Dr. F, an oncologist in a small community practice, has been asked by a pharmaceutical company to conduct early-phase clinical trials involving several new investigational chemotherapeutic agents that do not yet have FDA approval. These would be very small phase I trials, with the possibility of conducting some phase I/II and phase II trials in the future as well. The reimbursement he will receive for the research will substantially increase the income of his practice, provided that he is able to recruit and retain a sufficient number of subjects. “More importantly, though,” Dr. F thinks to himself, “I have so little to offer many of my sickest patients now. The best thing about doing clinical research is being able to offer them something new, that just might be their best hope.”
Dr. G treats patients with hemophilia. Although treatments have improved dramatically in recent years, hemophilia is a devastating, and devastatingly expensive, chronic disease. Because she has high hopes about promising experimental technologies, she also conducts research. She prides herself on the research partnerships she develops with patient–subjects who seek to contribute to the development of better treatments. Recently, however, she has received inquiries from patients with hemophilia from around the world who want to enroll in her research because the experimental interventions are provided free of charge. These patients tell her that they cannot afford standard therapies, and that enrolling in her research is their only hope for treatment. Dr. G is troubled by this reasoning, and discusses it with a colleague, who responds, “Lots of people enroll in research to get treatment. Almost all pediatric cancer patients are enrolled in research. And when HIV-positive patients can’t afford the drugs, my colleagues in the ID clinic immediately look to see what trials they might qualify for – even before they look for a free drug program they could use. Why are you worried about patients who want to be in research to get treated?”
The differences between the role of physician and that of researcher become readily apparent in the definitions of “clinical research” and “medical practice.” The Belmont Report distinguished between these activities as follows: “For the most part, the term ‘practice’ refers to interventions that are designed solely to enhance the well-being of an individual patient … and that have a reasonable expectation of success … . By contrast, the term ‘research’ designates an activity designed to test an hypothesis, permit conclusions to be drawn, and thereby to develop or contribute to generalizable knowledge …” (National Commission for the Protection of Human Subjects of Biomedical and Behavioral Research, 1979). Clinical research enrolling patients as research subjects is essential for the efficient development of safer and more effective treatments, but because of the differences between research and clinical practice, clinical research is given special regulatory oversight and research subjects are afforded special protections.
When physicians who treat patients also engage in clinical research, they take on a dual role, because the activities in which they engage have elements of both the clinician and the researcher. These roles have important differences that can conflict or become blurred. The differences in the conduct and goals of clinical research and medical practice can affect informed consent, alter the rights and duties of patient–subjects and clinical investigators, and give rise to conflicts of interest.
It is increasingly common for physicians in private practice as well as in academic medicine to find themselves in the dual role. Failure to understand and appreciate the differences between research and treatment can have adverse effects on patients’ understanding and safety, and it could compromise the value and validity of research. It is, therefore, increasingly important for physicians to recognize and manage the dual role, both in order to conduct scientifically and ethically sound research and to avoid confusion, harm, and exploitation of patients and research subjects.
There are several key issues for clinicians considering involvement in research: (i) understanding the scientific justification for conducting clinical research and what counts as an acceptable balance of risks of harm and potential benefits in clinical research; (ii) avoiding exploitation of patient–subjects and promoting justice in recruitment, enrollment, and the post-trial care of patients; (iii) ensuring that patients’ participation in research is voluntary and adequately informed; and (iv) properly managing the dual roles of physician and researcher.
There is a great deal of ethical, legal, regulatory, and policy guidance addressing both clinical research and medical practice. However, little of this guidance addresses the particular challenges of the physician’s dual role.
A small number of scholarly articles directly addresses the problem of managing the dual roles of physician–investigator (Churchill, 1980; Glass and Waring, 2002; Lemmens and Miller, 2002; Brody and Miller, 2003). A much larger body of scholarly literature addresses the physician’s moral role and responsibilities in conducting clinical research. Most of this discussion focuses on phase III randomized controlled clinical trials (RCTs). How can physicians’ moral obligation of beneficence – the obligation to do the best for their patients – be compatible with conducting research that involves their patients? For phase III research – the last step in determining whether an experimental intervention is sufficiently safe and effective to be considered a treatment – if a researcher believes that the experimental intervention is better than standard treatment, her duty as a physician seems to be to provide the experimental intervention. If the researcher believes that standard treatment is better, her duty is to provide the standard treatment. Either way, the physician should not be doing the research.
This conundrum centers on the concept of “equipoise.” Research should develop evidence that answers questions about an unproven intervention; the novel intervention has to look promisingly safe and effective, but the evidence for its safety and effectiveness relative to the alternatives must be lacking. That is, the research itself should be necessary in order to “disturb” equipoise and cause the scientific community to say, “Yes, now we know that intervention A is/is not a better treatment for X disease.” Because individual physician–researchers cannot easily both be in equipoise and fulfill their duties to their patients, the philosopher Benjamin Freedman advocated for the concept of “clinical equipoise” (Freedman, 1987) – meaning that in order for RCTs to go forward there must be a collectively held equipoise within the field. Individual clinical researchers need not have equipoise themselves; they need only recognize that there is disagreement within medicine about whether standard treatment or the unproven intervention is really better. Importantly, proponents of clinical equipoise oppose the use of placebo controls in clinical research except in very limited circumstances (Freedman, 1990; Freedman et al., 1996a,b).
Recently, debate has emerged about whether equipoise is the right concept to employ when distinguishing clinical research from medical practice. Some bioethics scholars argue that focusing on clinical equipoise, because it addresses the question of which is the better treatment, makes the moral mistake of blurring research and treatment. These scholars reason that research should be viewed as a distinct enterprise, which does not, cannot, and should not in itself serve the goal of providing treatment to patients. “Which is the better treatment?” is a question that cannot be answered until the research is completed; therefore, clinical equipoise is a misleading moral compass for clinical research. The alternative proposed by these scholars is a shift away from the norm of beneficence to non-maleficence, that is, an effort to minimize or eliminate where possible harms from research participation. The failure to reduce or eliminate such harms results in exploitation of patient–subjects (Miller and Brody, 2002, 2003; Miller and Rosenstein, 2003).
Although adhering to the norm of non-maleficence and a goal of “non-exploitation” may promote clear distinctions between clinical research and medical practice, clinical equipoise appears to be a richer and more satisfying moral standard for physician–researchers, who often view themselves as treating patients through clinical research (Miller and Weijer, 2003). Whether clinical equipoise provides adequate guidance for physicians’ decision making about whether research participation can maximize potential benefit to particular patient–subjects is currently the subject of debate (Hellman, 2002; Evans and London, 2006). Clinical equipoise does not, however, adequately address the issues raised by the enrollment of patients in early-phase clinical trials, as its focus is on RCTs.
Related to the challenge of appropriately distinguishing research from medical practice is the problem of the “therapeutic misconception.” This name was coined by Paul Appelbaum and colleagues (Appelbaum et al., 1982, 1987; Appelbaum, 1996, 2002) to describe the observation that some patients enrolled as subjects in clinical research failed to recognize that they were participating in research and misinterpreted some key features of research protocols. For example, subjects might not understand that they could be randomized to receive standard treatment, an unproven intervention, or a placebo, believing instead that the intervention they received was selected by their physician as the best treatment for them. Another aspect of therapeutic misconception, according to at least some definitions, is an unrealistic expectation of direct benefit from the experimental intervention. For instance, subjects may expect to be cured of their disease if they take an unproven drug, even though the research goal is to determine a safe dose of the drug in the short term (Horng and Grady, 2003).
Although Appelbaum and colleagues posited that both patients and physician–researchers could be affected by therapeutic misconception, it has most often been described as a characteristic of research subjects. However, when researchers have great hope in an unproven intervention, or are excited by the promise of new science, their own misconception may not only blur their sense of the differences between research and medical practice but also affect the informed consent process, inappropriately influencing patients’ decisions about research participation (Churchill et al., 1998; Dresser, 2000, 2002; Miller, 2000).
Another moral tension in the dual roles of the physician–researcher becomes evident when physicians approach their own patients about research participation. The patient may be reluctant to say no, out of a sense of indebtedness to the physician or out of concern that a refusal could impair the relationship and adversely affect future treatment. When there is confusion between research and treatment, or about the role and responsibilities of the physician–investigator, these perceived pressures could also affect subjects’ understanding of the research, or unduly influence their enrollment or continued participation. This may be true whether the confusion is the patient’s or the physician’s (Kass et al., 1996; Levinsky, 2002; Chen and Miller, 2003).
There are relatively few court decisions about clinical research. Many of those decisions address plaintiffs’ claims that they were unaware they were participating in research or were led by researchers to believe that they were being treated when in fact they were enrolled in research (Toth v. Community Hospital at Glen Cove, 1968; Estrada v. Jacques, 1984; Re Cincinnati Radiation Litigation, 1995). Several recent decisions involving research with vulnerable populations have suggested that researchers have physician-like obligations of beneficence toward their subjects and would restrict research participation by vulnerable subjects to circumstances where benefit to them is expected (TD v. New York State Office of Mental Health, 1996; Grimes v. Kennedy Krieger Institute, 2001). Such a stringent standard may protect vulnerable patients at the expense of maintaining a meaningful distinction between clinical research and medical practice.
A recent and controversial decision by the US Court of Appeals has further reinforced a treatment-oriented perspective on clinical research by finding that, under some circumstances, terminally ill patients may have a constitutional right of non-trial access to experimental interventions that have been shown to be safe enough in phase I trials to proceed to further testing (Abigail Alliance v. Eschenbach, 2006). It remains to be seen whether these legal developments will make it more difficult for patients and physicians to distinguish research from treatment.
The precise nature and significance of the difference between clinical research and medical practice has been debated as a matter of policy since the creation of the Nuremberg Code, the document that emerged from the Doctors’ Trial at Nuremberg describing the duties of researchers conducting human experimentation (Nuremberg Code, 1949). In 1964, the World Medical Association promulgated the Declaration of Helsinki specifically to address medical research enrolling patients as subjects, to which many physicians believed the Nuremberg Code did not apply. The Declaration addresses the challenges of combining clinical research with medical care, framed in terms of the duties of physicians “to promote and safeguard the health of the people.”
Another important policy perspective on the dual role comes from the legal literature on clinical research. A key legal characteristic of the physician–patient relationship is its fiduciary character: patients repose trust in their physicians, who make recommendations based on the patients’ best interests, as mutually determined. It has recently been persuasively argued by several scholars that the researcher–subject relationship has at least some characteristics of a fiduciary relationship (Coleman, 2005; Miller and Weijer, 2006). Although the two relationships are not the same, clinical researchers have some duties that are similar to those of physicians: for example, to protect patient–subjects from undue harm, and to support their informed decision making about research participation. Precisely how to characterize the differences between the roles of physician and researcher and the duties that follow from them remains to be determined; nonetheless, the fiduciary model offers a promising way to think about dual-role problems in clinical research.
There has been a great deal of empirical research addressing “therapeutic misconception.” Each study employs a somewhat different definition of the concept, as well as different questions to measure it in research subjects, but all the literature acknowledges that many patients who are research subjects appear to lack an adequate understanding of the difference between clinical research and medical treatment (Gray, 1975; Daugherty et al., 2000; Joffe et al., 2001; Appelbaum et al., 2004; Henderson et al., 2006). Recently, some scholars have begun examining in more depth the views of physician–investigators and research coordinators, in addition to patient–subjects, in order to assess whether therapeutic misconception in other participants involved in research is related to that in patient–subjects (Daugherty et al., 1995; Joffe and Weeks, 2002; Henderson et al., 2004a,b). Some of this research also examines how both patient–subjects and clinical researchers view the dual roles of physician and researcher (Henderson et al., 2004b; Easter et al., 2007). Other research examines how research consent forms describe clinical trials and their potential benefits for patient–subjects. There is agreement that the consent form and process can be a source of misconception in patient–subjects, but consensus does not exist regarding recommendations and guidance based on these data (Horng et al., 2002, 2003; King et al., 2005).
It is essential that physicians think of clinical research as distinguishable from, and complementary to, medical treatment. All clinical research is essentially future oriented: after the research is completed, the data gathered from it will inform medical treatment for future patients. In contrast, the physician’s duty to act in the best interests of the patient is immediate. Yet the clinical researcher also has immediate duties to research subjects: to protect them from harm insofar as possible, to ensure that they are provided with adequate information about research participation, and sometimes – depending on the design of the research – to maximize the possibility of benefit to them.
A great deal of guidance exists to assist clinical researchers in addressing research ethics questions, but there are few resources specifically focused on the problem of dual roles. Some have suggested maintaining a clear demarcation between the roles, for example by donning a red coat instead of a white coat in research settings (Dresser, 2002) or explaining to patients “I have two hats; a doctor hat and a researcher hat. I’m going to take off my doctor hat now and put on my researcher hat, so I can talk to you about a research study.”
Regardless of how physician–investigators emphasize the distinction to patients and research subjects, the best opportunities for clarifying the differences lie in the consent form and process. As a classic article in the legal literature notes, informed consent “encourages self-scrutiny by the physician–investigator” (Capron, 1974): meaning that the duty to explain the research to potential subjects also provides physician–researchers with an opportunity to make their own views clear to themselves, through articulating them to others.
It is important to recognize that delineating the characteristics of clinical research and the differences between participating in clinical research and receiving medical treatment is an exercise that is ultimately specific to each research protocol as well as to the treatments available to patients who may become subjects. Physician–researchers, and physicians who refer patients to clinical researchers, should take the time necessary to consider the ethical implications of their roles and to undertake clear and careful discussion about the differences between research and treatment with patients.
To manage the dual role, physician–researchers should:
Dr. F wants very much to be able to help his patients but has little to offer that can meaningfully extend the lives of many of them. As a result, he believes, as many oncologists do, that experimental agents hold more promise of safety and efficacy than do standard treatments. If Dr. F’s research goal is to treat his patients through a trial, then he may have a therapeutic misconception. Although it is appropriate for him to consider experimental agents promising, in early-phase oncology trials the genuine potential for subjects to experience meaningful benefit is often very small (Horstmann et al., 2005). In addition, because research participation is not the only way to care for patients with advanced cancer, Dr. F should also discuss and offer palliative care and other support services to his patients and their families. It is important that Dr. F recognizes his mixture of motivations for conducting early-phase research. If he decides to proceed with a research program, his desire to recruit and retain subjects to improve his income could bias the information he provides to them, and he and his research colleagues will have to work hard to avoid that (Klanica, 2005).
Dr. G has become aware of some of the pressures on the healthcare system that can blur the distinction between research and treatment and affect patients’ willingness to become research subjects. Patients who enroll in research because they do not have access to standard treatments may or may not have TM, but because their goal is to maximize their chances of benefit, they may not have the same view of their participation as Dr. G hopes for. They may resist randomization, or try to break a study blind, or drop out of the research at the end of the intervention period, avoiding follow-up data-gathering visits as not beneficial to them. If Dr. G continues to enroll these patients in her studies, she must work especially hard to educate them about research participation and to develop research partnerships with them. Once they have completed participation in her research, many of these patients will continue to lack access to treatment. Asking whether clinical research, and clinical researchers, can or should ameliorate inequities in access to healthcare opens up a significant global social policy question. There are other avenues Dr. G could pursue for improving healthcare access for patients both at home and abroad.