In talking about herd immunity for the novel coronavirus, Dr. Albert Ko, professor of epidemiology and medicine and department chair at the Yale School of Public Health, laid out a hypothetical.

Let's say 50% of Connecticut's roughly 3.5 million residents have to get infected to reach herd immunity: that's 1.75 million people. Assuming a global case fatality rate of 0.5%, which Ko said would vary based on a country's elderly population, 8,750 people in Connecticut would die.

University of Chicago genomics researchers Luis Barreiro and Haley Randolph wrote in a paper published earlier this month in the medical journal Immunity, "In the absence of a vaccine, building up SARS-CoV-2 herd immunity through natural infection is theoretically possible. However, there is no straightforward, ethical path to reach this goal, as the societal consequences of achieving it are devastating." SARS-CoV-2 is the scientific name for the novel coronavirus.

Assuming a worldwide herd immunity threshold of 67% and a fatality rate of 0.6%, COVID-19 deaths across the globe would exceed 30 million.

They said that in the absence of a vaccine, establishing herd immunity should not be the ultimate goal but the eventual byproduct of measures that protect the most vulnerable groups.

The Day spoke about herd immunity with Ko; Paulo Verardi, associate professor of virology and vaccinology at the University of Connecticut; Summer Johnson McGee, dean of the University of New Haven School of Health Sciences; and Michael Rajkumar, director of infectious disease at Backus Hospital in Norwich.

The four agreed on a few things: The estimated herd immunity threshold for the novel coronavirus is at least 60%-70%; we're nowhere near that number; the United States shouldn't be doing what Sweden is doing, and there are still many unknowns, including whether people infected with COVID-19 are immune to further infections.

"If we didn't take a phased approach to reopening and everybody just left their house and didn't wear a mask and didn't socially distance, within two to three weeks our hospitals and health systems would just be completely overrun and the death toll considerable," McGee said, "so I think that's a risk that nobody is willing to take."

Two infectious diseases professors wrote in a viewpoint published on the Journal of the American Medical Association website on May 22, "Even at the current pace of new COVID-19 infections in the US. with more than 25,000 confirmed cases a day, it would be well into 2021 before the herd immunity threshold would be reached. If current daily death rates continue, more than half a million US residents would have died from COVID-19 by then. Thus, a strategy relying on herd immunity in absence of a vaccine carries an enormous potential human toll."

What is herd immunity?

The authors of the paper in Immunity define herd immunity as "the indirect protection from infection conferred to susceptible individuals when a sufficiently large proportion of immune individuals exist in a population."

This can either occur via vaccinations or natural herd immunity through infections. The more infectious a disease is, the higher the percentage of the population must be immune to achieve herd immunity. This percentage is called the herd immunity threshold.

For example, measles is a highly infectious disease with which one person will infect 12 to 18 others on average. Taking the higher number, a formula for calculating the threshold shows that nearly 95% of the population needs to be immune in order to prevent outbreaks.

The 60%-70% estimate for herd immunity for COVID-19 is based on the assumption that each person with the coronavirus will infect roughly 2.5 to 3.5 others — much lower than measles, but higher than MERS, the seasonal flu, the 2009 H1N1 outbreak, Ebola or SARS.

Verardi explained that the measles virus doesn't have much variation, so getting vaccinated against one strain will pretty much protect against all measles viruses. But the flu changes a lot, meaning people who have developed antibodies against one strain can become infected with another.

Verardi said the good news is that COVID-19 doesn't seem to have much variation, and he believes a working vaccine would be effective even on virus mutations. The bad news is we don't know if people who are infected will be immune to future infections.

Antibody testing leaves many unknowns

People who knew they had the coronavirus, as well as asymptomatic carriers who didn't know they had it, can test positive for antibodies. But even then, scientists and public health officials are left wondering: What does that mean? Antibodies are proteins the body produces in response to viruses.

"Who knows how long the immunity is going to be? Is it going to be three months, is it going to be six months? Can people be re-infected? We don't know," Rajkumar said.

Verardi said we "really didn't have time to develop good antibody tests, and developing good antibody tests is a long process." One issue is there are other coronaviruses, and a test could pick up one of them instead of this one, giving people with positive results a "false sense of security" that they're protected.

Ko brought up another issue: "Immunity could do two things: One is you could still get infected but it could prevent you from getting bad disease, or it could prevent you from getting infected, and we don't know which one that is."

Verardi said developing an antibody test "that's specific (and) very sensitive," and then doing randomized mass testing would give a better sense of when natural herd immunity might occur in the absence of a vaccine. But we're not there yet."

Ko said Connecticut hasn't done antibody test sampling of the general population, but Yale New Haven Hospital found that 3% of people who gave birth had antibodies and unpublished data from Yale show 10% of health care workers have antibodies.

Ko is co-chair of the Reopen Connecticut Advisory Group, which has recommended the state do a seroprevalence survey —  which evaluates the number of people in a population with a certain disease — with a random sample of Connecticut residents, but Ko noted that still won't tell us what protection antibodies give.

What about other cities and countries?

Other states, and cities in other countries, have conducted such surveys.

Individual study results released between April 20 and May 21 show that the estimated percentage of people with antibodies is 19.9% in New York City, 17.5% in London, 11.3% in Madrid, 10% in Wuhan, 9.9% in Boston and 7.1% in Barcelona, The New York Times reported Thursday.

The Colorado School of Public Health estimated that 2.9% of Coloradans have had COVID-19, the Colorado Springs Gazette reported. Testing done May 8-12 in Los Angeles showed only 2.1% of the 1,014 people tested there have antibodies, CBS Los Angeles reported.

What about Sweden, where schools and restaurants have remained open, and which some have held up as a model for the U.S.?

In late April, the chief epidemiologist of Sweden's Public Health Agency, Anders Tegnell, said sampling and modeling data indicated 20% of Stockholm's population is immune to the coronavirus. He estimated at the time the city might reach herd immunity "in a few weeks' time" but then told NPR on May 25 that won't happen, and immunity is likely lower than 30%.

Survey results from the Public Health Agency published May 20 show that just 7.3% of blood samples from people in Stockholm tested positive for antibodies.

According to the Coronavirus Resource Center at Johns Hopkins University, Sweden is seeing 42.72 deaths per 100,000 people, much higher than its neighbors with similar welfare systems but stricter lockdowns — 5.69 in Finland and 4.44 in Norway.

Sweden's deaths per 100,000 is lower than that of Belgium, Spain, the United Kingdom and Italy but higher than the 31.42 of the United States. This comes despite a low population density in Sweden that lends itself to natural social distancing: The second-largest city of Gothenburg has a population density slightly larger than that of Boise, Idaho, the podcast Science Vs. noted.

"I think the approach in Sweden has allowed for their economy to remain open, but it's come at a significant cost," McGee said, citing high death rates among the elderly. In the U.S., the coronavirus also has disproportionately hit communities of color.

McGee also noted that this "was a much more reasonable gamble" in Sweden than it would be in the U.S., because of the country's comprehensive national health system and better way to track people.

Another difference is more trust in government: Karin Olofsdotter, Sweden's ambassador to the U.S., told NPR "there's a fairly big trust between the population and the government" in Sweden, which isn't exactly the case here.

Verardi said he is not at all in favor of the Swedish model, because we're a different country with a different health care system, "so for me that will be criminal in the sense that it would just lead to too much mortality." Until we have a vaccine, he thinks maintaining suppression strategies like social distancing is the only way.

According to 2016-18 data from the Organisation for Economic Co-operation and Development, Sweden and the U.S. are similar when it comes to the number of hospital beds per 100,000: 2.2 and 2.8, respectively. Both are in the bottom third of the 42 countries OECD reported.

"Increasing critical care capacity allows population immunity to be accumulated more rapidly, reducing the overall duration of the pandemic and the total length of social distancing measures," five Harvard T.H. School of Public Health researchers wrote in a report published May 22 in the journal Science.

They said under current critical care capacities, the pandemic could last into 2022, requiring social distancing measures to be in place for 25%-75% of the time, depending on seasonality.

One-time social distancing efforts may push the pandemic peak into autumn, which could exacerbate the load on critical care resources if the spread of the disease increases in the winter, they wrote. "Intermittent social distancing might maintain critical care demand within current thresholds, but widespread surveillance will be required to time the distancing measures correctly and avoid overshooting critical care capacity."

e.moser@theday.com