Doctors and RCTs

One ethical question that is raised in randomised controlled trials (RCTs) is whether or not the duties of a doctor qua scientist conflict with those of a doctor qua doctor. This is especially true in single-blind, placebo trials.

To explain, an RCT is a trial in which certain subjects (the "test" group) are given a medication, and other subjects (the "control" group) are given a placebo or the standard treatment. In a single-blind trial, the subjects do not know which group they are in, and in the double-blind trial, no one knows which group they are in.

The problem arises about half-way through an experiment. The doctor doctor has a moral obligation to provide the best possible care for his or her patient. However, usually aboult half-way through a single-blind trial, the doctor knows which group is doing better. Even in a double-blind trial, the researcher can usually extrapolate results from control groups in similar experiments. At this point, the doctor has conflicting duties. Qua scientist, the doctor should keep the experiment going to achieve the results. Qua doctor, the doctor should switch all his or her subject to the medication (if it is working) or off of the medication (if it is not). Since individual duties to patients usually trump general duties to science, this is a threat to RCTs generally. In fact, because of this ethical problem, some people have suggested abolishing RCTs.

There is an ethical solution to when an RCT is showing positive results. Both the doctor qua scientist and the doctor qua doctor may behave the same way. In a Phase II or Phase III drug trial, the drug is still considered experimental. Therefore, a doctor qua doctor may not perscribe the drug, even if he or she wanted to do so, and a doctor qua scientist would not want to provide the drug to control subjects for fear of disrupting the experiment. As such, the doctor faces no dilemma, as a doctor is not under an obligation to provide unapproved drugs to control subjects.

When the RCT is showing negative results, the problem is more complex. The patient is faring worse than he or she would on either a placebo or under the standard treatment, both of which are available to the patient. Here, the critics of RCTs are correct. One cannot continue to keep an unaware patient on a treatment that is less effective than another readily available treatment or than no treatment at all. It is not a defence to say that the patient consented at the beginning of the trial; now, the researchers know more than the patient, and this information must be revealed. All of the patients (both test and blind) in the study must be told that there is reason to believe that the experimental medication is less effective than the standard treatment, and asked if they still wish to continue with the study. If sufficient numbers remain, the experiment can continue. If not, then another methodology must be developed.

As such, RCTs do not necessarily need to be banned. In the case where medications are working, the doctor qua doctor would have no access to that medication anyway, and faces no conflict. However, in cases where the medication is showing bad results, the subjects need to be informed.