American Journal of Law & Medicine

Informed consent and the elusive dichotomy between standard and experimental therapy.

I. INTRODUCTION

A rich academic literature exists about issues of informed consent in medical care, and, to a lesser extent, about a variety of issues posed by human experimentation. Most commentators regard patient autonomy as a desirable--though in practice often unattainable--goal, and near unanimity exists about the necessity for even fuller disclosure before experimenting on subjects. Although this Article intentionally side-steps the broader debate about informed consent, it challenges the conventional wisdom that special disclosure rules should apply in the experimental context.

Clinical trials have become big business. Estimates suggest that as many as twenty million Americans have enrolled in formal biomedical studies, (1) though, as a measure of the full scope of medical experimentation on humans, that figure may represent only the proverbial tip of the iceberg. Historically, sponsors of clinical trials recruited subjects informally, counting on word of mouth among physicians and also perhaps posting flyers around college campuses. Nowadays, sponsors run national television advertisements and host Internet sites, (2) and they also may offer financial incentives to family physicians for recruiting subjects from among their existing patients. (3)

Clinical trials have proliferated in recent years for a variety of reasons. In part, the total volume of biomedical research and development has increased substantially, resulting in a greater number and complexity of new investigational products; in part, federal regulatory requirements for proof of safety and effectiveness have expanded; (4) and, in part, patients have become more interested in participating as research subjects, either because their conditions have not responded to existing treatments or because they lack the insurance coverage and resources to afford standard treatments. (5) In 1997, reacting to these developments, Congress instructed the National Institutes of Health (NIH) to maintain an Internet site designed for patients seeking ongoing clinical trials in which to enroll. (6)

Do patients who decide to enroll in clinical trials really understand the nature of these experiments? Federal regulatory agencies have created elaborate protections for human subjects, and a requirement to secure informed consent occupies a central place in these systems. In contrast, outside of the context of clinical trials, federal regulatory agencies have done little to insist that physicians secure informed consent from their patients, instead leaving that task to state tort law. Over the last several decades, medical malpractice doctrine has expanded the disclosure obligations of physicians, but the courts have struggled when asked to apply these rules in the context of experimental procedures. This should come as no surprise, however, because no distinct line separates standard and experimental treatment.

To a greater or lesser extent, all medical interventions have an experimental quality to them. Physicians try things out on their patients all of the time. In many instances, we hope, they do so based on well-grounded confidence in the likely utility of a particular therapy, though even then the variability in patient response may disappoint our expectations. In far too many instances, unfortunately, physicians select interventions that remain poorly understood. Even for those therapeutic choices subject to federal licensing requirements, particularly pharmaceuticals, product approval does not define the point at which an investigational intervention passes the threshold into standard therapy. Instead, the research phase continues after licensure, both in the sense that more safety data accumulates and insofar as physicians may improvise when using a product in ways not originally contemplated. Conversely, an investigational product may become the standard of care even before federal regulators bestow their blessing on a particular use.

To the extent that we encounter a spectrum rather than a bright-line distinction between standard and experimental interventions, it becomes especially important to understand what might justify heightened informed consent requirements in the context of clinical trials and then decide whether to extend these to atypical experimental treatments. If, for instance, greater uncertainty about risks and benefits, or fears of conflicts of interest, account for demanding more thorough consent in the experimental context, then perhaps all encounters between physician-investigators and patient-subjects must account for these features of their relationship (in effect, informed consent on a sliding scale). Such an analysis casts serious doubts on the all-or-nothing approach of federal regulatory agencies, and, conversely, it challenges the largely undifferentiated rules applied by the courts when they resolve medical malpractice litigation.

II. APPLICATIONS OF THE TREATMENT-RESEARCH DICHOTOMY

For the most part, legal scholars who have addressed informed consent issues have paid scant attention to the special issues that arise in the research setting. (7) In contrast, these issues have attracted substantial commentary among those in the biomedical research community. This relative disinterest among legal scholars may reflect the fact that courts have not really given the topic separate consideration in the medical malpractice context, and that administrative agencies have done so only as one aspect of complex regulatory regimes for the supervision of human subject research. Juxtaposing these two systems may offer some valuable insights because, as it happens, the rules that evolved in these separate domains have begun to converge.

A. MEDICAL MALPRACTICE DOCTRINE

1. Basic Contours of the Duty to Secure Informed Consent

Informed consent doctrine emerged from the intentional tort of battery. In order to avoid the risk of liability for unlawfully intruding into someone's personal space, a healthcare professional would have to secure his or her patient's consent before performing a surgical or other therapeutic intervention. (8) With time, this type of claim became firmly rooted in the law of negligence, (9) though extreme failures, such as the so-called "wrong site" surgeries that have become disturbingly common in recent years, (10) may still lead to the imposition of liability for battery. (11)

Whatever its precise denomination, the duty to secure informed consent reflects a commitment to patient autonomy and self-determination. (12) A number of commentators have questioned whether the doctrine in practice can promote such goals. (13) A few even question whether it should do so. (14) Still, others complain that, although the law can and should promote the ideal of patient autonomy, the doctrine as presently configured fails to vindicate the dignitary interests that it purports to protect. (15) It also aspires to a form of dialogue entirely foreign to traditional medical practice. (16) Although leading physicians and clinical researchers have implored their colleagues to take this obligation seriously, the situation does not appear to have improved much. (17) To the extent that aspects of this broader debate relate to informed consent in the experimental context, they will receive further elaboration below where relevant.

At the most basic level, before subjecting a patient to a diagnostic or therapeutic intervention, physicians must describe the nature of the procedure, which may relate to questions such as its purpose, site, invasiveness, duration and instruments used. In a few jurisdictions that continue to focus on the origins of the doctrine in the tort of battery, informed consent applies only in cases of surgical and other invasive procedures, thereby excluding therapeutic interventions such as pharmaceutical products recommended by a physician. (18) For the most part, however, physicians also must secure informed consent when they recommend noninvasive treatments, (19) particularly when they prescribe or administer medical technologies. (20) In parallel with the common law duty to warn that has emerged in other contexts, physicians also must disclose to patients any significant risks known to accompany medical treatments. (21) This includes a continuing duty to disclose subsequently discovered risk information to former patients, sometimes years after the treatment. (22)

In addition, physicians generally must disclose reasonable alternative courses of action to the patient. (23) They also may have an obligation to describe the benefits of a diagnostic or therapeutic procedure--in other words, the health risks associated with the failure to follow the physician's recommendation--in order to ensure that, if the patient declines to consent, the decision is also a fully informed (even if unwise) one. (24) These additional obligations suggest the extent to which the duty to secure informed consent has departed from its origins in battery and has even moved beyond a duty to warn of risks to include a broader obligation to educate the patient. (25) In practice, however, the doctrine focuses on what physicians must tell patients, which provides a necessary but hardly sufficient condition for ensuring truly informed consent.

Physicians could not possibly disclose all information about the risks and benefits associated with different options for treating a particular condition, and, even if feasible, such comprehensive disclosures would not necessarily promote patient autonomy. Thus, the duty to secure informed consent only requires that physicians communicate "material" information to their patients. (26) Two standards have emerged for defining what information qualifies as material. (27) Traditionally, courts asked what a reasonable physician would have disclosed under the circumstances. (28) With time, this "professional standard" came under criticism for its excessive paternalism and the effective immunity that it granted to defendants in medical malpractice cases. (29) As a result, almost half of the states have shifted to a "patient standard" of informed consent.

Under the patient standard for defining materiality, many courts use an objective test, asking whether a reasonable person would regard the information as important. (30) A few other courts opt for a more subjective test, which inquiries about the perhaps idiosyncratic prior knowledge and preferences of the particular patient. (31) Under either test, courts require expert testimony as to what a physician should have known about the relative risks and benefits of the procedure, (32) but they do not ask, as under the professional standard, for testimony about what other physicians customarily would have disclosed. (33)

The courts have recognized several exceptions that may limit the scope of the duty to secure informed consent even under a patient standard for defining materiality. (34) Physicians may not have to get consent from a patient in cases of an emergency, incompetence, or waiver. In addition, the therapeutic privilege allows a physician to withhold material information if he or she thinks that it would have a negative impact on the patient's prognosis. (35) This exception harkens back to the tradition of medical paternalism, and, if read expansively, it could end up swallowing the informed consent rule. (36)

Whatever the standard used for judging materiality, plaintiffs also must satisfy causation requirements. In informed consent cases, courts engage in a counterfactual inquiry that typically asks whether a reasonable patient would have declined the treatment had the physician disclosed the additional information. (37) By framing causation in these terms, courts may sacrifice some of the dignitary aspects of the tort, (38) for instance by discounting the other ways--short of refusing a treatment--that a patient might respond to such information (e.g., watching for possible adverse reactions, or getting one's affairs in order), (39) or by failing to account for different degrees of risk aversion. (40) Thus, a few courts have adopted a subjective standard for judging causation. (41)

Many states have codified aspects of the informed consent doctrine. (42) Typically, they have adopted the professional standard for defining materiality, (43) and a few jurisdictions have gone even further by precisely defining the disclosures required for different types of medical procedures. (44) Notwithstanding serious doubts about the effectiveness of written consent forms, (45) many state statutes create safe harbors for physicians by treating signed documents that contain basic types of information as presumptively adequate. (46) Informed consent generally does not, however, amount to an express waiver of the right to sue for medical malpractice. (47) Full disclosure by physicians only satisfies their duty to warn and does not extinguish their separate obligation to provide care of a type and in a manner that accords with the standards of the profession.

2. Special Tort Rules for the Research Context

Although some commentators have encouraged courts to develop distinct informed consent duties contextualized for different settings, (48) the doctrine remains fairly undifferentiated. (49) Even so, case law has elaborated on the materiality standard in ways that may move in this direction, and several arguments support the imposition of more rigorous informed consent requirements in the research context: greater uncertainty about the risks associated with an intervention, which makes deference to medical expertise less appropriate; greater uncertainty about the therapeutic benefit, which undermines any presumption that the subject would have consented; and heightened concerns about conflicts of interest, which means that the researcher may have goals other than doing what is best for the subject. (50) In addition, some have argued that reseachers know less about a particular individual than their family physician would know from repeated therapeutic encounters. (51) As elaborated more fully below, these features hardly distinguish research from treatment. (52)

Is investigational status presumptively material? This question arose in the case that resulted in the first, though somewhat tentative, articulation of the negligence-based informed consent doctrine. In its opinion, the court rejected the plaintiff's argument that the use of a pharmaceutical product in a manner that deviated from the manufacturer's recommendations automatically qualified as an experimental use; the court noted that this use, though still novel to some extent, had achieved recognition among radiologists. (53) Although a number of other courts subsequently reached a similar conclusion, (54) recently a few courts have held that the lack of approval by the Food and Drug Administration (FDA) might itself represent material information that a physician would have to disclose to a patient. (55) No comparable distinction exists in the special products liability rules that apply to sellers of medical technologies. (56)

When an investigational product is used as part of a formal research protocol, the failure to reveal the nature of the study more clearly breaches the duty to secure informed consent. In Mink v. University of Chicago, (57) a federal court rejected the defendant's motion to dismiss a battery claim in such a case. During the early 1950s, more than 1,000 women received diethylstilbestrol (DES) as part of their prenatal care at the university's teaching hospital as part of a clinical trial sponsored by the manufacturer to determine whether DES prevented miscarriages. Until almost two decades later, the women did not know that they had received DES or that they were part of a clinical trial. (58) Fearing that their children faced an increased risk of cancer from in utero exposure to the drug, the women filed a class action lawsuit against the university and the manufacturer. The defendants argued that, by entering the hospital for prenatal care, the women had consented to any course of treatment that it decided to provide them, but the court held that the plaintiffs' allegations created a genuine question of fact for purposes of the battery claim. (59) Without defending the failure to secure informed consent in that case, one prominent British researcher provocatively suggested that this particular study was less disturbing from an ethical standpoint than the behavior of a far greater number of obstetricians at the time who uncritically had prescribed DES in an effort to prevent miscarriages in their patients, (60) presumably also without any disclosure about its unproven safety or effectiveness for this use.

In Burton v. Brooklyn Doctors Hospital, (61) a New York appellate court encountered a case that amply illustrated the potential for conflicts of interest in the research setting. At the time of the plaintiff's birth in 1953, a debate existed in the medical community about the risks and benefits of high concentrations of oxygen for premature infants: Most physicians viewed supplemental oxygen as essential for preventing brain damage and death, but evidence had begun to emerge that it caused retrolental fibroplasia (RLF), which often resulted in blindness. The hospital had conducted its own research into the issue, but, because the results were inconclusive, it decided to participate in a national study of the problem. (62) As part of the study, one out of every three premature infants would receive high concentrations of oxygen, while two out of three would receive low concentrations, and physicians would perform diagnostic checks on their eyes once a week over the course of a month.

Shortly after the plaintiff entered the hospital, the attending physician decided to reduce the concentration of oxygen in his incubator because of the plaintiff's good condition. The next day, without informing the attending physician or the plaintiff's parents, another doctor on the hospital's staff ordered a dramatic increase in his oxygen levels as part of the study. (63) Even though the plaintiff developed early signs of retinal problems, he remained on high oxygen for the full four-week period, which resulted in permanent blindness. (64) The court affirmed a judgment against the hospital and researcher for medical malpractice and failure to secure informed consent, rejecting the argument that the duty to provide detailed risk disclosures did not yet exist in the early 1950s. (65)

Although not discussed in conflict of interest terms, the court in Burton clearly objected to the fact that the researcher had countermanded the individualized treatment decision made by the attending physician and then failed to make any individualized decision to withdraw the plaintiff from the study once signs of injury emerged. (66)

 
   Dr. Engle, who was not plaintiff's physician, and who admitted that 
   she had neither examined the baby nor had any responsibility for the 
   care of premature infants, changed the oxygen supply and ordered a 
   drastic increase.... [This] was an administrative judgment, based 
   upon a random allocation of babies into one of two groups for 
   monitoring as part of the Cooperative Study. Neither the hospital 
   nor 
   Dr. Engle offered any medical reason for placing plaintiff in 
   routine 
   (increased) oxygen. (67) 

Although the plaintiff's parents had signed a general informed consent form that authorized any appropriate treatment at the hospital, the court noted that no one had specifically disclosed to them the suspected risk of blindness associated with high concentrations of oxygen. (68)

Fortunately, such complete failures to disclose participation in a clinical trial seem to have become less common with time, but the adequacy of disclosures to research subjects continues to trigger litigation. Even when the subject understands his or her role as a research subject, inadequate risk disclosures may lead to the imposition of tort liability if the subject then suffers an injury. In Gaston v. Hunter, (69) for example, the plaintiff sought treatment for a herniated disk. Instead of a laminectomy, which involves the surgical removal of the disk, the defendants selected a novel procedure known as "chemonucleolysis," which involved injecting Discase[R] (chymopapain)--an investigational new drug then undergoing clinical trials--into the protruding intervertebral disk. When this treatment resulted in serious complications, the plaintiff brought tort claims against her physicians and the manufacturer of the drug. After a lengthy trial, the jury returned verdicts for the defendants, which, with one limited exception, the Arizona appellate court affirmed. With respect to her informed consent claim, the plaintiff argued that, even though she had signed an informed consent form as required by FDA regulations, a heightened standard of disclosure should apply to experimental treatments because of the uncertainty about risks and benefits. (70) The court rejected this argument, though it did agree that a physician "must inform his patient of the novel or investigational nature of the procedure." (71) In short, apart from requiring the separate mention that a treatment remains experimental, the same disclosure standard would apply.

In most of the reported cases that address the question of informed consent to research, however, the plaintiff received a novel therapy as a patient rather than in the capacity of a subject enrolled in a formal study. In Estrada v. Jaques, (72) for example, a team of physicians diagnosed the plaintiff with a false aneurysm, a weakened spot in an arterial wall, which they recommended treating by means of a percutaneous steel coil embolization, a procedure that remained experimental. (73) Among other things, the court reversed the summary judgment granted to the surgeons on the plaintiff's informed consent claim, and it offered the following guidance for the trial judge on remand: "[W]here the health care provider offers an experimental procedure or treatment to a patient, the health care provider has a duty, in exercising reasonable care under the circumstances, to inform the patient of the experimental nature of the proposed procedure." (74)

The court in Estrada defended its holding in part by referring to the conflict of interest that may exist when physicians make use of an experimental procedure. (75) It also noted the uncertainty about risks and benefits that exists in this context: although the surgeons had advised the patient that the procedure might not succeed, they did so "without informing him of its experimental nature and their consequent lack of knowledge of the risks of whether it would fail or not." (76) Although only a few other courts have dealt with this question, for the most part they also have concluded that the experimental nature of a procedure--even if not used in connection with a formal experiment--represents a material fact that physicians should disclose in connection with securing informed consent. (77)

Analogously, some courts have required that physicians disclose their "batting averages" for particular procedures. (78) Although plaintiffs usually bring these sorts of claims against healthcare professionals who have enough of a track record to reveal poor performance with a certain intervention, (79) the courts also could require the disclosure of relative inexperience or, to put it less delicately, that the physician plans to practice or refine his technique on a particular patient. (80) The court in Gaston had suggested as much even though it rejected the plaintiff's argument that a heightened duty of disclosure applied in the case of experimental treatments. (81)

Apart from asking whether it makes sense to impose a heightened tort duty to secure informed consent when using experimental treatments or engaging in research, courts would have to decide where to draw the line between experimental and standard therapy. Sometimes that characterization turns on the regulatory status of a product or the novelty of a procedure, while in other instances an established product or procedure may become experimental simply because a research protocol aims to investigate its use. A few state statutes impose special informed consent requirements for particular products (e.g., laetrile) or procedures (e.g., acupuncture) that qualify as investigational. (82) California may have the most comprehensive legislation addressing this issue, entitled the Protection of Human Subjects in Medical Experimentation Act, (83) though the courts also have struggled in defining its reach. (84) In other jurisdictions, courts have invalidated as unconstitutionally vague state statutes that criminalize certain activities characterized as "experimental." (85)

The need to draw a line between standard and experimental therapy arises in other contexts. For instance, the informed consent duty to disclose alternatives does not include telling patients about the availability of experimental treatments. (86) In insurance coverage disputes involving novel treatments, courts have used a general acceptance standard in deciding whether a procedure remains experimental. (87) In some instances, physicians adopt innovative treatments without having any real evidence of efficacy, but courts may decide in favor of patients who seek insurance coverage if the new procedure or technology attracts a sufficiently wide following. (88) In medical malpractice cases premised on a departure from custom, courts have recognized a limited privilege for clinical innovation. (89)

Perhaps the disclosure of experimental status serves as a proxy for revealing the uncertainty about risk, much like decisions requiring that healthcare workers disclose their HIV positive status whether or not it creates any measurable risk of transmission to patients. (90) Perhaps disclosure serves a function similar to a disclaimer, reinforcing the message that a particular medical treatment comes with no guarantees of success. (91) Or perhaps it alerts patients to the need for exercising greater vigilance about the potential for conflicts of interest. (92) Nonetheless, apart from the duty recognized by some courts requiring physicians simply to reveal "experimental" status to the patient, no other special informed consent requirements apply in this setting. By comparison, in the context of "non-therapeutic" research--namely, studies involving healthy volunteers or physiological studies on existing patients not designed to test potential treatments--a few courts have decided to impose heightened informed consent requirements. (93)

B. FEDERAL REGULATORY REQUIREMENTS

Federal administrative agencies play a far more significant role than the courts in supervising experiments involving human subjects, but they play almost no role in defining informed consent obligations outside of the clinical trials context. (94) Although the emphasis differs, as do the tools at their disposal, regulatory agencies have to struggle with the same definitional quandary as the courts. Indeed, deciding when something qualifies as experimental takes on even greater importance in this arena because the answer often determines whether any federal requirements apply at all. (In contrast, for purposes of applying the tort law duty to secure informed consent, characterizing a medical intervention as experimental ends up having little consequence.) Although federal agencies have given the definitional question more sustained attention, they too have not arrived at entirely satisfactory answers. On the contrary, their responses have helped to reveal why it might represent the wrong way of framing the question in the first place.

1. Basic Disclosure Rules for Medical Technologies

With limited exceptions, the Food and Drug Administration does not require that suppliers of medical technologies communicate risk information directly to patients. (95) Nor does it typically, in designing professional labeling, direct physicians to secure informed consent when prescribing particular drugs. (96) Over the last three decades, however, the FDA has proposed mandating that information about prescription drugs be given directly to patients. In 1970, the agency required that sellers of oral contraceptives supply a "patient package insert" (PPI). (97) The FDA tried to emphasize, however, that it did not intend to displace the central role of the healthcare professional: "Patient labeling serves primarily as an informational adjunct to the physician-patient encounter and is intended to reinforce and augment oral information given by the physician to the patient at the time the drug is prescribed." (98) Although physicians did not entirely welcome these initiatives, (99) the agency subsequently issued regulations requiring PPIs for a handful of other products. (100)

In 1979, based on studies indicating the usefulness of supplying written information directly to patients, the FDA proposed regulations that would have required PPIs for most prescription drug products. (101) After receiving various complaints about the costs of such an undertaking, the FDA instead established a three-year pilot program mandating the distribution of PPIs for ten classes of drugs. (102) Even this scaled-down program was stayed by the agency in 1981 and revoked the following year. (103)

In the early 1990s, the FDA expressed renewed interest in requiring the distribution of prescription drug labeling directly to patients because of concerns that physicians still were doing a poor job of communicating risk information. (104) As part of its "MedGuide" initiative, the agency proposed guidelines for the development of PPIs for all prescription drug products. (105) Congress intervened, however, by placing a temporary moratorium on finalization of this proposal in order to give the industry time to develop its own voluntary program as a substitute for the FDA's initiative. (106) Even so, the agency has imposed PPI requirements on a case-by-case basis as a condition of new drug approval. (107) Moreover, in the course of approving particular drugs that pose special risks to women, the FDA has gone a step further and called on physicians to supply their patients with an enclosed informed consent form and request their signature, as it did in the case of Accutane[R] (isotretinoin), (108) Mifeprex[R] (mifepristone), (109) and Thalomid[R] (thalidomide). (110)

2. Informed Consent Rules for Biomedical Research

Although federal regulatory agencies rarely impose any duty to secure informed consent from a patient undergoing treatment, they have mandated extensive disclosures to subjects enrolled in clinical trials. Prompted by reactions to the atrocities that occurred in Nazi Germany during World War II, (111) and revelations about nonconsensual human experimentation conducted in the United States, (112) the federal government has created elaborate protections to govern the conduct of biomedical research. Not surprisingly, given its prominence in the Nuremberg Code, (113) informed consent has become the centerpiece of these efforts. (114)

In 1974, the Department of Health, Education, and Welfare (HEW), now the Department of Health and Human Services (HHS), issued the first formal regulations governing human research. (115) These regulations apply to all types of research conducted at institutions that receive federal funding. (116) Seven years later, the FDA promulgated separate regulations, which apply to all studies of investigational drugs, medical devices, and biological products. (117) The two sets of requirements are largely co-extensive, and, because most research institutions receive federal funding for at least some of the work being done under their roofs, institutional review boards (IRBs) generally hold all research protocols to the standards set out in these regulations.

Whenever a study is covered by FDA or HHS regulations governing human subject protection, sponsors and investigators must abide by a number of different requirements, most importantly an obligation to secure informed consent. (118) Subjects must receive a written consent form, (119) and the regulations specify a number of items that must appear in the form, foremost among them a "statement that the study involves research." (120) The demand for written consent forms may have more to do with providing an IRB something concrete and uniform to review in advance of a clinical trial than an assessment of the relative effectiveness of communicating in this fashion. (121) Indeed, IRBs may become preoccupied with reviewing the niceties of the consent form and perhaps less concerned about their separate obligation to make independent risk-benefit assessments about the research protocol, (122) confident that potential subjects can "vote with their feet" so long as the consent form contains all of the necessary information. (123) In many cases, however, the investigators know so little about an experimental therapy that, even if they have fully disclosed all that they know and do so in a way that the subjects can understand, some commentators have wondered whether it even makes sense to speak of "informed" consent in the research setting. …

Log in to your account to read this article – and millions more.