Researchers Applaud Spanish COVID-19 Serological Survey

See “How (Not) to Do an Antibody Survey for SARS-CoV-2”

The Spanish study, on the other hand, earned the praise of epidemiologists and virologists who say that both the test and the design of the survey were successful. Chris Sempos, a nutritional epidemiologist who works with the National Institutes of Health and recently published a preprint showing how data from serology tests can be analyzed to produce the most accurate estimates, says, “Everyone’s trying to do a good job, and they don’t have much time right now. . . . Overall, I was very impressed with the Spanish survey.” He adds that, based on the accuracy of the tests, it’s possible that the true prevalence of SARS-CoV-2 exposure in Spain is around 6 percent instead of 5 percent, but he “wouldn’t say that’s qualitatively different.”

False starts

This is not the first time Spain has attempted to make use of an antibody test to reveal the extent of the spread of SARS-Cov-2 through its population. At the end of March, the government returned tests it had purchased from a company called BioEasy that scientists found were only able to detect 30 percent of cases—70 percent of positive controls were categorized as negative. BioEasy sent replacement tests, but the country also deemed them to be inaccurate, and sent back the entire order of 640,000 tests near the end of April. It’s not clear how much money the government spent on the tests, but according to the newspaper El Pais, “the process has begun to recover the money that was paid for the kits.”

As companies raced to provide tests, regulatory agencies, including the US Food and Drug Administration, have relied on manufacturers to measure the validity of their own products. After the early mishaps in Spain and concerns about tests being used in other European countries, China cracked down and has approved a limited number of tests for export to other countries. The US also moved to require emergency use authorizations for antibody tests.

The test used in Spain’s latest survey, manufactured by Zhejiang Orient Gene Biotech, did not appear to be approved for export from China, according to NBC News, however, it has received a CE mark in Europe, indicating the product’s safety.

Caryn Bern, an epidemiologist at the University of California, San Francisco, who is working on the COVID-19 Testing Project to evaluate a variety of COVID19 antibody tests, says that she “wouldn’t be surprised if it’s just bureaucratic red tape,” or a backlog of approvals in China, that explains the lack of export approval. “This seems to be a good test, if you ask me.”

Accuracy of the test

The company manufacturing the test reached out to a research group in Sweden to verify its accuracy. A small study by researchers at Uppsala University found that the test’s sensitivity for IgG—an antibody that tends to show up a week or more after infection—was 93.1 percent and its specificity, or how accurately it could confirm a lack of IgG antibodies for SARS-Cov-2, was 99.2 percent. While the test can also recognize another antibody, IgM, the Spanish survey focused on IgG.

“We were then lucky to be asked to evaluate the rapid test Zhejiang Orient Gene/Healgen, and even more lucky that this test performed so well as shown by our published results.” Åke Lundkvist, a virologist at Uppsala University and the senior author of the study, tells The Scientist in an email. “The Dutch Public Health Agency in Amsterdam and the Erasmus University in Rotterdam have later made an evaluation of a number of Rapid tests and some ELISAs, and they confirmed the very good performance of the Zhejiang Orient Gene/Healgen Rapid test.”

After measuring the test’s validity themselves, The Carlos III Institute of Health and the National Statistics Institute in Spain recruited 60,897 volunteers willing to take the rapid test, which requires a quick finger prick, and answer a series of questions about whether they’d had symptoms or exposure to others with COVID-19. To assess the accuracy of the survey, the team collected additional venous blood samples from the arms of nearly 90 percent of the participants to be analyzed in the laboratory using a more precise automated assay. At the time of the report, the center had analyzed  16,953 of the blood samples and found that 97.3 percent of these results matched the results from the rapid test.

“I'm extremely impressed by how robust the epidemiology and coordination are for the Spanish National Sero survey,” says Bern. “I think that the way they went about sampling and validating the rapid test were exactly what I would have recommended.”