It's time to talk about Interim Analysis in the clinical trials of SARS-CoV-2 vaccines. Most all large Phase 3 clinical trials, and quite a few smaller Phase 2 trials, I've known had Interim Analysis built into their design. It allows the trial to be temporarily stopped at some predefined midpoint. In trials of cancer drugs that I knew well, the stopping point was usually when roughly 40-50% of patients had been treated and evaluated. In vaccine trials, most of the volunteers were quickly vaccinated. Interim analysis happens when a given number of "events" occurred – where the event is a confirmed SARS-CoV-2 infection.
One of the trials recently announced some protocol features that will be part of its Interim Analysis. I can't find that news article, so I'll rely on memory. After about 150 infections occurred, the trial's Data & Safety Monitoring Board met and looked at unblinded data for those 150 cases. They could see how many infections occurred on the experimental SARS-Cov-2 vaccine arm and how many were on the control or placebo* arm.
Obviously, they don't expect zero infections on the experimental arm and all the infections on the placebo arm. And, they hope they don't see it work out to about 50% infections on each arm. The goals were already written into the trial's protocol, stating how many infections on the experimental arm would result in stopping the trial early for futility, for successful efficacy, or would allow the trial to continue. It is rare for a trial to stop early because the Interim Analysis shows success. But this depends on high high or how low they set the bar.
In the case of one of the vaccines, if I remember correctly, the bar was set for success if the placebo arm has at least 50% more infections than the experimental arm. For 150 infections, if at least 90 were on the placebo arm, and no more than 60 were on the experimental arm, it could indicate an early sign of success. (Please note, I'm not certain I remember these numbers correctly. I'm only trying to explain how Interim Analysis proceeds.) On the other hand, if infections were roughly equal on both arms, it would indicate the experimental vaccine offered no protection.
My point here is that these very large trials don't require that results must reflect all the people on the trial. If, and only if, enough people are protected from viral infection during the Interim Analysis, that could be considered good enough reason to begin widespread public vaccinations. The trial would continue to evaluate safety & efficacy of all 30,000 people on the trial. But the vaccine would be given temporary approval, pending the final results of the trial.
All the SARS-CoV-2 vaccine trials have Interim Analysis built into the trial protocols. The details of where the bar is set may vary, but all the trials all have this feature. It really isn't necessary to deliberately infect vaccinated people to see the results.
* Instead of true placebo, people on the control arm were given an vaccine unrelated to SARS-CoV-2, such as the DTaP vaccine against diphtheria, tetanus, and pertussis. It's considered unethical to inject a blank placebo when it's well known that immunity against diphtheria, tetanus, and pertussis wears off most people after about 10 years.
