Standard drug trials give real drugs to one group of people (treatment) and placebo to the other (control) to isolate the effect of the drug. Subjects are assigned to each group randomly and initially they don't know which group they are in. If it is a drug against a life-threatening disease and considerably less people die in the treatment group, we can say that the drug is effective. This also means that people in the control group died because they had been given placebo.

Maybe the best concrete example right now are the clinical trials of the COVID-19 vaccines. Thanks to the trials, we know that the currently available vaccines make fewer people die from COVID. And we only know that because we let people die by not giving them an effective vaccine. Their deaths were a direct consequence of a coin flip made by a researcher.

Few people seem to be bothered by this, while things like human cloning and organ markets are very controversial. Even simple thought experiments such as the trolley problem seem to be more divisive than our total acceptance of clinical trials. Hence, my question: is there an ethical framework where this acceptance can be explained?

  • 3
    Make sure to insist on the brand placebos. They work much better than the generics!
    – user4894
    Dec 31, 2021 at 20:34
  • "And we only know that because we let people die by not giving them an effective vaccine." This statement seems to be unproven and highly controversial, and reason enough to delete this question.
    – tkruse
    Jan 4, 2022 at 5:47
  • Any new drug must be tested on some kind of group that's much smaller than the number of people who could potentially be helped by the drug if it's effective, no? If there are 100,000 people who might be helped by some drug and you give it 1,000 as part of a drug trial with no placebo group, how is this morally different from a trial with 2,000 people where 1,000 get the drug and 1,000 get a placebo? Either way, 1,000 are getting the drug and 99,000 people who might benefit from the drug aren't getting it, at least not right away.
    – Hypnosifl
    Jan 4, 2022 at 19:56

4 Answers 4


Every sensible system of ethics accepts trade-offs between present value and future value

The use of randomised controlled trials (RCTs) in medical research involves a trade-off between gathering useful medical information for the future versus optimising an outcome for a present set of patients. If we encounter a situation where we have some a priori belief about the benefit/harm of a treatment then the optimal choice for a set of study participants might be that they all go in the treatment group, or they all go in the control group, which would defeat the RCT entirely and yield no useful information for the future. So, if we want to obtain useful medical information to improve outcomes for future patients, we must accept that we may need to make a sub-optimal choice for present patients in the trials. **

In the case of RCTs in medicine, trial participants must consent to be included in the trial, and they are made aware that they will be randomised into a control/treatment group, and they will not know which group they are in until after the study has been completed (and maybe not even then, depending on the protocol). So assuming that standard ethical protocols are not breached, study participants will have agreed in advance to participate, knowing that they might get a placebo instead of the treatment under consideration. There are also usually protocols to stop the study prematurely if one of the groups suffers high levels of harmful outcomes during the trial period.

So, ethically speaking, we have a situation where researchers ask a group of people to consent to a medical treatment, where the treatment might be (randomly) sub-optimal or even harmful for them, so as to advance the goal of gaining useful information for future medical decisions. That is, people agree to obtain sub-optimal treatment in the present in order to advance medical science for the benefit of people in the future. Any sensible theory of ethics would ---in principle--- permit people with informed consent to make this kind of trade-off. One could frame it in utilitarian terms (though that is not the only ethical theory that allows such trade-offs) by saying that the long-term future value of the medical knowledge gained from RCTs outweighs the losses due to sub-optimal treatment of patients in the RCTs.

Of course, while any sensible ethical theory will allow this in principle, the devil is really in the details: Was the patient given adequate information about the study to have informed consent? What was their alternative if they refused participation? Could they gain access to the treatment without consenting to the RCT? Etc. Those latter questions are important ethical questions that bear on the ethical validity of such studies.

** Your question confuses this issue somewhat by conflating the time-line; prior to the trials we have very little information on whether the treatment is effective, so it is often only in hindsight that we can make a good judgment on what would have been better for the study partipants, the treatment group, or the control group.


placebos must be used in clinical trials because the placebo effect is so strong in humans.

Presently, standard practice in a placebo-controlled clinical study of a drug being considered for a life-threatening disease is to abort the trial as soon as the initial data comes in that proves the efficacy of the agent being tested, and immediately administer the drug to all participants in the study.


Please do not spread fake information about development of drugs and vaccines.

"Few people seem to be bothered by this, ..." - The medical staff involved in drug trials are very concerned about this.

"And we only know that because we let people die by not giving them an effective vaccine." - Nobody who participates in such a study is left to die. All participants in such studies are being given as little doses of the virus as possible to reduce risk, and they are being surveilled constantly to immediately give them the best medical treatment possible if they show the slightest signs of sickness.

The ethical framework around this involves:

  • Not using a placebo study for serious conditions when there exists a proven treatment, comparing against that treatment else
  • Not doing such studies unless the condition is wide-spread and threatening many lives
  • informing all participants beforehand about all details, and all known and unknown risks
  • asking for the candidates written consent
  • selecting among suitable participants those with lowest risk of serious health issues
  • doing anything possible to reduce known and unknown risks, such as preparing for emergency measures, and monitoring candidates 24/7
  • compensating participants financially (without paying them so much to allow exploitation)

This does not eliminate all ethical concerns about such trials, but it reduces most of the concerns, enough to allow certain such trials to be done. But each trial involves a careful ethical consideration beforehand.

(Note that historically, unethical research has also be done of course, such as during World War 2)


As an example, the initial Pfizer Covid trial involved 20,000 people getting a vaccine, and 20,000 getting a placebo in a double-blind test. In the USA alone, about 300 million people got neither a vaccine nor a placebo. That's 15,000 times as many as people getting a placebo.

If you argue that the 20,000 shouldn't have been given a placebo, but a vaccine, then you also argue that the 300 million should have been given an untested vaccine - which is obviously a bad idea.

Getting a placebo means you won't get any positive effect from the real vaccine, but you also won't suffer any damage that the real vaccine might cause. And that is one of the two things that a trial should find out: Whether the vaccine helps, and whether it hurts. Without giving 20,000 people a placebo, we can't know whether the vaccine helps, and we can't know whether it causes damage.

In addition, all the participants were volunteers. They all knew that the might get the vaccine, and this might either help them or hurt them if we are unlucky, or they might get the placebo, and are no worse off than without participating in the trial.

Another problem with not using placebos is that people who know they are vaccinated might take more risks. So it would be entirely possible that there are more cases among vaccinated than in the normal population, or that there is less reduction in cases than hoped for. By giving placebos to half the participants, this is avoided. Since nobody knows in which group they are both groups will change their behaviour equally.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .