” Ninety percent protection against Covid-19 with two doses and 80 percent with one “, this is the promising result of a vaccination campaign among American caregivers with Pfizer and Moderna vaccines, made public Monday, March 29. These data, in line with what the two laboratories had previously announced in their trials, are reassuring. Indeed, an impression of a communication war has prevailed in recent months, with each laboratory wielding an ever greater efficiency rate – Pfizer even advanced the figure of 100% efficiency in adolescents on Wednesday, March 31.
In order not to simplify anything, some laboratories had to retroped. Challenged by the US regulator for incorporating “outdated” data into the results of a clinical trial, the Swedish-British laboratory AstraZeneca has revised down the effectiveness rate of its vaccine on symptomatic cases.
Questions about these figures, how they are calculated, controlled and how, possibly, they evolve. Here are four points to understand the issues behind these numbers.
Different efficiency rates can be put forward by laboratories. BOB EDME / AP What is efficacy for a vaccine ?
When talking about a 90% effective vaccine against Covid-19, it means that such a vaccine reduces the risk of developing a symptomatic form of the disease by 90% compared to another control group that has not received the vaccine, at a time when the epidemic is present.
Different efficacy rates can be highlighted by laboratories : the efficacy of Johnson & Johnson vaccine is 66% in general, but it rises to 85% when focusing on the prevention of severe forms of the disease. In Israel, where real-life trials have been taking place for several months, the effectiveness rate of Pfizer-BioNTech’s formula against severe forms of the disease has risen to 97 %.
How is it calculated ?
It is from phase 3 of clinical trials that the effectiveness of vaccines can be evaluated, when they are tested on tens of thousands of people. Indeed, in order to obtain convincing projections on a population as a whole, a reasonable (statistically) basis of infected persons is required.
This rate is calculated by taking two groups of volunteers, one vaccinated, the other on placebo, and comparing them during a given epidemic period. The effectiveness figures are stopped when a predefined number of Covid-19 patients is reached among the volunteers. Another possibility may be to voluntarily infect previously vaccinated people, a “controlled infection”, and observe the result – a faster technique (there is no need to wait for the virus to circulate), but ethically questionable in the case of a pandemic like Covid-19.
In conventional trials, where the virus is allowed to circulate, each laboratory sets a goal (which can be revised upwards, to strengthen the solidity of the data) : for example, Pfizer-BioNTech wanted to reach 170 cases of patients, and Moderna 196. Targets that then allow to calculate the rate of effectiveness : at Pfizer, of the 170 people who contracted Covid-19, only eight belonged to the group of vaccinated, which gives a protection rate of 95 % (on average, without detailing severe cases or age groups). In Moderna, where eleven people fell ill among the 196 vaccinated group, a rate of 94% is obtained <
At a vaccination center in Almaty, Kazakhstan, on April 2, 2021. PAVEL MIKHEYEV / REUTERS Who controls these measures ?
In phase 3 of the most sensitive clinical trials, where lives or serious health effects are at stake, laboratories must work with a data and safety monitoring board to ensure that the rights of trial participants are preserved. Created to measure and composed of medical, statistical and ethical experts, the committee can waive the anonymity of data in case of blind studies. This monitoring is submitted to health officials, the US Food and Drug Administration (FDA), the European Medicines Agency and national institutions.
The oversight and monitoring committee ” provides independent advice to ensure that participants are not exposed to undue risk. It also makes recommendations regarding the continuation, modification or discontinuation of a trial, ” explains the Eupati professional group. Its members ” are selected and appointed by the sponsor, but they must not have any connection with the trial, the sponsor or any other activity or entity that may affect their objectivity “.
The case of AstraZeneca is a good example of the efficiency of these control mechanisms : after several exchanges in February and March with the laboratory, the controllers would have signified that the latest results showed a real effectiveness between 69% and 74 %, says the Washington Post, which had access to a letter sent by the control committee to the laboratory and to federal health officials. He reports that the committee “strongly recommended” that this information be made public in the press release of 22 March, rather than posting only the 79% resulting from an interim analysis. Whose act, three days later, with an erratum published by AstraZeneca.
” Countries like Britain and Finland vaccinate children first, because they are the ones that feed the epidemic the most. We will have to ask the question “, judges virologist Bruno Lina, member of the scientific council. JOE KLAMAR / AFP Why can the efficiency rate vary ?
The effectiveness of the same vaccine can vary according to several criteria, including the definition of the effectiveness chosen (depending on whether one focuses on the prevention of severe or not) but also the dose administered, or the interval between doses. To take an example from the AstraZeneca cohorts, the effectiveness reached 82.4 % in volunteers with an interval of twelve weeks or more between the two doses, compared to 54.9 % with less than six weeks.
Moreover, the vaccine effectiveness is not equivalent against the variants : the latest published phase 3 trials (AstraZeneca, Johnson & Johnson and Novavax) have confirmed that their vaccines retain almost all their protective power against the so-called British variant ; on the other hand, they lose some of their strength against the so-called South African variant. The data from Pfizer and Moderna could only be constructed a posteriori, by putting blood samples of vaccinated people in contact with variants (the latter had in fact not yet appeared during the trials carried out between July and November 2020) : they point to a risk of more rapid weakening of immunity. Since then, industry and academic research have been searching for a second-generation vaccine that can protect against multiple variants at once.
Article reserved for our subscribers Read also Covid – 19: who will win the battle of variants ?
Since the circulation of the virus feeds the production of the variants, they pose by ricochet the question, more difficult to calculate but equally strategic, of collective efficiency, and no longer only individual. “In France, we vaccinate against the flu especially the elderly, who are the most fragile. But it is also those in whom the vaccine works the least well. Other countries, such as Britain or Finland, vaccinate children first, because they are the ones that feed the epidemic the most. We will have to ask the question “, judges virologist Bruno Lina, member of the scientific council and director of the National Reference Center for respiratory viruses, in Lyon.
The “Behind the Digit” series of Decoders dissects statistics appearing in the news. Find all the articles in our dedicated section. Our selection of articles on Covid-19
Find all our articles on SARS-CoV-2 and Covid-19 in our section
On the epidemic :
Visualize the evolution of the epidemic in France and in the worldCovid-19 : rise, rise, rise, rise stabilization or decline, where is the epidemic in your department ?What are the differences between vaccines ? And their side effects ?The figures on the Covid-19 pandemic are of interest, but they do not explain everything
And also :
Shopping centers, curfews, local confinements, what is allowed, restricted or prohibited in Francecompanies, schools, hospitals… which masks are sufficient or not ?Video: Covid – 19: which places are most at risk ?Beware of bad advice and false rumors: we help you to do the triInfo practices, checks…, follow the developments of the health crisis on WhatsApp See more See less
Contribute Sharing Sharing disabled sharing disabled sharing disabled Send by e-mail Sharing disabled sharing disabled sharing disabled sharing disabled sharing disabled
