The concept of the standard of care has figured prominently in recent controversies over the use of placebos in the design of randomized controlled trials (RCTs) conducted in the United States (1) and developing countries (2–4) as well as control group selection in critical care RCTs conducted in the United States (5, 6). Missing from these debates in the medical literature has been a critical analysis of the ethical significance of the medical standard of care for the design of clinical research. In this article we will analyze the ethical relevance of the standard of care in RCTs with particular attention to the design of critical care trials. The dispute regarding international RCTs, funded by sponsors in developed countries and conducted in developing countries, raises complex issues of justice, which we do not address (7).
RCTs differ fundamentally from medical care in their purpose, methods, and justification of risks (8). The RCT is primarily an experiment designed to answer a scientific question regarding treatment efficacy in groups of patients, with the aim of developing knowledge that can contribute to improving medical care. Although participants usually receive experimental or standard treatment, the RCT is not a form of personal therapy selected for particular patients. Procedures foreign to medical care are routine in RCTs—randomization, masked assignment of treatment, protocol-defined restrictions on treatment, placebo controls, and clinically unnecessary procedures and tests implemented to measure study outcomes. In RCTs, the risks of experimental interventions and standard treatments evaluated are justified primarily by the prospect of benefits to participants, and the risks of research procedures that are administered only to produce valid data are justified by the anticipated value of the knowledge to be gained from the research. In contrast, the risks of medical interventions administered to patients in clinical practice can be justified only by the potential medical benefits to them. It follows from these differences that the ethical principles that govern clinical research are not the same as those that primarily apply to medical care.
Ethical objections to particular RCTs because their design deviates from the standard of care therefore beg a fundamental question in the ethics of clinical research. If clinical trials are ethically distinct from medical care, why should the use, for example, of a control group that deviates from the medical standard of care count as an ethical objection to trial design? Some commentators have argued that placebo controls may be justified despite the existence of proven effective treatment, thus contravening the medical standard of care, provided that use of placebo is methodologically indicated and short-term assignment to placebo does not pose undue risks of serious harm (9, 10). Supporting this view are recent revisions to leading codes of research ethics, including a “Note of Clarification” to the Declaration of Helsinki (11) and the International Ethical Guidelines for Biomedical Research Involving Human Subjects of the Council for International Organizations of the Medical Sciences (12). Examples of justifiable trials with designs contrary to the standard of care include placebo-controlled trials evaluating new treatments for allergic rhinitis, heart burn, minor headaches, mild hypertension, and, more controversially, anxiety, depression, migraine, and chronic stable angina (10, 13, 14).
If deviation from the standard of care in clinical trials is considered to be unethical per se, then it would be unethical to evaluate novel experimental treatments in RCTs. Such experimental treatments, which have not been validated as safe and effective, depart from the medical standard of care. In view of this consideration, the former language of the Declaration of Helsinki concerning ethical trial design has been criticized as obviously and necessarily mistaken: “In every medical study, every patient—including those of a control group, if any—should be assured of the best proven diagnostic and therapeutic method” (15–17). Randomizing patients to an experimental therapy would violate this principle because it would not qualify as a “best proven” treatment.
Consequently, deviation from the standard of care in the design of RCTs is not, in itself, unethical. However, this does not mean that the concept of the standard of care is irrelevant to the ethics of clinical trials. Critical assessment of the relevance of the standard of care in research ethics requires its explication within the context of medical care.
Although basic to medical ethics, the concept of the standard of care is not internal to medical practice; rather, it derives from the law relating to medical malpractice. The traditional legal understanding of the standard of care to which physicians are held accountable refers to the typical or customary practice of physicians in the professional community, as evidenced by the testimony of expert witnesses (18). Customary medical practice has been treated as legally normative. Academic medical opinion, however, has traditionally distinguished between what doctors typically do and what constitutes the ethically normative standard of care. For example, many physicians do not treat patients with β-blockers after myocardial infarction, despite rigorous evidence showing that such treatment can reduce future myocardial infarctions (19). Accordingly, the medical “standard of care” is used in two different senses. In a descriptive sense, it refers to prevailing or routine practice patterns within a given medical community. In a normative sense, it refers to what is regarded as “best” practice in view of current scientific knowledge and expert opinion.
Recently, the development of evidence-based medicine has led to the view that the results of well-designed RCTs should govern the standard of care (20). When evidence from RCTs is lacking with respect to a given area of medical practice, some might argue that there is no standard of care (21). A more accurate view would be that there is no validated standard of care. In this case, routine practice reflects current understanding of pathophysiology, clinical experience, observational studies, and expert judgment but has not been validated by “the gold standard” of the RCT.
Speaking broadly, two basic moral considerations govern the ethics of designing RCTs: (1) pursuing clinically valuable knowledge by developing scientifically valid data to answer a question regarding treatment efficacy for a given condition and (2) protecting research subjects from harm or exploitation. The standard of medical care is relevant to both. This can be illustrated by examining the ethical controversy surrounding a recent critical care RCT evaluating two contrasting methods of mechanical ventilation for patients with acute lung injury (ALI) and acute respiratory distress syndrome (ARDS).
In 1996 the ARDS Network launched a clinical trial sponsored by the NHLBI randomizing patients with ALI and ARDS to either a low Vt intervention (6 ml/kg of predicted ideal body weight) or a higher Vt intervention (12 ml/kg of predicted ideal body weight) (5). The two study arms represented contrasting strategies for providing ventilatory support for patients with these conditions. Subsequent to publication in 2000 of the trial results demonstrating that the low Vt produced superior outcomes (22), the trial design was criticized by some critical care experts for evaluating extremes of practice and for failing to include a control group representing the prevailing standard of care for ventilation management of ALI and ARDS (23–26). It was argued that lacking such a control group, the trial was not designed optimally to improve the standard of care and that it failed to provide adequate protection for trial participants. The ethical criticism prompted vigorous objection by leaders of the ARDS Network, review of the study design by two expert panels, a temporary suspension by the Office of Human Research Protections of a comparable ongoing RCT comparing contrasting strategies for fluid management of patients with ALI and ARDS, and a series of articles in the medical literature describing and taking sides in this controversy (5, 21).
Three contextual factors are especially relevant to ethical assessment of study design for the ARDS Network clinical trials and comparable critical care RCTs: (1) the medical condition under investigation is rapidly lethal and the primary study outcome is mortality, (2) current routine practice is characterized by a variable quantitative pattern of treatment selection—in this case, ventilation management along a range of Vt, and (3) this routine practice has not been validated by evidence from RCTs. The defenders of the study design in the Vt trial emphasized the third factor (27). They argued that in the absence of a rigorously validated standard of care for ventilation of patients with ALI and ARDS, it was clinically valuable and ethically legitimate to evaluate two contrasting strategies for Vt management within the range of standard practice. Clinical equipoise—a state of uncertainty in the expert community regarding therapeutic merit (28)—existed between the two study arms and between either of the ventilation strategies and the unvalidated standard of care (29). Accordingly, patients were not randomized to treatment known to be inferior, and clinical equipoise did not dictate the inclusion of a standard of care control group reflecting the routine practice of intensivists. Although most, but not all (30), commentators hold that some form of equipoise is ethically necessary, equipoise is not sufficient to immunize trial design from legitimate ethical criticism.
Despite the fact that the Vt trial was not designed to determine whether either of the contrasting treatment strategies was better than the prevailing standard of care (31), the question whether patients have better outcomes with a low or higher Vt treatment strategy was certainly clinically relevant. However, it is hard to dispute that the trial would have had greater potential clinical value if it had included a control group representing the standard of care—the normal range of Vts for patients with ALI and ARDS selected by expert clinicians. Although this standard of care had not been validated by evidence from RCTs, it remained important to determine whether a particular strategy of ventilation management is better than current practice of intensivists. In the absence of a standard of care control group, it is difficult to conclude from the results of this ventilation trial alone that the lower Vt should be used in patients with ALI and ARDS.
Including a third standard of care control group would have required enrolling more subjects, adding to the cost and perhaps reducing the feasibility of completing the study. With respect to the prospective clinical value of the trial, the key ethical issue is whether the specific question that it was designed to answer had adequate value to justify the risks posed to research subjects. The absence of a standard of care control group is relevant, but not decisive, for this issue.
The claim that a standard of care control group was needed to adequately protect research subjects is more complex and difficult to evaluate. The impetus for this claim is the theoretical possibility that patients randomized to both the contrasting Vt treatment strategies could have worse mortality outcomes than comparable patients receiving standard ventilation management (23). For example, suppose hypothetically that the mortality rates for the low and higher Vt treatment strategies would be 40 and 50%, respectively. It is possible that trial-eligible patients randomized to receiving treatment by intensivists according to the current standard of care would have a lower mortality rate of 35%. Adding a standard of care control group as a third arm obviously would not obviate this possibility; however, it would permit a data and safety monitoring committee to detect this worst-case scenario and stop the trial early in light of emerging data, indicating a statistically significant difference in mortality between one or both study arms and the standard of care control group. The standard of care control group thus could promote the safety and reduce the risks of research subjects in the event that the worst-case scenario materialized.
How probable is this worst-case scenario? Certainly, no precise answer would be a possible antecedent to conducting the trial. For if the mortality rates of the contrasting ventilation strategies and the prevailing standard of care were known, then there would have been no point in conducting the trial. Although the answer to this question was not precisely knowable, the question is, and was, ethically relevant. Design for trials that exhibit the three contextual factors outlined previously should address this worst-case scenario. Any available or easily collectible data and expert opinion that would bear on judging its likelihood merit careful consideration and assessment in review and approval of study design.
The issue of whether a standard of care control group should have been included in this critical care trial is relevant to the major ethical requirement of minimizing risks to subjects. In an editorial concerning the controversy over this trial, Drazen (29) contended that in view of the study results, “…the mortality in the group receiving the larger tidal volume (the group that did less well) was similar to the mortality that would be expected on the basis of the known severity of their illness. This finding vitiates the argument that patients in the completed trial were put at increased risk.” This conclusion, however, is a non sequitur. It is not known that the mortality of those receiving the larger Vt was no greater than what comparable patients would have experienced if they had received standard ventilation management. Moreover, ethical assessment of trial design must be prospective. As Beecher (32) noted in 1966, “An experiment is ethical or not at its inception. It does not become ethical post hoc.”
Whether and when standard of care control groups are ethically required in critical care trials evaluating contrasting treatment strategies remains a matter of judgment about which reasonable and informed people may differ. In any case, it is important to recognize that if it is judged that the Vt trial should have included a third arm representing the prevailing standard of care, the trial would still have deviated from the standard of care by enrolling patients in the low and higher Vt arms. Because the standard of care for ventilatory management had not been validated, it was ethically justifiable to randomize patients to treatment strategies that they would not have received if they had been managed according to standard practice.
Clinical trials differ from medical care in ethically significant ways. To answer clinically valuable questions, it is often necessary for clinical trials to randomize subjects to interventions that deviate from the standard of care in medical practice. Nevertheless, control groups that represent the standard of care are often required to promote the clinical value of randomized trials and to protect research subjects. In the case of critical care trials, the question whether RCTs should include a control group representing the unvalidated standard of care raises complex scientific and ethical issues that call for careful assessment and judgment (6).
The authors thank Ezekiel Emanuel for helpful comments on a previous draft.
|1.||Rothman KJ, Michels B. The continuing unethical use of placebo controls. N Engl J Med 1994;331:394–398.|
|2.||Lurie P, Wolfe SM. Unethical trials of interventions to reduce perinatal transmission of the human immunodeficiency virus in developing countries. N Engl J Med 1997;337:853–856.|
|3.||Varmus H, Satcher D. Ethical complexities of conducting research in developing countries. N Engl J Med 1997;337:1003–1005.|
|4.||Angell M. The ethics of clinical research in the third world. N Engl J Med 1997;337:847–849.|
|5.||Steinbrook R. How best to ventilate? Trial design and patient safety in studies of the acute respiratory distress syndrome. N Engl J Med 2003;348:1393–1401.|
|6.||Silverman HJ, Miller FG. Control group selection in critical care randomized controlled trials evaluating interventional strategies: an ethical assessment. Crit Care Med (In press)|
|7.||Crouch RA, Arras JD. AZT trials and tribulations. Hastings Cent Rep 1998;28:26–34.|
|8.||Miller FG, Rosenstein DL. The therapeutic orientation to clinical trials. N Engl J Med 2003;348:1383–1386.|
|9.||Temple R, Ellenberg SE. Placebo-controlled trials and active-control trials in the evaluation of new treatments. Part 1: ethical and scientific issues. Ann Intern Med 2000;133:455–463.|
|10.||Emanuel EJ, Miller FG. The ethics of placebo-controlled trials: a middle ground. N Engl J Med 2001;345:915–919.|
|11.||World Medical Association: Declaration of Helsinki: Note of Clarification on Placebo-Controlled Trials [Accessed November 19, 2003]. Available at:|
|12.||Council for International Organizations of Medical Sciences. International ethical guidelines for biomedical research involving human subjects. Geneva: Council for International Organizations of Medical Sciences; 2002.|
|13.||Al-Khatib SM, Califf RM, Hasselblad V, Alexander JH, McCrory DC, Sugarman J. Placebo-controls in short-term clinical trials of hypertension. Science 2001;292:2013–2015.|
|14.||Miller FG. Placebo-controlled trials in psychiatric research: an ethical perspective. Biol Psychiatry 2000;47:707–716.|
|15.||World Medical Association. Declaration of Helsinki. JAMA 1997;277:925–926.|
|16.||Lasagna L. The Helsinki Declaration: timeless guide or irrelevant anachronism? J Clin Psychopharmacol 1995;15:96–98.|
|17.||Levine RJ. The need to revise the Declaration of Helsinki. N Engl J Med 1999;341:531–534.|
|18.||Mello MM. Of swords and shields: the role of clinical practice guidelines in medical malpractice litigation. Univ Penn Law Rev 2001;149:645–710.|
|19.||Simpson E, Beck C, Richard H, Eisenberg MJ, Pilote L. Drug prescriptions after acute myocardial infarction: dosage, compliance, and persistence. Am Heart J 2003;145:438–444.|
|20.||Sackett DL, Rosenberg WMC. The need for evidence-based medicine. J R Soc Med 1995;88:620–624.|
|21.||Steinbrook R. Trial design and patient safety: the debate continues. N Engl J Med 2003;349:629–630.|
|22.||The Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000;342:1301–1308.|
|23.||Eichacker PQ, Gerstenberger EP, Banks SM, Cui X, Natanson C. Meta-analysis of acute lung injury and acute respiratory distress syndrome trials testing low tidal volumes. Am J Respir Crit Care Med 2002;166:1510–1514.|
|24.||Oba Y, Salzman GA. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury. N Engl J Med 2000;343:813.|
|25.||Parshuram C, Kavanagh B. Meta-analysis of tidal volumes in ARDS. Am J Respir Crit Care Med 2003;167:798.|
|26.||Eichacker P, Banks S, Natanson C. Meta-analysis of tidal volumes in ARDS. Am J Respir Crit Care Med 2003;167:798–800.|
|27.||Brower RG, Matthay M, Schoenfeld D. Meta-analysis of acute lung injury and acute respiratory distress syndrome trials. Am J Respir Crit Care Med 2002;166:1515–1517.|
|28.||Freedman B. Equipoise and the ethics of clinical research. N Engl J Med 1987;317:141–145.|
|29.||Drazen JM. Controlling research trials. N Engl J Med 2003;348:1377–1380.|
|30.||Miller FG, Brody H. A critique of clinical equipoise: therapeutic misconception in the ethics of clinical trials. Hastings Cent Rep 2003;33:19–28.|
|31.||Brower RG, Rubenfeld G, Thompson BT. Meta-analysis of tidal volumes in ARDS. Am J Respir Crit Care Med 2003;168:255–256.|
|32.||Beecher HK. Ethics and clinical research. N Engl J Med 1966;274:1354–1360.|