813 P.2d 89 (1991)

Ilo Marie GRUNDBERG, individually, and Janice Gray, as personal representative of the Estate of Mildred Lucille Coats, deceased, Plaintiffs,
v.
The UPJOHN COMPANY, a Delaware corporation, Defendant.

No. 900573.

Supreme Court of Utah.

May 14, 1991. Rehearing Denied June 26, 1991.

Gary W. Pendleton, St. George, Earle F. Lasseter, Thomas R. Vuksinick, H. Ross *90 Workman, Salt Lake City, and C. Neal Pope, Daniel W. Sigelman, Steven W. Saccoccia, Edward H. Kellogg, Atlanta, Ga., for plaintiffs.

Merlin O. Baker, Paul S. Felt, Thomas L. Kay, Steven J. Aeschbacher, Salt Lake City, and Lane D. Bauer, Stephen E. Scheve, Laura D. Stith, Kansas City, Mo., for defendant.

DURHAM, Justice:

This case comes to us pursuant to rule 41 of the Utah Rules of Appellate Procedure as a question certified from the United States District Court for the District of Utah. The issue before us is whether Utah adopts the "unavoidably unsafe products" exception to strict products liability as set forth in comment k to section 402A of the Restatement (Second) of Torts (1965) ("comment k"). This question presents an unanswered issue of law for original disposition by this court.

We hold that a drug approved by the United States Food and Drug Administration ("FDA"), properly prepared, compounded, packaged, and distributed, cannot as a matter of law be "defective" in the absence of proof of inaccurate, incomplete, misleading, or fraudulent information furnished by the manufacturer in connection with FDA approval. We acknowledge that by characterizing all FDA-approved prescription medications as "unavoidably unsafe," we are expanding the literal interpretation of comment k.

The following facts are taken from the federal district court's certification order. Mildred Lucille Coats died at age 83 from gunshot wounds inflicted by her daughter, Ilo Grundberg, on June 19, 1988. Grundberg and Janice Gray, the personal representative of Coat's estate, brought this action, alleging that Grundberg shot her mother as a result of ingesting the drug Halcion, a prescription drug manufactured by defendant Upjohn to treat insomnia.^[1]

Plaintiffs allege that Grundberg took a .5 milligram dose of Halcion the day she shot her mother. They allege that this dose was recommended by her physician and was consistent with Upjohn's recommended dosage. Plaintiffs assert that Grundberg shot her mother while in a state of Halcion-induced intoxication, which allegedly included side effects such as depression, psychosis, depersonalization, aggressive assaultive behavior, and homicidal compulsion.

Plaintiffs' complaint states several causes of action, including common law negligence and strict liability. Plaintiffs claim that Upjohn failed to adequately warn about certain adverse side effects of Halcion and that Halcion was defectively designed. The failure-to-warn claim is scheduled for trial. The strict liability claim based on design defect is the subject of Upjohn's pending summary judgment motion, the outcome of which depends on this court's resolution of the certified question.

The parties agree that the Restatement (Second) of Torts section 402A, comment k (1965) and the principles it embodies provide an exemption from strict liability for a claimed design defect in the case of products that are "unavoidably unsafe." In moving for partial summary judgment, Upjohn argued that public policy supporting the research and development of new drugs requires a holding that all FDA-approved prescription medications are "unavoidably unsafe products" under comment k and, as such, manufacturers of those drugs would not be liable for a claim based on defective design. Plaintiffs argue that whether a drug is "unavoidably unsafe" must be determined on a case-by-case basis, with a determination in each case of whether the specific drug's benefit exceeded its risk at the time it was distributed. The district court found this to be a controlling question of law and certified it to this court.

Specifically, the issues we address at the request of the federal court are:

1. Does Utah adopt the "unavoidably unsafe products" exception to strict products liability as set forth in comment k to *91 section 402A of the Restatement (Second) of Torts (1965)?

(a) If Utah does adopt comment k, should FDA-approved prescription drugs be deemed as a matter of law to have satisfied the "unavoidably unsafe" prerequisite to the comment k exception, or should that determination be made on a case-by-case basis?

(b) If Utah does adopt comment k, and if it is further determined that its application to FDA-approved prescription drugs ought to be made on a case-by-case basis, is such determination a threshold question for the trial court or a question properly to be presented to the jury?

(c) If it is determined that comment k is to be applied to FDA-approved prescription drugs on a case-by-case basis, is evidence pertaining to adverse side-effects from the drug which are not alleged to have been personally suffered by the plaintiff relevant to the "unavoidably unsafe" determination?

I. UTAH LAW ON STRICT LIABILITY

Section 402A of the Restatement (Second) of Torts (1965) addresses the strict liability of sellers of products. This court adopted section 402A, of which comment k is one provision, in Ernest W. Hahn, Inc. v. Armco Steel Co., 601 P.2d 152, 158 (Utah 1979). Since then, we have adhered to section 402A and to at least one of its accompanying comments. See Mulherin v. Ingersoll-Rand Co., 628 P.2d 1301 (Utah 1981) (applying section 402A); Dowland v. Lyman Products for Shooters, 642 P.2d 380, 381 n. 2 (Utah 1982) (applying comment g). We have not addressed the application of comment k in the context of prescription drugs or otherwise.^[2] Although two Utah statutes address the liability of product and drug manufacturers, they do not directly address the comment k issues. See Utah Code Ann. § 78-15-6(3) (1987) (rebuttable presumption that product was not defective if manufactured according to industry standards), § 78-18-2 (punitive damages unavailable if drug was approved by FDA).

In its entirety, comment k reads:

k. Unavoidably unsafe products. There are some products which, in the present state of human knowledge, are quite incapable of being made safe for their intended and ordinary use. These are especially common in the field of drugs. An outstanding example is the vaccine for the Pasteur treatment of rabies, which not uncommonly leads to very serious and damaging consequences when it is injected. Since the disease itself invariably leads to a dreadful death, both the marketing and the use of the vaccine are fully justified, notwithstanding the unavoidable high degree of risk which they involve. Such a product, properly prepared, and accompanied by proper directions and warning, is not defective, nor is it unreasonably dangerous. The same is true of many other drugs, vaccines, and the like, many of which for this very reason cannot legally be sold except to physicians, or under the prescription of a physician. It is also true in particular of many new or experimental drugs as to which, because of lack of time and opportunity for sufficient medical experience, there can be no assurance of safety, or perhaps even of purity of ingredients, but such experience as there is justifies the marketing and use of the drug notwithstanding a medically recognizable risk. The seller of such products, again with the qualification that they are properly prepared and marketed, and proper warning is given, where the situation calls for it, is not to be held to strict liability for unfortunate consequences attending their use, merely because he has undertaken to supply the public with an apparently useful and desirable product, attended with a known but apparently reasonable risk.

Comment k establishes an exception to the strict products liability section 402A *92 imposes on "[o]ne who sells any product in a defective condition unreasonably dangerous to the user or consumer or to his [or her] property... ." § 402A(1). This liability applies whether or not "the seller has exercised all possible care in the preparation and sale of his product... ." § 402A(2)(a). Comment g defines a "defective condition" as a condition "not contemplated by the ultimate consumer which will be unreasonably dangerous to [that consumer]." Comment k, however, defines a category of "unavoidably unsafe" products that "when properly prepared, and accompanied by proper directions and warning, [are] not defective, nor ... unreasonably dangerous." (Emphasis in original.)

We agree with comment k's basic proposition — that there are some products that have dangers associated with their use even though they are used as intended. We also agree that the seller of such products, when the products are properly prepared and marketed and distributed with appropriate warnings, should not be held strictly liable for the "unfortunate consequences" attending their use. Thus, we adopt comment k's basic policy as the law to be applied in this state and must now turn to the issue of how to apply that policy.

II. APPLICATION OF COMMENT K

As a condition to its application, comment k requires that the product be "properly prepared, and accompanied by proper directions and warning... ." There are three types of product defects: manufacturing flaws, design defects, and inadequate warnings regarding use. See Prosser & Keeton, The Law of Torts § 99, at 695-98 (5th ed. 1984); Savina v. Sterling Drug, Inc., 247 Kan. 105, 795 P.2d 915, 923 (1990). By its terms, comment k excepts unavoidably unsafe products from strict liability only to the extent that the plaintiff alleges a design defect; comment k's immunity from strict liability does not extend to strict liability claims based on a manufacturing flaw or an inadequate warning. The purpose of comment k is to protect from strict liability products that cannot be designed more safely. If, however, such products are mismanufactured or unaccompanied by adequate warnings, the seller may be liable even if the plaintiff cannot establish the seller's negligence. Toner v. Lederle Laboratories, 112 Idaho 328, 732 P.2d 297, 305 (1987). Both parties agree in this case that the prerequisite to a comment k exemption — that the drug "was properly prepared and accompanied by warnings of its dangerous propensities" — must be established on a case-by-case basis. This limitation on the scope of comment k immunity is universally recognized.^[3]

Even in the case of a clearly alleged design defect, however, comment k is unclear on the scope of its protection.^[4] Until recently, most courts refrained from applying a design defect theory to products liability cases involving prescription drugs. Beginning with Brochu v. Ortho Pharmaceutical Corp., 642 F.2d 652 (1st Cir.1981), however, and more recently in Savina v. Sterling Drug, Inc., 247 Kan. 105, 795 P.2d 915 (1990), some states have permitted recovery for a strict liability claim based on the theory that the drug was defectively designed.

*93 Some courts have applied comment k on a case-by-case basis, conditioning application of the exemption on a finding that the drug is in fact "unavoidably unsafe." See Savina, 795 P.2d 915; Toner, 732 P.2d 297; see also Feldman v. Lederle Laboratories, 97 N.J. 429, 479 A.2d 374, 382-83 (1984) (involving allegations of failure to warn, but stating, "Whether a drug is unavoidably unsafe should be decided on a case-by-case basis ... ."); Collins v. Eli Lilly Co., 116 Wis.2d 166, 342 N.W.2d 37, 52 (1984) (comment k applicable only if drug in question was placed on market without adequate testing because of exigent circumstances); Patten v. Lederle Laboratories, 676 F. Supp. 233 (D.Utah 1987) (federal district court predicting that Utah would adopt comment k).

California was the first state to fashion a risk/benefit test to determine which drugs are entitled to comment k protection. In Kearl v. Lederle Laboratories, 172 Cal. App.3d 812, 218 Cal. Rptr. 453 (1985),^[5] the California Court of Appeal specifically discussed the problems society would face by subjecting drugs to the same accountability as other products, allowing unlimited redress for plaintiffs injured by pharmaceutical products. Such problems, the court noted, include delayed availability of needed drugs and imposition of the costs of research, development, and marketing of new products beyond that which manufacturers, especially small manufacturers, might be willing to risk. 218 Cal. Rptr. at 459 (quoting Feldman, 460 A.2d at 209).

The Kearl court expressed discomfort, however, with the "mechanical" method by which many appellate courts had concluded that drugs are entitled to special treatment. 218 Cal. Rptr. at 463. Thus, Kearl set forth a risk/benefit analysis to be carried out by the trial court on a case-by-case basis. Id. at 463-64. Under this approach, a product may be deemed unavoidably unsafe and thus exempt from a strict liability design defect cause of action only if the court concludes that (1) the product was intended to provide an exceptionally important benefit, and (2) the risk posed was substantial and unavoidable when distributed. Id. at 464.^[6]

Idaho adopted and to some extent refined the Kearl approach in Toner v. Lederle Laboratories, 112 Idaho 328, 732 P.2d 297 (1987), a case addressing a suit against the manufacturer of a vaccine to immunize against diphtheria, pertussis, and tetanus ("DPT"). Toner required the drug manufacturer to prove at trial, on a case-by-case basis, that the benefits of the drug outweighed the risks at the time of marketing. 732 P.2d at 305-09. To qualify as an "unavoidably unsafe product" under this approach, "there must be, at the time of the subject products' distribution, no feasible alternative design which on balance accomplishes the subject product's purpose with a lesser risk." Id. at 306 (citing Belle Bonfils Memorial Blood Bank v. Hansen, 665 P.2d 118, 123 (Colo. 1983)); Kearl, 218 Cal. Rptr. at 464. If there were alternative drug product designs that could have effectively achieved the same purpose, the court reasoned, the risk would not be "unavoidable" or "apparently reasonable" and the marketing and use of the product would not be justified. Toner, 732 P.2d at 306.

In direct contrast to those courts applying comment k's immunity on a case-by-case basis are courts holding that all prescription drugs are entitled as a matter of law to the exemption from strict liability claims based on design defect. In Brown v. Superior Court, 44 Cal.3d 1049, 245 *94 Cal. Rptr. 412, 751 P.2d 470 (1988), the court addressed claims brought by plaintiffs who sued drug companies for injuries allegedly arising from their mothers' in utero exposure to diethystilbestrol, a synthetic hormone marketed for use during pregnancy. The court weighed the problem of whether imposing strict liability on drug manufacturers comports with the traditional goals of tort law, namely, deterrence and cost distribution. 751 P.2d at 478. The court acknowledged that a drug might be safer if pharmaceutical companies withheld it from the market until scientific skill and knowledge advanced to the point where all dangerous side effects could be discovered. Id. at 479. There was concern, however, that this delay, when added to the delay normally required for the FDA to approve a new drug, would not serve the public welfare. The court cited examples of several potentially useful drugs being withdrawn from the market or their availability seriously curtailed because of the liability crisis. Id. at 479-80.

The Brown court acknowledged the appeal of the Kearl cost/benefit approach, yet found the "mini-trial" procedure unworkable because of its negative impact on the development and marketing of new drugs. Brown, 751 P.2d at 481. Another of the Brown court's objections to Kearl was that it left the trial court to hear and resolve mixed questions of law and fact, placing the trial court in the role of fact finder. Brown, 751 P.2d at 481-82. The court found the cost/benefit test too openended and predicted that it would lead to disparate treatment of the same drug by different judges. Id. at 482.

The Brown court stressed three public policies mitigating against imposing strict liability for prescription drugs. First, drug manufacturers might stop producing valuable drugs because of lost profits resulting from lawsuits or the inability to secure adequate insurance. Id. at 479-80. Second, consumers have a vested interest in prompt availability of new pharmaceutical products. Imposing strict liability for design defects might cause manufacturers to delay placing new products on the market, even after those products receive FDA approval. Id. at 479. Finally, the added expense of insuring against strict liability and additional research programs might cause the cost of medication to increase to the extent that it would no longer be affordable to consumers. Id.

The plaintiffs and amici curiae in Brown asserted that the language of comment k cannot be interpreted to grant blanket immunity from strict liability to all prescription drugs. Rather, they asserted that only those drugs that are "unavoidably dangerous" are eligible for such protection. Id. at 482 n. 11. The court, although conceding that the comment is not entirely clear on this point, noted that "the comment was intended to and should apply to all prescription drugs." Id. (emphasis added). Brown concluded that "because of the public interest in the development, availability, and reasonable price of drugs, the appropriate test for determining responsibility is the test stated in comment k... ." Id. at 477.

In Castrignano v. E.R. Squibb & Sons, Inc., 546 A.2d 775 (R.I. 1988), the Rhode Island Supreme Court had the opportunity to review Toner's case-by-case risk/benefit analysis and Brown's broad exemption approach when formulating its own approach to drug products liability. The Rhode Island court opted for the more restrictive case-by-case approach, id. at 781, and developed a directed verdict standard to balance the roles of the judge and the jury in applying comment k. Id. at 781-82. Under the Castrignano test, if at the time of marketing the apparent benefit outweighed the apparent risk, comment k applies and recovery for design defect is precluded. Id. If a trial judge concludes that reasonable minds could not differ in deciding that a drug's benefit exceeds its risk, then as a matter of law, the trial judge can extend comment k protection. Id. at 782. If the judge feels that reasonable minds could differ on the question, the judge must submit the issue to the jury. Id.

In reviewing the approaches of other jurisdictions toward strict products liability for design defects in drug products, we are troubled by the lack of uniformity and certainty *95 inherent in the case-by-case approach and fear the resulting disincentive for pharmaceutical manufacturers to develop new products. Toner's attempt to clarify the "unreasonably dangerous" standard seriously curtails the defendants' chances of success in establishing comment k immunity as a matter of law. One commentator notes that a defendant would have an easier time rebutting a plaintiff's prima facie case of design defect under the traditional standard than meeting the tough burden of "earning" the comment k exemption. See Reilly, The Erosion of Comment k, 14 U. Dayton L.Rev. 255, 266 (1989).

Toner applied a very literal and restrictive interpretation of comment k. For example, the comment cites examples of certain drugs and vaccines as products that "supply the public with an apparently useful and desirable product, attended with a known but apparently reasonable risk." Based on this language, Toner opined that (1) comment k "[c]learly ... contemplates a weighing of the benefit of the product against its risk" and (2) was never intended to exempt all drugs from strict liability. Toner, 732 P.2d 297, 306. Even if we agree with the court in Toner that the comment contemplates a "weighing" of the drug's risks and benefits, we find it unnecessary to conclude that a court is the proper body to engage in that weighing process. Furthermore, we need not be bound by the specific language of comment k and may adopt and apply its fundamental policy without restricting ourselves to what we perceive to be its literal interpretation.

Castrignano's somewhat nebulous standard, designed to control when the question of comment k's application reaches the jury, leaves a great deal of room for plaintiffs to manuever cases to the jury. Additionally, like the traditional case-by-case approach, this standard ignores the effectiveness of the FDA's regulatory process.

We agree with Brown that the case-by-case method first articulated in Kearl is unworkable, even in light of Toner's refinement of the test. We find the Brown result more in line with the public policy considerations in the important area of pharmaceutical product design. We do not agree, however, with the Brown court's apparent attempt to use the plain language of comment k as the vehicle for exempting all prescription drugs from strict liability rather than relying on the policies underlying that comment.

The American Law Institute's restatements are drafted by legal scholars who attempt to summarize the state of the law in a given area, predict how the law is changing, and suggest the direction the law should take. The restatement serves an appropriate advisory role to courts in approaching unsettled areas of law. We emphasize, however, that section 402A of the Restatement (Second) of Torts, as drafted in 1965, is not binding on our decision in this case except insofar as we explicitly adopt its various doctrinal principles. We agree with the principle comment k embodies, that manufacturers of unavoidably dangerous products should not be liable for a claim of design defect. We are persuaded that all prescription drugs should be classified as unavoidably dangerous in design because of their unique nature and value, the elaborate regulatory system overseen by the FDA, the difficulties of relying on individual lawsuits as a forum in which to review a prescription drug's design, and the significant public policy considerations noted in Brown. We therefore reach the same conclusion as did the California Supreme Court in Brown, albeit pursuant to a slightly different rationale.

III. UNIQUE CHARACTERISTICS OF DRUGS

Because prescription drugs are chemical compounds designed to interact with the chemical and physiological processes of the human body, they will almost always pose some risk of side effects in certain individuals. Despite these risks, new drugs are continually approved by the FDA because of their social benefit in saving lives and alleviating human suffering. The health care system and general standard of living in this country, for example, would be seriously impaired without such essential drug *96 products as antibiotics that allow quick recovery from ailments that were once debilitating or even fatal. See 37 Food, Drug & Cosmetic L.J. 15 (1982).

In addition, because the expansion of tort liability is justified primarily on the basis of deterring or transferring the cost of injuries, it is appropriate to consider the increased costs that could result from the curtailment of the production of prescription drugs. One commentator notes:

[D]rugs are our most cost-effective input in supplying the demand for health. A ten-dollar prescription is frequently a substitute for $2,000 worth of hospital services — a substitute that produces a positive outcome with much higher frequency than hospital care... . If we are serious about minimizing costs, our best bet is to increase the number of drug innovations.

Brozen, Statements, Drugs & Health: Economic Issues and Policy Objectives, 305 (Helms ed. 1981) (emphasis added).

Despite inherent risks, and in contrast to any other product, society has determined that prescription medications provide a unique benefit and so should be available to physicians with appropriate warnings and guidance as to use. The federal government has established an elaborate regulatory system, overseen by the FDA, to control the approval and distribution of these drugs. See 21 U.S.C. §§ 301-393. No other class of products is subject to such special restrictions or protections in our society.

IV. FDA REGULATION

Congress created the FDA to "protect consumers from dangerous products." United States v. Sullivan, 332 U.S. 689, 696, 68 S.Ct. 331, 335, 92 L.Ed. 297 (1948). In its role as "both a health promoter ... and ... a public protector," the FDA employs a comprehensive scheme of premarket screening and post-market surveillance to ensure the safety and efficacy of all licensed medications. 50 Fed.Reg. 7452 (1985).

Before licensing a new medication, the FDA employs an extensive screening mechanism to ensure that the potential benefits of the product outweigh any associated risks. The manufacturer initiates the review by submitting an Investigational New Drug Application ("IND"), containing information about the drug's chemistry, manufacturing, pharmacology, and toxicology. See 21 U.S.C. § 355(b)(1) (Supp. 1991); 21 C.F.R. § 312.21 (1990). If the FDA approves the IND, the drug's sponsor may gather data on clinical safety and efficacy needed for a New Drug Application ("NDA"), the formal license application. The NDA must include very detailed reports of all animal studies and clinical testing performed with the drug, reports of any adverse reactions, and any other pertinent information from world-wide scientific literature. 21 U.S.C. § 355(b) (Supp. 1991); 21 C.F.R. § 314.50 (1990).

The new drug approval process can require years of testing and review. By the time an NDA is submitted, it often consists of thousands of pages of material describing studies of the drug in several hundred to several thousand patients. See 47 Fed. Reg. 46626 (Oct. 19, 1982). The FDA carefully scrutinizes the data supporting the NDA, requiring "substantial evidence" consisting of adequate and well-controlled investigations. 21 U.S.C. § 355(d) (Supp. 1991). The application is reviewed by physicians, pharmacologists, chemists, microbiologists, statisticians, and other professionals within the FDA's National Center for Drugs and Biologics who are experienced in evaluating new drugs. 47 Fed. Reg. 46626 (Oct. 19, 1982). Recommendations by those professionals are then reviewed by management personnel within the National Center for Drugs and Biologics before the FDA makes a final determination to approve or reject the new drug application. Id.

Elaborate premarket screening, however, does not ensure review of approved prescription medications where adverse reactions may appear after extensive preapproval testing. For this reason, the FDA also conducts extensive post-market surveillance. All reports of adverse drug reactions ("ADRs") must be reported to the *97 FDA, regardless of whether the physician, the manufacturer, or others believe the reaction to be drug-related. 21 C.F.R. § 314.80(b). The manufacturer must also periodically submit reports as to what actions it took in response to ADRs and must submit data from any post-marketing studies, reports in the scientific literature, and foreign marketing experience. 21 C.F.R. §§ 314.80(b), .80(c). The FDA has authority to enforce these reporting requirements; any failure to comply may subject a manufacturer to civil and criminal penalties. 21 U.S.C. §§ 332-34 (1972 & Supp. 1991). In response to its surveillance findings, the FDA may require labeling changes or if necessary withdraw NDA approval and thereby revoke the license to market the medication. Id. at § 355(e).

We find this extensive regulatory scheme capable of and appropriate for making the preliminary determination regarding whether a prescription drug's benefits outweigh its risks. The structured follow-up program imposed by law ensures that drugs are not placed on the market without continued monitoring for adverse consequences that would render the FDA's initial risk/benefit analysis invalid. Allowing individual courts and/or juries to continually reevaluate a drug's risks and benefits ignores the processes of this expert regulatory body and the other avenues of recovery available to plaintiffs.

We note that the Utah Legislature has recognized the value of the FDA approval process and the public interest in the availability and affordability of prescription drugs by restricting the extent of liability for injuries resulting from the use of those drugs. Utah Code Ann. § 78-18-2(1) (Supp. 1990) states that "punitive damages may not be awarded if a drug causing the claimant's harm: (a) received premarket approval or licensure by the Federal Food, Drug, and Cosmetic Act, 21 U.S.C. Section 301 et seq... ." This policy, designed to avoid discouraging manufacturers from marketing FDA-approved drugs, applies even to drugs marketed with inadequate warnings.

The legislature has also acknowledged the important role of governmental standards in Utah Code Ann. section 78-15-6(3).^[7] In that section, the legislature declared that there is a rebuttable presumption that a product which fully complies with the applicable government standards at the time of marketing is not defective.^[8]

Our prior case law supports this approach as well. In Barson v. E.R. Squibb & Sons, Inc., 682 P.2d 832 (Utah 1984), we addressed the sufficiency of evidence for a claim that a drug manufacturer negligently failed to warn of risks associated with its product. We held that even after meeting governmental requirements, if there are dangers about which the drug manufacturer knew or should have known, the manufacturer may be subject to liability. Id. at 836. Thus, consistent with our holding in this case, if a manufacturer knows or should know of a risk associated with its product, it is directly liable to the patient if it fails to adequately warn the medical profession of that danger. Id. at 835.

*98 Moreover, the standard in Utah to which drug manufacturers must adhere to establish an adequate warning is very strict. In Barson, we stated:

In determining whether a manufacturer has breached that duty [to adequately warn] and the extent to which a manufacturer is required to know of dangers inherent in its drug, it is important to point out that the drug manufacturer is held to be an expert in its particular field and is under a "continuous duty ... to keep abreast of scientific developments touching upon the manufacturer's product and notify the medical profession of any additional side effects discovered from its use." The drug manufacturer is responsible therefore for not only "actual knowledge gained from research and adverse reaction reports," but also for constructive knowledge as measured by scientific literature and other available means of communication.

Barson, 682 P.2d at 835-36 (emphasis added).

V. PROPER FORUM FOR RISK/BENEFIT ANALYSIS

Finally, we do not believe that a trial court in the context of a products liability action is the proper forum to determine whether, as a whole, a particular prescription drug's benefits outweighed its risks at the time of distribution. In a case-by-case analysis, one court or jury's determination that a particular drug is or is not "defectively designed" has no bearing on any future case. As a result, differences of opinion among courts in differing jurisdictions leaves unsettled a drug manufacturer's liability for any given drug. Although the FDA may have internal differences of opinion regarding whether a particular new drug application should be approved, the individuals making the ultimate judgment will have the benefit of years of experience in reviewing such products, scientific expertise in the area, and access to the volumes of data they can compel manufacturers to produce. Nor is the FDA subject to the inherent limitations of the trial process, such as the rules of evidence, restrictions on expert testimony, and scheduling demands.^[9]

One commentator has argued that courts as a whole are unsuited to render responsible judgments in the design defect area generally. See Henderson, Judicial Review of Manufacturers' Conscious Design Choices: The Limits of Adjudication, 73 Colum.L.Rev. 1531 (1973). He argues that decisions in this area are arbitrary due to their "polycentric" nature in which "each point for decision is related to all the others as are the strands of a spider web." Id. at 1536. These issues are difficult to litigate because

[i]f one strand is pulled, a complex pattern of readjustments will occur throughout the entire web. If another strand is pulled, the relationships among all the strands will again be readjusted. A lawyer seeking to base [an] argument upon established principle and required to address himself in discourse to each of a dozen strands, or issues, would find [the] task frustratingly impossible.

Id.

Although we do not accept the notion that courts are unsuited to address design defect claims in any products liability action, we do agree that prescription drug design presents precisely this type of "polycentric" problem. A drug is designed to be effectively administered to specific individuals *99 for one or a number of indications. To determine whether a drug's benefit outweighs its risk is inherently complex because of the manufacturer's conscious design choices regarding the numerous chemical properties of the product and their relationship to the vast physiologic idiosyncracies of each consumer for whom the drug is designed. Society has recognized this complexity and in response has reposed regulatory authority in the FDA. Relying on the FDA's screening and surveillance standards enables courts to find liability under circumstances of inadequate warning, mismanufacture, improper marketing, or misinforming the FDA — avenues for which courts are better suited. Although this approach denies plaintiffs one potential theory on which to rely in a drug products liability action, the benefits to society in promoting the development, availability, and reasonable price of drugs justifies this conclusion.

In light of the strong public interest in the availability and affordability of prescription medications, the extensive regulatory system of the FDA, and the avenues of recovery still available to plaintiffs by claiming inadequate warning, mismanufacture, improper marketing, or misrepresenting information to the FDA, we conclude that a broad grant of immunity from strict liability claims based on design defects should be extended to FDA-approved prescription drugs in Utah.^[10]

HALL, C.J., and ZIMMERMAN, J., concur.

HOWE, Associate Chief Justice: (dissenting).

I dissent. I would limit the comment k exception to its literal wording as advocated by Justice Stewart. I would apply it only to a narrow class of prescription drugs which have lifesaving potential, drugs which have attendant risks but which are administered because they have the potential of saving the life of a person who is probably doomed to death by a disease, an epidemic, or a fatal condition. I believe that is what is meant by "unavoidably" unsafe. The exception would thus apply only to a small number of drugs — not the full gambit of drugs approved by the FDA. While I recognize the problem associated with a case-by-case application, I do not think the problem would be as great as portrayed by the majority opinion since only a relatively small number of drugs would fall into the class entitled to the exemption.

The Supreme Court of Wisconsin in Collins v. Eli Lilly Co., 116 Wis.2d 166, 342 N.W.2d 37 (1984), adopted a narrow application of comment k:

Drug companies like other sellers or manufacturers, have a duty to produce and market reasonably safe products. We recognize in some exigent circumstances it may be necessary to place a drug on the market before adequate testing may be done. Insofar as these circumstances exist, we agree with the comment [k] that strict liability should not be imposed. However, we find no exigent circumstances which would excuse DES producers or marketers from adequately testing DES before