The ethics of some placebo-controlled randomized controlled trials

Abstract

It has been well-recognized that a randomized controlled trial (RCT) represents the best investigative format to establish medical evidence. In an RCT, essential variables can be controlled and bias minimized. When testing for the efficacy and safety of a medication, it is often compared to a blinded placebo. There is little argument to challenge the role of an RCT in this regard. When a subject is enrolled in an RCT investigating the therapeutic value of a medication for the treatment of a condition, it is often compared to a placebo, so as to best measure whatever benefit this new medication can have, independent of a possible placebo or contextual effect.¹ There are many examples of RCTs performed for the testing of a medication being measured for its efficacy in treating a condition when compared to a placebo. However, when there is a known effective medication that can be used with which that comparison can be made, it may be difficult to justify the use of a placebo in that instance, and there are many examples of this circumstance in all medical specialties, including as referenced in these gynecologic examples.^{2, 3} The putative medications that can be used for such a comparison (instead of placebo), have themselves been compared with placebo in prior Investigational New Drug (IND) trials performed for Federal Drug Administration (FDA) approval.

It is fair to recognize that the ethical principle of nonmaleficence (“first, do no harm”) may be violated by the physicians involved in that RCT.⁴ Simply put, if the comparison being made in this sort of a study was between a standard medication and a newer medication (possibly being better), then that would certainly be acceptable. Many studies are not performed in that way, however. As a physician-scientist, duty-bound to provide standard-of-care medications to patients and also to see innovative medical treatment be properly tested in order to advance evidence-based medical science, a placebo-controlled investigation fails in this regard, especially for conditions with currently available therapies of proven medical benefit. To not provide available therapy to a patient presenting with a properly documented condition may be an example of harm to that patient.

According to a principle of Logic (Transitive Property of Equality), if A = B, and B = C, then A = C.⁵ While that tenet may refer to an objective mathematical concept, it may still be useful to apply to the circumstance being described, in that if a FDA-approved medication was objectively found to be efficacious for treating a particular condition in a placebo-controlled investigation, it can itself be used (rather than a placebo) in future studies. Of course, this logical principle may only apply for those investigations showing a medication found to be either more effective or noninferior when compared with that same FDA-approved medication. However, there would not need to have a placebo-controlled trial to find a new medication to be efficacious. Nonetheless, RCTs continue to be conducted in which a patient having a documented disturbing condition (e.g., endometriosis) is provided a placebo by a physician researcher (∼50 % of the time), rather than using a known-to-be therapeutic medication. It should be recognized that this action (i.e., denial of a proper medication to use in that circumstance) represents a violation of the ethical principle of nonmaleficence. This occurs regularly in RCTs. The pursuit of logically derived medical evidence is desirable, but it should not be done while abandoning the ethical principle of nonmaleficence.

Perhaps it is time to re-examine the standard conduct of the placebo controlled RCT, with regard to the role of Medical Ethics and Logic.