Robinson v. McNeil Consumer Healthcare

In the United States Court of Appeals For the Seventh Circuit No. 09-4011 K AREN R OBINSON, et al., Plaintiffs-Appellants, v. M C N EIL C ONSUMER H EALTHCARE, a division of M C N EIL-PPC, INC., Defendant-Appellee. Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 07 C 5603—James F. Holderman, Chief Judge. A RGUED JUNE 2, 2010—D ECIDED A UGUST 11, 2010 Before E ASTERBROOK, Chief Judge, and P OSNER and K ANNE, Circuit Judges. P OSNER, Circuit Judge. Karen Robinson and her husband (suing for loss of consortium) brought a products liability suit against McNeil Consumer Health- care in an Illinois state court. The case was removed to a federal district court in Illinois under the diversity jurisdiction. McNeil’s parent, Johnson & Johnson, was 2 No. 09-4011 a defendant in the district court, but the jury found in its favor and the appellants do not challenge the finding, so it is out of the case. The district judge ruled that Virginia law governed the substantive issues in the case. That law both rejects strict liability as a basis for a products liability suit, Harris v. T.I., Inc., 413 S.E.2d 605, 609-10 (Va. 1992); Sensenbrenner v. Rust, Orling & Neale, Architects, Inc., 374 S.E.2d 55, 57 n. 4 (Va. 1988); Lust v. Clark Equipment Co., 792 F.2d 436, 439-40 (4th Cir. 1986) (applying Virginia law); compare Restate- ment (Second) of Torts § 402A (1965), so that a plaintiff has to prove negligence; and deems contributory negli- gence a complete defense to a claim of negligence. E.g., Litchford v. Hancock, 352 S.E.2d 335, 337 (Va. 1987); Fein v. Wade, 61 S.E.2d 29, 31-32 (Va. 1950). After a six-day trial the jury found that McNeil had been negligent, and calculated Mrs. Robinson’s compensa- tory damages at $3.5 million. But the jury also found that she had been contributorily negligent, and so— since contributory negligence is a complete defense to negligence under the law of Virginia—the judge entered judgment for McNeil. The Robinsons appeal, but since Mr. Robinson’s claim is derivative from his wife’s, we needn’t discuss it, and for the sake of simplicity we’ll pretend that his wife is the only plaintiff. McNeil manufactures and sells Children’s Motrin, an over-the-counter drug (though there’s also a prescrip- tion version, as we’ll have occasion to note). The active ingredient is ibuprofen, a non-steroidal anti-inflamma- tory drug (NSAID), used primarily to alleviate pain and No. 09-4011 3 fever, that is also the active ingredient in Advil. The “Warnings” section on the label of the bottle of Motrin that Mrs. Robinson bought for her child begins: “Allergy alert: Ibuprofen may cause a severe allergic reac- tion which may include: hives, facial swelling, asthma (wheezing), shock.” After additional warnings of side effects the label says: “Stop use and see a doctor if an allergic reaction occurs.” She read the warnings before buying the drug. In September 2005, four or five months later, she awoke in the middle of the night with a headache and took two teaspoonfuls of the Children’s Motrin that she had bought—the dose suggested for a child six to eight years old—without reading the warnings (the specifics of which she had forgotten) on the label. When she awoke the next morning she noticed a rash on her chest. The rash worsened throughout the day. That night she woke up with a fever and took two more teaspoonfuls of the Motrin, again without reading the warnings. The next morning she went to see her doctor, who gave her a dose of Benadryl and prescribed a Medrol pack, both being drugs for treating allergic reactions. She mentioned that she had taken Children’s Motrin; he did not react. Later that day she noticed that the rash on her chest was sprouting blisters, and her fever increased. After waking up late that night she again took two teaspoonfuls of the Motrin without reading the warnings. The next morning, with her condition deteriorating rapidly, she went back to her doctor, who immediately ordered her hospitalized. She was diagnosed with TEN (toxic epidermal necrolysis), 4 No. 09-4011 an especially severe form of SJS (Stevens-Johnson syn- drome). TEN is a rare but life-threatening disease that causes severe blistering and consequent sloughing off of skin over much of the body, together with serious damage to the mouth, eyes, throat, and esophagus. Jean- Claude Roujeau, Robert S. Stern & Bruce U. Wintroub, “Cutaneous Drug Reactions,” Harrison’s Principles of Internal Medicine 343, 346 (Anthony S. Fauci et al. eds., 17th ed. 2008); Pierre-Dominique Ghislain & Jean-Claude Roujeau, “Treatment of Severe Drug Reactions—Stevens- Johnson Syndrome and Toxic Epidermal Necrolysis,” www.sjsupport.org/pdf/tsdr.pdf (visited July 21, 2010). The treatment for the disease is similar to that given burn victims. Mrs. Robinson survived, but sixty percent of her skin had sloughed off, and she lost the vision in one eye and has only limited vision in the other, which requires con- stant medical treatment; she is expected to go blind eventually. She has required multiple operations on her throat and esophagus as a result of the damage to those organs caused by the disease. The initial legal question presented by these unhappy facts is choice of law. Virginia as we said makes con- tributory negligence a complete defense to liability for negligence. Today that is distinctly a minority position, Restatement (Third) of Torts: Apportionment of Liability § 7 comment a (2000), contrary to the prediction in Pen- nsylvania R.R. v. Aspell, 23 Pa. 147, 149-50 (1854), that a rule of law “not likely to be changed in all time to come [is] that there can be no recovery for an injury caused by No. 09-4011 5 the mutual default of both parties.” Illinois makes the victim’s negligence a partial defense under the rubric of “comparative fault,” which merely reduces the damages awarded the plaintiff unless the plaintiff’s negligence exceeds the defendant’s, in which event the plaintiff’s negligence is a complete defense. 735 ILCS 5/2-1116(c); Board of Trustees of Community College District No. 508 v. Coopers & Lybrand, 803 N.E.2d 460, 465 (Ill. 2003); Miller v. Illinois Central R.R., 474 F.3d 951, 957 (7th Cir. 2007) (Illinois law). The jury was not asked to decide whether Mrs. Robinson’s negligence exceeded McNeil’s, as it should have been asked if, as the plaintiff argues, the tort law of Illinois rather than of Virginia governs the case. Several states have a connection to the events giving rise to Mrs. Robinson’s claim and therefore a potential, though for most of the states an attenuated, interest in the application of their law. She had bought the bottle of Children’s Motrin in Georgia but was living in Virginia when she took the drug and her initial medical treatment was administered there, after which she spent a month in the burn unit in a hospital in Balti- more where she was diagnosed with TEN. She moved with her husband and child to Illinois the following year. McNeil is a New Jersey corporation headquartered in Pennsylvania; we do not know where the drug was manufactured or the label composed and affixed to the bottle. The applicable conflicts rule is of course that of Illinois, the forum state, and it uses the common but loose “most significant relationship” test. Barbara’s Sales, Inc. v. Intel 6 No. 09-4011 Corp., 879 N.E.2d 910, 919-20 (Ill. 2007); Townsend v. Sears, Roebuck & Co., 879 N.E.2d 893, 898-99 (Ill. 2007). That test points presumptively to the law of the jurisdiction in which the tort occurred (that is, the lex loci delicti). Id. (and cases cited there). This means—because there is no tort without an injury, e.g., Rozenfeld v. Medical Protective Co., 73 F.3d 154, 156 (7th Cir. 1996) (Illinois law); Rice v. Nova Biomedical Corp., 38 F.3d 909, 915-16 (7th Cir. 1994) (same), and so a tort can’t be said to occur until an injury is produced—the place where the injury was inflicted. Kamelgard v. Macura, 585 F.3d 334, 341 (7th Cir. 2009) (Illinois law); cf. Townsend v. Sears, Roebuck & Co., supra, 879 N.E.2d at 905-06; Esser v. McIntyre, 661 N.E.2d 1138, 1141 (Ill. 1996); Miller v. Long-Airdox Co., 914 F.2d 976, 978 (7th Cir. 1990) (Illinois law); Restatement (Second) of Conflict of Laws § 146 (1971). That was Virginia. But because the injury is a continuing one, it is being experi- enced in Illinois. Indeed, Mrs. Robinson’s condition has deteriorated since she moved to Illinois and prob- ably will continue to deteriorate; therefore, the plaintiff argues, Illinois is the real site of the injury. The lingering or worsening of an injury over a con- siderable period is common in personal injury cases, rather than an exceptional circumstance that might justify a departure from the ordinary principles of choice of law. If you suffer permanent brain damage in a motorcycle accident in Virginia and later move to Illinois, your suffering and treatment will continue and your condition may deteriorate. But to make the con- tinuation or exacerbation of an injury a basis for applying Illinois tort law to your case would open vistas No. 09-4011 7 of forum shopping. Severely injured persons would move to the state whose law was most favorable to their tort claim and argue that that state had the “most signifi- cant relationship” to the injury because the plaintiff’s aggregate suffering and perhaps expense of medical treatment would be greatest there. To avoid this incen- tive to forum shop, the initial place of the injury is properly deemed the place in which the injury occurred. This was clearest when lex loci delicti was the rule gov- erning choice of law in tort cases. See Restatement (First) of Conflict of Laws § 377 (1934). Under the “most significant relationship” test, which demoted lex loci delicti to a presumption, a party can argue for applying the law of the state in which the greatest costs of the injury were incurred. But normally the argument should fail because of the encouragement that accepting it would give to forum shopping. See Reich v. Purcell, 432 P.2d 727, 730 (Cal. 1967) (Traynor, C.J.) (“if the choice of law were made to turn on events happening after the accident, forum shopping would be encouraged”); but see Fisher v. Professional Compounding Centers of America, Inc., 311 F. Supp. 2d 1008, 1015 (D. Nev. 2004) (Nevada law). A person who was severely handicapped for life as a result of a tort could choose among 50 states’ tort laws. Concern with forum shopping animates the parallel rule that makes venue proper in the jurisdiction in which the plaintiff’s claim accrues, which is usually the initial place of the injury. Quaid v. Baxter Healthcare Corp., 910 N.E.2d 1236, 1243-45 (Ill. App. 2009); Green v. North Arundel Hospital Ass’n, Inc., 730 A.2d 221, 229 n. 8 (Md. App. 1999). 8 No. 09-4011 There is, however, a potential ambiguity in the con- cept of “injury.” Suppose Mrs. Robinson’s symptoms had first appeared after she left Virginia. It is common for a disease to have a latency period, which is to say an interval between the infection or other trauma and when the first symptoms appear, and maybe in such a case the place where they first appear should, by analogy to the discovery rule in statutes of limitations, see, e.g., Ross v. Johns-Manville Corp., 766 F.2d 823, 827-28 (3d Cir. 1985); Wilson v. Johns-Manville Sales Corp., 684 F.2d 111, 115- 17 (D.C. Cir. 1982), be deemed the place of injury. Illinois’s intermediate appellate court so held in Mllar-Mintz v. Abbott Laboratories, 645 N.E.2d 278, 282 (Ill. App. 1994); cf. Montgomery v. Wyeth, 580 F.3d 455, 459-61 (6th Cir. 2009) (Tennessee law), but we have expressed doubt that the Supreme Court of Illinois would agree. Pittway Corp. v. Lockheed Aircraft Corp., 641 F.2d 524, 527-28 n. 5 (7th Cir. 1981) (Illinois law). It is true that the latency period of most diseases is too short, even if the victim realizes he has the disease before any symptoms appear, to prompt him to move to the state whose law would be most fa- vorable to his claim. But there are important excep- tions, as for asbestosis and DES, and in such cases the spur to forum shopping would be great if the law of the state of the first symptoms would govern. In any event the present case is not one of latency; Mrs. Robinson had symptoms of TEN before she left Virginia. So we think Virginia law does govern this case, and dooms her appeal because there was, as we shall see, at least some evidence of contributory negligence. But No. 09-4011 9 applying Illinois rather than Virginia tort law would not change the outcome of the appeal. Exactly what it means to say that a plaintiff’s contribu- tory negligence did (or did not) exceed the defendant’s, a determination required by Illinois’s comparative- fault rule, is unclear. 4 Fowler V. Harper, Fleming James, Jr. & Oscar S. Gray, Harper, James and Gray on Torts § 22.16, pp. 471, 475-76 (3d ed. 2007). But Davis v. United States, 716 F.2d 418, 429 (7th Cir. 1983) (Illinois law), suggests that it means that “though both parties to the accident, injurer and victim, could have avoided the accident at reasonable cost—for otherwise both would not have been negligent and the need to compare their negligence would not arise—one of them could have avoided the accident at a lower cost than the other, and he should bear a larger share of the burden of having failed to do so.” See also Wassell v. Adams, 865 F.2d 849, 854 (7th Cir. 1989) (also Illinois law). Mrs. Robinson appears to have been the party who could have avoided the injury at lower cost, assuming for the moment that her developing TEN was caused by the Motrin. For the evidence of McNeil’s negligence in selling Children’s Motrin, with or without a prescription and with or with- out additional warnings, was slight. Ibuprofen is an immensely popular drug for relief of pain and fever. Although the plaintiff contends that it’s unreasonably dangerous, or defective in design, either contention implying (if accepted) that the drug should be taken off the market, at the oral argument of the ap- peal her lawyer said “we’re not saying take it off the 10 No. 09-4011 market . . . we’re not saying it should be banned.” So maybe the plaintiff just wants it sold by prescription only. But this would increase the drug’s cost (because of the prescribing physician’s fee and the time required by the patient to obtain and fill the prescription) and thus reduce its availability. And how dangerous is ibuprofen, anyway? Adverse reactions (which include both allergic reactions and other side effects) to painkilling drugs are common, see N. Franklin Adkinson, Jr., “Drug Allergy,” Allergy: Principles & Practice 1212, 1216 (Elliott Middleton, Jr., et al. eds., 5th ed. 1998); Fred E. Karch & Louis Lasagna, “Adverse Drug Reactions: A Critical Review,” 234 JAMA 1236, 1236- 37 (1975); Roujeau et al., supra, at 347; Donald D. Stevenson & Ronald A. Simon, “Sensitivity to Aspirin and Nonsteroidal Antiinflammatory Drugs,” in Allergy, supra, at 1225. Ibuprofen does not appear to carry a risk of serious allergic reactions or other serious side effects that is greater than is created by drugs that might be substituted for it. Compare “Ibuprofen—Adverse Reactions,” www.merck.com/mmpe/lexicomp/ibuprofen. html#NF1185, to “Naproxen—Adverse Reactions,” www. merck.com/mmpe/lexicomp/naproxen.html#N143371, and to “Aspirin—Adverse Reactions,” www.merck.com/mmpe/ lexicomp/aspirin.html#N36183; see also Daniel H. Solo- man, MD, MPH, “Nonselective NSAIDs: Overview of Adverse Effects,” www.uptodate.com/online/content/topic. do?topicKey=treatme/7262; “ACPA Consumer Guide to Pain Medication & Treatment,” www.theacpa.org/ documents/ACPA%20Consumer%20Guide% 202010%20010410.pdf (all visited July 29, 2010). There is No. 09-4011 11 also no evidence that if the risk from ibuprofen is greater, it is not offset by the drug’s therapeutic prop- erties. Should aspirin be salable by prescription only? How about peanuts? The prevalence of TEN from all causes is estimated to be only between .4 and 1.2 cases per million users of the drug, and what fraction of that slight probability is due to ibuprofen is unknown and may be zero. For while it is true that SJS/TEN is primarily and perhaps ex- clusively caused by allergic reactions to drugs, there is considerable doubt, expressed by the defendant’s expert witness, Dr. Maja Mockenhaupt, one of the world’s leading authorities on the disease, that ibuprofen is one of them. There is unquestionably an association between SJS/TEN and ibuprofen because headache and fever, which are symptoms of the diseases, are commonly treated with drugs containing ibuprofen. But in such cases the causation runs from SJS/TEN to ibuprofen rather than vice versa. When a drug is claimed to cause the very symptoms for which it is a designated treat- ment, determining the direction of causation is difficult at best. Dr. Mockenhaupt was confident that, if there is a causal connection, still one dose could not cause the disease. That statement was important to the defense of contributory negligence because a reasonable jury could not have found Mrs. Robinson negligent in taking the first dose. It was the middle of the night and she had a headache and she was taking just the child’s dose. Furthermore, because headache is one of the earliest 12 No. 09-4011 symptoms of the disease and Mrs. Robinson had both headache and rash before she took the second dose of Motrin, and because, according to Mockenhaupt, the disease is usually latent for four days to four weeks after the patient first develops it, the fact that Mrs. Robin- son exhibited symptoms within a day of taking the first dose of Motrin, and the further fact that she had taken ibuprofen in other forms for years without an allergic reaction, convinced Dr. Mockenhaupt that Mrs. Robinson had developed TEN before she took the first dose. It might seem that the danger of developing TEN from taking ibuprofen (if there is such a danger) would be reduced if it were a prescription drug. Although there don’t appear to be any factors that predispose a person to have such a reaction to ibuprofen—factors that a doctor might elicit from a patient in deciding whether to write a prescription—the warning accompanying a prescrip- tion drug, because it is addressed to the prescribing doctor, can be more detailed than a warning for an over- the-counter drug, which is read only by the consumer. And indeed the FDA requires that the package insert for prescription ibuprofen drugs warn of the risk (slight as it is—maybe even nonexistent) that taking such a drug might cause SJS/TEN. But if Children’s Motrin were a prescription drug, Mrs. Robinson’s doctor would have prescribed it for her child because there is nothing to suggest that the child had some condition that might make a prescribing physician worry that the child might develop SJS/TEN No. 09-4011 13 from taking the drug. If it were prescribed, this would tend to allay rather than heighten any fears that Mrs. Robinson might have about taking the drug herself (in fact she had no fears). And anyway the warning about SJS/TEN, being directed to the physician and thus written in technical language, would mean nothing to her. A row in the 1996 package insert for prescription Children’s Motrin (which contains the same amount of ibuprofen as the over-the-counter version), headed “skin and appendages,” states that there is a 3 to 9 percent probability of developing a rash from taking the drug; that the incidence of toxic epidermal necrosis (TEN) in persons taking it is unknown but is less than 1 percent; and that there may be a causal relation between the drug and the disease. Letter from Willie D. Pagsuyuin, Director, Regulatory Affairs, McNeil Consumer Products Company, to Susan Raigrodski, Division of Anti-Inflamma- tory, Analgesic and Ophthalmologic Drug Products (HFD- 550) 12-13 (Apr. 16, 1996); see also “Children’s Motrin Ingredients,” www.motrin.com/page.jhtml?id=/motrin/ products/1_2_1.inc&sec=ingredients (visited July 29, 2010). (The current insert states that it’s unknown whether there is any causal relation between the drug and TEN. “Motrin,” www.pfizer.com/files/products/uspi_motrin.pdf (visited July 28, 2010).) What would Mrs. Robinson have learned from reading such a warning? The decision whether to permit a drug to be sold over the counter rather than just by prescription is for the FDA to make. U.S. Food and Drug Administration, “Nonprescription Drug Products: What We Do,” ww w.fda.gov/AboutFD A /C en tersO ffices/C D E R/ 14 No. 09-4011 ucm106342.htm (visited July 21, 2010); “Drug Applica- tions for Over-the-Counter Drugs,” www.fda.gov/Drugs/ DevelopmentApprovalProcess/HowDrugsareDeveloped andApproved/ApprovalA pplications/Over-the- CounterDrugs/default.htm (visited July 21, 2010); Dan R. Harlow, “The FDA’s OTC Drug Review: The Develop- ment and an Analysis of Some Aspects of the Procedure,” 32 Food Drug Cosmetic L.J. 248, 250-53 (1977); see also Weinberger v. Bentex Pharmaceuticals, Inc., 412 U.S. 645, 650- 51, 653-54 (1973). The agency bases its decision on whether the drug is safe and effective for use without a doctor’s permission, 21 C.F.R. § 310.200(b); and it has decided not to require that drugs containing ibuprofen be sold by prescription only. Granted, that doesn’t bar a court from holding that state law requires warnings on the label of an over-the-counter drug beyond what the FDA has required. Wyeth v. Levine, 129 S. Ct. 1187, 1198 (2009); Mason v. SmithKline Beecham Corp., 596 F.3d 387, 390-91, 394-95 (7th Cir. 2010); Demahy v. Actavis, Inc., 593 F.3d 428, 433 (5th Cir. 2010). The plaintiff argues that the label on the bottle of Children’s Motrin should have added “rash” to the other allergic reactions warned against and should have mentioned SJS/TEN as one of the possible allergic reactions and (since virtually no consumer who was not a physician would have heard of the disease) recited its horrific consequences. But then the label would have had to describe as well every other serious disease that might, however infrequently, be caused, or even just arguably caused (for it is unclear whether ibuprofen can cause SJS/TEN), by ibuprofen. And it would have to recite the No. 09-4011 15 symptoms of the disease if it was rare. The resulting information overload would make label warnings worth- less to consumers. See Troy A. Paredes, “Information Overload and its Consequences for Securities Regulation,” 81 Wash. U. L.Q. 417, 440-43 (2003); Howard Latin, “ ‘Good’ Warnings, Bad Products, and Cognitive Limitations,” 41 UCLA L. Rev. 1193, 1211-15 (1994); cf. Richard Craswell, “Taking Information Seriously: Misrepresentation and Nondisclosure in Contract Law and Elsewhere,” 92 Va. L. Rev. 565, 583-85 (2006); Mark Geistfeld, “Inadequate Product Warnings and Causation,” 30 U. Mich. J.L. Reform 309, 322 (1997). So forbidding the sale of ibuprofen over the counter would be very costly—probably more costly than letting consumers, having been warned by the label of possible allergic reactions, decide to stop taking the drug when symptoms that might have been caused by it appear. See Kelso v. Bayer Corp., 398 F.3d 640, 642-43 (7th Cir. 2005). The FDA directed McNeil (after the accident, however) to add “skin reddening, rash, and blisters” to the warnings on the label for Children’s Motrin. But though later requested to do so, the agency decided not to require mention of SJS/TEN (or SJS/TEN plus its horrible symptoms), believing with reason that the addition would confuse rather than enlighten. In a letter from Steven K. Galson, Director, FDA Center for Drug Evalu- ation and Research, to Roger E. Salisbury, No. 2005P- 0072/CP1 (June 22, 2006), the FDA stated: “We believe that the overall benefit versus risk profile for ibuprofen products remains very favorable when they are used according to the labeled instructions. It is in the interest 16 No. 09-4011 of the public health to maintain in the pediatric OTC market a range of therapeutic options for the short-term relief of pain. Further . . . other available OTC drugs for short-term relief of pain and fever can also be associated with serious, potentially life-threatening adverse events in certain settings and patient populations.” See also W. Page Keeton et al., Prosser & Keeton on the Law of Torts § 96, pp. 685, 687 (5th ed. 1984). And would the word “rash” on the label have dissuaded Mrs. Robinson from taking the drug? She testified and we accept that if the label, which she read when she bought the drug, had been really scary—had it men- tioned SJS/TEN as a possible allergic reaction and listed all the dreadful symptoms of TEN in particular—she would not have bought it. But it is implausible that if instead of saying “Allergy alert: Ibuprofen may cause a severe allergic reaction which may include: hives, facial swelling, asthma (wheezing), shock,” it had said “Allergy alert: Ibuprofen may cause a severe allergic reaction which may include: hives, rash, facial swelling, asthma (wheezing), shock,” she would not have bought it. If facial swelling, asthma, and shock are not sufficient deterrents, rash is not likely to be (or skin reddening or blisters, but the plaintiff’s particular concern is the absence of the word “rash” from the list of allergy symp- toms). Anyway she didn’t read or remember the warnings before taking the Motrin, so it wouldn’t matter what the label had said unless it had contained truly terrifying warnings that the state of medical knowledge would not have justified. No. 09-4011 17 Her doctor may have lulled her into thinking that Motrin couldn’t have been the source of her symptoms because he didn’t react to her telling him that she was taking the drug. She told him she’d taken it and he thought that her symptoms were caused by an allergy, yet he merely gave her drugs to combat the allergic reaction; he didn’t tell her to stop taking Motrin even though allergic reactions to NSAIDs are well known by the medical profession. See Roujeau et al., supra, at 347; Stevenson & Simon, supra, at 1225; Larkin v. Pfizer, Inc., 153 S.W.3d 758, 759-60 (Ky. 2004); see generally Jones v. Detroit Medical Center, No. 288710, 2010 WL 2010003 (Mich. App. May 20, 2010) (per curiam). Yet she didn’t sue him for malpractice, and she doesn’t argue that it was reasonable for her (and therefore not contributorily negligent) to rely on the doctor’s failure to warn her that Motrin might be the cause of her symp- toms. See DiLeo v. Nugent, 592 A.2d 1126, 1133 (Md. App. 1991); Angelo v. Diamontoni, 871 A.2d 1276, 1281-82 (Pa. Super. 2005). Even if she had argued this, all she could have gotten from the jury would have been a finding that she was not negligent to take the third dose of Motrin. A determination that she had been negligent in taking the second dose would have stood (we said that her taking the first dose was not negligent), and under Virginia’s rule of contributory negligence would have barred her claim. She argues, contrary to Dr. Mockenhaupt’s testimony, that the first dose caused her to develop SJS/TEN, and if this is right then any carelessness in taking the second 18 No. 09-4011 and third doses at worst aggravated the condition, just as if she’d delayed visiting the doctor. And then she would face a defense not of contributory negligence but only of failure to avoid avoidable consequences, which is the tort counterpart to the better-known contract doctrine of mitigation of damages. (Virginia uses the latter term for tort cases as well, Monahan v. Obici Medical Manage- ment Services, Inc., 628 S.E.2d 330, 336 (Va. 2006).) Such a failure reduces but does not preclude a damages award. Id. at 337; Sawyer v. Comerci, 563 S.E.2d 748, 754 (Va. 2002); Lawrence v. Wirth, 309 S.E.2d 315, 317-18 (Va. 1983); Barron v. Ford Motor Co. of Canada Ltd., 965 F.2d 195, 199 (7th Cir. 1992); Ellerman Lines, Ltd. v. S.S. President Harding, 288 F.2d 288, 289-90 (2d Cir. 1961) (Friendly, J.). But this argument has been forfeited because the plaintiff did not propose jury instructions that would have asked the jury to determine whether she had merely failed to avoid avoidable consequences, rather than having been contributorily negligent. In any event, the argument would have been unlikely to persuade the jury. There was conflicting testimony on whether the first dose by itself caused her to develop TEN. While testifying that ibuprofen does not cause SJS/TEN, Dr. Mockenhaupt was emphatic that in any event the first dose would not have done so. The jury must have disbe- lieved the first assertion because it found McNeil to have been negligent, implying that the ibuprofen in Motrin caused the plaintiff’s TEN. (“ ‘Proof of negligence in the air . . . will not do,’ ” Palsgraf v. Long Island R.R., 162 N.E. 99, 99 (N.Y. 1928)—that is, an injury is required even for prima facie liability, since as we said earlier there is no tort No. 09-4011 19 without an injury.) But the jury’s further finding that the plaintiff was contributorily negligent implies that it believed Mockenhaupt’s second assertion (that the first dose had not caused the plaintiff’s TEN), as it was en- titled to do for she is, as we said, one of the world’s leading experts on SJS/TEN, as none of the plaintiff’s experts was. The two assertions were not inconsistent. There is no inconsistency in testifying that a dog cannot walk on its hind legs but that if that’s wrong and it can, still it can’t (unless perhaps its name is Faith, see “Faith the Dog’s Official Website,” www.faiththedog.info (visited July 29, 2010)) walk on its hind legs for an hour at a time. A dog can walk on its hind legs, with some training; and a witness who denies this might be thought less credible, in testifying that a dog can’t walk on its hind legs for an hour at a time, than if she’d known that a dog can walk on its hind legs. But such a tension is for a jury to weigh. To summarize, there was enough evidence that the plaintiff was contributorily negligent to bar her claim under Virginia law, and enough evidence that her contributory negligence exceeded the defendant’s negligence to bar her claim even if Illinois rather than Virginia law applied. But the plaintiff makes the further argument that in closing argument McNeil’s lawyer made a “judicial admission” that the plaintiff had not been contributorily negligent. What the lawyer said was: “We are, of course, not blaming Karen Robin- son for her own injuries. We never have. We never will.” What he meant—for he was speaking to laypersons—was that McNeil was not contending that Mrs. Robinson had been justly punished for being careless. The suffering 20 No. 09-4011 she has experienced from SJS/TEN has been dispropor- tionate to her failure to exercise due care in consuming McNeil’s product. It would be monstrous to suggest otherwise, and McNeil’s lawyer didn’t want to leave the jury with the impression that he was trying to insinuate such a thought—an impression that the plain- tiff’s lawyer had sought to create by stating that “they [the defendants] blame her for being misled.” A judicial admission is a statement, normally in a pleading, that negates a factual claim that the party making the statement might have made or considered making. Were the plaintiff’s conception of judicial admis- sions accepted, statements made by lawyers in opening and closing arguments, in making objections, at side bars, and in questioning witnesses would be treated as pleadings and searched for remarks that might be con- strued as admissions though neither intended nor under- stood as such. Trials would be turned into minefields. That is why “in order to qualify as judicial admissions, an attorney’s statements must be deliberate, clear and unam- biguous.” MacDonald v. General Motors Corp., 110 F.3d 337, 340 (6th Cir. 1997); see also Best v. District of Columbia, 291 U.S. 411, 415-17 (1934); Oscanyan v. Arms Co., 103 U.S. 261, 263-64 (1880); McCaskill v. SCI Management Corp., 298 F.3d 677, 680 (7th Cir. 2002); Butynski v. Springfield Terminal R.R., 592 F.3d 272, 277-78 (1st Cir. 2010). That standard has not been satisfied in this case. The plaintiff complains finally of the district court’s refusal to allow an amendment to the final pretrial order that would have added a claim, under Virginia law, of breach of implied warranty. In their proposed final No. 09-4011 21 pretrial order, the parties had agreed that the plaintiff would not be pursuing such a claim even though it was in the complaint. Between then and the filing of the final pretrial order the district court decided that Virginia law would apply rather than Illinois law, and Virginia law as we know does not recognize strict liability as a ground for products liability. The plaintiff moved promptly to amend the final pretrial order to reinstate the warranty claim, but with trial scheduled to begin in just a few days the judge refused to allow the amendment. The plaintiff’s lawyer says he had every reason to think that Illinois law would apply and that’s why he dropped the warranty claim. But there was no reason to think that—indeed it was unlikely that Illinois law would apply, as we have explained, or that McNeil would fail to argue for applying Virginia law, which was more favorable to it than Illinois law. McNeil had already proposed jury instructions based on Virginia law, for example an instruction—the key instruction in the case—that contributory negligence if proved would be a complete defense to the plaintiff’s negligence claim. So the plaintiff dropped the ball; and if the fault was, as undoubtedly it was, her lawyer’s, she may have a remedy against him. But given the proximity of trial, the judge was within his authority in holding the lawyer to his waiver of the breach of warranty claim—and for the further reason that such a claim was unlikely to succeed, except possibly in confusing the jury. It can’t be argued seriously that McNeil implicitly warranted that 22 No. 09-4011 Children’s Motrin will not cause SJS/TEN. That would imply that the company had a duty to guarantee against every conceivable adverse consequence of taking the drug, however remote, esoteric, or even conjectural; and that is not the law. Featherall v. Firestone Tire & Rubber Co., 252 S.E.2d 358, 366-67 (Va. 1979); Adelman-Tremblay v. Jewel Cos., 859 F.2d 517, 521-22 (7th Cir. 1988); Keeton et al., supra, § 96, p. 687. The FDA decided not to require such a warning because it would confuse rather than inform; and a court cannot order a drug company to place on a label a warning if there is “clear evidence” that the FDA would not approve it. Wyeth v. Levine, 129 S. Ct. 1187, 1198 (2009); Lofton v. McNeil Consumer & Specialty Phar- maceuticals, 682 F. Supp. 2d 662, 678 (N.D. Tex. 2010). The “clear evidence” in this case is the agency’s refusal to require a reference to SJS/TEN on the label of over-the- counter drugs containing ibuprofen, when it had been asked to do so in the submission to which the agency was responding. And it would be odd to think that McNeil had a legal duty to guarantee against a risk that the FDA thought not worth warning against. Even if the plaintiff had been permitted to try to estab- lish a breach of implied warranty, and had succeeded in making a prima facie case, she would have been unlikely to prevail at trial. Although there is no defense of con- tributory negligence, as such, to breach of implied war- ranty, the consumer’s conduct is not irrelevant to the seller’s liability. Law often describes the same things by different names. We gave an example earlier— “mitiga- tion of damages” in a contract case and “avoidable con- sequences” in a tort case. Even when a defendant’s No. 09-4011 23 liability is strict, as in implied warranty and strict products liability cases, a plaintiff who fails to avoid a danger that is either open, in the sense of visible, or obvious (“open and obvious” is the conventional name of this defense), or who misuses the product, is barred from relief. Wood v. Bass Pro Shops, Inc., 462 S.E.2d 101, 103 (Va. 1995); Harris-Teeter, Inc. v. Burroughs, 399 S.E.2d 801, 802-03 (Va. 1991); Mesman v. Crane Pro Services, 512 F.3d 352, 358-59 (7th Cir. 2008); Freeman v. Case Corp., 118 F.3d 1011, 1014 (4th Cir. 1997) (Virginia law). Both are apt descriptions (the first more so) of the plaintiff’s conduct in continuing to dose herself with Motrin after experiencing an allergic reaction that began shortly after she took the first dose. A FFIRMED. 8-11-10