[ad_1]
The number of deaths left by Covid-19 in Brazil exceeded 150 thousand people this Saturday (10).
However, seven and a half months after the first case in the country, the real dimension of the disease among Brazilians is unknown.
Even the first researchers who tried to estimate the total number of infected people, that is, who tried to discover seroprevalence, the percentage of infected people, with or without symptoms, were eluded by the new coronavirus.
The amount of antibodies in infected people begins to decrease after a few weeks. It boils down to the point that the tests generally used in these investigations no longer capture these signs of infection. After a while, a person who was once infected may appear to have never been invaded by the new coronavirus. For this reason, these first major surveys likely indicated relative stability or a drop in the infection rate since June – they underestimated the size of the epidemic. The number of officially confirmed cases exceeds 5 million.
Knowing the size of the infection can help define health policies, including vaccination. It can indicate when a certain percentage of infected people alone leads to a reduction in the number of new cases, the so-called herd immunity. Correction of these surveys may be necessary to find out if the infection leaves people immune to the virus, for how long, and with what level of resistance to the disease. Even without detectable antibodies, there may be immunity, but antibodies are no guarantee of protection.
New studies attempt to correct the underestimation. In tests carried out by the São Paulo City Council with samples from the population of São Paulo, the percentage of infected varied between 9.8% and 13.9% from July to August. This small variation is not compatible with the number of patients and deaths. The work of Brazilian researchers published in September indicates that the city of São Paulo could have more than 22% infected in August, so the total number of contaminated would be 10 times higher than what appears in the official count of confirmed cases. In Manaus, up to 66% would be infected, compared to 27.5% of the results “without correction”.
Serological surveys, such as the one in the city of São Paulo and Epicovid, national, try to obtain the general number of infected people in the population by examining a sample of a part of this same group of people. Those selected are examined with so-called rapid tests, which detect if there are antibodies in the blood collected by means of a finger prick.
These tests are cheaper and easier to apply, without which it would not be possible to test tens of thousands of people. Before use, its sensitivity, the ability to detect infections, is compared to that of more accurate tests. However, these tests are calibrated with more severe patients with the disease. In other words, apparently, the tests end up being more sensitive to infections in which there was a greater defense reaction of the organism and, therefore, more production of antibodies.
In the case of an asymptomatic or less aggressive disease, there would be a lower production of antibodies and a more rapid deterioration, indicate the studies, the immunologists and epidemiologists heard by the sheet.
“The total of antibodies always decreases. In the case of the new coronavirus, this seems to be faster than in the case of Sars or Mers, ”says immunologist Gabriel Victora, professor and head of the Lymphocyte Dynamics Laboratory at Rockefeller University in the United States. In the case of Sars and Mers, two coronaviruses that are very lethal but spread very little at the beginning of the century, the antibodies last one to three years. In coronaviruses that cause common colds, less than a year and up to three months.
However, the defense cells have a memory of the crown or other invader, producing more antibodies in the event of a new infection; otherwise, the immune reaction depends on a complex and coordinated reaction of various types of defense cells. It remains to be seen what level of protection this defense system offers against a new viral attack. On the other hand, reinfections with the same type of virus could cause milder illness, even though waves of contamination are recurring, such as in cold outbreaks, Victora says.
A new infection can also occur because the virus undergoes mutations, because immunity against one type of corona is not enough to protect against a similar one, or because the defense of those once infected is gradually broken. It is not very well known, even in the case of ancient coronaviruses. Even less the new.
In epidemiological studies, the attempt remains to give a dimension to the spread of the coronavirus. In the Manaus and São Paulo study, instead of visiting and evaluating the population, the researchers used samples of blood donations stored between February and August in blood centers. It is less expensive and allows better testing, says immunologist Ester Sabino, one of the study’s authors, a professor of medicine at USP and a member of the group that sequenced the genome of the “Brazilian” coronavirus.
In these samples, the antibody level of donors infected for a longer time may also decrease. Then, Esther’s group followed up with people infected by Covid to measure that rate.
Decrease and correct the results.
With statistical adjustments for demography, for the sensitivity of the test and this correction, they reached an estimate of real infection in Manaus: from 44% to 66% in August, going from 60% in early June. Since the number of new cases has since decreased in the city, it is assumed that the threshold of herd immunity would be there.
Since the pandemic began, there have been controversies about herd immunity. At first, with simple mathematical models and intrinsic premises, it would be 60%, it was estimated. Taking into account the heterogeneity of social contacts and the fact that few infected people are more polluting than most, the threshold for herd immunity could be 20% to 40%. However, the figures in Manaus and the difficulty of estimating the true size of the epidemic cast doubt on all these accounts.
At least two other Brazilian studies are trying to find out how long an infected person can be caught in the mass test network. One of them, in Rio de Janeiro, from Fiocruz, takes blood and throat and nose samples (with swabs, those big cotton swabs) in a row and accompanies the infected. This is how the epidemiologist Claudio Struchiner, professor of applied mathematics at the Getúlio Vargas Foundation, tells us who worked with the researchers who carried out the national study, Epicovid, one of the first in the world, initiated by personnel from the Federal University of Pelotas.
“We noticed the drop in June, especially in the north. One hypothesis is that the test used begins to lose sensitivity in two or three weeks, those tests that were designed when the dynamics of the virus were not well known. Now let’s try to find out what they actually measure, ”says Struchiner. Similar work is being carried out in Pelotas (RS), says Pedro Hallal, epidemiologist and dean of Ufpel, who coordinates Epicovid.
It was also in June that the first studies began to appear to measure the level and rate of breakdown in the amount of antibodies. But these studies in China, the United Kingdom, and the United States have obtained different figures for the rate of antibody reduction. Furthermore, it is not known if this rate will be constant and will continue indefinitely. To complicate matters, a large study in Iceland found that the level of antibodies there only drops slightly, but remains stable for at least four months.
One of the Epicovid researchers, Aluísio Barros, an epidemiologist at Ufpel, points out other difficulties. The level of detectable antibodies depends on the severity of the infection, in addition to other characteristics of the individuals. As it is not known how the severity of the disease is distributed among those infected, it would also be difficult to estimate the amount and the general rate of deterioration in the level of antibodies and, therefore, a more precise number of infection rates.
In the study by Ester Sabino’s group, caution is required when extrapolating the results. The sample of blood donors is not representative of the general population (it only contains asymptomatic people and people between 16 and 69 years old, for example), although it has been corrected for the general characteristics of the inhabitants of Manaus and São Paulo. Furthermore, the raw data (without any correction) on the prevalence of infection among donors from São Paulo and those from the São Paulo city hall sample, in the general population, are very similar, which would reinforce the research findings .
Barros from Ufpel believes we may be in the dark for other reasons. There are still two dozen suspected cases of reinfection with the new coronavirus around the world. It is not known whether this is good news or a sign of relative ignorance. A person may have had an asymptomatic infection, rarely detected, and a reinfection with mild symptoms or vice versa, says Barros.
Another problem with blood donor studies is that few cities have blood centers with infrastructure capable of allowing storage and good research, says Ester. In addition to the work already published on Manaus and São Paulo, his group studies Belo Horizonte, Curitiba, Fortaleza, Recife, Rio de Janeiro and Salvador.
Among the many controversies is the ability of different surveys to capture the number of people infected even at any given time. At the end of June, Epicovid tests indicated a prevalence of 1.2% in the city of São Paulo; in the city’s survey, the infection rate was nearly 10% in early July.
All these difficulties call into question the idea of ”immunity passports” (whoever has the virus can circulate freely) or estimates of herd immunity. They make it more difficult to separate the effect of the (possible) immunity rate from the influence of behavior (distance, masks) in slowing down the epidemic. If herd immunity is indeed close or has been reached, it is possible to relax the distance measures. But we still don’t know.
