Stan Grossfeld / The Boston Globe by Pictures of Getty
Mounting evidence suggests the coronavirus is less common and less deadly than it first appeared.
The evidence comes from tests that detect coronavirus antibodies in a person’s blood, rather than the virus itself.
The tests found many people in the United States were infected, but never became seriously ill. And when these serious infections are included in the coronavirus statistics, the virus appears less dangerous.
“The current best estimates for the risk of death by infection are between 0.5% and 1%,” says Caitlin Rivers, an epidemiologist at the Johns Hopkins Center for Health Security.
This is in contrast to death rates of 5% or more based on calculations that include only those who are sick enough to be diagnosed with tests that have detected the presence of a virus in a person’s body.
And revised estimates support an early prediction by Dr. Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases and a leading member of the White House Coronavirus Task Force. In an editorial published in late March on New England Journal of Medicine, Fauci and colleagues wrote that the death rate for COVID-19 “could be less than 1%.”
But even a virus with a death rate of less than 1% presents a serious threat, Rivers said. “That’s many times more deadly than seasonal influenza,” he said.
The new evidence comes from places like Indiana, which completed the first phase of a massive test effort early in May.
The Indiana program began shortly after coronavirus cases began to appear in the state. The governor’s office contacted Nir Menachemi, who heads the health policy and management department at Indiana University’s Richard M. Fairbanks School of Public Health.
The governor wants basic information, such as how many people are infected, and how many are going to die.
At the time, “it was really hard to know for sure,” Menachemi said. “And frankly, not just in our state, but in any state.”
That’s because health officials only know about people who are sick enough to be tested for the virus. And that number can be misleading, Menachemi says.
“It does not capture the vast number of people there who may be infected but are not seeking medical care,” he said.
So beginning in late April, Menachemi, working with the Indiana State Department of Health, led a study of more than 4,600 people statewide. Most are selected at random.
Participants received two trials. The first is the standard test that looks for the virus itself. It shows if you have an active infection. The second is a test that looks for antibodies to the virus in a person’s blood. It saw people infected but recovering.
Preliminary results show that coronavirus is infected by about 3% of the state’s population, or 188,000 people.
“That’s 188,000 people who represent about 11 times more people than the conventional selective test has met in the state up to that point,” Menachemi said.
And 45% of infected people reported no symptoms.
For Menachemi and his team, it was like finally seeing the whole glimpse of the coronavirus iceberg, rather than just above the water.
And the data allows them to calculate something called the death rate of the infection – the odds of an infected person dying. In the past, scientists have relied on what is known as a case-by-case study, which calculates the likelihood of dying someone with symptoms.
The death rate of infections in Indiana has been about 0.58%, or approximately one death for every 172 people infected.
And the results in Indiana are similar to those proposed by antibody studies in many other areas. In New York, for example, a state-of-the-art antibody study has an infection mortality rate of around 0.5%.
Studies in Florida and California have suggested even lower death rates, but the results are less certain, Rivers said.
“They may have recorded people who were more likely to be infected than would be fine,” he said, which would lead to an excessive amount of infections and a small amount of the death rate of the infection.
An antibody study in Santa Clara County, Calif., Used ads on Facebook to find participants – a tactic that is unlikely to attract a random sample. Also antibody studies are becoming more accurate when conducted in areas where the prevalence of infections is low.
Calculating infection rates in the United States is extremely helpful for researchers, but less so for individuals infected, Rivers says.
“Fortunately, children and adolescents are at risk of serious illness and death,” he said. “But older adults are at high risk.”
Studies suggest a healthy probability of a teenager dying from an infection is less than 1 in 1,000. But for someone in poor health in their 90s, it can be over 1 in 10.
And that means different states in the United States should expect different rates of infection, says Juliette Unwin, a fellow researcher at Imperial College, London.
“In areas such as Maine and Florida, we have found that the fatality ratio of infection is higher than in other areas where the demographic is younger,” he said.
Unwin is part of a large team in the U.K. tracking both infection and death from coronavirus in the United States. The team sets the infection death rate for the U.S. somewhere between 0.7% and 1.2%.
“It is subject to change and change, just like everything else in science,” said Samir Bhatt, a senior lecturer at Imperial College London. “But I don’t think we have an order of magnitude out.”
In order to get a more accurate estimate of infections and the rate of infection deaths nationwide, the National Institutes of Health has launched an antibody study that will test 10,000 people. The results will be released on a circular basis. The study is expected to wrap up by early 2022.