Editor’s Note: The Atlantic is making vital coverage of the coronavirus available to all readers. Find the collection here.
For months, the lack of testing for the novel coronavirus in the United States has allowed, ironically, a small glimmer of hope: The official number of COVID-19 cases, currently 957,875, is almost certainly too low.
If people who want to get tested can’t get tested, and if people who are asymptomatic never think to get tested, then it stands to reason that many more Americans must be immune to COVID-19 than presently known. The only way to find the true number is by looking for immunity, which requires testing for proteins called antibodies in the blood that are evidence of past infection. Such antibody tests have been seen as key to reopening the country; other countries have even proposed using them for “immunity certificates” that would allow the immune to return to work.
But if the first results from antibody surveys, also known as “serosurveys,” in the U.S. are anything to judge by, simply not enough people are immune. Too many Americans are still vulnerable to COVID-19 infection for these tests to be the “game changer” that many were hoping for.
A pair of controversial surveys in the Bay Area and Los Angeles County found antibodies in 2.5 to 4 percent of the population—and even those numbers may be overestimates due to methodological flaws. In New York City, the country’s COVID-19 epicenter, 24.7 percent of people tested positive for antibodies. (The statewide number is 14.9 percent.) These rates do translate to many times more cases than officially documented, to be clear, but they are still a far cry from the 70 percent scientists believe is necessary to reach herd immunity and stop disease transmission. And if only a small fraction of the population can return to work without fear of getting the coronavirus, a return to something resembling normal is still a long way off.
“I think there was a lot of hope that we would do the antibody testing or do the serosurveys and then we would see there was a huge amount of immunity built up in the population,” Natalie Dean, a biostatistician at the University of Florida, told me. But earlier data from outside the U.S. have suggested otherwise. One study in the hard-hit municipality of Gangelt, considered “Germany’s Wuhan,” found 14 percent of people testing positive for antibodies. “It was clear from that point, for me, that we weren’t going to see big numbers,” Dean said.
The idea of reopening the country based on antibody tests runs into the technical limits of the tests themselves. Scientists don’t know exactly how long immunity to COVID-19 will last and what level of antibodies confers immunity. “There are just unknowns for an individual,” Gigi Gronvall, a senior scholar at the Johns Hopkins Center for Health Security and a co-author of a recent report on antibody testing, told me. “The ‘get out of jail free’ card we’re hoping for, it’s just not how it should be sold,” she said. The World Health Organization on Friday warned that issuing “immunity certificates” based on antibody testing would be premature.
Meanwhile, a flood of new test kits with varying rates of accuracy is now hitting the market. In the U.K., for example, Prime Minister Boris Johnson touted finger-prick antibody tests as a “game changer,” only for the government to realize that the 3.5 million tests it bought from China were not reliable enough to use. Of particular concern here is the false-positive rate: If the prevalence of COVID-19 is quite low in the population—say, 5 percent—and a test can identify people who are truly negative with 95 percent reliability, half of the “positives” it returns will be false positives. In other words, half of the people the test says have antibodies wouldn’t actually have them. “I wouldn’t want to tell a nurse or physician ‘Go back to work’ based on that,” Michael Osterholm, the director of the Center for Infectious Disease Research and Policy at the University of Minnesota, told me.
Moreover, Osterholm said, antibody tests don’t give a snapshot of the present. It can take two weeks for a patient to develop a detectable amount of antibodies in their blood, so antibody surveys are necessarily backward-looking. But when public-health officials are deciding whether schools or businesses are safe to reopen, the key piece of information is the number of people currently infected. “I want to know what is happening now,” Osterholm said, “and antibody testing will not get that to you.”
The key strategies for stopping the disease are still the same ones experts have been promoting from the beginning: testing, contact tracing, isolating for those who test positive for COVID-19, and social distancing for everyone else. “There’s going to have to be some level of new normal for a while,” Dean said. Ongoing antibody surveys will help clarify the true scope of the pandemic and the true proportion of asymptomatic carriers, and those data can indeed help inform public-health decisions. But as far as antibody testing goes, Gronvall said, “it’s not the silver bullet for everything.”
Current antibody surveys are revealing, furthermore, that immunity to COVID-19 can vary widely from location to location. Take New York City, which had a relatively high antibody positive rate of 24.7 percent, while the rest of New York State was at just 3.2 percent. The pandemic may be global but, as Yonatan Grad, an immunologist at Harvard University, told me, “it is made up of hyperlocal epidemics that are differentially impacting communities.” If neighboring cities, states, or countries are at very different points in their outbreak trajectory, it could create difficult questions about when and how to reopen. The places that have best succeeded in stopping COVID-19 will be the ones most vulnerable to infections in the future. Singapore, for example, succeeded in containing the virus early on, only to see a huge surge of cases in March and April. “At some point,” Grad said, “we’re going to need to think about How do we all get to the same place?”
We want to hear what you think about this article. Submit a letter to the editor or write to email@example.com.