Official case counts often substantially underestimate the number of coronavirus infections. But in new studies that test the population more broadly, the percentage of people who have been infected so far is still in the single digits.
The numbers are a fraction of the threshold known as herd immunity, at which the virus can no longer spread widely. The precise herd immunity threshold for the novel coronavirus is not yet clear; but several experts said they believed it would be higher than 60%.
Even in some of the hardest-hit cities in the world, the studies suggest, a vast majority of people still remain vulnerable to the virus.
Some countries — notably Sweden, and briefly Britain — have experimented with limited lockdowns in an effort to build up immunity in their populations. But even in these places, recent studies indicate that no more than 7%-17% of people have been infected so far.
Covid-19: Live updates
In New York City, which has had the largest coronavirus outbreak in the United States, around 20% of the city’s residents have been infected by the virus as of early May, according to a survey of people in grocery stores and community centers released by the governor’s office.
Similar surveys are underway in China, where the coronavirus first emerged, but results have not yet been reported. A study from a single hospital in the city of Wuhan found that about 10% of people seeking to go back to work had been infected with the virus.
Viewed together, the studies show herd immunity protection is unlikely to be reached “any time soon,” said Michael Mina, an epidemiologist at the Harvard TH Chan School of Public Health.
More on Covid-19
The herd immunity threshold for this new disease is still uncertain, but many epidemiologists believe it will be reached when between 60% and 80% of the population has been infected and develops resistance. A lower level of immunity in the population can slow the spread of a disease somewhat, but the herd immunity number represents the point where infections are substantially less likely to turn into large outbreaks.
“We don’t have a good way to safely build it up, to be honest, not in the short term,” Mina said. “Unless we’re going to let the virus run rampant again — but I think society has decided that is not an approach available to us.”
The new studies look for antibodies in people’s blood, proteins produced by the immune system that indicate a past infection. An advantage of this test is that it can capture people who may have been asymptomatic and didn’t know they were sick. A disadvantage is that the tests are sometimes wrong — and several studies, including a notable one in California, have been criticized for not accounting for possible inaccurate results or for not representing the whole population.
Studies that use these tests to examine a cross section of a population, often called serology surveys, are being undertaken around the country and the world.
These studies are far from perfect, said Carl Bergstrom, a professor of biology at the University of Washington. But in aggregate, he said, they give a better sense of how far the coronavirus has truly spread — and its potential for spreading further.
The herd immunity threshold may differ from place to place, depending on factors like density and social interaction, he said. But, on average, experts say it will require at least 60% immunity in the population. If the disease spreads more easily than is currently believed, the number could be higher. If there is a lot of variation in people’s likelihood of becoming infected when they are exposed, that could push the number down.
All estimates of herd immunity assume that a past infection will protect people from becoming sick a second time. There is suggestive evidence that people do achieve immunity to the coronavirus, but it is not yet certain whether that is true in all cases; how robust the immunity may be; or how long it will last.
Mina of Harvard suggested thinking about population immunity as a firebreak, slowing the spread of the disease.
If you are infected with the virus and walk into a room where everyone is susceptible to it, he said, you might infect two or three other people on average.
“On the other hand, if you go in and three out of four people are already immune, then on average you will infect one person or fewer in that room,” he said. That person in turn would be able to infect fewer new people, too. And that makes it much less likely that a large outbreak can bloom.
Even with herd immunity, some people will still get sick. “Your own risk, if exposed, is the same,” said Gypsyamber D’Souza, a professor of epidemiology at Johns Hopkins University. “You just become much less likely to be exposed.”
Diseases like measles and chickenpox, once very common among children, are now extremely rare in the United States because vaccines have helped build enough herd immunity to contain outbreaks. There is no vaccine for the coronavirus, so getting to herd immunity without a new and more effective treatment could mean many more infections and many more deaths.
If it is assumed that herd protection could be achieved at 60%, that means New York City is only one-third of the way there. And, so far, nearly 250 of every 100,000 city residents has died. New York City still has millions of residents vulnerable to catching and spreading this disease, and tens of thousands more who are at risk of dying.
Infections have not been evenly distributed throughout the population, with low-income and minority communities in the United States bearing a greater burden. On Thursday, Gov. Andrew Cuomo of New York announced that antibody testing showed that some neighborhoods in the Bronx and Brooklyn had double the infection rate of New York City in general. Those areas are already approaching the herd immunity threshold, when new outbreaks become less likely. But because they are not isolated from the city at large, where immunity rates are much lower, residents are still at risk.
In other cities, serology surveys are showing much smaller shares of people with antibodies. The quality of these studies is somewhat varied, either because the samples weren’t random or because the tests were not accurate enough. But the range of studies shows that most places would have to see 10 or more times as many illnesses — and possibly, deaths — to reach the point where an outbreak would not be able to take off.
The serology studies can also help scientists determine how deadly the virus really is. Currently, estimates for what’s called the infection fatality rate are rough. To calculate them precisely, it is important to know how many people in a place died from the virus versus how many were infected. Official case rates, which rely on testing, undercount the true extent of infections in the population. Serology helps us see the true footprint.
In New York City, where 20% of people were infected with the virus by May 2, according to antibody testing, and where more than 18,000 had died by then, the infection fatality rate appears to be around 1%.
For comparison, the infection fatality rate for influenza is estimated at 0.1%-0.2%. But the way the government estimates flu cases every year is less precise than using serology tests and tends to undercount the number of infections, skewing the fatality number higher.
But even if the fatality rates were identical, Covid-19 would be a much more dangerous disease than influenza. It has to do with the number of people who are at risk of getting sick and dying as the disease spreads.
With the flu, only about half the population is at risk of getting sick in a given flu season. Many people have some immunity already, either because they have been sick with a similar strain of flu, or because they got a flu shot that was a good match for the version of the virus they encountered that year.
That number isn’t high enough to fully reach herd immunity — and the flu still circulates every year.
Covid-19, unlike influenza, is a brand-new disease. Before this year, no one in the world had any immunity to it at all. And that means that, even if infection fatality rates were similar, it has the potential to kill many more people, as 1% percent of a large number is bigger than 1% of a smaller number.
“There aren’t 328 million Americans who are susceptible to the flu every fall at the beginning of the flu season,” said Andrew Noymer, an associate professor of public health at the University of California, Irvine. “But there are 328 million Americans who were susceptible to this when this started.”