Benjamin Freedman’s account of clinical equipoise, widely accepted as a standard principle for research, intends to make meaningful ethical appraisal of the involvement of human subjects in clinical trials on a case-by-case basis [1]. As a principle, however, equipoise relies heavily on epistemic considerations: the beliefs held by and knowledge, and evidence available to the researchers, as well as consensus among the biomedical community. In order to maintain the balance Freedman argues is necessary for an ethical determination of whether it is appropriate to conduct a particular clinical trial using human subjects.

According to this principle, two conditions should be present to justify the involvement of human subjects [2]. First, enough uncertainty about the most efficient way to treat a given problem should exist in order to substantiate the trial itself. Second, Freedman argues, for equipoise to be sustained, whether the trial is “randomized, placebo-controlled, or blinded,” the researchers must believe that there is no evidence to show the superiority of one treatment over the other. Interestingly enough, equipoise is rarely addressed as an epistemological problem, and, perhaps because the ethical implications of equipoise are seen as its foremost priority, epistemological claims about equipoise are given liberty not offered to other claims about belief, evidence, or knowledge.

Freedman argues that two treatments being tested must be in equipoise in order for the trial to be said to be in the best interest of the patient. In that sense, equipoise takes in to account all of the evidence supporting the efficacy of treatments A and B, assuming these are involved in the trial. Clinical equipoise treats the welfare of the human subjects involved in the trial as its guiding principle. The constraint on equipoise’s reach seems to be at odds with its view toward human welfare. Why limit equipoise to only the trial at hand?

Suppose that equipoise is granted for the time being between treatment A and B, but not between A and C, which is not involved in the study but has been proven to be more effective, but also, say, more costly. Should clinicians be asked to forget that they know that treatment C is more effective for the sake of comparing treatments A and B? And in doing so, can we really argue that the clinicians’ beliefs are, in fact, justified? Further, can we really assume the equipoise is possible, as Freedman argued, in all clinical trials? Particularly in placebo-trials, it seems hard to believe that clinicians can hold justified beliefs that treatments A and B are in equipoise, though it isn’t clear that on Freedman’s account such beliefs need be justified [3].

Ethicists that argue that such constraints are appropriate for the sake of justifying a clinical trial are by implication granting the prioritization of research over the subjects, and, in that case why appeal to equipoise at all? Hellman [4] describes the ethical tension associated with equipoise as being “a conflict between the interest of patients who are sick today, and the interests of the group of people who will become sick in the future” [4]. If this is the case, however, it seems that informed consent for participation in a trial, rather than equipoise, would be sufficient for ethical guidelines. In a sense, simply telling patients that clinicians were unsure of the relative benefits and harm that could be brought about by the trial would have the same effect as trying to weigh the clinician’s beliefs about the various treatments.

Generally speaking, this reliance on the beliefs of researchers raises a number of epistemological questions about who ought to share beliefs that the treatments are in equipoise in order to consider the clinical train ethical. Appeals to the “medical community” do not determine how many clinicians must share particular beliefs in order for a trial to be considered in equipoise. With regard to the beliefholders, Freedman’s formulation of equipoise also fails to address the problem of dissent, contrary hypotheses, and views. Clinical equipoise is also vague with respect to the justification and evidence necessary for beliefs that maintain or disrupt equipoise. Initiatives like open access, which aim to provide much less restrictive access to scientific information, have made important strides toward making important clinical information available. Still, questions about how to effectively use clinical information remain unanswered by Freedman and in the subsequent literature on equipoise.

Freedman attempts to answer these questions by contrasting clinical equipoise to Fried’s [5] earlier conception of individual equipoise, but it is not clear that the consensus clause added by Freedman makes this reliance on belief any less problematic [5]. Further, questions about what criteria must be met for distinctions to be made between evidence, belief, knowledge, opinion, and so on, arise when we rely on clinicians’ beliefs to inform research guidelines [6].

The line between hypothesis and evidence seems to blur when judgments about clinical equipoise are made. Researchers cannot possibly conduct a clinical trial without any beliefs about the treatments that human subjects are to receive. More likely, the beliefs are considered hypotheses but not evidentiary, or, are dismissed as being sufficient for individual equipoise but not for clinical equipoise [5]. Freedman [1] was successful in making a distinction between these sorts of equipoise, but this seems to serve only to force clinicians to operate under a veil of ignorance and ignore their own individual beliefs about the trial, or to so dilute clinical equipoise that it operates as a theoretical yet impractical sort of consensus. Again, the vagueness over belief in Freedman’s [1] account leaves equipoise in a seemingly fragile position.