How will we know if a COVID-19 vaccine works?

Ideally, Phase III studies to assess COVID-19 vaccine efficacy are double blinded RCTs where participants are actively followed post-vaccination to identify COVID-19 infection and the outcomes of infection.

In the image below, numbers of cases of infection (measured by regular swabbing of participants when they are asymptomatic) are the same between the vaccine and placebo groups, suggesting the vaccine does not reduce the likelihood of infection.

However, the data suggests that infected vaccinees are less likely to develop symptomatic infection than those participants who received the placebo, suggesting the vaccine may be efficacious against symptomatic disease.

If you need a text version of the image above, this is available as a PDF.

Participant numbers

In this example, none of the infected individuals in the study required hospital admission or died, meaning the study was unable to provide information as to whether the vaccine reduced the likelihood of these outcomes.

It is essential that such studies include sufficient numbers of participants that reach particular efficacy endpoints if they are able to provide a reliable measure of efficacy.

If incidence of SARS-CoV-2 infection is low, then greater numbers of participants are needed in order to have sufficient numbers of participants infected to provide efficacy data.

If all participants are young adults, who are unlikely to develop severe disease or die, then the study is unlikely to be able to provide efficacy data for these endpoints.

Pooling data

It is currently expected that individual Phase III studies will not be able to include sufficient numbers of people in order to provide data for vaccine efficacy against severe disease and death. However, by pooling data from multiple studies, it may be possible to calculate this.

It is possible that safe COVID-19 vaccines shown to induce strong immune responses will be licensed, with effectiveness data collected via subsequent Phase IV studies.

Not all participants will be exposed to SARS-CoV-2 during a study’s follow-up, not all exposed individuals become infected, and not all infected individuals become symptomatic, require hospitalisation or die.

There are multiple factors which determine these outcomes, as demonstrated in the image below. This is a reason why it is important that participants are randomised to receive study vaccine or placebo in efficacy studies, and why large numbers of participants are required to conduct appropriately powered trials.

Figure reproduced from Hodgson et al, 2020. An expandable version is available as a PDF.