Clinical studies of the efficacy of a vaccine evaluate not only whether an immunizer is effective or not, but the size of that efficacy. If this rate is low, it is not usually worthwhile, in epidemiological and financial terms, to adopt the product as a prevention strategy in the context of public health.
In most cases, regulatory agencies, such as Anvisa, require 70% effectiveness of a vaccine for which it receives the license. In the case of a possible vaccine against covid, it is already assumed that it can be approved in Brazil and other countries if it is at least 50% effective. But how is this index calculated?
According to experts, the indicator will be discovered from calculations that will consider the number of infected with covid-19 in the two groups participating in the clinical study: the one who took the vaccine and the one who took the placebo. “First, we check how many infections occurred in each group and calculate the corresponding incidence in the total group. Then, we have a formula for calculating effectiveness: it is the incidence in the placebo group minus the incidence in the vaccinated group divided by the incidence in the placebo group. . The result is multiplied by 100 to reach the percentage index of effectiveness “, explains Denise Garrett, vice president of Sabin Institute of Vaccines.
The explanation is a simplification of the calculations made by the researchers, who use sophisticated statistical programs for such analyzes. But giving an example in numbers: if scientists find that in the placebo group the incidence of covid was 20% and, in the vaccinated group, it was 10%, the count would be 20 minus 10, and the result (10) would be divided by 20 , reaching the index of 0.5, which multiplied by 100 would reach an efficiency of 50%.
Of course, in practice, many other data are considered. For this result to have the so-called “statistical power”, it is necessary to have a minimum number of infected people to make the sample reliable. In other words, if these calculations are made with a very small group of people, there is a chance that the results found are the result of chance and not really the power of the vaccine, so scientists say that the differences in incidence found in the two groups need be “statistically significant”.
In the case of coronavac, this minimum sample for a first partial analysis is 61 infected, but it may be that the difference found between the vaccinated group and the placebo is still not robust enough to prove the effectiveness in a definitive way. And then we move on to the second analysis, in which the minimum sample will be 151 infections.
It is because of this need for statistically robust data that partial data analysis is not always sufficient to prove effectiveness and register a product. “There is a chance that the effectiveness rate measured in the first analysis will decrease in a final analysis. Here in the United States, for example, it was determined that vaccine researchers will have to wait two months after applying the second dose to make any application for registration, precisely to have a more robust sample and more follow-up time “, explains Denise.
Partial analyzes, called interim, are common in clinical studies, regardless of pandemic. But in normal situations, they are usually used only to confirm the safety of the product and to confirm the continuity of the study in new volunteers. Only in exceptional cases, such as rare diseases without treatment or in the context of the covid pandemic, can these analyzes yield results in order to support a registration application.
Efficacy analyzes in a vaccine study can measure not only how much the product was able to reduce the number of infections, but also other results associated with preventing the disease. “What exactly will be measured depends on the design of the study. Researchers may want to measure the number of infections, but they can also check whether there has been a reduction in mortality, whether there has been a reduction in serious cases. It may be that a vaccine is not as effective for prevent contamination, but it is able to reduce the complications caused by the virus. Generally, a research protocol monitors several of these outcomes “, explains Gustavo Mendes, general manager of medicines at Anvisa.
According to information from the Clinicaltrials.gov website, the US government portal that gathers information from clinical trials, the two primary outcomes that will be measured in the study of coronavac are the incidence of covid in volunteers two weeks after the application of the two doses and the occurrence adverse events. But the research also foresees monitoring seven other secondary outcomes, among them the incidence of severe cases of covid, the incidence of the disease with only one dose of the immunizing agent, the seroconversion rate (when antibodies against the disease become detectable by exams) , among others.
Coronavac is now the most advanced research vaccine in Brazil, with 9,000 of the 13,000 volunteers already immunized. The immunizer developed by Oxford in partnership with Astrazeneca is in a similar stage, with 6,000 of the estimated 10,000 volunteers already immunized with at least one dose, according to information from the Federal University of São Paulo (Unifesp), which coordinates the national research arm . Vaccines from Pfizer and yes Janssen. The latter interrupted global testing temporarily to investigate side effects in a foreign volunteer.