from Dr. Ivie addressed to the insurance carrier in the workers’ compensation proceeding. Sterling’s contention that the letters were admissible under Fed.R.Evid. 803(1) and 803(4) is without merit.
III.
Sterling asserts that the trial court erred in admitting evidence of a change in the wording of the package insert which was made after June 3, 1974, the last date on which DeLuryea used Talwin. Sterling contends that the change was irrelevant and thus inadmissible under Fed.R.Evid. 4022 and furthermore was a subsequent remedial measure inadmissible under Fed.R. Evid. 407.3
The package insert available in June of 1974 contained a warning that tissue damage at injection sites had occurred and that *228“rotation of injection sites is, therefore, recommended.” DeLuryea introduced evidence that Sterling, in its October, 1975 package insert, changed the statement that rotation of injection sites was “recommended” to read that such rotation was “essential.”
DeLuryea argues that the evidence was clearly relevant to the issue of adequacy of the prior warning and that Rule 407 is inapplicable to actions based on strict liability. She further argues, somewhat inconsistently, that even if Rule 407 precludes admission of subsequent changes in strict liability actions, the admission herein would not be error because the change in wording was not significant.
This court has determined that the rule against admitting evidence of subsequent remedial measures does not apply to actions based on strict liability. Robbins v. Farmers Union Grain Terminal Association, 552 F.2d 788 (8th Cir.1977); Farner v. Paccar, Inc., 562 F.2d 518 (8th Cir.1977); Unterburger v. Snow Company, Inc., 630 F.2d 599 (8th Cir.1980). See also McGowne v. Challenge-Cook Bros., Inc., 672 F.2d 652, 665 (8th Cir.1982) (deferred ruling on issue to district court on remand). The court, however, never has addressed the issue in a prescription drug case. Robbins v. Farmers Union Grain Terminal Association, 552 F.2d at 789 (cattle feed supplement); Farner v. Paccar, Inc., 562 F.2d at 521-22 (tractor-trailer); Unterburger v. Snow Company, Inc., 630 F.2d at 601, (grain auger).4
The plaintiffs in Robbins claimed that defendant failed to give adequate warnings for the safe and effective use of defendant’s cattle feed supplement. The district court admitted evidence of a subsequent change in instructions warning that the product should not be fed to newly arrived feedlot cattle. This court affirmed the jury verdict for plaintiffs, holding that Rule 407 did not apply to products liability actions based on strict liability.
This court in Robbins based its decision on two grounds, following Ault v. International Harvester Company, 13 Cal.3d 113, 117 Cal.Rptr. 812, 528 P.2d 1148 (1974). First, proof of negligence or culpable conduct is not required in a strict liability action because the issue is the product’s defect rather than the manufacturer’s conduct. The exclusionary rule by its own terms excludes evidence of subsequent remedial measures only when such evidence is offered to prove negligence or culpable conduct. Second, the “public policy” assumption justifying the rule—that modifications would not be made in the absence of the rule—is invalid in products liability actions because the manufacturer of mass-produced goods is motivated by economic self-interest to make the product safer.
However, the conclusion in Robbins—that the doctrine of strict liability by its nature does not include negligence or culpable conduct—does not apply to the circumstances in this case. The strict liability instruction given by the district court contained the following language:
A drug may be supplied in a defective condition which renders it unreasonably dangerous when it is distributed by the manufacturer without reasonable and adequate warnings of dangers inherent or reasonably foreseeable in its use for a purpose and in a manner which the manufacturer should reasonably foresee.
The emphasized language in the passage above also appears in nearly identical form in the negligence instruction. The strict liability instruction also makes specific reference to the duty to warn in the negligence instruction. The issue of due care in giving adequate warning, essentially in the *229same terms used to define negligence, is placed squarely in the strict liability instruction. Foreseeability is an inherent consideration in determining negligence, and the instruction makes it so with respect to strict liability.
This makes particularly relevant these words of the Fourth Circuit in Werner v. Upjohn Co., 628 F.2d 848, 858 (4th Cir.1980):
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? [Citations omitted.]
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. The standard for liability under strict liability and negligence is essentially the same.
Werner held that Rule 407 applied even though the action was based on strict liability and that introduction of subsequent warnings was error. See also, Lindsay v. Ortho Pharmaceutical Corp., 637 F.2d 87 (2d Cir.1980); Smyth v. Upjohn Co., 529 F.2d 803 (2d Cir.1975); Chambers v. G.D. Searle & Co., 441 F.Supp. 377 (E.D.Md.1975).
The evidence in this case points to Talwin being an unavoidably unsafe product.5 Dr. DeMere testified that there is no drug on the market capable of doing good that isn’t also capable of doing somebody harm. Dr. McShane stated that all prescription drugs have a potential for harm. In addition, Talwin was a new drug when DeLuryea first started using it. This is further reason for concluding, as in Werner, that the standards for liability under strict liability and negligence are essentially the same. See also, Basko v. Sterling Drug, Inc., 416 F.2d 417, 425 (2d Cir.1969); Restatement of Torts (Second) § 402A, comment k. Defendant’s conduct in giving the warning is in issue. Consequently, the reasoning in Robbins, that strict liability does not include negligence or culpable conduct, does not apply to the circumstances of this case.6 Rule 407 requires exclusion of evidence of subsequent remedial changes in Sterling’s warning literature. The exhibits, documents, articles, package inserts, and Physicians’ Desk Reference warnings after June 1974 were thus inadmissible. Because we reverse on other grounds, we need not examine the prejudicial effect of the admission of this evidence.
IV.
Sterling claims that the district court erred in instructing the jury on the issue of breach of express warranty and in failing to require a finding of reliance on the part of the buyer. The instruction given is patterned after Arkansas Model Jury Instructions, AMI 1011. Consistent with AMI 1011, it makes no reference to reliance. The instruction, however, makes a significant omission from AMI 1011. In the paragraph defining express warranty, the instruction of the court states:
*230Any affirmation of fact or promise made by Winthrop Laboratories to the buyer which relates to the product sold (Talwin)* creates an express warranty that the goods shall conform to the affirmation or promise.
AMI 1011, by contrast, includes the following language after the asterisk above: “and becomes part of the basis of the bargain.” The language in AMI 1011 is adapted from Ark.Stat.Ann. § 85-2-313 (1961). DeLuryea argues that the objection made by Sterling, that “the instruction itself does not require that anyone rely upon the alleged express warranty,” is not sufficiently broad to cover the argument based on the omission of the words “basis of the bargain.” Because we reverse on other grounds, we need not reach these troublesome issues which may not recur on retrial.
V.
DeLuryea argues that the district court erred in refusing to submit the issue of punitive damages. In reviewing this ruling we must consider the evidence in the light most favorable to DeLuryea and give her the benefit of all favorable inferences reasonably to be drawn from the evidence. If the evidence is such that reasonable persons might differ, then the issue is for the jury. Farner v. Paccar, Inc., 562 F.2d at 522.
Viewed in the light most favorable to DeLuryea, the evidence shows that Sterling knew that Talwin caused tissue damage and drug dependence and failed to adequately warn of these dangers. The evidence establishes, however, that warnings were given concerning both tissue damage and drug dependence.
The package inserts beginning in September, 1968 discussed nodules and ulcerations at injection sites and recommended that injection sites be rotated.7 On August 28, 1969 a letter was addressed to all doctors discussing drug dependence from Talwin.8
Our review of the entire record leads us to conclude that the district court did not *231err in ruling that there was no evidence to support punitive damages and that there was no indication of malice, wantonness, or reckless indifference to the consequences from which malice could be inferred. De-Luryea relies on Hoffman v. Sterling Drugs, 485 F.2d 132 (3d Cir.1973). The court in Hoffman, however, detailed some nineteen articles or published letters and some thirteen letters or memoranda directed to defendant or internally circulated that concerned severe retinal changes caused by defendant’s drug. The evidence showed that defendant, knowing that its drug could cause serious retinal changes, knew or should have known that its attempted warnings would not effectively reach the medical profession. The appellate court held that whether this failure to take action reasonably calculated to warn physicians of a risk of great magnitude was in reckless disregard of the public’s health should have been submitted to the jury. The evidence in Hoffman was substantially different from that presented in this case. DeLuryea relied almost exclusively on expert opinion evidence and did not present the devastating documentary evidence presented in Hoffman. DeLuryea cites no Arkansas case discussing the issue of punitive damages in a drug liability case.
Sterling raises another issue concerning the admission of statements allegedly made by Sterling’s detail men to doctors in North Carolina. We do not find that the district court abused its discretion in admitting this testimony. Haynes v. American Motors Corp., 691 F.2d 1268 at 1271-72 (8th Cir.1982); Auto-Owners Insurance Co. v. Jensen, 667 F.2d 714, 722 (8th Cir.1981).
Because of the error in excluding Dr. Ivie’s deposition, we reverse the judgment of the district court and remand the case for new trial on the issues of liability and compensatory damages.
. Rule 403 provides:
Although relevant, evidence may be excluded if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence.
. Fed.R.Evid. 402 provides in part, “evidence which is not relevant is not admissible.”
. Fed.R.Evid. 407 provides:
When, after an event, measures are taken which, if taken previously, would have made the event less likely to occur, evidence of the subsequent measures is not admissible to prove negligence or culpable conduct in connection with the event. This rule does not require the exclusion of evidence of subsequent measures when offered for another purpose, such as proving ownership, control, or feasibility of precautionary measures, if controverted, or impeachment.
. It is not clear whether state or federal law applies to this issue in a diversity case. See Oberst v. International Harvester, Inc., 640 F.2d 863, 867 n. 2 (7th Cir.1980) (Swygert, J. dissenting) (concludes that Fed.R.Evid. 407 “probably” applies). Arkansas, however, has adopted Uniform Rule 407, which differs only slightly from the federal rule. Ark.Stat.Ann. § 28-1001-407 (1979). Arkansas has not addressed the issue of the applicability of Rule 407 to strict liability actions, but follows the general rule that subsequent remedial measures are inadmissible to prove negligence. Gist v. Meredith Marine Sales & Service, Inc., 272 Ark. 489, 615 S.W.2d 365 (1981).
. Talwin was held to be an unavoidably unsafe product in Crocker v. Winthrop Laboratories, 514 S.W.2d 429, 432-33 (Tex.1974).
. The other rationale behind Robbins and Ault —that admission of a subsequent remedial measure would not discourage improvements by the manufacturer of mass-produced goods— has been criticized because it may be equally applicable to negligence actions. Werner v. Upjohn Co., 628 F.2d 848, 857-58 (4th Cir.1980).
. The September, 1968 package insert stated: Rarely reported reactions include ... nodules and ulceration at injection site....
In December, 1971 this language was changed to read:
Rarely reported reactions include ... soft tissue induration, nodules, cutaneous depression, and ulceration (sloughing) at the injection site....
In June, 1973 this language was changed to read:
Rarely reported reactions include ... dermatologic: soft tissue induration, nodules, cutaneous depression, ulceration (sloughing) at the injection site, toxic epidermal necrolysis....
In May, 1974 the following new section was added after the paragraph listing the most common adverse reactions:
Dermatologic Reactions: Soft tissue induration, nodules, and cutaneous depression can occur at injection sites. Ulceration (sloughing) and severe sclerosis of the skin and subcutaneous tissues (and, rarely, underlying muscle) have been reported after multiple doses.
Toxic epidermal necrolysis was still listed under rarely reported reactions. Under “PRECAUTIONS,” the following was added:
Tissue Damage at Injection Sites. Severe sclerosis of the skin, subcutaneous tissues, and underlying muscle have occurred at the injection sites of patients who have received multiple doses of pentazocine as lactate. Constant rotation of injection sites is, therefore, recommended. In addition, animal studies have suggested that Talwin may be tolerated less well subcutaneously than intramuscularly.
Beginning in September, 1968 the following language appeared concerning dosage and administration:
As with most parenteral drugs, when frequent daily injections are needed over a prolonged period, intramuscular administration is preferable to subcutaneous. In addition, constant rotation of sites (e.g. the upper outer quadrants of the buttocks, mid-lateral aspects of the thighs, and the deltoid areas) is recommended.
In October, 1975 this language was changed to read:
The subcutaneous route of administration should be used only when necessary because of possible severe tissue damage at injection sites (see WARNINGS). When frequent injections are needed, the drug should be administered intramuscularly. In addition, constant rotation of injection sites (e.g., the upper outer quadrants of the buttocks, mid-lateral aspects of the thighs, and the deltoid areas) is essential.
. The letter stated:
Drug Dependence: Patients with a history of drug abuse should be under close supervi*231sion. There have been instances of psychological and physical dependence on Talwin in patients with such a history and, rarely, in patients without such a history. Abrupt discontinuance following the extended use of parenteral Talwin has resulted in symptoms such as abdominal cramps, elevated temperature, rhinorrhea, restlessness, anxiety and lacrimation. Even when these occurred, discontinuance has been accomplished with minimal difficulty. In the rare patient in whom more than minor difficulty has been encountered, reinstitution of parenteral Talwin with gradual withdrawal has ameliorated the patient’s symptoms. Substituting methadone or other narcotics for Talwin in the treatment of the Talwin abstinence syndrome should be avoided.