• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

The Case of Placebo Controlled Trials

Posted on: Monday, 3 October 2005, 12:00 CDT

By Glass, Kathleen Cranley; Waring, Duff

This article addresses medical negligence as a cause of action for harm to a research participant occasioned by withholding established, effective therapy in the placebo arm of a clinical trial, and the insufficiency of consent as a defense to such a claim.

Some authors argue that the ethics of medical care and the ethics of research differ, and that it is a mistake to conflate the two.1 They propose "that medical research and medical treatment are two distinct forms of activities, governed by different ethical principles."2 This raises the question of whether physicians who are also clinical investigators may separate their role as physician from that of researcher when they are involved in clinical trials, thereby avoiding the obligations required in the physicianpatient relationship. Miller and Brody suggest that medical training is to blame for what they believe is the physicians' tendency to confound the obligations of research and practice. "Physician-investigators, after all, went to medical school" they explain. They believe that considering research with patients outside the ethical framework of the physician-patient relationship may be "difficult and threatening" to physicians who have "psychological needs" to consider the ethical obligations flowing from their relationship with their patients. They appear to suggest that if physicians went to "investigator's school" instead of medical school, they would not have such "psychological needs."3

The authors criticize what they call the "mainstream ethical approach" to clinical trials under which physicians retain the therapeutic obligations of beneficence and non-maleficence governing the clinical practice of medicine. Their ethical basis for distinguishing research from clinical care is that the two enterprises have different objectives. The former is designed to answer a research question, while the latter is intended to provide care in the patient's interest. The result is a claim that the ethics of the traditional physician/patient relationship do not apply to research because a clinical trial is not "a form of therapy." Consequently, they allege, the investigator's obligation is only one of not exploiting research participants by exposing them to excessive risk, as opposed to acting in their best interests.4

The discussion of a physician investigator's obligations is important in the context of placebo controlled trials, where the choice of treatments for control patients has been recognized as both a methodological and an ethical issue. Some have argued that placebo and not active controls are required methodologically in order to demonstrate whether an unproven treatment is effective, even when an established effective intervention, or standard therapy, is available.5 We take issue with this contention, and have written to this effect elsewhere.6 Many, ourselves included, have written about the ethics of using placebo controls in clinical trials and the issues have been well laid out.7

However, little consideration has been given to important legal questions arising out of such trials. In particular, questions of potential legal liability for harm to a research participant occasioned by withholding available treatment in the placebo arm of a trial have received little attention beyond our own work.8

One response to this apparent lack of interest may be that there have not been any legal judgments involving harm from participation in a placebo controlled trial, so why worry?9 Yet this ignores the general rise in litigation in human research over the past several years. Research cases may have special appeal to plaintiffs' attorneys since, unlike traditional malpractice cases where we usually assume that the physician intended to help the patient, the physicianinvestigator's interests may not be completely aligned with the patient's.10 Issues beyond the patient's interest may be at play in research. Physicians may receive financial compensation for recruiting participants or conducting a trial. There may be institutional pressures to conduct research. Improvements in professional stature and prestige or the wellbeing of future patients may also serve as motivations for involvement in research.11

In looking for legal guidance on the placebo issue, there is no legislation or case law directly on point. However, using an analysis based on common law principles of legal liability and relevant American and Canadian non-placebo research cases, we believe there is a sound legal argument that a research participant who is harmed by the denial of established, effective treatment in the placebo arm of a trial might have an action in negligence against the physician-investigator. In this article we look at a narrow "slice" of possible legal issues: 1) the potential for medical negligence as a cause of action for harm to a research participant occasioned by withholding established, effective therapy in the placebo arm of a trial and 2) whether consent to participate can provide a sufficient defense to such a claim.

Civil Liability for Medical Negligence

Claims for medical malpractice are usually brought in the tort of negligence.12 Regimes of medical negligence apply to all areas of medicine and medical ' research, whether a trial tests a new therapy against placebo or active treatment. While a review of the fundamentals of medical negligence might seem elementary, they play a key role in responding to those who tell us that physician- investigators have no therapeutic obligations towards their patients.

Duty

In all tort cases, a number of elements must be established: that a duty of care was owed; that the duty was breached, and that the breach was the cause of injury.13 To establish medical negligence, therefore, research participants who are harmed must first prove that the physician/investigator owed them a duty of care. Only then will the alleged negligence be considered. This is important in the context of proposais offered by those such as Miller and Brody, cited above, who argue that physician/investigators do not owe their patients a duty to act in their best interests, but rather they have the duty to avoid exploiting research participants by exposing them to excessive risk.14 But such an ethical analysis conflicts with established legal principle, which relies on the notion of "holding out." Did the physician/investigator hold him or herself out as ready and willing to diagnose, treat, or refer the patient/ participant? If so, a doctorpatient relationship has been established, and the physician owes the patient a duty of care.15

Physicians themselves acknowledge the existence of the physician/ patient relationship when they are involved in research to test treatment. The American Medical Association's (AMA) Code of Medical Ethics states that in research designed to test the efficacy of treatment, the investigator "must recognize that the physician- patient relationship exists and that professional judgment and skill must be exercised in the best interest of the patient."16 The AMAs Current Opinions on its Code of Medical Ethics states that "in a clinical investigation primarily for treatment," the physician should evince the same concern and caution for the welfare, safety and comfort of the research participant as is required of a physician who provides medical care to a patient outside of any clinical investigation."17 Indeed, in research designed to test the efficacy of treatment, the physician-investigator "must recognize that the patient-physician relationship exists and that professional judgment and skill must be exercised in the best interest of the patient."18 The Preamble to the World Medical Association's Declaration of Helsinki recognizes the doctor-patient relationship with the phrase: "The health of my patient will be my first consideration."19 It states further that "considerations related to the well-being of the human subject should take precedence over the interests of science and society."20 Documents such as the AMA Code of Ethics or the Declaration of Helsinki leave no room for a distinction between the physician as researcher and the physician as clinician when it comes to the patient's interest.

Breaching the Standard of Care

Once a doctor-patient relationship has been established, by what standard will the court judge the physician/investigator's behavior? This question raises a number of interesting points and goes to the heart of the issue of the investigator's obligations. To establish negligence, the plaintiff must prove that the defendant failed to meet the established standard oi care, that is, failed to act with the skill and care of a reasonable practitioner of the same experience and standing. U.S. case law has expressed the reasonable care standard as a legally enforceable, non-delegable duty of physicians to render professional services. These services must be consistent with the objectively ascertained, minimally acceptable level of competence that physicians can be expected to apply given their qualifications and expertise and the circumstances of the case. In sum, the physician's duty of care is to t\reat each patient "with such reasonable diligence, skill, competence, and prudence as are practiced by minimally competent physicians in the same specialty or general field of practice throughout the United States, who have available to them the same general facilities, services, equipment and options."21

In Canada, a physician who treats a patient must use "a reasonable degree of skill and knowledge and must exercise a reasonable degree of care," meeting the standard of the "normal, prudent practitioner of the same experience and standing."22 The test of whether a physician meets this standard is an objective one. The provision of care is measured against a reasonable physician who possesses and exercises the skill, knowledge, and judgment of the normal, prudent practitioner of his or her special group. In making this comparison, reference is made to the particular circumstances at the material time.2'3 The standard of care is influenced by the foreseeable risk: the greater the risk, the higher the standard of care.

Given the above, would an argument hold up under legal scrutiny that a physician/investigator does not owe the duty that ordinarily arises out of a physicianpatient relationship, if the patient is a research subject? Will the relationship between physician- investigators and research participants be judged differently, i.e., would investigators only have a duty to avoid exploiting research participants by exposing them to excessive risk, as has been contended?21 We believe that the weight of legal evidence goes against this argument.

Although research and therapy can be distinguished, they often occur together. Ordinarily, potential research participants are approached as patients when they are seeking medical assistance in hospitals, clinics, or private physicians' offices. If an individual seeks medical services from a physician, and a doctorpatient relationship is established, by what mechanism would a different (and lower, given it would allow leaving some patients untreated), standard of care be established?

Physician-investigators have never been granted legal immunity from the obligations of clinical care. While there are no judgments exactly on point about placebo controlled trials, in most major jurisdictions the general rules of medical law are equally applicable to experimental procedures, whether carried out with a therapeutic purpose or not. There is no separate regime of liability for medical research with human participants.25 But even if there were, it is unlikely that it s purpose would be to limit a physician's obligation to a patient seeking care. In fact, there is commentary to the effect that the highest standard of care is expected of a physician using a new or experimental procedure or treatment.26 There can be "an increased and extended duty of care" to the research participant.27

The most thorough analysis of medical research litigation has been done by Haavi Morreim, who calls for a new category of "negligent research injuries," stating that "courts need to recognize clinical research as a distinct area of medical activity and to attune tort doctrine specifically to its nuances."28 In so doing, she emphasizes the difference in objectives of the investigator and the physician, though noting that "courts must understand that when research takes place in the context of also providing medical care, it is important to distinguish which activities are part of the patient's medical treatment and which are part of the research."29 The former, she believes, "can rightly be addressed under traditional malpractice theories, but research should be addressed distinctively."30 However, Morreim also argues that research cannot be judged by the standards of practice, since it is, by definition, a deviation from those standards. Research, she claims, must be judged by the content and implementation of the study protocol, leaving the difficult question of whether a physician can deviate from the standards of practice to the quality of the protocol itself.31 She avoids directly addressing the difficult question concerning the obligations of physician/ investigators we pose here, stating that she will not address "whether or how the mixed physician/patient and investigator/ subject relationship should be permitted."32 Left open is whether a physician who is an investigator may act solely as an investigator and recruit patients seeking care into protocols involving experimental treatment interventions without creating a physician/ patient relationship and its resultant obligations. Unfortunately, we gain no insight into whether she would argue that the difference in objectives between research and therapy legally justify a physician/investigator to randomize a patient to a clinical trial arm in which he or she could receive less than the medical standard of care, i.e., no therapy, or a placebo, when established effective therapy exists. We contend that it would not.33

Morreim bases her arguments on the nature of research and how she believes courts should act, not how they do act. We found no line of cases in either the U.S. or Canada indicating that an injured participant could not seek compensation in medical negligence in the same way an ordinary patient could.

Randomizing patients to the placebo arm of a trial prevents them from receiving either established therapy or a novel therapy that is thought by a sufficient number in the expert clinical community to be equivalent and worthy of a clinical trial. Would this be considered an action below the standard of care of a medical practitioner as defined above? We need to ask, "Would a normal, prudent practitioner have offered an established effective treatment to a patient with this condition?" Unless there are special circumstances (e.g., where no proven effective treatment exists, where a patient is allergic to such treatment, where a patient has previously rejected established treatment because of negative side effects, or when the trial is for an "add on" therapy in which all patients receive standard treatment plus placebo), we suggest that replacing effective treatment with placebo controls would constitute negligence.

Is it possible to argue that placebo-controlled trials do meet the legal standard of care, even if established effective therapy is withheld? After all, in an active control trial, half the patients, those on the experimental arm, also have established effective therapy withheld. Further, they are exposed to an unapproved therapy that might carry risks, including the risk that it will be ineffective for the condition under study. However, for trials of new agents, there must be sufficient pre-trial information to create uncertainty about the comparative merits of each arm of the trial as the preferred intervention in a defined population.34 Such information includes animal studies, tests on healthy volunteers, case studies or information from similar pharmacological entities. In the best judgment of those designing the trial, participants should have an equivalent opportunity to benefit no matter which arm they are in. Both placebo and active control trials that do not meet this standard might be found to be"substandard medicine" by a court, with investigators not meeting the appropriate legal standard of care.

Relevance of Fiduciary Duty to the Standard of Care While there is hesitancy in the United States to open up another cause of action for negligent medical care,35 both U.S. and Canadian courts and commentators have recognized the fiduciary nature of the physician/ patient relationship. '6 Fiduciary law can be thought of not only as a separate source of legal duties, but instead as creating a "legal status that heightens or alters ordinary contract and tort duties."37

A fiduciary is defined by law as a person entrusted with power or property to be used for the benefit of another and is legally held to the highest standard of conduct.38 Physicians have both knowledge and skills needed by their patients, causing an imbalance in power between them. case law premises the fiduciary nature of the relationship on this "asymmetry of information" between physicians and their patients. Patients are intended to take advantage of the physicians' superior information and expertise with the expectation that it will be used in the patient's best interests.39 The patient's vulnerability may also be heightened by anxiety, physical discomfort, or distress and feelings of dependency.

The fiduciary relationship between physicians and patients involves "every element of trust, confidence and good faith."40 As fiduciaries, physicians must act in the best interests of their patients, refrain from exploiting their vulnerabilities, and not allow their own interests to come in conflict with those of their patients.41 If physician/ investigators are held to have a doctor- patient relationship, they may well be judged according to "heightened" duties, with an obligation to use their superior information and expertise for their patients' interest, not their own. How might a physician/investigator breach such duties in a randomized controlled trial? Allowing ill patient/participants to deteriorate in a placebo arm might amount to a subordination of their patients' health in favor of scientific, professional, or financial interests. We discuss this further below.

Viewing the investigator as a fiduciary may become important in an action for injury in placebo-controlled trials because malpractice law frequently ignores financial conflicts of interest. Physicians may receive financial benefits well beyond remuneration for professional services if they recruit patient-participants or conduct trials. They may also have a financial relationship with the sponsor of the trial. Professional rewards, such as publications, promotion, and enhancement of reputation in the research community are often competing interests.42These interests create the potential for an investigator's conflict with the primary duty of care to the patient. A randomized controlled trial with a placebo arm is not designed to benefit patients in that trial. It is designed for the benefit of others, whether they are future patients, investigators, sponsors, or investors. Therefore, patients injured from lack of treatment in the placebo arm may be able to argue that the physician/ investigator is a fiduciary with "heightened" duties to act in a way that favors the well-being of their patients participating in clinical trials.

Studies have looked at the role of trust in patients' decisions to participate in research. They show that patients trusted their physicians to never endorse options that were not in their best interests.43 Keep in mind that potential research participants are most often approached as patients when seeking medical attention in hospitals, clinics, or private physicians' offices. These issues clarify the importance of physician/investigators' fiduciary obligations in clinical trials, and their obligation to live up to them.

Given all of the above, we see no legal foundation to an argument that physician/investigators are absolved from their obligations as physicians and will not be liable to claims of medical negligence when engaging in clinical research.44 We have not found any legal mechanism by which a physician may shrug off those duties flowing from the status of physician or the physician/patient relationship. The duties are clear to us in medical negligence law, and are heightened when that relationship is viewed through the fiduciary lens.

Harm

Civil liability of the physician-investigator for administering placebos in randomized controlled trials depends on whether the patient has suffered damage and whether it can be proven that an appropriate intervention, whether a pharmacologically active substance, or a diagnostic or surgical intervention, "would have produced better effects."45 Without a centralized system for reporting adverse events, it is difficult to generalize about the effects of placebo-controlled trials. Yet the issue is not merely theoretical. Examples of potential harm in such trials have been studied extensively in psychiatry. This is particularly important given the large number of placebo-controlled trials in this area.

Research into the withdrawal of neuroleptics suggests serious clinical risk to patients whose medications are abruptly withdrawn in placebo controlled drug trials. Gilbert et al found uncommon risks of serious adverse effects such as hematemesis, neuroleptic malignant syndrome, emergent dyskinesia, tardive akathisia, and progressive Parkinsonism.46

Viguera and Baldessarini et al conducted a study involving 1,210 schizophrenic patients, most of whom were participants in the placebo arms of RCTs from 1969 to 1995. They compared continued versus discontinued anti-psychotic treatment and concluded that there is a high early risk of psychotic morbidity after the abrupt withdrawal of oral neuroleptics.47 "Appreciable increases in relapses can be detected even within days of discontinuation of medication in some particularly vulnerable or treatment-sensitive persons."48

Although Gilbert et al noted that patients who deteriorated after neuroleptic withdrawal often recompensated quickly when treatment was resumed,49 some evidence suggests that others "may have a difficult time returning to their previous level of function." Abrupt withdrawal may actually worsen the patient's condition by contributing to "treatment nonresponsiveness."50 The clinical risks of morbidity might occasionally exceed those associated with the natural history of the untreated illness. There is a "distinct probability" that schizophrenia becomes more difficult to treat with each relapse.51 Drawing on studies of both schizophrenia and depression, Viguera and Baldessarini claim that "worsening of primary disorders often follows the rapid removal of long-term psychotropic treatment."52 Thus the withdrawal of medication required for randomization in a placebo-controlled trial may worsen some patients' conditions, i.e., make them more difficult to treat after medication is resumed. If so, then the resulting harm may not be limited to the symptoms endured during the trial. Those symptoms can be harmful enough, although proponents of placebo-controlled trials often overlook them.

Symptoms that might be expected for participants in the placebo control arm of an anti-psychotic drug trial might include: increased paranoia, delusions, profoundly confused thinking, agitation, impaired self-care, and increased risk of aggressive behavior and harm to others.33 These are the direct harms of schizophrenia that initially lead patients to seek treatment. There can be painful consequences to this "clinical turmoil," including increased risk of legal confinement "with attendant psychosocial and financial costs, increased risk of suicide, and severe disruption of the lives of the patient and family."54

There is suggestive evidence that slow, tapered discontinuation might limit or delay relapse in schizophrenia. An emerging opinion is that a slow taper over a period of weeks or months is clinically recommended.55 Even so, there is no evidence that a slow taper followed by a placebo will prevent relapse in patients who responded favorably to standard anti-psychotics before enrolling in the trial. As for newly diagnosed patients, there is no precedent in the current clinical management of schizophrenia or depression for exacerbating a patient's condition by substituting placebos for effective treatment. Even if there are patients whom physicians would leave temporarily untreated, random assignment to treatment versus placebo is unknown in clinical care.

While definitions of schizophrenic relapse vary, they usually involve clinical assessment or the use of rating scales scores "to indicate the worsening of psychotic symptoms severe enough to warrant hospitalization or reinstitution of antipsychotic treatment."36 If a patient/participant must prove injury in a negligence action, then exacerbation of the psychotic symptoms that initially required treatment should constitute direct harm. Some have claimed that the import of this harm can be minimized in the interests of science. It has been suggested that there is "negligible evidence for any lasting disadvantage even in the face of substantial contemporaneous symptomatic disadvantage."57 second, it has been noted that risk is limited to the time the participant is enrolled in the trial.58 We fail to see how "substantial symptomatic disadvantage," let alone relapse, is rendered acceptable if it is so limited. These arguments might limit the quantum of damages in a negligence action, but they would not obviate the suffering on which the plaintiff's claim is based.

Similar concerns might be raised about antidepressant drug trials that use placebo controls. Geddes et al pooled data from thirty-one placebo-controlled, randomized trials involving 4,410 participants who had responded favorably to antidepressants. The aim of the study was to determine whether continuing treatment with antidepressants reduces the risk of relapse (i.e., "the return of symptoms during a period of remission") or recurrence (i.e., a new episode during a period of recovery) of depressive symptoms. The authors concluded that continued antidepressant therapy for at least another year would benefit many patients who remain at appreciable risk of recurrence after four to six months. This continued treatment could reduce the risk of relapse by 70% as opposed to treatment discontinuation. They noted, however, that the average rate of relapse on placebo was 41% compared to 18% on active treatment.59

The problems of clinical disadvantage from delayed treatment or use of placebo are not limited to psychiatry. Evidence exists that delay of treatment or failure to treat may modify the long-term course of illness on chronic diseases such as hypertension and rheumatoid arthritis. Short-term treatment of severe hypertension with placebo and long-term treatment of mild hypertension with placebo are associated with increased morbidity and mortality. In the case of rheumatoid arthritis, although many patients who meet the criteria for the disease may have self-limited or mild disease, cumulative evidence has shown that the course of rheumatoid arthritis in most patients seen at treatment centers (where clinical trials are more likely to take place) is not benign but is associated with functional decline, work disability, and premature death.60 Patients seen over time in treatment settings and those who enter clinical trials have persistent inflammatory symmetrical arthritis and experience radiological progression and functional declines.61

Evidence has also evolved that, although not inducing remission in most patients, disease-modifying anti-rheumatic drugs do have meaningful effects both when used alone and in combination - that may modify short-term and long-term outcomes of rheumatoid arthritis.62

For these diseases, a period of poor disease control may result in long-term, irreversible organ damage. With chronic conditions, therefore, ethical decisions about placebo controls should include consideration of the severity of the disease, the rate at which the dis ease causes irreversible damage, and the demonstrated capacity of treatment to alter the outcome of the disease."63

Causation

Causation can often be difficult to prove in medical negligence trials. There is no reason to believe that this would be different for placebo-controlled trials. In the case of a placebo-controlled trial for an anti-psychotic, the plaintiff would have to prove that, but for the assignment to the placebo arm of a trial, and the consequent withholding or withdrawing of treatment, the relapse or exacerbation of symptoms would no\t have occurred. The plaintiff would thus have to demonstrate that the provision of standard effective treatment would, on the balance of probabilities, have reduced or eliminated the harm. The same would hold for any functional decline, work disability, or premature death experienced by patient/participants receiving placebo in a trial for hypertension or rheumatoid arthritis.

Consent as a Defense

A person who is harmed by participation in a placebo-controlled trial may have a cause of action against an investigator. But there are potential defenses available to an investigator. Chief amongst them is an appeal to the autonomy of the patient in choosing to participate in the trial. Both law and medicine put a high premium on individual autonomy. Some therefore claim that so long as patients are competent, well informed, and can act freely, the choice to participate should be theirs.64 A substantial amount of contemporary medical case law has involved the notion of informed consent and the importance of insuring that any risks involved in medical interventions are assumed in an informed, voluntary fashion. The law further allows for a voluntary assumption of risk, in which case the plaintiff can waive the defendant's duty to observe a required standard of care. The notion of allowing altruistic patients to take on extra risk as research participants for the benefit of future patients has a certain appeal.

Yet there are a number of arguments against an unlimited "appeal to liberty" in the arena of medical care. The law allows for voluntary assumption of risk, but only in very limited circumstances and with limits on allowable risk. Some statutes specifically disallow the waiver of liability for negligent infliction of bodily harm. In the case of placebo-controlled clinical trials involving patients seeking medical care, the issue would be whether one can consent to the provision of care that does not meet the professional standard. If the same rules of medical law apply to research that evaluates therapeutic interventions on ill patients, treatment consistent with competent medical practice cannot be sacrificed. This will apply for all clinical research offered to patients for whom treatment is appropriate, whether the issue is introduction of an experimental drug in a clinical trial or use of placebo in the control arm. There is no such thing as "contracting" for what would otherwise be considered negligent practice. The law protects individuals from making such poor health-care choices because patients may be vulnerable. Such vulnerability is particularly possible because of illness. Patients have a relationship of trust with their physicians and are in a situation of power imbalance since physicians have greater medical knowledge. Consent of the weaker party constitutes a feeble defense for the bearer of a fiduciary obligation.65

We have found no legal precedent allowing physicians to "opt out" of their professional obligations because they are researchers in addition to being physicians. In fact, being a researcher adds obligations to those existing already by creating a heightened standard of care.66 A recent analysis of other legal jurisdictions has come to the same conclusion.67

Placebo and Public Policy

A public policy argument can also be made that asking patients to waive physicians' professional obligations to treat, will have a negative impact upon the practice of medicine and public health. People do not have unlimited discretion to choose whatever medical treatment they wish, including substandard care. The law protects people from making certain poor choices on the theory that people are vulnerable to making such choices when it comes to health matters.68

Some authors call for a clear distinction between research and therapy, proposing "that medical research and medical treatment are two distinct forms of activities." They claim that they are governed by different ethical principles.69 They ignore that the relevant participants in clinical trials are persons seeking medical care. They overlook the fact that for patients, care in clinical trials is not a "different form of activity." The fact that health services are part of a protocol designed to answer a research question cannot convert them into something other than health services even if they are experiments.

Patients should always understand the distinction between therapy and research and the implications research participation has for them. But one cannot honestly argue that obligations associated with the provision of clinical care can be ignored as if care were not being provided at all. Franklin Miller and his colleagues allege that those who object to the use of placebo controls when established effective therapy is available "conflat[e] the ethics of clinical trials with the ethics of therapeutic medicine."70 But neither moral theory nor legal principle distinguishes the primary obligations physicians have to those under their care, nor permits making patients' interests secondary to answering the research question. The fact that physician/investigators may benefit financially or professionally requires that they be held to an even higher standard than that of ordinary medical practice. Research participation need not exploit participants if the trial design assures genuine uncertainty about the various arms of trial.71 By enrolling in the clinical trial of a new therapeutic agent, individuals are agreeing to bear some risk, since less will be known about the new therapy than about treatments that have already been studied. However, to initiate a clinical trial, there must be sufficient pre-trial information (e.g., animal studies, tests on healthy volunteers, case studies) to create genuine uncertainty on the part of the expert clinical community about the comparative merits of each arm of the trial as the preferred intervention in a defined population. Benjamin Freedman referred to this as a state of "clinical equipoise" in which either arm of a clinical trial can be randomly offered as equivalent.72 It recognizes that the normative evaluation of medical practice by way of law, regulation, or ethics occurs by reference to an expert community standard.73 Freedman's notion of clinical equipoise thus provides a moral foundation for clinical trials in which the ethics of practice and research are combined.74 We believe it embodies the principles found in law as well. In other words, people seeking treatment should not, by enrolling in a trial, be agreeing to medical attention that is known to be inferior to current clinical practice. With limited exceptions, this forecloses the use of placebos when established, effective treatments exist.75

Conclusion

The ethical requirement that a trial proceed from a state of clinical equipoise can thus be seen as a fiduciary "vehicle"76 for evaluating the physician-investigator's adherence to professional standards of patientcentered care. Since principles of civil liability apply to patient care and to biomedical research, we contend that a patient who is harmed by receiving a placebo could claim for damages against the physician/investigator. While a patient may freely assume the risks of competently practiced medical interventions, physicians cannot escape the professional duty to practice competent medicine by eliciting their patients' informed consent to do so. This duty is premised on making the health of the patient the physician's first concern. We argue that this duty is not trumped by a competing interest in the advancement of research. Consequently, it also applies to physician/investigators who enroll patients in clinical trials.

Using the important distinction between research and therapy, intended to protect research participants, as a rationale for weakening physicians' obligations to their patients is a disturbing development. A physician's primary duty to patients is not optional, and cannot be ignored when that patient receives care in a clinical trial. Neither law nor ethics allows physicians to "opt out" of their professional obligations because they are also researchers.

Arguments claiming a "conflation" of the notions of research and therapy unfortunately have come at a time of increasing skepticism in the general population about the motives of researchers and research sponsors, and the ability of governments to protect participants. Clinical trials recruiting patients seeking clinical care are not dealing with hypothetical patients who have no medical needs and who fully understand that that their best interests take a back seat to the benefit of, at best, future patients, and at worst, the bottom line for sponsors looking to capture a share of the market

No one would suggest that a clear distinction should not be made between medical care provided in the context of research and medical care provided outside this context. Even with careful pretrial studies, there are unknown risks in research as well as extra procedures and tests. But this distinction does not provide a legal or moral justification for providing substandard treatment for trial participants.

In this analysis we have not explored all possible aspects of potential physician/investigator legal liability in placebo- controlled trials when established effective therapy exists. Nor have we evaluated all possible defenses that might be raised. However, we do make the case that physician/investigators may be liable for harm to patient/participants who are randomized to the placebo arm of a trial, and that consent may not provide an adequate defense. Physician/investigators, institutions, Institutional Review Board members, regulators, and sponsors should take note of the potential for legal liability.

In all tort cases, a number of elements must be established: that a duty of care was owed; that the duty was breached, and that the breach was the cause of injury.

Viewing the invest\igator as a fiduciary may become important in an action for injury in placebo-controlled trials because malpractice law frequently ignores financial conflicts of interest. Physicians may receive financial benefits well beyond remuneration for professional services if they recruit patient-participants or conduct trials.

The problems of clinical disadvantage from delayed treatment or use of placebo are not limited to psychiatry. Evidence exists that delay of treatment or failure to treat may modify the long-term course of illness on chronic diseases such as hypertension and rheumatoid arthritis.

We have found no legal precedent allowing physicians to "opt out" of their professional obligations because they are researchers in addition to being physicians.

Acknowledgements

Thanks to Miriam Brouillet for her research assistance on this paper. This work was funded by the Canadian Institutes of Health Research and the Social Sciences and Humanities Council of Canada.

References

1. F. G. Miller and H. J. Silverman, "The Ethical Relevance of the Standard of Care in the Design of Clinical Trials," American Journal of Respiratory Critical Care Medicine 169, no. 5 (2004): 562- 64; F. G. Miller and H. Brody, "The Clinician-investigator: Unavoidable but Manageable Tension," Kennedy Institute Ethics Journal 13, no. 4 (2003): 329-46; F. G. Miller and H. Brody, "What Makes Placebo-controlled Trials Unethical?" American Journal of Bioethics 2, no. 2 (2002): 3-9.

2. F. G. Miller and H. Brody, "A Critique of Clinical Equipoise," Hastings Center Report 33, no. 3 (2003): 19-28 at 20, 24-25.

3. Id. at 24.

4. Id. at 26-27.

5. R. Temple, "Government Viewpoint of Clinical Trials," DrMg Information Journal (1982): 10-17; R. Temple, "Problems in Interpreting Active Control Equivalence Trials," Accountability 4 (1996): 267-75; R. Temple, and S. S. Ellenberg, "PlaceboControlled Trials and Active-Controlled Trials in the Evaluation of New Treatments," Annals of Internal Medicine 133 (2001): 455-63.

6. K. C. Glass and D. Waring, "Effective Trial Design Need Not Conflict with Good Patient Care" American Journal of Bioethics 1, no. 2 (2002): 25-26; D. Fergusson, K. C. Glass, D. Waring, and S. Shapiro, "Turning a Blind Eye: The Success of Blinding Reported in a Random Sample of Randomized, Placebo-Controlled Trials," British Medical Journal, 328, no.7437 (2004): 432-40.

7. Id.; see also F. Benjamin, "Placebo-Controlled Trials and the Logic of Clinical Purpose," IRB: A Review of Human Subjects Research 12 (1990): 1-6; K. J. Rothman and K. B. Michels, "The Continuing Unethical Use of Placebo Controls," N. Engl. J. Med., 331 (1994): 394-98; B. Freedman, K. C. Glass and C. Weijer, "Placebo Orthodoxy in Clinical Research II: Ethical, Legal, and Regulatory Myths," Journal of Law, Medicine & Ethics 24 (1996): 252-59, at 256; K. C. Glass and D. Waring, "Effective Trial Design Need Not Conflict With Good Patient Care" American Journal of Bioethics 2 (2002): 25-26, at 26; F. G. Miller and H. Brody, "What Makes Placebo-Controlled Trials Unethical?" American Journal of Bioethics 2 (2002): 3-9, at 2-5; E. Emanuel and F. G. Miller, "The Ethics of Placebo-Controlled Trials: A Middle Ground," N. Engl. J. Med., 345 (2001): 915-19; National Placebo Initiative, Health Canada & Canadian Institute of Health Research, Draft Report of the National Placebo Working Committee (October, 2004).

8. see B. Freedman, K. C. Glass and C. Weijer, supra note 7; K. C. Glass and D. Waring, supra note 6; D. Waring, and K. C. Glass,, "Legal Liability for Harm to Research Subjects in Placebo Controlled Trials" in New Directions in Biomedical Research: Regulation, Conflict of Interest and Liability, T. Lemmens and D. Waring (eds) (Toronto: University of Toronto Press, in press).

9. We are aware of no such judgments. However, a Complaint and Demand for Jury Trial was filed in the Orange County, North Carolina Superior Court by a patient/ plaintiff randomized to the placebo arm of a clinical trial testing a new psoriasis treatment. The plaintiff claimed damages for the development of debilitating psoriatic arthritis resulting from participation in the trial. The trial protocol required discontinuation of his current medication and withholding of the experimental drug. William Hamlet v. Genentech, Inc; Xorna, Ltd.; Western Institutional Review Board, Inc.; Parexel International, LLC; Mark. S. Fradin, M.D.; and Chapel Hill Dermatology, P.A., Superior Court of Justice, Superior Court division 03 CVS 1161, Orange County, North Carolina (July, 2003). Five of the six defendants have settled. Personal communication, Alan Milstein, attorney for the plaintiff, June 8, 2005.

10. M. M. Mello, D. M. Studdart and T. Brennan, "The Rise of Litigation in Human Subjects Research" Annals of Internal Medicine 139 (2003): 40-45.

11. Bok, Universities in the Marketplace (NJ: Princeton, 2003): at 144-50; K. C. Glass and T. M. Lemmens, "Conflict of Interest and Commercialization of Biomedical Research: What is the Role of Research Ethics Review?" in T. Caulfield and B. Williams-Jones, eds., The Commercialization of Genetic Research: Ethical, Legal and Policy Issues (New York: Plenum, 1999): 79-99 at 86-88.

12. D. B. Dobbs, The Law of Torts, vol. 1 (St. Paul, MN: West Publishing Co., 2001): at 269; G. Robertson, "Negligence and Malpractice," in J. Downie, T. Caufield, and C. Flood, eds., Canadian Health Law and Policy, 2d. ed. (Markham, Ontario: Butterworths Canada Ltd., 2002): 91-109, at 91; M. A. Jones, Medical Negligence (London: Sweet and Maxwell, 1991): at 20; D. Giesen, International Medical Malpractice Law: A Comparative Law Study of Civil Liability Arising From Medical Care (Dordrecht: Nijhoff, 1988): at 25, 30-31; J. H. King Jr., The Law of Medical Malpractice in a Nutshell, 2d. ed. (St. Paul, MN: West Publishing Co., 1986): at 9.

13. E. I. Picard and G. B. Robertson, Legal Liability of Doctors and Hospitals in Canada, 3d. ed. (Toronto: Carswell, 1996); Restatement (second) of Torts 282 (1965).

14. Miller and Brody, supra note 1, at 3-5; see also Emanuel and Miller, supra note 7, at 915-19.

15. Picard and Robertson, supra note 13, at 1-8.

16. American Medical Association Council on Ethical and Judicial Affairs, Code of Medical Ethics, at (last visited June 22, 2005).

17. Id.

18. Id. at Article E-2.07, Clinical Investigation.

19. Declaration of Helsinki (Helsinki: World Medical Association, 52nd General Assembly, 2000).

20. Id. at Articles 3,5,10, and 15; see also Medical Research Council of Canada, Natural Sciences and Engineering Research Council of Canada, Social Sciences and Humanities Research Council of Canada, Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (Ottawa: Public Works and Government Services Canada, 1998): at i.4.

21. Hall v. Hilburn, 466 So. 2d 856, 858 (MS 1985).

22. Crits v. Sylvester, 1 D.L.R. (2d.) 502, 508, (1956), aff'd [1956] S.C.R. 991.

23. Picard and Robertson, supra note 13, at 186.

24. Miller and Brody, supra note 1, at 3-5; see also Emanuel and Miller, supra note 7, at 915-19.

25. Giesen, supra note 12, at 50.

26. Id. at 193-195.

27. D. Giesen, "Civil Liability of Physicians for New Methods of Treatment and Experimentation: A Comparative Examination," Medical Law Review 3 (1995): 22-52, at 30.

28. E. H. Morreim, "Medical Malpractice Litigation and Malpractice Tort Doctrines: Courts on a Learning Curve," Houston Journal of Health Law and Policy 4 (2003): 1-92, at 92. Professor Morreim published a further iteration of these arguments in "Litigation in Clinical Research: Malpractice Doctrines Versus Research Realities," Journal of Law, Medicine & Ethics 32 no. 3 (2004): 474-484.

29. Id. at 34

30. Id.

31. Id. at 41.

32. Id. at 52.

33. Professor Morreim does allow in her discussion of fiduciary duties that there could be a relationship in a "pure research setting, in which the patient's only relationship with an investigator is as a research subject." Id. at 50-51. We believe that this could only be the case for "patients" who are not seeking or needing care, e.g., recruitment of persons to donate blood for a study unrelated to their own care. This situation is easily distinguishable from one in which a patient seeks advice and care from a physician who is an investigator and is recruited into a trial for a therapy addressing the patient's needs, including those with a placebo control.

34. B. Freedman, "Equipoise and the Ethics of Clinical Research," N. Engl. J. Med. 317 (1987): 141-45.

35. M. A. Rodwin, "Strains in the Fiduciary Metaphor: Divided Physician Loyalties and Obligations in a Changing Health Care System," American Journal of Lam and Medicine 21 (1995): 241-57, at 243.

36. M. A. Hall, M. A. Bobinski and D. Orentlicher, Health Care Law and Ethics, 6th ed. (New York: Aspen Publishers, 2003): at 163- 64; Lockett v. Goodill, 430 P.2d 589, 591 (WA, 1967); Moore v. Regents of the University of California, 793 P.2d 479 (1990); Mclnerney v. MacDonald (1992) 93 D.L.R. (4th) 415; Norberg v. Wynrib, (1992) 2 S.C.R. 227; Rodriguez v. British Columbia (Attorney General), (1993) 3 S.C.R. 519.

37. Hall, Bobinski and Orentlicher, supra, at 163-64.

38. W. L. Presser and P. Keeton, The Law of Torts, 5th ed. (St. Paul. MN: West Publishing Co, 1984).

39. M. J. Mehlman, "Fiduciary Contracting Limitations on Bargaining Between Patients and Health Care Providers," University of Pittsburgh Law Review 51 (1989-1990): 365-417, at 390.

40. Lockett v. Goodill, 430 P.2d 589, 591 (WA, 1967), as cited by Hall, Bobinski and Orenticher, supra note 35.

41. G. Robertson, Negligence and Malpractice, in J. Downie, T. Caulfield and C. Flood, eds., Canadian Health Law and Policy, 2nd ed. (Markham, Ontario: Butterworths, 2002): 91-109; Morreim, supra note 26, at 48; Moore v. Regents of the University of California, 793 P.2d 479, 483, 488 (1990). The Supreme Court of California has determine\d that an interest in research can conflict with an interest in the patient-participant's health and that a physician- investigator's eagerness to promote the advancement of science may result in riskier experimental treatments. The physician- investigator must disclose to a participant "any personal interests unrelated to the patient's health, whether research or economic, that may affect the physician's professional judgment." An action for conversion was dismissed even though the physician converted the patient-participant's blood products into a marketable commodity without consent. But the action for breach of the fiduciary duty to disclose financial interests in the research was accepted.

42. Glass and Lemmens, supra note 11.

42. Advisory Committee on Human Radiation Experiments, The Human Radiation Experiments (New York: Oxford University Press, 1996).

44. Gomez v. Comit excutif du Conseil des mdecins, dentistes et pharmaciens de l'Hpital universitaire de Qubec, (2001) J. Q.. No. 5544. This was dealt with specifically by the Quebec Court of Appeal. At issue was a complaint filed with a university hospital medical disciplinary committee by the family of a man who died after participating in a research study on prostate cancer. The complainant argued that the patient/research subject had not received the proper standard of care, and that as a result of the negligence of the physicians involved in the research project, he had not been diagnosed in a timely fashion. The researchers argued that medical research escapes the traditional disciplinary rules of the profession and is only submitted to the authority of specialized research ethics committees. The Court stated that research is part of the mission of a university hospital; that in this type of research subjects could reasonably have expected diagnostic services and exceptional care; and that the terms "medical act" and "research" cannot be strictly separated. Discussing the latter, the Court points out that clinical research is an integral part of medicine, and is undertaken by physicians whose first professional duty is the protection of the health and well-being of individuals. Biomedical research and medical acts are not in opposition to one another, the Court states. In this court's view, research subjects can rightly expect that when research activities undertaken in medical research centers involve medical procedures, such as diagnostic services, the interventions they undergo will meet the standard of care physicians owe to their patients.

45. Giesen, supra note 12, at 595-96, 251.

46 P. L. Gilbert, J. Harris et al, "Neuroleptic Withdrawal in Schizophrenic Patients," Archives of General Psychiatry 52 (1995): 173-88, at 173,175,182-85; B. Spivak et al, "Neuroleptic Malignant Syndrome During Abrupt Reduction of Neuroleptic Treatment," Acta Psychiatrica Scandinavia 81 (1990): 168-69.

47. A. C. Viguera and R. J. Baldessarini et al, "Clinical Risk Following Abrupt and Gradual Withdrawal of Maintenance Neuroleptic Treatment," Archives of General Psychiatry 54 (1997): 49-55, at 49- 54.

48. R. J. Baldessarini, A. C. Viguera, "Medication Removal and Research in Psychotic Disorders," Archives of General Psychiatry 55 (March, 1998): 281-82, at 282.

49. Gilbert, Harris et al., supra note 46, at 182.

50. R. J. Wyatt, "Neuroleptics and the Natural Course of Schizophrenia," Schizophrenia Bulletin 17 (1991): 325-51; J. R Greden and R. Tandon, "Long-Term Treatment for Lifetime Disorders?" Archives of General Psychiatry 52 (1995): 197-99, at 198.

51. Greden and Tandon, supra, at 198.

52. Viguera, and Baldessarini supra note 47, at 51.

53. Greden and Tandon, supra note 51, at 198; C. Elliott and C. Weijer, "Cruel and Unusual Treatment," Saturday Night 110, no. 10 (1995): 31-34, at 32.

54. Greden and Tandon, supra note 51, at 198. There is anecdotal evidence that details such as relapse rates have been kept out of the informed consent process because they were "overly frightening." Put another way, such details might discourage enrollment in RCTs. The altruism that supposedly motivates some psychiatric patients to enroll in placebo-controlled trials may not endure an understanding of the potential for adverse effects. see also D. King, "Debatable Forms of Consent," The Boston Globe, November 16, 1998, at AOl. King quotes Wesley Acorn, then president of the national consumer council for the U.S. National Alliance for the Mentally 111, as saying, "Do you think people say 'Gee, I'll sign up for more suffering. Many of us suffer enough on our own.'"

55. Baldessarini and Viguera, supra note 46, at 282; Viguera, Baldessarini et al., supra note 45, at 54; R. J. Baldessarini, "Risks and Implications of Interrupting Maintenance Psychotic Drug Therapy," Psychotherapy and Psychosomatik 63 (1995): 137-41, at 139; D. V. Jeste, P. L. Gilbert et al., "Considering Neuroleptic Maintenance and Taper on a Continuum," Archives of General Psychiatry 52 (1995): 209-12, at 210; R. Wyatt, "Risks of Withdrawing Antipsychotic Medications," Archives of General Psychiatry, 52 (1995): 205-08, at 207.

56. A. C. Viguera, R. J. Baldessarini et al., supra note 47, at 50.

57. W. T. Carpenter and R. Coney, "Sense and Nonsense: An Essay on Schizophrenia Research Ethics," Sdzophrenia Research 35 (1999): 219-25, at 222.

58. W. T. Carpenter, P. S. Appelbaum and R. J. Levine, "The Declaration of Helsinki and Clinical Trials: A Focus on PlaceboControled Trials in Schizophrenia," American Journal of Psychiatry 160 (2003): 356-62, at 359.

59. J. R. Geddes, S. M. Carey, C. Davis and T. A. Furukawa et al., "Relapse Prevention with Antidepressant Drug Treatment in Depressive Disorders: A Systematic Review," The Lancet 361 (2003): 653-61.

60. T. Pincus and L. F. Callahan, "The 'Side Effects' of Rheumatoid Arthritis: Joint Destruction, Disability and Early Mortality," British Journal of Rheumatology 32, suppl. 1 (1993): 28- 37; F. Wolfe and M. A. Cathey, "The Assessment and Prediction of Functional Disability in Rheumatoid Arthritis," Journal of Rheumatology 18 (1991): 1298-1306; P. A. Reilly, J. A. Cosh, P. J. Maddison, J. J. Rasker and A. Silman, "Mortality and Survival in Rheumatoid Arthritis: a 25 Year Prospective Study of 100 Patients," Annals of the Rheumatic Diseases 49 (1990): 363-69.

61. P. Emery, "The Roche Rheumatology Prize Lecture: The Optimal Management of Early Rheumatoid Disease: The Key to Preventing Disability," British Journal of Rheumatology 33 (1994): 765-68; T. Pincus and L. E. Callahan, "How Many Types of Patients Meet Classification Criteria for Rheumatoid Arthritis?" Rheumatology 21 (1994): 1385-89.

62. J. F. Fries, C. A. Williams, D. Morfeld, G. Singh and J. Sibley, "Reduction in Long-Term Disability in Patients with Rheumatoid Arthritis by Disease-Modifying Anti-rheumatic Drug-Based Treatment Strategies," Arthritis & Rheumatism 39 (1996): 616-22; P. L. Van Riel, D. M. van der Heijde, I. H. Nuver-Zwart and L. B. van de Putte, "Radiographic Progression in Rheumatoid Arthritis: Results of 3 Comparative Trials," Journal of Rheumatology 22 (1995): 1797- 99; O. Forre, "Radiologic Evidence of Disease Modification in Rheumatoid Arthritis Patients Treated with Cyclosporine: Results of a 48-week Multicenter Study Comparing Low-Dose Cyclosporine with Placebo: Norwegian Arthritis Study Group," Arthritis & Rheumatism 37 (1994): 1506-12.

63. C. M. Stern and T. Pincvis, "Placebo-Controlled Studies in Rheumatoid Arthritis: Ethical Issues," The Lancet 353 (1999): 400- 03, at 400.

64. Id; Temple and Ellenberg, supra note 5.

65. T. Lemmens, D. Sprumont, H. Nys, J. Singh, and K. C. Glass, "CIOMS' Placebo Rule and the Promotion of Negligent Medical Practice" European Journal of Health Law 11 (2004): 153-74, at 163.

66. Neufield v. McQuitty, 18 AR 271(1979); Halushka v. University of Saskatchewan 53 D.L.R. (2d) 436 (1965); Cryderman v. Ringrose, 3 W.W.R. 481 (Alta. CA, 1977); Coughlin v. Kuntz 17 B.C.L.R. 365 (1987); Civil Code of Quebec, S.Q,, 1991, Articles 1474,1477; Grimes v. Kennedy Kreiger Institute, 782 A.2d 807, 834 (2001); Gomez v. Comit excutif du Conseil des mdecins, dentistes et pharmaciens de l'Hpital universitaire de Qubec, supra note 43.

67. T. Lemmens et al, supra note 65.

68. J. Menikov, Law and Bioethics: An Introduction (Washington, DC: Georgetown University Press, 2001).

69. Miller and Brody, supra note 1.

70. Miller, supra note 2.

71. Freedman, supra note 34.

72. Id.

73. B. Freedman, "Ethics and Placebo-Controlled Thrombolytic Trials: The Future," Coronary Artery Disease 2, no. 7 (1991): 849- 52, at 850.

74. C. Weijer, "When Argument Fails," American Journal of Bioethics 2 (2002): 10-11, at 10.

75. The National Placebo Working Committee Report on the Appropriate Use of Placebo in Clinical Trials (Ottawa: Health Canada and Canadian Institutes of Health Research, 2004).

76. B. Freedman, K. C. Glass and C. Weijer supra note 7; K. C. Glass and D. Waring supra note 7, at 26. M. Oberman, "Mothers and Doctor's Orders: Unmasking the Doctor's Fiduciary Role in Maternal- Fetal Conflicts," Northwestern University Law Review 94 (2000): 451- 501, at 459.

Kathleen Cranley Glass, A.B., LL.B., D.C.L., is Director ofMcGiIl University's Biomedical Ethics Unit, Associate Professor in the Departments of Human Genetics andBioethics and Clinical Ethidst at The Montreal Children's Hospital. DuffWaring, LL.B., Ph.D., is an Assistant Professor at York University in Canada

Copyright American Society of Law and Medicine, Incorporated Fall 2005

Source: Journal of Law, Medicine & Ethics, The

More News in this Category