Health1and the Law
`
`The Broader Message of Accutane
`SHEILA R. SHULMAN, LLB, MPH
`Who has the responsibility for conveying to patients the
`and malformations of the heart and brain.4
`risks associated with prescription drugs? This question was
`The FDA approved Accutane in September 1982. The
`raised in a powerful way in April 1988 by the disclosure that
`drug's serious teratogenic potential was fully recognized and
`severe birth defects in 62 infants had been attributed to the
`acknowledged at that time. The FDA-approved labeling for
`use of the drug Accutane during pregnancy. The Epidemiol-
`Accutane specified prescribing restrictions, contraindica-
`ogy Branch within the Food and Drug Administration (FDA)
`tions, and warnings. It was to be prescribed for only one
`subsequently estimated that between the years 1982 and 1986
`indication-the treatment of severe recalcitrant cystic acne-
`the drug may have caused as many as 1,300 birth defects
`and only for those patients who were unresponsive to other
`nationwide.1 The agency also estimated that women who
`forms of therapy. The labeling, which constitutes the pre-
`became pregnant while on the drug had spontaneous abor-
`scribing information prepared specifically for physicians,
`tions at close to three times the rate in the general population.
`warned of the potential for major human fetal abnormalities
`In response to these disclosures, the FDA sought the
`should the drug be used during or just prior to pregnancy.
`opinion of its Dermatologic Drugs Advisory Committee on
`Since 1984, specific birth defects have been cited. The
`the issue of whether Accutane should be removed from the
`message to physicians was and continues to be clear and
`market and, if not, what measures could be instituted to
`forceful. How then does that message make its way to the
`eliminate inappropriate use of the drug. Following a meeting
`women of childbearing years who not only comprise 40
`on April 26, 1988, the Committee recommended the contin-
`percent of all Accutane consumers, but who can exercise a
`ued marketing of Accutane with specific changes in the
`substantial measure of control over the conditions that may
`labeling directed to physicians and patients.2 Factors influ-
`render this drug safe?
`encing this recommendation included evidence of the sub-
`stantial benefits that the drug offers to a narrowly defined
`group of patients, and the existence of steps that could be
`taken to circumscribe the risks associated with the use of the
`drug. The Committee's action seems appropriate. Little
`justification can be found for depriving patients of an effective
`drug for which there is no equivalent alternative until all
`reasonable steps have been exhausted to effectively bring
`home to the patient the known risks and recommended
`precautions. In addition, concerns were expressed by some
`who appeared before the Committee that even if the drug
`were removed from the market, patients would obtain illicit
`access to Accutane and would use it without medical super-
`vision.
`Accutane offers a classic example of a highly effective
`drug which also carries a significant risk of harm. While it is
`currently the only approved drug that can provide prolonged
`remission or, for some, permanent relief from a severe and
`disfiguring form of acne, it is also a potent human teratogen.3
`The birth defect syndrome most commonly associated with
`Accutane includes enlarged and misshapen head, cleft palate,
`tiny or unformed ears, facial paralysis, abnormally smalljaw,
`
`The Physician as Learned Intermediary
`The distribution of prescription drugs is governed by a
`complex network of legislative and legal rules. The key
`players are the FDA, the drug manufacturer, the physician,
`the pharmacist, and the patient. It is the physician, however,
`standing between the manufacturer and the patient, who
`serves as legal gatekeeper, controlling both access to pre-
`scription drugs and access to information about those drugs.
`The authority which physicians now exercise over this
`aspect of health care has evolved from statutory and common
`law sources. The shift to prescription drug status and man-
`dated medical supervision which began in 1938 was grounded
`primarily in a concern about public safety. However, the
`regulatory provisions that moved physicians into a position of
`control also required the removal of patient labeling infor-
`mation, a step which fostered the distancing and eventually
`the exclusion of the patient from the prescribing process. The
`assumptions were that certain drugs could be dangerous if
`access were permitted without the supervision of a physician
`and that adequate directions for lay use were impossible to
`write. The 1938 regulatory changes abandoned attempts to
`communicate information to patients and instead linked
`prescription drug designation to restrictive labeling
`requirements.5 Even appropriate uses for the drug were to
`be written in medical terms "not likely to be understood by
`the ordinary individual."5
`In 1951, the Durham-Humphrey Amendment formally
`authorized the FDA to determine the prescription status of all
`drugs.6 The Amendment, which now appears in section 503
`
`Address reprint requests to Sheila R. Shulman, LLB, Research Associate,
`Center for the Study of Drug Development, Tufts University, Boston, MA
`02111.
`This paper was accepted for publication by the editor of the Public Health
`& the Law, George J. Annas, JD, MPH, Chief, Health Law Section, Boston
`University, Schools of Public Health and Medicine, 80 East Concord Street,
`Boston, MA 02118.
`
`© 1989 American Journal of Public Health 0090-0036/89$1.50
`
`AJPH November 1989, Vol. 79, No. 11
`
`1 565
`
`Wockhardt 1009
`
`

`
`PUBLIC HEALTH AND THE LAW
`
`of the Federal Food, Drug & Cosmetic Act, requires the
`involvement of "a practitioner licensed by law to administer
`such drug" and expressly exempts prescription drugs from
`the statutory requirement that the drug labeling contain both
`adequate directions for use and adequate warnings.7 If
`patients were to receive information about the drugs pre-
`scribed for them, that too, it appeared, was to come from the
`prescribing physician.
`Court decisions have reinforced the central role played
`by physicians. In most instances where federal law requires
`a prescription, the courts have relieved pharmaceutical
`manufacturers of the obligation imposed on manufacturers in
`general to communicate directly with patients about the risks
`inherent in their products. At common law, the manufactur-
`er's duty to warn is owed instead to the physician. Adequate
`and timely information about adverse reactions of which the
`manufacturer is or should be aware must be conveyed to
`physicians. In the courts' view, the physician then functions
`as a learned intermediary between the manufacturer and the
`patient and has an affirmative duty to be knowledgable about
`the drugs he or she prescribes.8 The rationale is a familiar
`one. The physician is considered to be in the best position to
`weigh the risks and benefits of a specific drug for individual
`patients. The courts also are persuaded by arguments that
`direct communications from manufacturer to consumer may
`be too difficult, could unduly interfere with the doctor-patient
`relationship, and might frighten or confuse the patient,
`discouraging compliance with the prescribed therapy.9
`Two exceptions to the learned intermediary rule have
`emerged. In both, the lack or perceived lack of a physician's
`involvement was critical. The first involves mass immuniza-
`tion programs. Since physicians are absent or only periph-
`erally involved in such programs, a personal medical deter-
`mination or a balancing of the risks for the individual patient
`is impossible. Vaccine manufacturers therefore must provide
`patients with a warning of the risks associated with the
`vaccination allowing patients themselves to weigh the ben-
`efits and risks.'" Second, in MacDonald v. Ortho Pharma-
`ceutical Corp.," the Supreme Judicial Court of Massachu-
`setts imposed a common law duty on manufacturers of oral
`contraceptives to communicate risk information directly to
`the user. The court found oral contraceptives unique for a
`number of reasons, including the substantial nature of the
`risks associated with the drug. Central to the decision,
`however, was the argument that patients take an active role
`in selecting this method of birth control, often deciding to
`take the pill before consulting a physician. This, the court
`determined, had the effect of relegating the prescribing
`physician "to a relatively passive role." The possibility was
`acknowledged that oral communications between the physi-
`cian and the patient could be too scanty to provide an
`adequate warning. Therefore, the court concluded, the man-
`ufacturer's obligation encompassed a duty to warn the
`ultimate user. "
`
`The Learned Intermediary-Who Benefits?
`While the learned intermediary doctrine sets out a clear
`expectation of full disclosure from manufacturer to physi-
`cian, there is no such expectation between physician and
`patient. The courts have acknowledged that the physician as
`the learned intermediary has complete discretion to deter-
`mine what facts, if any, should be shared with the patient.8
`The Supreme Court of Illinois has characterized the extent of
`disclosure as "a matter of medicaljudgment.2 This discre-
`
`tion, however, is in direct conflict with, and arguably should
`be displaced by, the doctrine of informed consent which
`incorporates elements of broad disclosure by the physician
`and active decision-making by the patient. In this instance,
`the physician has an overriding affirmative duty of disclosure
`which encompasses all material risks associated with pro-
`posed medical treatment including prescription drug therapy.
`The interplay of these two legal doctrines, one requiring
`physicians to know about the drugs they prescribe and the
`other requiring them to share this information with their
`patients, should work to the patients' benefit. However, the
`divergence of legal theory and medical practice in the
`prescription drug context is striking. The criticism that the
`law has substantially failed to provide the competent patient
`with a meaningful opportunity to participate in medical
`decision-making on a voluntary, informed and understanding
`basis is perhaps most relevant to this aspect of medical
`care. 13
`The perfunctory nature of drug prescribing has been
`acknowledged by the FDA, prompted in large part by an
`increasing awareness of serious adverse drug reactions and a
`recognition that patients want more information about their
`drugs, a finding supported by the FDA's own studies. '4 In
`spite of the agency's assertion that patients have "both a right
`and a need to know about the drugs they use," and its
`conclusion that physicians generally do an inadequate job of
`informing their patients about prescribed drugs, little has
`been done to alter the situation. 5 The FDA has the authority
`to require a drug manufacturer to include labeling written in
`nontechnical language directed specifically to patients. It has
`done so, however, in only seven instances.* In 1979, a
`comprehensive patient package insert program for a wide
`range of prescription drugs was proposed by the FDA. 14 Prior
`to the effective date of the initial implementation phase in
`1982, the program was revoked in favor of voluntary private
`sector initiatives.'5
`
`Response to the Accutane Crisis
`The experience with Accutane forces confrontation with
`the ineffectiveness of the current system. The FDA-approved
`labeling or prescribing information prepared by Hoffmann-La
`Roche, the manufacturer of Accutane, is published in the
`Physicians' Desk Reference (PDR), a resource used by most
`physicians. Since 1985, the contraindications have been
`boxed off and stand out from the remainder of the text. They
`are printed in bold type and convey a strong, clear warning
`of the drug's teratogenic potential. A March 1984 labeling
`change added a recommendation that a pregnancy test be
`performed two weeks prior to prescribing the drug for all
`women of childbearing age. The physician also has been
`urged to counsel the patient about the potential risk to the
`fetus and, in the event pregnancy occurs during the course of
`treatment, to discuss the desirability of continuing the preg-
`nancy. It appears therefore that, to the extent the PDR
`warning alerted physicians to the risks posed to the fetus, the
`manufacturer fulfilled its obligation to warn the physician.
`It is less easy to verify whether an adequatejob was done
`at the physician-patient level. We simply do not know how
`
`*Oral contraceptives 21 C.F.R. 310.501(a) (1987); Estrogens 21 C.F.R.
`310.515 (1987); Progestational drug products 21 C.F.R. 310.516 (1987); Depo-
`provera 21 C.F.R. 310.501a (1987); Diethylstilbesterol 21 C.F.R. 310.501(b)
`(1987); Bronchodilators containing isoproterenol 21 C.F.R. 201.305 (1987);
`Intrauterine devices 21 C.F.R. 21 C.F.R. 310.502 (1987).
`
`1 566
`
`AJPH November 1989, Vol. 79, No. 11
`
`Wockhardt 1009
`
`

`
`well the risks of Accutane were communicated by physicians
`to young women desperate for relief from a socially stigma-
`tizing condition. We do, however, have the statistics on birth
`defects cited earlier. Questions on this point also arise in the
`medical literature. The January 1988 issue of the Canadian
`Medical Association Journal describes a survey ofpatients in
`the Province of Saskatchewan who received four months of
`treatment with Accutane. 6 One-third ofthe women surveyed
`reported that they used no contraception during treatment.
`An earlier report in the New England Journal of Medicine
`provides results of an investigation of 154 pregnancies with
`fetal exposure to the drug.4 Sixty-seven percent of the
`exposed pregnancies were among women who were either
`pregnant when therapy began or were not using contracep-
`tion during treatment.
`How well were the risks of Accutane shared with these
`women? Could a more diligent job have been done by the
`prescribing physicians? There is no question that recently
`adopted prescribing restrictions for Accutane have been
`designed to compel physician-patient interaction. The re-
`sponse of both the FDA and Hoffmann-La Roche to the
`Accutane crisis zeroed in on the physician and the events at
`the time of prescribing. The FDA recommended a doubling
`of the print size in the boxed contraindications section for
`Accutane in the PDR. The language of the warning has been
`further strengthened. While consideration was given to
`restricting prescribing authority to dermatologists, Hoffman-
`LaRoche opted intead to include a cautionary statement in
`the prescribing information suggesting that Accutane be
`prescribed only by physicians who have special comptence in
`the diagnosis and treatment of severe cystic acne, and who
`understand the risk of teratogenicity. The physician is re-
`quired to ensure that female patients meet eight specific
`criteria, including a determination by the physician that the
`individual patient is "reliable in understanding and carrying
`out instructions." All criteria must be satisfied before a
`woman can be considered a suitable patient for treatment
`with Accutane. Within the two weeks prior to starting a
`course of therapy, the patient must have a negative serum
`pregnancy test. A written consent form must be signed by
`both the patient and the physician before the drug is pre-
`scribed. The patient is asked to acknowledge that the phy-
`sician has disclosed, and that she understands, the risks of
`Accutane and the importance of using effective birth control.
`An additional reminder of the warnings now appears in the
`form of a red "avoid pregnancy" symbol on the back of the
`new blister-pack container for each capsule of the drug.
`Hoffmann-La Roche has assembled a Pregnancy Pre-
`vention Program Kit for distribution to physicians who
`prescribe Accutane. The kit includes a large file box with
`forms and brochures, to be used and distributed to the patient
`by the physician during the consultation. A video tape
`accompanies the box. The 10-minute tape explains how to
`use the boxed materials and further cautions the physician
`about appropriate prescribing procedures. In addition to the
`consent document, the box contains a written test which the
`patient must complete in the physician's office prior to
`signing the consent form; the 10 true-false questions are
`intended to evaluate the patient's comprehension of the risks
`of Accutane. A brochure providing detailed birth control
`information and a general patient information brochure,
`which was introduced voluntarily by Hoffmann-La Roche in
`1982, are also provided for distribution to the patient. The
`new steps in the prescribing procedure clearly raise questions
`as to their enforceability; however, in the event of litigation,
`
`AJPH November 1989, Vol. 79, No. 11
`
`PUBLIC HEALTH AND THE LAW
`
`the prescribing physician would want to produce copies of
`the consent and test documents as evidence of his compliance
`with the recommended prescribing regimen. On this point,
`the "Physician's Guide to Consent" which is part of the
`Pregnancy Prevention Kit suggests that "good medical man-
`agement practices" require that physicians retain copies of
`these documents.
`The new requirements and guidelines for Accutane are
`impressive and present a stark contrast to traditional pre-
`scribing routines. One aspect of the program, however,
`deserves more discussion. The 1988 version of the patient
`information brochure includes a line drawing of an infant with
`the characteristic deformities associated with exposure to
`Accutane. A larger version of this drawing is provided in the
`boxed materials along with the suggestion that it be used at
`the physician's discretion during the counseling of potential
`Accutane patients. No such leeway is allowed the physician
`with other parts of the program such as the true-false test or
`the signing of the consent document. Of all the boxed
`materials, this diagram seems to offer the most promising tool
`for educating patients, many of whom are teenagers, about
`the risks of this drug. It is a jarring attention-getter and may
`be exactly what is needed for these patients to appreciate and
`remember more than the promise the drug holds for their
`skin. The discretion allowed here is a remnant ofthe past, the
`old fear of frightening or upsetting the patient. Since the
`drawing also appears in the patient information brochure, it
`will be seen by the patient eventually. It seems more
`constructive to require the physician and patient to examine
`the illustration together allowing discussion of any fears or
`concerns it might generate.
`In addition, the brochure currently carries vivid before
`and after photographs of Accutane patients which demon-
`strate how effective the drug can be. The photographs convey
`a strong promotional message detracting from the purpose of
`the brochure which presumably is to help patients stay alert
`to the risks of a potent drug. If photographs such as these are
`to be included, a balancing of risks and benefits would be
`better achieved by replacing the current line drawing of an
`affected infant with an equally vivid photograph depicting
`actual birth defects.
`
`Conclusion
`
`Possibly the most striking element of the Accutane crisis
`and the activity it has generated is that none of the risk
`information is new. Since the approval of the drug, the
`warning language directed to physicians in the PDR has been
`forceful and unambiguous, and seemed to demand an equally
`forceful intervention by physicians. Further, 90 percent of all
`Accutane prescriptions are written by dermatologists who
`should be intimately knowledgable about the drug.
`The Florida Supreme Court recently ruled on the ques-
`tion of the adequacy of the Accutane warning provided to
`physicians.17 In a wrongful death action brought by the
`mother of a child born with severe birth defects associated
`with the ingestion of Accutane during pregnancy, the court
`affirmed a lower court decision in which the manufacturer's
`warning as far back as 1982 was found to be "accurate, clear,
`and unambiguous." The court also reasoned that any inad-
`equacy in the warning could not have been the proximate
`cause of the birth defects in this case since the prescribing
`physician admitted that he fully understood the warnings and
`had prior knowledge of the drug's inherent dangers.
`
`1567
`
`Wockhardt 1009
`
`

`
`PUBLIC HEALTH AND THE LAW
`
`The concept of a single learned intermediary is called
`into question by the Accutane experience. To restrict the
`disclosure of information to one party within the drug
`distribution system is artificial and inefficient. Manufacturers
`of non-pharmaceutical products have a duty to warn con-
`sumers directly about risks associated with the use of their
`products. The obligation should be no less for those who
`manufacture pharmaceuticals. The FDA could begin by
`reviving the patient package insert program it proposed in
`1979 requiring drug manufacturers to funnel their expertise
`directly to consumers of prescription drugs. Clearly written
`materials for distribution to patients by physicians and
`pharmacists would supplement the information offered by the
`physician. The brochure prepared for Accutane is evidence
`that direct manufacturer to patient communications are not
`the impossible task so often assumed. Recent studies provide
`some indication that printed materials used as an adjunct to
`verbal counseling increase patients' knowledge about a
`specific drug and improve the ability to take precautionary
`steps when side-effect symptoms occur."'20 An additional
`point of reference will be provided by a follow-up study with
`female Accutane users being conducted by the Slone Epi-
`demiology Unit at the Boston University School of Public
`Health which will assess the effectiveness of the new mea-
`sures now linked to the prescribing of Accutane.
`In their role as learned intermediaries, physicians have
`been stingy with information about the drugs they prescribe.
`Too often their involvement is as passive or detached as that
`of the physicians in the oral contraceptive and vaccine cases
`discussed earlier. This is where another critical change must
`take place, not just for Accutane but for all prescription
`drugs. All drugs carry some measure of risk. Patients need
`the help of their physicians to appreciate those risks. More
`time can certainly be devoted to this aspect of the physician-
`patient encounter. Specific risks must be disclosed to the
`
`patient as part of a preexisting requirement for informed
`consent to all medical treatment. If the physician is casual
`about drug therapy, we should not be surprised when patients
`reflect this same attitude.
`
`REFERENCES
`1. Dr. David Graham, Epidemiology and Biostatistics Division of the FDA:
`Memo presented to the Dermatologic Drugs Advisory Committee Hearing,
`April 26, 1988.
`2. Marwick C: FDA ponders approaches to curbing adverse effects of drug
`used against cystic acne. JAMA 1988; 259:3225.
`3. Stern RS: When a uniquely effective drug is teratogenic: The case of
`isotretinoin. N EngI J Med 1989; 320:1007-1009.
`4. Lamer EJ, Chen DT, Hoar RM: Retinoic Acid Embryopathy. N EngI J
`Med 1985; 313:837-841.
`5. 3 Fed. Reg. 3168 (December 28, 1938).
`6. 65 Stat. 648 (1951).
`7. 21 U.S.C. s.353(b) (1982).
`8. Terhune v. A. H. Robins Co. 577 P.2d 975,978 (Wash. 1978).
`9. In re Certified Questions 358 N.W.2d 873,882-883 (Mich. 1984).
`10. Reyes v. Wyeth Laboratories 498 F.2d 1264,1276 (5th Cir.), cert. denied,
`419 U.S. 1096 (1974).
`11. 475 N.E.2d 65 (Mass. 1985).
`12. Kirk v. Michael Reese Hospital and Medical Center 513 N.E.2d 387,393.
`(Ill. 1987).
`13. Weisbard A. Informed Consent: The Law's Uneasy Compromise With
`Ethical Theory, 65 Nebraska Law Review 749, 751 (1986).
`14. 44 Fed. Reg. 40,016 et. seq. (1979).
`15. 47 Fed. Reg. 39,148 (1982).
`16. Hogan DJ, Strand LM, Lane PR: Isotretinoin therapy for acne: A
`population-based study. Can Med Assoc J 1988; 138:47-50.
`17. Felix v. Hoffman-LaRoche, 540 So.2d 102 (Florida Supreme Court, 1989).
`18. Johnson MW, Mitch WE, Sherwood J, et al: The impact of a drug
`information sheet on the understanding and attitude of patients about
`drugs. JAMA 1986; 256:2722-2724.
`19. Regner MJ, Hermann F, Ried LD: Effectiveness of a printed leaflet for
`enabling patients to use digoxin side-effect information. Drug Intell Clin
`Pharm 1987; 21:200-204.
`20. Sand CD, Robinson JD, Orlando JB: The oral contraceptive PPI: Its effect
`on patient knowledge, feelings, and behavior. Drug Intell Clin Pharm 1984;
`18:730-735.
`
`I
`
`EPA Journal Seeks Subscribers
`The US Environmental Protection Agency is attempting to broaden the readership of the EPA
`Journal, a bi-monthly magazine that covers current developments in air and water pollution control,
`solid waste management, hazardous waste disposal, Superfund, pesticides, and ground water contam-
`ination. The journal discusses new government regulations and standards, new technologies, and
`various approaches to environmental management.
`Potential subscribers should send their name, address, and $8 (domestic) or $10 (foreign) to the
`Superintendent of Documents, Government Printing Office, Washington, DC 20402, and include a
`statement that you wish to subscribe to the EPA Journal.
`
`1568
`
`AJPH November 1989, Vol. 79, No. 1 1
`
`Wockhardt 1009