Some explanations on the clinical trials of the vaccine against Covid-19.
To verify the efficacy and safety of experimental vaccines against Covid-19, clinical trials are carried out on tens of thousands of volunteers, generally divided half between a placebo group and a treated group, blind. This is the case for the Pfizer / BioNTech and Moderna trials, which each recently reported very high efficacy.
Who conducts clinical trials?
The American group Pfizer is conducting and financing the clinical trial of its vaccine itself, on 44,000 people in the United States, Brazil, Germany, South Africa and Turkey. Participants volunteer and, if accepted, receive two doses three weeks apart, and are monitored regularly.
Moderna is leading the trial in partnership with the National Institutes of Health, America’s largest public research organization, which is co-funding the study and co-developed the vaccine, also in two doses, spaced four weeks apart. It takes place only in the United States, with 30,000 volunteers.
Do participants know if they are receiving the placebo or the vaccine?
No. The nurse giving them the injection also ignores it. Placebo is a simple physiological serum (saline solution), without any therapeutic effect.
How do you know if the vaccine is working?
All the volunteers in the trials lead their lives normally: after their injections they go home, and work, study and live like everyone else, with the same advice on caution – or confinement or wearing a mask – as the rest of the world. population.
Over time, a number of participants will naturally catch Covid-19. We will know this because they must regularly inform researchers of their symptoms, and any suspected case will be diagnosed.
If the vaccine is effective, the number of cases in half of participants who received the real vaccine will be lower than in the other half, the one who only had the placebo.
The goal is for the difference to be large enough to exclude it being the result of chance. Statistical methods come into play to achieve a predetermined level of certainty. 100% efficiency would mean that there were no cases among the vaccinated, and several in the placebo group.
Note that the primary purpose of vaccines is not to prevent contamination by the coronavirus: it is to prevent people from developing the disease caused by the virus, that is to say the Covid-19. It will be considered particularly effective if, a fortiori, it prevents severe forms of Covid-19.
The idea is that if a vaccine prevents getting sick, the public health goal will be met, even if asymptomatic forms continue to spread.
Are participants voluntarily exposed to the coronavirus?
No. We simply observe whether, “in real life”, there is a difference in the incidence of Covid-19 between people who have been vaccinated and those who have not.
As the disease does not have a 100% effective cure, it would not be ethical to deliberately infect people with the coronavirus, although some researchers have proposed doing so on young and healthy people.
But why have a placebo group?
If we only followed 10,000 vaccinated people, and 100 fell ill with Covid-19, how would we know the degree of effectiveness of the vaccine? Maybe without the vaccine there would be as many sick, or double, or ten times as many.
The only way to determine the effectiveness is to compare to an equivalent number of unvaccinated people, therefore with a placebo group.
With a very large sample of tens of thousands of people, representative of the population, it is almost certain that the two groups will have the same diversity of profiles and behaviors, which makes it possible to compare them.
There is indeed no reason why all young adults in one group, for example, should refuse to wear a mask, but not those in the other group. Especially since everyone does not know if they have had the placebo or the vaccine.
Who analyzes the data?
It is not the companies that analyze the data directly, but independent expert committees often called “data and safety monitoring boards” (DSMB), and whose members’ names are kept secret to protect them from any political or other pressure. .
In the case of Pfizer, this committee is made up of five people.
In the case of Moderna, AstraZeneca / Oxford and Johnson & Johnson, a single committee of independent experts has been set up by the National Institutes of Health, with 10 to 15 members, according to Kaiser Health News.
At regular intervals provided for in the trial protocols, these committees lift the veil on the accumulated data to see to which groups the participants belong. They can then inform the manufacturer in case of conclusive results, and the manufacturer can then use the data to apply or not for a marketing authorization.
In addition to efficiency, the committees closely monitor safety. They analyze the severity and frequency of side effects, which will be decisive in authorizing or not the product.
170 cases, isn’t that a small sample?
Pfizer reported observing 170 cases of Covid-19 within seven days of the second dose, which seems small compared to the total number of participants (44,000).
But statistically, this result is sufficient. In fact, 162 cases were found in the placebo group, and only 8 in the vaccinated group. The difference is so great that it is extremely unlikely to be a fluke.
The calculated efficacy of 95% means that vaccinated people were 95% less likely to catch Covid-19 compared to unvaccinated people.
Will the placebo volunteers be able to be vaccinated at the end?
This is the great ethical dilemma. Usually, when a drug or vaccine proves its effectiveness, participants who received the placebo are informed that they have received the placebo and are offered, if they wish, to have the real product.
In all cases, nothing would prevent participants in doubt from going independently to the pharmacy to be vaccinated.
The problem is that the trials are supposed to last two years or more, in order to determine the duration of protection conferred by the vaccine, and to watch for the appearance of long-term side effects.
So far, health authorities, questioned by manufacturers, have not made a recommendation to resolve the problem.