Expert: it is not yet possible to accurately identify the owner of a fake certificate

As Novye Izvestia has already reported, the other day the head of the Gamaleya Institute, which developed the Russian Sputnik vaccine, Academician Gintsburg, said that his staff had created a test that would distinguish the vaccinated from the unvaccinated. It is suggested to use the test in hospitals where people with moderate and severe covid course are brought. Such a patient says: "Your vaccine does not work, so I was vaccinated, but I still lie now in a prone position and on oxygen." And the doctor said to him: “Come on, dear man, let's check how you were vaccinated” - and, hop, the test shows that the dear person lied, he was not vaccinated with anything, but bought a certificate.

Molecular biologist and science journalist Irina Yakutenko commented on the words of the academician, given the new data on coronavirus and vaccines:

“The idea is tempting, given that the scale of purchase of certificates in Russia is apparently unpleasantly significant. But how is it possible to test a person, because vaccines do not remain in the body and after a few days no injected substance can no longer be detected - hello to anti-vaccine users, who are sure that the vaccine gets stuck in the body forever and after N years will do something bad to this body? The vaccine does not remain, but the antibodies developed after interacting with it remain - and it is their new test that uses as evidence of the patient's honesty or deceit. Targeted vaccine antibodies target the coronavirus spike protein, and they are, plus or minus, the same in vaccinated and, for example, those who have been ill. In fact, this is not entirely true, and in those vaccinated, especially with mRNA vaccines, antibodies are of higher quality, since they are directed to the most vulnerable spots of the spike protein, plus, they do not have unnecessary antibodies to other coronavirus proteins, for example, to N- a protein, of which there is a lot in the viral particle, but antibodies against which do not help in any way to stop the multiplication of the virus. However, these minor differences cannot be caught with a test: effective antibodies are produced even after an illness, and a comparison of their number cannot be considered unambiguous evidence of anything.

But in the case of Sputnik, there is one clue. This vaccine delivers the coronavirus spike protein gene to cells using another virus called a vector virus. This in itself harmless virus infects cells, forcing them to synthesize a spike protein - and its own proteins encoded in the genome of the vector virus. Accordingly, the vaccinated body will develop antibodies not only to the SARS-CoV-2 spike protein, but also to the proteins of the vector virus. And it is these antibodies that the test developed at the Gamaleya Institute catches. Nice and unambiguous - yes, but not quite.

As a vector virus, Sputnik uses two types of weakened adenoviruses, viruses that cause common colds. In the first dose, the spike protein gene delivers the twenty-sixth type (Ad26) adenovirus to the cells, and in the second, the fifth type (Ad5). These viruses are not bred in the laboratory, they are quite widespread in the human population, especially Ad5. Adenoviruses were numbered as they were discovered (in the West, of course, in Africa their frequency is different), so, roughly, the lower the number in the name, the more this adenovirus walks around people, which means that the larger percentage will have antibodies to it. About 80% of the adult population has met with Ad5 in Europe, so the presence of antibodies to it in patients with covid who are suspected of buying a certificate does not mean anything. We get colds every year, and very often the virus that makes you sneeze and cough is type 5 adenovirus. It is for this reason that Ad5 is used as a vector in the second dose of Sputnik, and not in the first, so that numerous antibodies to it do not interfere with the production of antibodies to the coronavirus spike protein.

Ad26 is much less common. For Russia, as usual, there are no reliable data, however, for Europe, too. In some works, they are not seen at all, however, in a fresh article in the Lancet magazine devoted to Sputnik Light, antibodies were found in as many as 20 out of 110 volunteers (18% is a lot, although, of course, to draw far-reaching conclusions on 110 volunteers unreasonably). For comparison, in the United States, the prevalence of Ad26 is, according to various sources, from 2-4 to 10%, in China - about 35%. In other words, if you randomly select 100 people, 10-15 of them will test positive for antibodies to Ad26, even if they were not vaccinated with Sputnik.

That is, these people will be tested for deceit: doctors will decide that a person got sick after being vaccinated, although in fact he is not vaccinated. In other words, the test invented at the Gamaleya Institute will give a systematic (constantly present) error of 10-15%, worse for vaccination. In order to draw individual conclusions about a specific person, 15% is too much, especially when it comes to some kind of sanctions. In fact, the error is even greater, since not all the more crowded people develop an adequate response to the vaccine - and these people will just be in the group of those who are considered cheaters. That is, it is a mistake in the other direction.

For collecting statistics, these conditions are quite ok: knowing about the presence of a systematic error, we can take it into account in our estimates and estimate where the true value should be. In this case, we can find out the lower limit of how many people do buy certificates, and, for example, what percentage of those who were admitted to intensive care with covid and claim to be vaccinated were actually vaccinated.

To calculate the lower bound for the share of those who buy vaccination documents, you need to randomly select a large group of people with certificates (a large one, since we assume that they are still bought by a minority, albeit a significant one, therefore, in order for the conclusions to be statistically reliable, this minority should get a lot), and conduct a test for all of them for antibodies to Ad26 (the question, however, is how to convince them to take the test: obviously, citizens with fake certificates will not be eager to participate in the study). Among those vaccinated, antibodies to Ad26 should be about 100% (approximately - because, I repeat, some people do not develop an adequate response to vaccination - by the way, it would be good to estimate how many there are).

Suppose, among the holders of certificates, the proportion of people with antibodies to Ad26 will be less than 100% - let it be 80%. So, 20% are guaranteed to have cheated the state and bought certificates. But among those who have antibodies, a certain percentage must have them outside of the vaccination, that is, we cannot say whether they were vaccinated or bought a certificate. Thus, we will have a lower estimate - a little less than 20% of purchased certificates (a little less - because of those who have not developed an answer to the vaccine). Important: we are talking only about certificates of vaccination with Sputnik, the test does not apply to two other Russian vaccines, they do not give antibodies to adenovirus.

By the way, having a test in hand, it would finally be possible to determine the prevalence of Ad26 in the Russian population. And then about SARS-CoV-2, we learned three orders of magnitude more in a year and a half than we know about ordinary coronaviruses that have been with us for hundreds and thousands of years. It would be great if the creators of the test decided to conduct such a large-scale population study related not only to covid. But so far they have not announced such intentions..."