CLINICAL TRIALS

Clinical trials are systematic investigations on human subjects testing the safety and efficacy of novel medical interventions, including drug, surgical, or behavioral treatments. Conventionally clinical trials are divided into four types or phases. In a phase I clinical trial, typically involving tens of subjects, a novel procedure is tested for the first time in human beings and data is collected on safety. In a phase II trial, which may involve hundreds of patients, evidence is sought that a novel intervention has a therapeutic effect on the disease of interest. In a phase III clinical trial, often involving thousands of patients, the novel intervention is compared to a standard intervention or placebo. In a phase IV trial, called a post-marketing study, information is collected on the long-term safety and efficacy of the intervention from patients receiving the intervention in clinical practice and measured against a control treatment. The rigorous evaluation of novel medical interventions in clinical trials is a foundation of evidence-based medicine.

Historical Development

The randomized clinical trial is one of the most important advances in medicine in the twentieth century. Prior to its development, treatments were adopted on the basis of the publication of a series of cases in which their use had proved helpful. Due to numerous sources of potential bias, including variation in expertise from one clinician to the next, and the selection of patients more likely to recover for inclusion in the study, case series often led to misleading results. Clinicians were faced with numerous treatments from which to choose, and little evidentiary basis upon which to ground a choice. For example, Richard Doll, a well-known British clinical trialist, described how at the start of his research into the treatment of peptic ulcer in 1948, he was able to list purported treatments beginning with each letter of the alphabet.

In clinical trials until mid-twentieth century, two treatments for comparison were allocated to alternating patients. This method was flawed by the fact that physicians could anticipate the treatment assignment, and thereby select which treatment a particular patient would receive by changing the patient's position in the queue. It was not until mid-century that R.A. Fisher's allocation strategy using random numbers, developed in 1926 for agricultural experiments, was used in clinical trials. Allocation using random numbers countered bias in selection and provided statisticians with an estimate of random error, a key component of modern statistical analysis. The first clinical trial to randomly allocate treatments to patients using random numbers was the United Kingdom Medical Research Council (MRC) whooping cough immunization trial initiated in 1946. The better-known MRC streptomycin trial in tuberculosis started a few months later, but published its results before the whooping cough trial in 1948.

Since the mid-twentieth century, the clinical trial has undergone a dramatic increase in use for the evaluation of the safety and efficacy of novel medical interventions. A variety of social and political factors supported this trend. The period following the Second World War witnessed an unprecedented public investment in health research. In 1945, the United States National Institutes of Health budgetary appropriation was $700,000; in 1970 its appropriation was $1.5 billion. Drugs regulation underwent significant changes in this period as well. From 1938 to 1962, the U.S. Food and Drug Administration (FDA) was only empowered to require that new drugs be tested for safety. Following on the heels of the thalidomide tragedy, in which hundreds of infants were born with congenital malformations after exposure to thalidomide in utero, legislative reform dramatically increased the FDA's power. The 1962 Kefauver-Harris Act expanded the FDA's mandate to test new drugs for both safety and efficacy.

The testing of new drugs for safety and efficacy in clinical trials occurs in an increasingly international environment. Cooperation among drugs regulators and manufacturers seeks to standardize the conduct of clinical trials and their review by drugs regulators. The International Conference on Harmonization Good Clinical Practice guidelines are a key instantiation of this effort. The protection of human subjects in research is similarly seen as a matter of global concern. Perhaps the most influential ethics document in the international forum is the World Medical Association's Declaration of Helsinki. The declaration requires that clinical trials be reviewed by appropriately constituted research ethics committees; research be free of misconduct; the consent of human subjects be obtained; study participation pose a favorable balance of benefits to harms; and subjects be selected equitably.

Ethical Issues

Some of the most important ethical challenges of clinical trials stem from conflicting duties of the physician-researcher. Physicians have fiduciary obligations to patients, including a duty to provide competent personal care. Researchers, by contrast, have obligations to science and society, including duties to provide treatment as prescribed in the trial protocol, ensure that patients comply with treatment, and encourage them to stay in the study. Given that these duties may conflict, the central moral question of the clinical trial is: When may physicians legitimately offer patients enrollment in a clinical trial? While a variety of answers have been provided to this question, the most widely accepted is that of clinical equipoise. According to clinical equipoise, physicians may legitimately offer patients enrollment in a clinical trial only if the medical interventions within the study are consistent with competent medical care. More formally, it requires that at the start of the study there exists a state of honest, professional disagreement as to the preferred treatment. The consequences of clinical equipoise for the design of clinical trials are far reaching.

Two issues in respect to the design of clinical trials have dominated research ethics literature since the 1990s. The first is the proper role of placebo controls in the new drug approval process in developed countries. Drug regulatory agencies in developed countries, such as the FDA, have long required that new drugs prove superior to placebos in at least two clinical trials before licensure. The practice in the United States is rooted in legislation that requires the FDA to ensure new drugs are efficacious, that is, that they have some effect in treating the condition of interest, but generally restricts its ability to demand evidence of comparative effectiveness. According to clinical equipoise, placebo-controlled clinical trials are unproblematic when there is no proven treatment for the condition of interest.

Criticism has focused on the use of placebo controls in clinical trials testing novel interventions for treatable medical conditions, such as severe depression and schizophrenia. The use of placebos in these cases is impermissible, because no competent physician would fail to offer a patient treatment and, accordingly, clinical equipoise is violated.

The 2002 revision of the Declaration of Helsinki sets aside this fundamental moral requirement, and for the first time permits the use of a placebo control when "compelling and scientifically sound methodological reasons" exist. This change seems to violate a core provision of the declaration requiring that "[i]n medical research on human subjects, considerations related to the well-being of the human subject should take precedence over the interests of science and society." Whether there are in fact scientifically sound methodological reasons to prefer a placebo control over a standard treatment control remains an open question.

The second clinical trial design issue to receive considerable attention in the literature is the choice of control treatment in clinical trials of new and affordable treatments for developing countries. Disagreement was originally sparked by clinical trials testing the efficacy of short-course zidovudine against placebos for the prevention of transmission of HIV from mother to child. Critics of the clinical trials pointed to the existence of an effective prevention regimen called ACTG 076 used in developed countries. Denying subjects in the clinical trials conducted in developing countries access to this prevention regimen, they claimed, constitutes an ethical double standard between developed and developing countries.

Proponents of the clinical trials countered that the ACTG 076 regimen is not suited to administration in many developing countries and the cost is prohibitive. Changes in international regulation have tended to entrench rather than resolve the dispute. The Declaration of Helsinki proscribes placebo controlled trials in developing countries when effective treatment exists in developed countries saying that "[t]he benefits, risks, burdens and effectiveness of a new method should be tested against those of the best current prophylactic, diagnostic, and therapeutic methods." Yet the "International Ethical Guidelines for Biomedical Research Involving Human Subjects" permit placebo controlled trials under these circumstances provided the clinical trial is "responsive to the health needs of the population from which the research subjects are recruited and there [is] assurance that, if it proves to be safe and effective, it will be made reasonably available to that population."

OPEN QUESTIONS. The interface between the ethics and science of clinical trials is replete with challenging questions yet to be addressed adequately. What ought the role be for adaptive designs, for instance, clinical trials in which the probability of being assigned to one treatment or another is dynamic in an attempt to minimize the number of subjects who receive the treatment that turns out to be inferior? Can alternative medical treatments be evaluated rigorously in clinical trials? Alternative practitioners may claim that alternative treatments cannot be removed from a holistic treatment context, a substantial obstacle to the rigorous assessment of the treatment's efficacy. How will pharmacogenetic testing impact the conduct of clinical trials? Proponents of pharmacogenetics suggest that identification by genetic testing of those likely to respond to treatments and those likely to suffer adverse events would increase the efficiency and safety of clinical trials. Critics wonder if the gains from such testing will be as large as promised and what impact it will have on the generalizability of clinical trial results.

While ethical issues in the design of clinical trials are the subject of ongoing scholarship, ethical aspects of the conduct and reporting of clinical trials are relatively ignored. As clinical trials accumulate data on outcomes, disparities may emerge between the treatments in the clinical trial raising questions as to whether the trial ought to be stopped early. It is generally agreed that when clinical trials use outcome measures of mortality or serious morbidity an independent data monitoring committee should be established to periodically review accumulating data. A satisfactory moral framework to guide the decisions of data monitoring committees has yet to be developed.

Ethical issues in the reporting of clinical trial results also deserve attention. If researchers fail to report the results of a negative clinical trial, subjects in the trial were exposed to risk for naught and the problem of publication bias is compounded. While this seems problematic intuitively, a moral basis for an obligation to publicize clinical trial results has yet to be articulated.

CHARLES WEIJER

SEE ALSO Complementary and Alternative Medicine;Drugs;Human Subjects Research.

BIBLIOGRAPHY

Angell, Marcia. (1997). "The Ethics of Clinical Research in the Third World." New England Journal of Medicine 337: 847–849. A stinging critique of placebo controlled clinical trials testing the efficacy of short course zidovudine in the prevention of perinatal transmission of HIV in developing countries.

Crouch, Robert, and John Arras. (1998). "AZT Trials and Tribulations." Hastings Center Report 28(6): 26–34. A sober response to Marcia Angell's critique.

Doll, Richard. (1998). "Controlled trials: The 1948 Watershed." British Medical Journal 317: 1217-1220. A historical review of changes to clinical trials at mid-century by one of the central figures in its development.

Emanuel, Ezekiel; Robert Crouch; John Arras; et al., eds. (2003). Ethical and Regulatory Aspects of Clinical Research: Readings and Commentary. Baltimore, MD: Johns Hopkins University Press. The definitive collection of readings in the ethics of human experimentation.

Miller, Paul, and Charles Weijer. (2003). "Rehabilitating Equipoise." Kennedy Institute of Ethics Journal 13: 93–118. A philosophical piece examining the conflicting duties of physician and researcher, and how these conflicts might be resolved.

Piantadosi, Steve. (1997). Clinical Trials: A Methodologic Perspective. New York: John Wiley & Sons. An authoritative methodological resource for the design of clinical trials.

INTERNET RESOURCES

Council for International Organizations of Medical Sciences. "International Ethical Guidelines for Biomedical Research Involving Human Subjects." Available from http://www.cioms.ch/frame_guidelines_nov_2002.htm. Guidelines for the conduct of research in resource-poor nations.

National Placebo Working Committee. "National Placebo Initiative: Draft Report of the National Placebo Working Committee." Canadian Institutes of Health Research. Available from http://www.cihr-irsc.gc.ca/e/services/pdf_19320.htm. A thorough document setting out the scientific, ethical, legal, and policy case for and against the widespread use of placebo controls in clinical trials.

World Medical Association. "Declaration of Helsinki." Available from http://www.wma.net/e/policy/b3.htm. The most influential international statement of moral standards for the conduct of clinical research.