There is an ever-expanding group of people who have been infected by the coronavirus but are no longer symptomatic, if they ever were—the convalescents. Governor Andrew Cuomo, of New York, has advocated for widespread blood tests as a way for restarting the economy in a way that is consistent with a responsible public-health strategy. “I believe, once we get that test, you’re going to find hundreds of thousands of people who have had the coronavirus and resolved. Let the younger people go back to work. Let the recovered people go back to work.” Yet there have been a number of reports, from China and Japan, about patients who have recovered from COVID-19 but subsequently fell ill again. With the U.S. economy in a state of near-collapse, is it worth the risk to invite convalescents back to work?

With other diseases, convalescents have proved invaluable. “We’ve seen this in previous epidemics where survivors have a special role,” Dr. Seema Yasmin, a former officer in the Epidemic Intelligence Service who now teaches science journalism and clinical-communication skills at Stanford University, said. She observed the phenomenon firsthand in Liberia, in December, 2015, when she was reporting on an Ebola outbreak and talked to survivors who were caring for stricken family members. “People who have been infected but recovered have antibodies in their system that allow them to be out and about and be exposed to the virus, knowing that it’s unlikely that they’ll become reinfected,” she said, citing a 2017 study in the Journal of Infectious Diseases, which documented Ebola survivors who still had antibodies forty years after infection. “Of course, with this virus, we also have to learn how long protective antibodies will stay in your body,” Yasmin said. “It’s such a new pathogen, we haven’t had a lot of time to say, ‘Six months after infection, you still have protection.’ It’s just too soon.”

If the virus does have the capacity to sicken people who have already become ill from it, devising a vaccine might have little impact on future infections. Dr. Barney Graham, who is the deputy director of the Vaccine Research Center at the National Institute of Allergy and Infectious Diseases, points out that “even common viruses like influenza or respiratory syncytial virus can also reinfect”; these viruses, fortunately, “tend to cause progressively less disease on subsequent infections.” With COVID-19, Graham said, “survivors may develop enough immunity to keep the virus from getting deeply into the lung, confining it to the upper airways.” In that case, a repeat infection of COVID-19 would be more like a cold than pneumonia.

Recently, researchers in China exposed four rhesus macaques to the virus that causes COVID-19. All became ill. One of the four monkeys was euthanized a week later, to determine its level of infection; it had mild to moderate pneumonia. The remaining monkeys recovered, producing high levels of antibodies. By the twenty-eighth day, a viral load was undetectable in the three surviving monkeys. The monkeys were then reinfected with the virus. They each showed a brief, moderate rise in temperature and no other symptoms. Another monkey was sacrificed, and no virus was found in its tissues. No further evidence of disease was found in the two surviving monkeys. It’s a hopeful report, but very limited.

Dr. Tom Frieden, a former director of the Centers for Disease Control and Prevention who now runs Resolve to Save Lives, an initiative to prevent epidemics and cardiovascular disease, cautions COVID-19 convalescents not to think that they are totally immune from future infections. “People can go back to work, but what I don’t want is for them to assume that they are protected by that prior infection,” he told me. “Don’t assume that we have these new supermen and superwomen whom we can send out onto the battlefield, because the bullets aren’t going to penetrate their armor. We don’t know if that’s going to be the case.”

A close friend of Freiden’s, Dr. Barron Lerner, a practicing internist at Bellevue Hospital and a professor of population and health at N.Y.U. Langone Hospital, was infected with COVID-19 after being exposed to a symptomatic patient, on March 11th. Five days later, Lerner was running a fever and experiencing body aches. By March 23rd, he was feeling better, and he returned to work, although he is still suffering a lingering fatigue. “Health-care workers need to be brought back as quickly as possible,” he told me. He’s not been seeing patients directly, given that there’s an unresolved question of reinfection. He told Frieden, “I’m rooting for the immunity.” Frieden responded, “You can root for it, but don’t count on it.”

Immunity varies with different diseases. Frieden gave the example of an epidemic of measles that took place on one of the Faroe Islands, an archipelago two hundred miles north of Scotland, in 1846, before a vaccine was available. “Everyone got it, a hundred per cent, because it is one of the most infectious diseases,” Frieden said. After the outbreak, there were no cases of measles on the island for sixty-five years. “Then measles came back, and again everyone on the island was infected—except those over sixty-five. They were all still immune.”

There are also examples of viruses that don’t provide immunity after infection, or only protect against one strain of a disease and not another. A 1908 study documented multiple reinfections of smallpox during an epidemic in Trinidad at the beginning of the twentieth century. DNA viruses, like the one that causes smallpox, are relatively stable, but the singular characteristic of RNA viruses, such as influenza and coronaviruses, is that they mutate. That’s why we can get frequent colds, and why it’s so difficult for scientists to decide which strains of influenza should be targeted by the annual vaccine. Having one strain of dengue fever, for instance, doesn’t guarantee that the next one will be less virulent. “You get dengue once, and it’s not so bad,” Frieden told me. “You get it a second time, with a different strain, and you can die.” So far, SARS-CoV-2, as the COVID-19 virus is officially called, has not mutated. “But it could,” Frieden said.

According to Marc Lipsitch, the director of Harvard University’s Center for Communicable Disease Dynamics, previous studies of coronaviruses show that complete immunity “doesn’t last long,” but that subsequent infections are less severe and, probably, less contagious. “That should be our guiding hypothesis,” he told me.

There are hundreds of coronaviruses that infect bats and which might pass into the human population. The outbreak of Severe Acute Respiratory Virus, or SARS, in 2002, derived from horseshoe bats dwelling in a Chinese cave. It killed about ten per cent of the people it infected. In 2012, another coronavirus, Middle East Respiratory Syndrome, or MERS, arose in Saudi Arabia, apparently originating with Egyptian tomb bats before spreading into camels and then into humans. Among those who tested positive for the disease, it had a frightening mortality rate of more than thirty per cent, but it was slow to spread. Both of these predecessors are closely related to the virus that causes COVID-19. So far, no successful vaccine or treatment has been developed for either of these predecessor diseases. It is likely that future coronaviruses, perhaps more mortal or contagious, will emerge.