Imprimir Republish


Brazil begins testing

To better understand the spreading pandemic, research initiatives aim to increase the scale of COVID-19 diagnoses throughout the country

Blood is collected in Pelotas, Rio Grande do Sul, to conduct a rapid test as part of the EPICOVID19-BR study

Daniela Xu / UFPEL

Brazil is trying, finally, to determine the size of its novel coronavirus pandemic, which saw its first reported case more than three months ago, on February 26. Over the last few weeks, mass-testing programs have been launched and the initial results from epidemiological studies have begun to be released, with the potential to help the country understand the dynamics of the disease and devise strategies to combat it. There are three standouts among the largest Brazilian research projects: the study “Evolving Prevalence of COVID-19 Infection in Brazil” (EPICOVID19-BR), coordinated by the Federal University of Pelotas (UFPel) in Rio Grande do Sul; a pilot study in the capital of São Paulo led by Grupo Fleury; and a large-scale testing program being conducted by the government of São Paulo.

Experts are unanimous in stating that these initiatives are important because Brazil is among the countries that have done the least amount of testing for COVID19. According to the Ministry of Health, the testing rate at the end of May was 4,200 tests per one million inhabitants, the lowest rate among nations with high case numbers. At that time, 871,800 Real Time-PCR molecular assays—which detect fragments of coronavirus RNA in samples taken from deep in the nose or throat—had been performed in Brazil. Of these, 460,100 were processed in public laboratories and 411,700 in the country’s five largest private labs. The agency did not report the number of rapid serological tests, which detect the presence of antibodies in the blood.

“Brazil should be testing on a large scale in order to reduce the basic reproduction number [a measure of transmission rate between people]. We should also follow protocols aimed at identifying infected individuals, then isolate them and enable contact tracing, along with adopting physical-distancing measures, social isolation, and quarantine,” says epidemiologist Hélio Neves, from the Santa Casa de São Paulo School of Medical Sciences. “It’s taken a long time for us to have RT-PCR tests available in Brazil in a reasonably sufficient number, which still hasn’t been completely resolved.”

Researchers say the delay in getting tests processed is another bottleneck. “Although there’s talk about the large capacity our dozens of public and private labs have to perform molecular tests, that’s not what’s happening in practice. Many municipalities don’t yet have the capacity or training to conduct these tests,” says Neves. At the end of May, more than 2 million RT-PCR tests were awaiting processing in the public healthcare system. Even more troubling, the lack of a trivial resource—cotton swabs for collecting samples in cases of suspected infection—also contributed to the low number of tests being done throughout Brazil.

In an attempt to set a new pace for testing, in May the Brazilian Ministry of Health launched its Diagnosticar para Cuidar (Diagnosing for Caring) strategy, which plans to conduct 46.2 million tests by the end of September. Of this total, 24.2 million will be RT-PCR molecular (RNA) diagnoses, with another 22 million serological rapid tests. In the first weeks of the program, 8 million test kits had been passed on to the states. The ministry’s plan is to do the RNA tests within seven days of the onset of symptoms, when the virus is active in the patient’s system. However, rapid testing, which identifies the body’s immune response to infection within a few minutes, must be performed after the eighth day of symptom onset.

The strategy adopted in São Paulo will follow a similar path. The intention of the São Paulo government is to complete 1.3 million RT-PCR exams and 2 million rapid tests by the end of August. According to a statement by hematologist Dimas Tadeu Covas, director of the Butantan Institute and coordinator of the Coronavirus Contingency Center in São Paulo, with this increase in diagnoses, the state should reach a ratio of 27,000 tests per one million residents, close to the levels of testing in some European countries. As of mid-May the state of Ceará had the highest testing rate, at 5,400 tests per one million people.

The São Paulo program will begin testing people with only mild COVID symptoms using the RT-PCR exam, a test that had previously only been used with hospitalized patients, and healthcare professionals experiencing symptoms. Meanwhile, the rapid serological tests are focused on law enforcement personnel and their cohabitants during the first phase, beginning with those working in the city of São Paulo. This will be followed by testing of healthcare professionals and families of people who have had contact with patients.

The poor accuracy of rapid tests can produce inaccurate results and mislead people

However, rapid tests—performed by collecting a drop of blood with a fingertip lancet—have one big drawback. If sampled during the first few days of infection when the body hasn’t yet produced IgM and IgG antibodies, the test will show negative results, even though the person is already infected. The production of IgM only begins after the seventh to tenth day of infection, while longer-lasting IgG antibodies, which should conceivably protect a person from future infections, are produced after about 15 to 20 days of infection.

Furthermore, scientists warn that some rapid diagnostic kits being used in Brazil are unreliable. A study by the Oswaldo Cruz Foundation (FIOCRUZ) in Rio de Janeiro, led by hepatologist Hugo Perazzo, revealed that the 16 test kits analyzed at the end of March were of low sensitivity, i.e., they showed reduced ability to correctly identify sick individuals.

“These tests can be useful in a medical emergency during the COVID-19 pandemic in Brazil,” pointed out the study, which was published in The Brazilian Journal of Infectious Diseases. “However, it’s important to note that the rate of false negative results from tests that detect IgM antibodies for SARS-CoV-2, which is used for diagnosing COVID-19 in its acute phase, ranged from 10% to 44%.” The team analyzed all 16 test kits approved by the National Health Surveillance Agency (ANVISA) up to that time—today there are more than 100.

The study was based on sensitivity and specificity data (the test’s ability to correctly identify uninfected individuals, avoiding false positives), supplied by the companies themselves, when requesting product authorization from ANVISA. This is because the testing kits are being approved by simply analyzing the information provided by the registrants, without prior laboratory evaluation.

“ANVISA relaxed some rules to allow for faster evaluations of registration requests for products aimed at controlling COVID-19. This includes these tests, which haven’t been previously analyzed in the lab, but that are being monitored in partnership with the INCQS [National Institute for Quality Control in Health, at FIOCRUZ] and the Ministry of Health,” says Leandro Rodrigues, general manager of Health Products at ANVISA. As of the end of May, two of the 20 tests monitored by INCQS had been found to be “unsatisfactory.”

One of the kits analyzed by FIOCRUZ is manufactured by the Chinese company Wondfo. The federal government purchased a batch of 5 million units of their test kit. The test’s sensitivity rate is 86%, which presents a 14% risk of false negatives, a high rate for use in clinical practice. For this reason, experts say that rapid serological tests are not ideal for making diagnoses or indicating who has COVID-19, nor are they useful in determining who may be released from quarantine under the presumption they have already acquired antibodies against the virus—the highly touted “immunity passport,” which is still highly questioned.

Since the end of April, pharmacies have been authorized to carry out rapid tests on the population to detect antibodies to the SARS-CoV-2 virus. The ANVISA decision was aimed at expanding the network and availability of testing, and at reducing the demand on public health services. The measure was criticized by the experts. “Because of their poor accuracy, rapid serological tests should not be used in pharmacies. They just mislead people,” says biologist Natalia Pasternak, founder of the Science Question Institute and collaborating researcher at the Institute of Biomedical Sciences of the University of São Paulo (ICB-USP).

According to the researcher, these tests are indicated only for use in epidemiological studies, in which the false-negative rate can be compensated for using statistical tools. This is how they are being used by the UFPel team that is coordinating the EPICOVID19-BR national virus prevalence research. Scheduled to be carried out in 133 Brazilian municipalities, the survey uses the Wondfo lab’s diagnostic kit. In total, 99,750 people will be tested by teams from IBOPE [Brazilian Institute of Public Opinion and Statistics], in three stages, separated by intervals of 21 days between each phase.

Léo Ramos Chaves Drive-thru testing is conducted at a shopping mall in São Paulo to avoid lines and crowdsLéo Ramos Chaves

“In the beginning one of our biggest concerns was choosing the test. Although rapid serological tests are not as accurate as those processed in the laboratory, they are the obvious choice for large-scale studies such as ours,” says epidemiologist Aluísio Barros, from the UFPel Epidemiological Research Center and one of the coordinators of the EPICOVID19-BR project. “The matter of field operations in a survey with national coverage is fundamental. So we opted for an easy-to-apply test. In terms of balancing logistics and validity, this test is quite good.”

The results of the first phase of the survey, carried out in May, showed a COVID-19 prevalence of 1.4% in the population at the sampled locations, a rate that was seven times higher than the official statistics had indicated. Extrapolating to the overall Brazilian populace, the EPICOVID19-BR estimate suggests that the number of contaminated people in the country is around 2.5 million, much more than the 365,000 confirmed cases recorded at the time.

The study also showed that the country was still far from achieving collective, or herd immunity, reached if 70% of the population contracts the virus. In this scenario, the high number of immune individuals impedes transmission of the virus because there are so few susceptible people left, and the epidemic is effectively under control. Until that point, however, there would be an excessive number of deaths.

The city of São Paulo, an epicenter for the pandemic in Brazil, has also initiated a sample survey to determine the number of people with antibodies against the new coronavirus. Coordinated by Grupo Fleury, IBOPE, and Instituto Semeia, the mapping was carried out in six districts across the city, the three with the most deaths, and the three with the highest incidence of the disease. The results of the first phase showed that 5.2% of the population in these regions has already had contact with the virus.

“One advantage with our study is that the teams collect blood from veins—not at the fingertip, as in the case of rapid testing,” notes infectologist Celso Granato, clinical director of Grupo Fleury and a professor at the Federal University of São Paulo (UNIFESP). “The test, from the Chinese firm Snibe, uses a chemiluminescence analysis methodology, and has a sensitivity and specificity of around 90%.”

Granato explains that the study results give a momentary X-ray of the percentage of population that may be immune to the virus in the neighborhoods sampled. By conducting the second stage of visits, when they intend to collect blood in every district of the city, it will be possible to calculate how quickly the virus is spreading. “This is how we’re going to understand the dynamics of the epidemic among us,” Granato says.

Physicist Domingos Alves, a specialist in theoretical and computational healthcare modeling and a researcher at the Center for Information and Informatics in Health (CIIS) at USP’s Ribeirão Preto Medical School (FMRP-USP), emphasizes the importance of epidemiological studies, such as those conducted by UFPel and Grupo Fleury. “They reveal a more realistic portrait of the pandemic. Today, we really have no idea of what’s happening in this country, since we only test cases that require hospitalization. The underreporting is alarming,” Alves states. “Mass testing enables monitoring the epidemic as it evolves, isolating the sick, and reducing the circulation of symptomatic and asymptomatic patients. And it can help determine the easing of quarantine measures.”