With potential COVID-19 vaccine trials beginning around the globe, one vaccine approach is to induce the creation of antibodies in the participants.

The first phase of the trial will check whether the vaccine candidate induces the creation of antibodies to kill the coronavirus that causes COVID-19, and whether it provokes any serious adverse side effects.

Either the potential vaccine or a placebo will be given to participants.

Two doses of COVAX-19, or a placebo, will be injected into 40 healthy people between the ages of 18 and 65

As the aim of the trial is to check for the creation of antibodies, what is the purpose of having a placebo group? Would it be a reasonable assumption to conclude that it's biologically impossible for the antibody to be created in the participant on placebo?

A standard example of the placebo effect would be in pain management - wherein participants will be given a painkiller or a sugar pill. There do exist known mechanisms for these effects, however in the case of creating an antibody - how can that be possible?

  • 4
    "and whether it provokes any serious adverse side effects" -- how do you know if there are side-effects with out a control group??
    – Hot Licks
    Jul 6, 2020 at 22:01
  • 1
    @Hot Licks: I don't think the question is whether you need a control group, I think the question is whether you need to administer a placebo injection to the control group. To which the answer is "yes", but the reason is not obvious to the questioner. Jul 7, 2020 at 1:29
  • 4
    @SteveJessop - How would you do a double-blind study otherwise?
    – Hot Licks
    Jul 7, 2020 at 1:30
  • 1
    @HotLicks: I'm not sure what the expected level of ability is on this site, whether it should be self-evident to questioners that double-blind trials are superior to single-blind or non-blinded trials. If not then your comment should be an answer (indeed, it is BlueRaja's answer), if so then I guess the question is off-topic? Jul 7, 2020 at 1:32
  • In one of the studies in question, the placebo group was not given simply saline, but a meningitis vaccine, which would be capable of inducing 'sore arm'.
    – Neil_UK
    Jul 9, 2020 at 8:43

4 Answers 4


People could develop antibodies from natural exposure to the virus. The vaccine is trying to cause antibodies to exist in more people (and/or more strongly) than would express them naturally, therefore a good comparison group is a sample taken randomly in the same way as those getting the vaccine: a placebo group.

At the same time, these trials tend to assess safety outcomes; again, to assess safety you want to know that effects are no worse than those in some comparison population. Comparing to placebo is typically a gold standard for this comparison.

You're right that a placebo wouldn't be as necessary if we could prevent random exposure, but in the middle of a pandemic that isn't feasible.

  • 20
    In addition, people may encouter side effects, this is known to happen even for placebo treatments (see nocebo). So in general, the placebo arm is also needed for this aspect. Jul 6, 2020 at 15:15
  • 4
    @cbeleitesunhappywithSX Indeed, that's part of the "assess safety outcomes"; you want to know side effects are no worse than placebo, not that they are no worse than "no treatment".
    – Bryan Krause
    Jul 6, 2020 at 16:25
  • 2
    @SteveJessop Same as with any other drug or treatment - you don't want patients (or study staff) to know which group they are in. Could affect loss to follow up or compliance with the study overall if people know they are in the placebo group, people could report side effects differently if they know they might have gotten the treatment (whereas if you give no injection, those people know for sure that any 'side effects' they experience aren't actual side effects). Allows you to blind the study personnel. Wikipedia mentions a bunch of these reasons: en.wikipedia.org/wiki/Placebo
    – Bryan Krause
    Jul 7, 2020 at 1:40
  • Indeed, there's the issue of exposure associated with the trial itself, as it's possible that simply visiting a doctor's office to get an injection could be associated with some level of exposure. Without the control arm, you'd never be able to tell if vaccinated individuals developed more antibodies just because they were naturally exposed during their trip to the doctor, or if they indeed developed more antibodies than other people who visited a doctor. Dec 8, 2020 at 20:28

People in the control group will behave differently if they know they're in the control group.

For example, it's not unreasonable to expect people will do less social distancing once they're vaccinated. This will increase their chances of being exposed to the virus.

  • I hope they told both groups that there is a control group, so they don't both get less cautious... Jul 7, 2020 at 14:19
  • 6
    @user253751: Clinical Trials will have extensive consent documents which outline, in non-specialist terms, how the trial is to be run, who will have access to the data, etc. For a blinded control trial it will say "you will be given either the study drug or a placebo; neither you nor your doctors will know which you have received. You will be given a card with a phone number which can be called if, in a medical emergency, this information needs to be known". If that last bit comes to pass, your data will be de-blinded and shared with your doctors.
    – CCTO
    Jul 7, 2020 at 17:40

Good question. A phase I trial normally has small numbers, is purely sized to test safety, and it would have no placebo arm. With side effects and antibodies are end points, and a placebo group, this would more often be called a combined phase I/II trial.

The purpose of the placebo should be to compare the proportion of side effects and antibodies in the vaccine arm and those in the placebo arm. However with only 20 people in each arm, this seems a bit, ehm, optimistic.

Imagine if 4 of the vaccinated people report a sore arm, 2 a headache and sore throat, and 2 a fever, that might be unacceptable. But headache and sore throat are common, and as well as fever during winter-time, when colds and influenza are running around. So I expect the reason for the placebo group is to be able to partially compensate for influenza-like illnesses in the community, so they could suggest that some of the reported vaccine side effects aren't from the vaccine. (Note, you can't just subtract one from the other -- with that few people in each group, random variation will be high, so it may be that more of the "placebo" folks got unlucky... but the vaccine still causes fevers in 10% of the vaccinated folks).

As far as antibodies, you're right -- assuming that in Adelaide, Australia the rates of COVID-19 are very low, then only the vaccinated volunteers could develop antibodies. But what if one of the vaccinated people gets a mild infection naturally? Then you'd better hope that 1-2 of the placebo volunteers do as well, so that you can know the antibody response isn't necessarily from the vaccine.

Or what if the antibody test is not quite accurate? What if it has some false positives? Then some of the placebo volunteers might test positive as well, and be a flag to take the positive results with a grain of salt.

  • 2
    This, the rilevation of advrse effect is what makes the placebo necessary. Jul 6, 2020 at 14:17
  • 2
    Yes, where there is no current "standard of care" treatment, placebo-controlled blinded trials are necessary to determine if the new stuff is any better than, literally, nothing. Antibodies, side-effects, remission and other measured outcomes will have some non-zero rate in the population, and you need to see the difference the trial drug causes. Once a standard of care exists, it becomes mostly unethical to use placebos (depending on severity of disease, and unless there are people who cannot use the standard of care) and the trials of new drugs will be in comparison to the standard.
    – CCTO
    Jul 6, 2020 at 15:27

Antibodies only reflect one part of the immune response: the humoral immunity.

There is another part of the immunity that can't be detected by dosing the antibodies: the cell-mediated immunity. It produces no antibodies. It only creates effector T-cells that can detect the presence of viruses (or some intra-cellular bacteria) inside body cells and destroy these cells or trigger apoptosis.

This is one reason why checking the clinical outcomes of a vaccine is interesting. The results will be compared to antibody titers for interpretation.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.