My piece yesterday went through some of the evidence, but new information keeps flooding in:
First, researchers at Stanford gave antibody tests — which detect whether someone had COVID in the past, not just whether they’re currently infected — to more than 3,000 people in Santa Clara County, Calif. (also known as Silicon Valley). Only about 3 to 4 percent had antibodies, but this suggests the county was undercounting cases by 50 to 85 times. If we’re undercounting that badly nationwide, more than 10 percent of the country could already have had the illness and gained some immunity.
However, there are important limits here. It’s just one county, of course, and the study’s participants were not a random sample of that jurisdiction’s residents:
Our sampling strategy selected for members of Santa Clara County with access to Facebook [because that’s where they were recruited from] and a car to attend drive-through testing sites. This resulted in an overrepresentation of white women between the ages of 19 and 64, and an under-representation of Hispanic and Asian populations, relative to our community. Those imbalances were partly addressed by weighting our sample population by zip code, race, and sex to match the county. We did not account for age imbalance in our sample, and could not ascertain representativeness of SARS-CoV-2 antibodies in homeless populations. Other biases, such as bias favoring individuals in good health capable of attending our testing sites, or bias favoring those with prior COVID-like illnesses seeking antibody confirmation are also possible. The overall effect of such biases is hard to ascertain.
The possibility that high-risk individuals sought to participate in the study to find out if they’d had it seems strong to me, but we really don’t know how closely the sample approximates reality. There have been more technical criticisms as well. Nonetheless, this is a great piece of the puzzle to have, considering the sketchy nature of the other evidence available.
Meanwhile, other researchers patrolled the streets of Chelsea, Mass., asking if they could take blood samples, and holy wow:
Nearly one third of 200 Chelsea residents who gave a drop of blood to researchers on the street this week tested positive for antibodies linked to COVID-19, a startling indication of how widespread infections have been in the densely populated city.
Sixty-four residents who had a finger pricked in Bellingham Square on Tuesday and Wednesday had antibodies that the immune system makes to fight off the coronavirus, according to Massachusetts General Hospital physicians who ran the pilot study.
The 200 participants generally appeared healthy, but about half told the doctors they had had at least one symptom of COVID-19 in the past four weeks.
Of course, though, this has similar problems. It focuses on people who are out and about, while those holed up at home are (A) less likely to acquire the disease but (B) sometimes holed up precisely because they have it and don’t want to spread it. And this is a hard-hit area that doesn’t represent the rest of the country. Still, like the Stanford study, it gives us a little piece of information we didn’t have before.
The data out of Washington State are a little less promising. A local surveillance system, which stresses it “is not yet achieving a representative sample of the population in the greater Seattle and King County region,” estimates that “the average population prevalence between March 23 and April 9 was 0.24%.”
Still less promising are samples collected from hospital workers and visitors in Wuhan, who really ought to have sky-high COVID-19 rates. Just 2.5 percent tested positive for antibodies.
Lastly, as numerous folks pointed out on Twitter throughout the day, New York City now has fatalities totaling more than 0.1 percent of its entire population — 0.1 percent being an important threshold because it’s the fatality rate of the flu. In other words, even if 100 percent of New York is infected and no one dies from here on out, the virus is more dangerous than the flu. So, while the fatality rate will likely be lower than expected, it won’t be too low.
(A random thought about flu comparisons, by the way: Two diseases with the same fatality rate can kill very different numbers of people. The flu kills tens of thousands each year despite the fact that about 60 percent of kids and 40 percent of adults are vaccinated and some other people have leftover immunity from previous years. If COVID-19 has a fatality rate of 0.1 percent and two-thirds of Americans get it because no one is immune, that will entail 220,000 deaths.)
The gold standard here will be a nationally representative, random antibody test with a high participation rate. Till then, we’re stuck trying to extrapolate from less representative efforts such as these, which are still pointing in many directions at once.
Editor’s note: This piece has been emended since its original posting.