Margaret Sharon Worsham v. A.H. Robins Company

CLARK, Circuit Judge: This products liability action was brought against A.H. Robins (Robins) for serious illness from infection suffered by Margaret Worsham while she was wearing a Daikon Shield, an intrauterine contraceptive device (IUD) manufactured by the company. Ms. Worsham claimed that the Daikon Shield that she had been wearing caused her to suffer a serious form of pelvic inflammatory disease (PID) known as a tubo-ovarian abscess. *679 In 1978, Margaret Worsham was an unmarried, 34-year-old registered nurse when the events leading up to the lawsuit occurred. Ms. Worsham became ill and went to the emergency room on June 16. Although she was advised twice, at the emergency room and later by her gynecologist, to enter the hospital, she delayed doing so until June 22. Once in the hospital, Ms. Worsham’s condition (pelvic infection made serious by an abscess) worsened and she underwent a complete hysterectomy (removal of the ovaries, uterus and fallopian tubes). The Daikon Shield worn by Ms. Worsham was removed and discarded at the hospital. It was not introduced into evidence at trial. Evidence was presented to the jury which could and evidently did lead them to reach the following conclusions. The tuboovarian abscess suffered by Ms. Worsham was caused by the Daikon Shield. There were design and manufacturing problems with the Daikon Shield which made it a possible source for PID. Once inserted, the Daikon Shield, like other IUDs, sits in the uterus of the wearer. The tailstring, which allows the wearer to tell if the Shield is in place and provides the method of removal, hangs from the uterus into the vagina. The Daikon Shield tailstring, unlike other IUDs sold in the 1970s which used monofilament tailstrings, was a multifilament suture. This bundle of individual nylon fibers was enclosed by a nylon sheath. The multifilament tailstring had a “wicking tendency,” a tendency to carry fluids and bacteria up through the fibers. A double knot tied at the head of the tailstring was included to prevent the movement of bacteria into the uterus. Evidence was presented, however, to show that the tailstring sheath had a tendency to lose integrity just below the double knot. This loss of integrity could come at the time the tailstring was tied to the Shield or if the sheath deteriorated over time in situ (in place in the body). When the Shield was in situ, the part of the tailstring just below the double knot was in the uterus. A break in the sheath would allow bacteria growing in the tailstring to escape directly into the uterus. Bacteria normally grow in the vagina. A constant stream of bacteria going from the vagina through the tailstring into the uterus, a sterile or bacteria free organ, can lead to infection. There is evidence that Robins knew of the wicking tendency in 1971 and that the sheath would crack or deteriorate over time in 1972. Robins never did any testing on when the string would deteriorate in situ. Nevertheless, in a question and answer brochure written for wearers of the product (1972), Robins indicated the Shield could protect them from conceiving for several years and that some women had been protected for five years or longer. Ms. Worsham had worn her Shield for a little over five years when she became ill in 1978. By 1975, Robins knew of tests which indicated that bacteria grew in the tailstrings and that a sample of some strings showed breaks in the nylon sheath. Robins received reports, in addition to the small number of adverse reaction complaints it had on file, that there was an association between pelvic infection and the Daikon Shield tailstring. Robins never considered the complaints sufficient to require it to take any action other than listing “infection” as a problem in 1972 labeling and later PID as an adverse effect in its October 1973 labeling. After Robins became aware of an association between septic abortion and the IUD in 1973-74, the company advised doctors to remove the Shield from patients who became pregnant. In June 1974, Robins withdrew the product from the market and decided not to remarket. Since that time, Robins issued no warnings to women still wearing Daikon Shields that they should discontinue use of them. The causes of tubo-ovarian abscess in women of Ms. Worsham’s age group were identified as gonorrhea (the most common cause), endometriosis, appendicitis, gunshot, stab and knife wounds, trauma to the abdomen, surgical complications and infections following delivery and radiation treatment for various forms of cancer in the reproductive organs. None of these causes appeared likely with regard to Ms. Wors *680 ham while there was testimony that the Shield was responsible. As a result of this infection, Ms. Worsham underwent surgery which left her without the ability to bear children. A jury awarded Ms. Worsham compensatory damages of $2,500,000 and punitive damages of $1,000,000. On February 19, 1982, a final judgment was entered on the jury verdict in the amount of $1,750,000 in compensatory damages (the jury had found the plaintiff to be 30% responsible for the injuries she suffered due to her failure to timely seek hospital care) and $1,000,000 in punitive damages. Robins moved for post-trial relief including judgment notwithstanding the verdict or a new trial or other appropriate relief, including remittitur. The trial judge ruled that a new trial would be granted unless the plaintiff remitted all compensatory damages above $950,000 and all punitive damages above $500,000. An amended final judgment reflecting the remittitur was entered on August 4, 1982. Claiming that the relief afforded through remittitur below was inadequate, Robins filed a notice of appeal and raised eight issues: (1) the plaintiff failed as a matter of law to prove that there was a defect in her Daikon Shield and therefore the claims upon which plaintiff prevailed should not have gone to the jury; (2) in submitting the question of defect to the jury, the court erred in its instructions; (3) evidence was insufficient as a matter of law to show that Robins failed to act as a reasonably prudent pharmaceutical company; (4) — (6) the court improperly admitted evidence (hearsay documents, one exhibit at the close of evidence after defendant had relied on a ruling that the document was inadmissible, evidence of other lawsuits and claims against the company); (7) the claim for punitive damages failed as a matter of law; and (8) the court erred by amending a jury interrogatory after deliberations had begun. Finding the evidence sufficient to support the jury verdict and the alleged trial errors nonexistent or harmless, we affirm. I. Defect and Causation Ms. Worsham’s case was submitted to the jury on four theories of liability. 1 The first three theories required a showing that the product in question was defective. The first claim in strict liability alleged that the Shield was defective and unreasonably dangerous. The second and third claims, of negligent design and manufacture and negligent failure to warn, could not have been reached unless the jury found defect. The jury found for the plaintiff on all three theories. Robins claims that the plaintiff as a matter of law failed to prove such a defect. By making this argument, Robins, in effect, asks this court to rule that the plaintiff’s first three claims should not have gone to the jury. At the close of evidence, Robins moved for a directed verdict on the question of the defective product because the plaintiff had not shown that there was a break in the sheath of her Daikon Shield and therefore had not shown causation. In response to that motion, the plaintiff argued that she had disproved any possible alternative to infection resulting from a cracked Daikon Shield. The key issue is the determination of whether plaintiff was required to prove a break in the sheath which had been discarded during surgery. The standard for determining whether there is sufficient evidence to submit a case to the jury following a motion for a directed verdict has long been estab *681 lished. The court should consider all evidence, not just that which supports the non-mover’s case, in the light and with all reasonable inferences most favorable to the party opposed to the motion. Boeing Co. v. Shipman, 411 F.2d 365, 374-75 (5th Cir.1969) (emphasis added). The case should go to the jury if there is substantial evidence, evidence of such quality and weight that reasonable and fair-minded jurors in the exercise of impartial judgment might reach different conclusions, opposed to the motion for directed verdict. A mere scintilla of evidence cannot present a question for the jury. There must be a conflict in substantial evidence to create a jury question. Id. This case presents a typical jury question of how to determine what caused an injury and whether that cause was a defective product. Reviewing the motion for directed verdict made at trial leads us to conclude that Robins is arguing that there was a lack of substantial evidence. Robins claims that the jury could find causation only through speculation. Robins’ position is that the plaintiff’s theory of defect centered on the claim that the plastic sheath surrounding the tailstring of the Shield would break and thereby permit bacteria to travel up the tailstring and infect the person wearing the Shield. The plaintiff, however, failed to prove that her sheath was broken and evidence introduced to support the possibility that this might occur indicated that most of the sheaths showed no breaks. 2 The jury was therefore, according to Robins, left without proof that it was more likely than not that the plaintiff’s theory of defect accurately presented the cause of Ms. Worsham’s infection. Robins claims that the plaintiff could not and did not refute possible alternative causes. Evidence was presented by plaintiff and defendant that the annual incidence rate of PID in women who do not wear IUDs was 2-3%. Robins also put on evidence through its expert witness that approximately 85% of pelvic infections are in women who do not wear IUDs. 3 Throughout the trial, Robins offered alternative causation theories. For example, Robins argued that the plaintiff could be among the group of women who would have gotten a pelvic infection regardless of the presence of an IUD. The company also offered testimony positing that sexual transmission of the bacteria, exposure to gonorrhea, and appendicitis were likely causes of Ms. Worsham’s infection. In response, the plaintiff presented a theory of defect, that a characteristic of the product resulted in the injury suffered, and offered specific evidence to disprove alternative causation. According to the plaintiff, the Daikon Shield tailstring was defective because of the design of the tailstring. Evidence was introduced which argued that the tailstring sheath had a tendency to crack and to deteriorate over a period of time in the body. Evidence was adduced which showed that the multifilament fibers in the suture would wick (carry) fluids. According to the plaintiff’s theory of defect, a break in sheath would allow bacteria to travel through the tailstring into the uterus thereby causing infection. Plaintiff then presented evidence that these defects, rather than alternatives, caused her injury. One expert negated all possible causes of tubo-ovarian abscess with respect to the plaintiff’s case and testified that there was an association be *682 tween the use of IUDs and this severe form of infection. That same expert and another testified that the theory of defect argued by the plaintiff was the cause of the infection. Evidence was introduced that the bacteria found in Ms. Worsham was similar to that found in the tailstrings of Daikon Shields which had breaks in them. In addition, other physician experts testified that the causes singled out by Robins, gonorrhea, appendicitis and sexual transmission, were highly unlikely. What a review of the evidence reveals is that the jury was presented with a battle of experts on the question of causation. As Boeing teaches us, “it is the function of the jury as the traditional finder of facts, and not the Court, to weigh conflicting evidence and inferences, and determine the credibility of witnesses.” 411 F.2d at 375. A thorough review of the evidence presented at trial reveals that both sides presented substantial evidence on the question of causation. A jury question was, therefore, created on whether the Daikon Shield was defective. II. Jury Instruction On Defect In addition to arguing that the evidence on defect was insufficient to go to the jury, Robins argues that the district court gave an instruction on defect (plaintiffs requested instruction No. 4) which improperly allowed the jury to speculate on the presence of a defect in the Daikon Shield worn by Ms. Worsham. The instruction complained about relates to the issue of how the plaintiff must prove the alleged defect in a product. The instruction given by the district court reads as follows: To prove her case that the Daikon Shield is defective in design and manufacture, it is not necessary for Sharon Worsham to produce the Daikon Shield which she actually wore, and to point to a specific defect in it. • The presence of defects in the shield she wore may be inferred from the testimony and evidence regarding defects in other Daikon Shields or Daikon Shield strings, if the evidence in the case warrants this inference. With regard to the first sentence of the instruction, Robins claims that since the plaintiff set out to prove a specific defect in the tailstring of her Shield, she is not entitled to an instruction which would allow the jury to find for her whether or not she proved that specific defect. Robins also argues that the plaintiff cannot properly lay claim to the inference in the second sentence of the instruction. As the source for the contested jury instruction, the plaintiff offered Cassisi v. Maytag, 396 So.2d 1140, 1151 (Fla.Dist.Ct.App.1981), which held that an inference of defect sufficient to send the case to the jury could be created by proof of a probable product malfunction during normal usage coupled with refutation of other causes of the accident. Robins contends that the Cassisi inference was developed to assist a plaintiff who can show that a product malfunctioned during normal usage but who cannot prove the specific source of the malfunction. Application of the Cassisi inference under the facts of this case is therefore improper since the plaintiff never established the threshold requirement of proving a malfunction during normal usage. Offering proof of sudden illness with the Daikon Shield in place does not prove a malfunction according to Robins. In addition, the instruction fails to correctly set forth the inference established in Cassisi. As it reads, the instruction does not require the jury to find that the plaintiff proved an injury resulting from-product malfunction during normal usage. The plaintiff argues that the requirements were met for the Cassisi inference to come into play. First, the plaintiff was not required to prove a defect in her Daikon Shield. She did offer proof that the design of the Shield was defective and that the problem, cracks in the sheath surrounding the tailstring, could be caused in several ways. The Cassisi inference is properly applied in cases where, as here, the product has been lost or destroyed. 396 So.2d at 1150-51. The Cassisi inference also frees the plaintiff from having to disprove all alternative causation theories in order for the case to go to the jury. In this case, *683 however, the plaintiff argues that evidence was offered to disprove each of the alternative theories offered by Robins. Finally, the plaintiff contends that the proof of her sudden illness with the Shield in place was adequate proof of a malfunction during normal usage to satisfy the Cassisi threshold. We believe both parties fail to understand the complete implications of Cassisi. The issue in Cassisi was whether the trial court had erred in granting summary judgment to the manufacturer of a clothes dryer. The case must be considered in that posture. In Cassisi, plaintiffs were able to show that their clothes dryer had been in normal use for over a year when a fire apparently began inside the dryer and spread to their home. The expert evidence offered on causation indicated that the malfunction was inherent in the product because the fire originated within the dryer. The expert, however, was unable to negate other possible theories of causation and the trial court granted judgment to defendant. 396 So.2d at 1142-43. The Florida District Court of Appeals allowed the question of defect to go to the jury if the evidence showed that the origin of the fire was probably in the dryer (although not a certainty) so that the jury could weigh that evidence against other possible causes of the fire. Cassisi relied principally on Greco v. Bucciconi, 283 F.Supp. 978 (W.D.Pa.1967), aff 'd, 407 F.2d 87 (3d Cir.1969), for the proposition that when a product malfunctions that would not malfunction but for a defect, a plaintiff is entitled to an inference of a defect. But Cassisi went further than Robins would have us believe when it argues that the plaintiff here loses because she never established that the IUD malfunctioned. In reversing summary judgment for the manufacturer, the court in Cassisi held that absolute positive proof of product malfunction 4 is not necessary where the product is destroyed; provided plaintiff can point to evidence that the cause of the accident most probably originated in the product. Evidence to so establish a defect can include facts that negate other causes of the accident and at the same time point to the product as the most logical cause. This permits proof of product defect when the product is destroyed (Cassisi), or lost, as here. It is left to a jury to hear the facts and infer a defect in a product if the evidence warrants the inference. The instruction on defect given by the district court is supported by other Florida law allowing defect and causation to be proven by circumstantial evidence. See 41 Fla.Jur.2d 551. To prove that the product she used was defective, the plaintiff did not have to produce it at trial. C.R. Bard, Inc. v. Mason, 247 So.2d 471 (1971); McCarthy v. Florida Ladder Co., 295 So.2d 707, 709-10 (1974). The jury could have properly inferred from the evidence offered about the problems with the tailstring, how a break in the sheath could occur and how such a break would allow bacteria to wick up the multifilament string, that the tailstring on the plaintiffs Shield was defective. The part of the instruction dealing with permissible inferences has a basis in Florida products liability law. See Alpert, Products Liability: The Law in Florida § 1-9 (1979). Proof of defect in this case was established in a manner similar to that used in C.R Bard, Inc. v. Mason, supra. In Bard, an intravenous catheter inserted into the vein of the plaintiff severed and part of it traveled through the vein to the heart causing injuries requiring surgery. The plaintiff offered testimony that a small percentage of the catheters contained burrs which could result in the severance even if the catheter was properly inserted. The jury found that this had happened in the case before it. Bard then challenged this finding arguing that the verdict was founded on an inference; one inference being the discharged catheter *684 was faulty and the further inference that the portion which reached the patient’s heart was severed as a result of the burr. The court rejected that argument noting that a finding that the catheter severed as a result of faulty manufacture was based on an inference. While there was no direct evidence on that issue, however, all of the other possibilities were excluded by direct evidence which the jury could and did believe. 247 So.2d at 472. In the present case, the finding that the tailstring was broken was based on an inference. But the plaintiff did offer expert testimony on the question of defectiveness as well as evidence establishing that the plaintiff’s illness was not caused by the alternatives put forward by Robins. Elimination of alternative causes is one of several accepted types of proof for establishing product defect. 5 Since the jury instruction on defect had its basis in Florida products liability law, we find it to be a proper instruction. III. Sufficiency of Evidence on Claims 2 and 3 To return a verdict for Ms. Worsham on claims 2, 3 and 4, 6 the jury had to find that *685 Robins did not act as a reasonably prudent pharmaceutical company under the circumstances. The jury instructions on the second and third claims, negligent design and manufacture and negligent failure to warn, informed the jury that they must consider the defendant’s conduct according to this standard. The jury was also informed that the conduct of the company was being judged against the state of scientific and medical knowledge as of March 1973 for claim two and as of a reasonable time prior to June 1978 with respect to claim three. Robins’ position is that whether the company should have acted differently under the circumstances is a matter beyond the knowledge of lay jurors. What was required, according to Robins, to prove a breach of the standard of care and thereby present jury issues on the negligence claims was expert testimony. Without such evidence, the plaintiff could not demonstrate the standard of care a reasonably prudent pharmaceutical company would have used in designing its IUD prior to 1973 (claim two) or show that Robins’ receipt of adverse reports about PID and the Shield placed any duty to warn on the company (claim three). There was a total lack of testimony by plaintiff’s experts that the acts of the company violated the standard of care applicable to a reasonably prudent pharmaceutical company. By contrast, all of the defense experts testified that Robins behaved in a prudent and proper manner. In particular, Robins points to testimony by one of its experts that all drug companies receive complaints when they distribute a new product. Robins saw no need to act differently since the number of complaints was small. Both Salem v. United States Lines Co., 370 U.S. 31, 82 S.Ct. 1119, 8 L.Ed.2d 313 (1962), and Fed.R.Evid. 702 present the general rule as to the need for expert testimony. Salem states that expert testimony is not necessary if the primary facts are accurately presented to the jury and the jurors are as capable of understanding and drawing correct conclusions from the facts as an expert witness. 370 U.S. at 35, 82 S.Ct. at 1122, 8 L.Ed.2d at 317. Rule 702 provides that if scientific, technical, or other specialized knowledge will help the trier of fact understand the evidence or determine a fact, an expert witness may testify. 7 The relevant inquiry regarding admission or exclusion of expert evidence is whether such testimony will aid the jury in resolving issues. If such testimony will assist the jury, it may be offered into evidence. As both Salem and Rule 702 make clear, there is no rule requiring expert testimony on certain issues. It is true that there are cases in which a case would fail without expert testimony because the technical and scientific aspects of the case would result in a jury’s inability to comprehend the issues. Robins, however, has not shown that the standard of care issue is, by definition, a matter beyond the jurors or that this is “one of the rare causes of action in which the law predicates recovery upon expert testimony.” Salem, supra, 370 U.S. at 35, 82 S.Ct. at 1122, 8 L.Ed.2d at 317. 8 *686 With regard to the claim dealing with negligent design and manufacture, the jury was instructed to look at the scientific, medical and technical knowledge available during the first few years the Shield was on the market. The plaintiff did introduce evidence that Robins was aware, through its own employees as well as other sources, of problems with the Shield. A jury could have properly decided that in light of this knowledge Robins was negligent in its design and manufacture. The jury could have also determined that in light of Robins’ knowledge of possible problems with the Shield that the reports of PID it received confirming its own information were sufficient to trigger some duty to warn. Expert testimony labeling Robins’ conduct as negligent was not required. IV. Evidentiary Issues Robins argues that the trial court made significant errors in admitting evidence containing references to complaints about PID and the Daikon Shield and to other lawsuits. Three particular rulings, according to Robins, created the atmosphere for a trial by rumor. First, the court allowed hearsay reports of adverse reactions into evidence. Second, the trial court admitted a reference to other suits after Robins had relied on a ruling that the document was inadmissible. Third, the trial court allowed testimony to be read to the jury which made reference to other lawsuits and claims against the company. The trial judge has wide discretion when determining the admissibility of evidence. Such rulings will not be disturbed on appeal absent a clear showing of abuse of discretion. Vallot v. Central Gulf Lines, 641 F.2d 347, 350 (5th Cir.1981). What this court must undertake is a search through the trial record to see if the judge’s decisions about the relevance and materiality of the evidence were arguably proper. Our standard of review requires more than a judgment that we would have ruled differently. Page v. Barko Hydraulics, 673 F.2d 134, 140 (5th Cir.1982). The first alleged error by the trial judge was his admission of ten documents read to the jury which consisted of reports by doctors and company field representatives accompanied by responses from the Robins’ doctor in charge of the Shield, which spoke about adverse reactions associated with Daikon Shield use. Robins claims that these reports compromised the fairness of the trial by exposing the jury to the opinion of “experts” who could not be cross-examined and to a cataloguing of complaints about the Shield which bore no relationship to the problems suffered by Ms. Worsham. A review of the ten exhibits in question reveals that all of them refer to pelvic infection and/or structural problems with the Daikon Shield sheath and tailstring. Since these were the problems stressed by the plaintiff in her case on defect, any prejudice that might have arisen from mention of other problems in two or three of the documents appears to be limited. During the lengthy sessions held on the admission of individual exhibits, portions of irrelevant information had already been deleted. In addition, attempts were made to avoid introducing hearsay into the trial. The plaintiff argued admissibility for the purpose of showing that Robins had received notice that the Shield was connected with severe pelvic infection. Before the documents in question were read to the jury, the trial judge gave a limiting instruction to this effect requested by the company. The court informed the jury that the documents were not to be taken as evidence that the reports were true but rather were allowed into evidence for the purpose of showing notice. The jury was also told that the court was not suggesting that the reports constituted adequate notice. The position taken by Robins is that notice was not a controverted issue in the case. Robins argues that it freely admitted to receiving adverse reaction reports in excess to those illustrated through the documents introduced. What the plaintiff argued in presenting the documents was the medical significance of the reports to a pharmaceutical company. For these re *687 ports to be relevant on that issue, according to Robins, the plaintiff would have had to introduce expert testimonial foundation showing that these documents would have had significance to a pharmaceutical company in monitoring its product. The position taken by the plaintiff-appellee is that what was in issue was the question of adequate notice. According to Ms. Worsham, Robins consistently took the position that it never received any adverse information of a degree which would have required the company to act differently. The exhibits were introduced to show that Robins was receiving notice which, combined with what the company already knew about structural problems with the Shield, should have caused them to take action. The question about whether the documents could be admitted to show notice was hotly debated before the court. The trial judge properly refused to keep the plaintiffs proof on this issue from the jury. The issue in dispute at trial was not whether Robins knew PID could occur from use of the Shield but rather whether the company had received notice about defect and its possible consequences of a type sufficient to make it take action. It was not the fact of notice but the adequacy of notice that the parties argued over. A reading of the response sent by Robins to all those complaining of adverse effects amply confirms this impression of the issue. The letters indicate that the company did not believe that there was a serious problem with PID or the design of the Shield. In one exhibit which confirmed knowledge that Robins possessed about possible defects, the doctor admitted that the design might produce problems but did not believe them to be frequent. The documents were properly submitted and admitted into evidence as proof of Robins’ state of mind and therefore did not constitute hearsay. Webb v. Fuller Brush Co., 378 F.2d 500, 502 (3d Cir.1967); Weinstein & Berger, Weinstein’s Evidence ¶ 801(c)[01] at 801-70— 801-71. The second alleged evidentiary error concerned the admission of Plaintiff’s Exhibit 56. That document contained a statement made by Robins’ house counsel in 1974 that if the Shield were taken off the market it would be a confession of liability and Robins would lose many of the pending lawsuits. Robins objects both to the manner in which the statement was introduced and its content. At an early discussion on admission of evidence, plaintiff’s counsel tried to introduce the exhibit, Robins objected and the court reserved ruling. Later while plaintiff’s counsel was reading numerous documents to the jury, the court admitted this exhibit as part of a group of exhibits. Believing that the document had been properly admitted, plaintiff’s counsel showed the document to opposing counsel and told him the parts he intended to read to the jury, which he then did. Shortly afterwards at a side-bar conference, counsel for Robins indicated that he was worried about the mention of other pending lawsuits. When the trial judge examined Exhibit 56 again, he expressed surprise that it had been admitted. The judge decided not to do anything about the document at that time. After the jury retired for the day, counsel for Robins expressed the same concern about the mention of pending lawsuits. The judge said that he thought that he had not admitted the document. Counsel for Ms. Worsham pointed out that he had and said that he had showed the portion of the statement to opposing counsel. The judge said that when he admitted Exhibit 56 he was looking at the first page (and obviously not at the portion which mentioned the lawsuits). He expressed concern about the “confession of liability” portion of the statement, although he accepted plaintiff’s argument for why the document was admissible for that purpose, and the reference to pending lawsuits. Plaintiff’s counsel had argued that the mention of “confession of liability” was properly admitted because this was an internal Robins’ communication which showed that the company’s concern for liability surpassed that of its concern for wearers of the Shield. Robins asked for a mistrial once it found out that the court *688 was concerned about the admission of the exhibit. The court refused to grant a mistrial but concern about the mention of “pending lawsuits” prompted the judge to offer a corrective instruction to the jury at that point in the trial. Robins refused the corrective instruction. After the closing of evidence, plaintiffs counsel argued that he should be allowed to refer to Exhibit 56 in his closing argument and to put it before the jury. The reason plaintiffs counsel gave was that the defense case consisted of claims of minimal adverse reaction reports and testimony that there was no indication from medical literature of a problem with the Daikon Shield. Robins objected to this, arguing that the document was improperly admitted and that the “advice” given in the statement was ignored and the product taken off the market. After receiving argument, the trial court decided to allow the exhibit in primarily because the complexion of the case was changed by the defense evidence and it had already come in. We find no reason to disturb the rulings on admissibility by the trial judge. Robins claims that it was prejudiced by this final decision because it could not put on any evidence to show that the company had disregarded the advice given in Exhibit 56. When considering the admissibility of the document at an earlier stage, however, the court did not say that the document was improperly admitted. Instead, the judge indicated that he had reservations about the mention of pending lawsuits not the part about the confession of liability if the product were taken off the market. The judge then offered a corrective instruction. After refusing that request, Robins proceeded to put on a case which urged that the company never received notice of a type which would have caused it to act differently. Later in the case, Robins was also on notice that the plaintiff planned to refer to the exhibit in his closing argument and that the court considered this to be proper. At that point, Robins could have contended the significance of the exhibit in its closing argument. There was already evidence before the jury that Robins had suspended marketing in the summer of 1974, very soon after the statement was made by Robins’ house counsel. Even if the mention of pending lawsuits was improper, admission of Exhibit 56 constituted harmless error. F.R.Civ.P. 61. The final evidentiary problem pointed out by Robins concerns the admission of testimony concerning, other lawsuits. The testimony complained about, that of a Robins’ employee Dr. Owen, was first heard outside the presence of the jury. Defense counsel asked Dr. Owen whether he had an opinion as to whether there was anything in the medical literature from 1974 to 1978 which would have required the company to remove the Daikon Shield from those women already wearing them. Dr. Owen testified that in his opinion there was nothing in the literature that would have suggested or required Robins to withdraw the product from the market or the women already wearing the Shield. Plaintiff-appellee argues that by eliciting this testimony Robins opened the door for cross-examination on whether Dr. Owen had received notice through other sources such as lawsuits. It was this questioning of Dr. Owen which the trial judge permitted to be read to the jury. Robins now argues that this testimony concerned lawsuits which were not shown to be similar to the plaintiff’s. According to Ms. Worsham, the portion of the testimony read back to the jury was restricted to questions about lawsuits involving the tailstring and tubo-ovarian abscesses. A review of the testimony reveals that Dr. Owen was asked only about suits involving a defective tailstring and tubo-ovarian abscesses. It is true that Dr. Owen testified that all he remembered about the lawsuits was what personal injuries were claimed and that there was nothing in the information which he received which noted a defective tailstring. When asked about tubo-ovarian cases, Dr. Owen testified that there were more than thirty but beyond that he did not know. Evidence of similar accidents and injuries is admissi *689 ble to prove the harmful tendency or capacity of a product and the defendant’s knowledge thereof. Atlantic Coast Line R. Co. v. Hadlock, 180 F.2d 105 (5th Cir.1950). For the evidence to be relevant, the proponent of it must show the similarity of conditions and that the occurrences are not too remote in time. Jones & Laughlin Steel Corp. v. Matherne, 348 F.2d 394, 400-01 (5th Cir.1965). The plaintiff appears to have met that burden with regard to Dr. Owen’s testimony. The time frame was restricted to the years during which the doctor said that there was nothing in the medical literature which required action. As for the similarity of circumstances, while it is true that the doctor disclaimed any knowledge of suits involving a defective tailstring, he did testify that there were suits dealing with tubo-ovarian abscesses. Robins did have notice of a connection between the Shield and the same type of injury suffered by the plaintiff. Notice was a contested issue throughout the trial. Since the probative value of the evidence was not substantially outweighed by any unfair prejudice to Robins, the testimony of Dr. Owen was properly admitted. F.R.Evid. 403. V. Change in Special Interrogatory Robins claims that it was prejudiced by the trial court’s changing a special interrogatory after it had relied on the interrogatory in structuring its closing argument and the jury had begun deliberations. The case was submitted to the jury with general jury instructions and a special verdict pursuant to 49(a) Fed.R.Civ.P. 9 Robins drafted a set of proposed interrogatories and submitted them to the court and to the plaintiff. Plaintiff’s counsel reviewed the interrogatories and made several objections but not to the interrogatory later corrected by the court. 10 While giving his closing argument, defense counsel did refer to the first interrogatory. At the end of the charge to the jury, the court read the first interrogatory to the jury and the plaintiff did not object. Subsequently, plaintiff’s counsel was asked to review the verdict form before it went to the jury and again did not object. Problems with the first interrogatory did not become obvious to the judge and plaintiff’s counsel until the jury had been in deliberation for several hours and sent out a question about a phrase in the interrogatory. The interrogatory which confused the jury dealt with the issue of defect. The portion which caused the court to alter the interrogatory is that underlined below. 1. Was the string material specified and utilized by A.H. Robins in the manufacture of the Daikon Shield defective and unreasonably dangerous so that the Daikon Shield was more likely to cause pelvic infection than other IUDs? (Verdict Form, Question 1). Once the jury informed the judge that they were having problems with the instruction, the court asked them for further detail on the problem and then conferred with both parties about what should be done. Plaintiff’s counsel admitted that he had overlooked the question before but argued that it was an inaccurate statement of the law on the issue of defect. Robins argued then, as it does now on appeal, that the interrogatory was tailored to the evidence put on at trial on the question of defect. The district court found that the interrogatory as submitted to the jury did not correspond with the instructions given on defect under the strict liability count and that it had the potential of confusing the jury about the issue it was to decide. Because the judge believed that the submission of the first *690 question constituted reversible error, he decided to bring the jury back in, told them that the first question did not conform to the instructions and that the proper question would not include the phrase about whether the Daikon Shield was more likely to cause pelvic infections than other IUDs. After the jury was so instructed, Robins moved for and the court refused to grant a mistrial. Both the question of whether to employ special interrogatories and the question of how those interrogatories are to be framed are matters within the discretion of the trial judge. Dreiling v. General Electric Co., 511 F.2d 768, 774 (5th Cir.1975). 11 Robins, however, claims that the action of the court in amending the original question left the jury free to construct a theory of defect without reference to an objective standard founded upon the evidence. Robins’ position is that interrogatory one accurately presented the plaintiffs theory of defect. Throughout the trial, the plaintiff tried to prove that the Daikon Shield was defective by comparing it to other IUDs. Putting in a phrase which would require a comparison of the Shield to other IUDs, Robins urges, simply guaranteed that the jury would have some objective standard to guide its determination of defect. The district court, however, amended the interrogatory because he correctly found it to be inconsistent with the jury instruction on defect. As the court noted, the defect instruction makes no mention of comparisons between the Daikon Shield and other IUDs. Instead, the instruction, taken from the Florida Standard Jury Instructions on products liability (PL 4, strict liability; PL 2, implied warranty of merchantability), tells the jury to find the product defective if it is in a condition unreasonably dangerous to the user or if it is not reasonably fit for the uses intended. A determination of whether a product is defective and unreasonably dangerous does not require a comparison of it to other products. It is true that evidence had been put on comparing the Shield to other IUDs. But the plaintiff introduced such evidence to point out that the unique design and manufacture of the Shield, the multifilament string and the nylon sheath, were what created the possibility for infection. The question which should have then gone to the jury was whether the product on trial was defective. Special interrogatories should be consistent with jury instructions. The interrogatories must pose the question presented by the case accurately and be stated in a fashion that “avoids the potential for confusing or misleading the jury.” Petes v. Hayes, 664 F.2d 523, 525 (5th Cir.1981). The trial court was properly concerned that the interrogatory as presented would lead the jury to answer no on the question as presented even if it found the Shield defective. Such a result might have occurred since the interrogatory could be interpreted to require the jury to find the product defective only if it was more likely to cause infection than other IUDs. The jurors were therefore correctly instructed on law and then given a possibly contradictory interrogatory. The trial judge was therefore correct in his assumption that giving the original interrogatory was reversible error. See Petes, supra, 664 F.2d at 526. We, therefore, find that the court was acting within its discretion when- it called the jury back and reinstructed it on the question of defect. Robins argues that the correctness of the verdict form is not the only issue presented by the trial court’s action. The company claims that this court must examine whether the trial judge could change the interrogatory in the middle of delibera *691 tions after a party had relied upon it. There is no arguing with the fact that plaintiffs counsel should have discovered the problem with the interrogatory earlier. It is also true that counsel should be advised of the judge’s plans for the use of interrogatories before closing argument so that he can “plan an effective argument whose objective is to translate persuasion into specific decisive action.” Clegg v. Hardware Mutual Casualty Co., 264 F.2d 152, 157 (5th Cir.1959). Defense counsel relied on the acceptance by the court of the interrogatory and made reference to it and what it meant in his closing argument. Any claim of prejudice arising from the court’s “changing the rules of the game,” however, must be examined in light of F.R. Civ.P. 61. Clegg, 264 F.2d at 157. Robins claims that it was severely prejudiced since the alteration of the interrogatory discredited the closing argument structured around the interrogatory. An examination of the closing argument, however, reveals that defense counsel spent the majority of his time arguing alternative causation, the duty to warn and punitive damages. Robin’s closing argument was not focused upon the comparison between the Daikon Shield and other IUDs suggested by the interrogatory. If there had been any error committed by the change of the interrogatory during deliberations, it would have been harmless. Since we find that there was enough evidence from which the jury could have determined that Robins so departed from the applicable standard of care as to raise an inference that the company acted “wilfully, wantonly, maliciously or with conscious disregard for the legal rights of the plaintiff,” we decline to review the award of punitive damages. The punitive damages issue is supposed to go to the jury if the court at the close of evidence decides that “there is a legal basis for such damages shown by any interpretation of the evidence favorable to the plaintiff.” Wackenhut Corp. v. Canty, 359 So.2d 430, 435 (Fla.1978). In conclusion, therefore, we hold that Ms. Worsham presented sufficient evidence from which a jury could reasonably find her Daikon Shield to be defective and the cause of her injuries. We also find that she presented sufficient evidence as to the applicable standard of care with regard to the negligence claims. Finally, we find that the trial court gave correct jury instruction on defect, made no errors in admitting evidence and properly changed the special interrogatory in the verdict form to conform to the jury instructions. Accordingly, the judgment of the district court is AFFIRMED. Footnotes [1] The fourth claim was that even if the Daikon Shield was not defective, and therefore not more dangerous than other IUDs sold in the same period, A.H. Robins was liable for failing to warn that wearers of IUDs were at greater risk of contracting PID than women not wearing IUDs. Since the jury found that there was a defective product, a requirement for liability on the first three claims, it was not required to reach a verdict on this fourth claim. Robins contends that by submitting this fourth claim the plaintiff conceded that it did not have enough evidence to prove a defect in the product. Actually, the fourth claim stands, as the verdict form makes clear, as an alternative theory of liability. The position taken by the plaintiff was that even if the jury found that the Shield was not defective it could still return a verdict for Ms. Worsham. [2] The plaintiff's chief expert on defect testified that he had done a study on about 600 Daikon Shield tailstrings. The tailstrings came from Daikon Shields which had been worn (there was no evidence about how long the Daikon Shields had been in place). In that study, the doctor found breaks in 4% of the black sheaths surrounding the tailstrings and in 13% of those with transparent sheaths. Robins also elicited admissions on cross-examination. Robins points to these results and testimony, that some of the samples might have been broken in the removal from the women, as proof that the likelihood that the sheath on the plaintiffs tailstring was cracked was negligible. One could also conclude from the evidence that the sheaths did have a propensity to break. [3] No evidence was adduced on the incidence rate of tubo-ovarian abscess, the more severe form of PID and the type of injury suffered by the plaintiff, in women who wear or who do not wear IUDs. [4] In Greco, there was such positive proof. Plaintiff was working with an automated machine that lowered and cut off his fingers when it should have lifted in its automated movement. Thus, malfunction equated to defect in the machine. [5] P. Sherman, Products Liability 325 (1981); Rheingold, Proof of Defect in Product Liability Cases, 38 Tenn.L.Rev. 325, 327-39 (1971). [6] The jury instructions given on Claims 2 and 3 relating to the standard of care were as follows: Claim 2: First, you cannot consider the claim for negligence in design and manufacture, unless you first determine that the product was defective. If you determine that the Daikon Shield was defective, you must then consider Sharon Worsham’s claim that A.H. Robins was negligent in its design and manufacture. Negligence means the failure to do an act which a reasonably prudent pharmaceutical company would do, or the doing of an act which a reasonably prudent pharmaceutical company would not do under the same or similar circumstances, to protect users of its product from bodily injury. Reasonable care is that degree of care which a reasonably prudent pharmaceutical company would use under the same or similar circumstances, and at the same point in time. In addition, to be unreasonably dangerous, the product, because of the defect in design, must have created a risk of harm to a user which an ordinarily competent obstetrician-gynecologist would not reasonably have expected or have created a greater risk of harm than such a medical practitioner would reasonably have expected. * * * * * * A product such as the Daikon Shield may be defective if its design so unreasonably increased the risks of injury from intrauterine contraceptive devices called IUD’s, that a reasonably prudent manufacturer, with knowledge of such risks, would not have marketed and sold it in March of 1973. In determine [sic] whether defendant’s design decisions were reasonable, you shall consider the following factors: The gravity of the danger posed by the design of the Daikon Shield; the likelihood that such a danger would occur; the feasibility of a safer alternative design; the financial cost of an improved design; and the adverse consequences to the product and the consumer that would result from an alternative design. (T. 3397-98). Claim 3: You are instructed that a pharmaceutical company has a continuing duty to monitor scientific or medical discoveries and knowledge. Consequently, a warning deemed legally adequate when the device left the control of the manufacturer may be rendered deficient by actual or constructive knowledge acquired subsequently, if a reasonable opportunity to warn of the new discoveries existed. You are instructed that the A.H. Robins Company would have had a duty to provide warnings to the medical profession between March 1973 and a reasonable time prior to June 1978, only if the following elements are established by the greater weight of the evidence. One: That the Daikon Shield is, in fact, defective. Two: That between March 1973 and a reasonable time prior to June 1978, that scientific or medical discoveries and knowledge during that period would have led a reasonably prudent pharmaceutical company to conclude that the Daikon Shield is defective. Three: In light of this knowledge, and in light of the medical profession, a reasonably prudent pharmaceutical company would have provided warnings of a defect. If you find by the greater weight of the evidence that A.H. Robins was under a duty to provide warnings, you must determine whether or not that duty was breached. If a manufacturer of a medical device knows or in the exercise of reasonable care, should have known, that the use of its device *685 may be harmful or injurious and such danger would not be obvious to an ordinary competent physician dispensing said device, then the manufacturer is obligated to use reasonable care to warn the medical profession of the danger and it is negligent if it fails to do so. (T. 3400-02). There is no need to consider the instruction on claim four since the jury did not decide it. [7] The complete rule reads as follows: Testimony by Experts If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in form of an opinion or otherwise. Rule 702, F.R.Evid. [8] Causes of action in which recovery is predicated upon expert testimony include medical malpractice actions, determinations as to insanity and proof of foreign law. 7 Wigmore, Evidence §§ 2090, 2090(a) (Chadbourn rev. 1978). There are also other areas acknowledged by the courts to be amenable only to expertise. For example, expert testimony is often required to establish defective design of a product. Weinstein & Berger, Weinstein's Evidence ¶ 702[02] at 702-9 n. 3. Both parties agree that the plaintiff in this case produced expert testimony about defective design. [9] The relevant portion of F.R.Civ.P. 49(a) reads as follows: The court may require a jury to return only a special verdict in the form of a special written finding upon each issue of fact____ [I]t may use such other method of submitting the issues and requiring the written findings thereon as it deems most appropriate. The court shall give to the jury such explanation and instruction concerning the matter thus submitted as may be necessary to enable the jury to make its findings upon each issue. [10] This court rejects, as did the district court, plaintiff-appellee's assertion that Robins somehow surreptitiously altered the special verdict form to include the extra phrase. [11] In determining the adequacy of forms of special interrogatories, this court considers several factors: 1) whether, when read as a whole and in conjunction with the general charge, the interrogatories adequately presented the contested issues to the jury; 2) whether the submission of the issues to the jury was fair; and 3) whether the “ultimate questions of fact” were clearly submitted to the jury. Dreiling v. General Electric Co., 511 F.2d 768, 775 (5th Cir.1975). Reviewing the special interrogatories of the present case in light of these factors, they appear adequate except for the additional phrase in the first question.