The Midas Problem with Randomized Control Trials

In public program evaluation, Randomized Control Trials (RCT) are exulted, commonly referred to as the “Gold Standard.” However, when applied to the field of Public Health, not all that is gold is good. Just ask King Midas, a King found in Greek mythology who was gifted with golden touch and then nearly starved himself and almost killed his daughter. The Gold Standard can lead to ethical concerns and troubling outcomes.

As the name suggests, RCTs involve randomly allocating participants into the treatment and control groups. The treatment group undergoes the program that is being tested and their outcomes are compared with the control group, who did not undergo the change. By randomly assigning groups we achieve statistical independence, a person’s decision to be treated verse not being treated has no effect on their outcome. RCTs allow for researchers to eliminate selection bias to see the treatment effects, the effects of the program that is being tested.

However, by making people’s decision to be treated verse not being treated not affect the outcome, researchers can take away individual’s ability to choose to be treated. A 2009 Bulletin of the World Health Organization highlights some of these ethical concerns, which they note are even more troubling when opting out of participating is difficult. This is particularly an issue in Group Randomized Control Trials, where documenting group consent is fraught with difficulty. The author’s outline steps that can be taken by researchers to combat these ethical concerns, such as not sacrificing individual consent for group consent.

Yet, focusing on individual consent has the potential to invalidate results. Ideally, participants should reflect the population the program is to target. However, if certain members of that targeted population are unwilling to participate, the effects of that population is unable to be fully analyzed. For instance, participants in a drug trial are more likely to have undergone all other known treatments. Therefore the treatment effects shown from a RCT may only be valid for the most extreme cases. External validity, or the generalizability of results, is therefore compromised.

Another concern with RCTs in health policy is that just as there are concerns with the effects of a new treatments, there also are concerns with denying treatments. To address this with drug trials, just as there are protocols for stopping drug trials as a result of negative outcomes, there are protocols for stopping trials as a result of positive results. For instance, in the case of vaccines, when a vaccine shows incredibly promising results, trails will be sped up or even cancelled in order to give the public access more quickly.

Despite RCTs being called the “Gold Standard” they are not without problems. In addition to difficulty conducting RCTs, ethically concerns are particularly concerning in health related RCTs. However, with this in mind, and proper protocols in place, RCTs do have their role in health policy program evaluation.