Blood tests for COVID-19 antibodies ramping up with reopening plans
Answers to the monthslong mystery of who may have secretly carried the coronavirus could be solved through blood tests by the state and private testing firms, but questions remain of just how accurate the profile may be.
New York State’s antibody tests, which are being conducted by the state Department of Health’s Wadsworth Center, are providing the largest sample to date, with more than 8,000 tested and results showing infection rates in some areas exceeding 20%.
The state's tests are called microsphere immunoassays, which are designed to detect specified antibodies in blood.
Blood is collected using a dried-blood spot card from a finger prick, then sent to the Wadsworth Center in Albany for testing.
The specific IgG antibodies the test detects usually develop three to four weeks after infection, though it’s still uncertain whether COVID-19-produced antibodies are produced at the same rate and are long lasting.
Those who receive the test get one of three results: reactive, nonreactive and indeterminate. A reactive result can mean a person either had a COVID-19 infection in the past or it can mean a subject is still infected, the state said.
Nonreactive results means no antibodies were detected, but doesn’t necessarily mean a person is not infected; a nasal-swab diagnostic test would be needed to make sure, the state said.
The state emphasizes that it’s still uncertain that those with antibodies are immune to the virus. That won’t be known until they are exposed again, and studies can be done on whether those people were reinfected.
“The million-dollar question is: What do these antibodies mean?” said Deborah Schron, medical director of Northwell Health Labs. “We generally feel that antibodies at least offer temporary or seasonal immunity, though we can’t prove that at this point. The virus has only been around a maximum of six months, and all this testing is brand new.”
Running parallel to antibody surveys by New York State is a nationwide study by the National Institutes of Health and local testing by Long Island hospital systems.
The NIH “serosurvey” will analyze blood samples from as many as 10,000 volunteers in a bid to reveal how widely the new coronavirus has spread around the country.
Meanwhile, Northwell Health is rolling out a testing program for the 70,000 employees of its health care system.
Schron said that Northwell’s antibody survey would ramp up to 10,000 tests per day and would begin at hospitals hardest hit by the outbreak, such as North Shore University Hospital in Manhasset, Long Island Jewish Medical Center in New Hyde Park and LIJ Forest Hills.
Dr. Elliott Bennett-Guerrero, vice chair for clinical research at Stony Brook Medicine’s department of anesthesiology, said that health care system is running a 1,500-person antibody study of health care workers and people who live with them.
Bennett-Guerrero said antibody testing plays a key role in understanding the virus by giving a clearer picture of how many people have been infected.
“The only way you know the true mortality rate is to know the total number of people infected,” he said.
Isaac Weisfuse, a medical epidemiologist at Cornell University and former deputy commissioner of health for New York City, urged a level of caution when drawing conclusions from the state antibody tests thus far.
“We shouldn’t run away with the numbers and tell ourselves they are an exact estimate of how many people are infected,” he said.
“You need to gather more data about particular populations and put the snapshots together to determine what’s going on," Weisfuse said. "Until then, I’m a little nervous about opening up. There are still a lot of high-risk, susceptible people out there."
Two conclusions that can be drawn from the tests, Weisfuse said, are that the virus has been "transmitted a lot" and that mortality rates may not be as high as originally estimated. Cuomo has said the death rate from COVID-19 evidenced by the antibody test is around 0.5%, much lower than early estimates that put it in single digits, though higher than that of the flu, which is around 0.1%.
What’s needed now, he said, are many more snapshots. “We need to test a wider array of populations” than those at grocery stores in particular neighborhoods, he said.
The idea of using the data to issue “immunity passports” to those with the antibodies, as is being contemplated elsewhere in the world, is too soon, Weisfuse said.
“It’s an attractive idea, but it may be premature to use for that particular use” because of unknowns about the specific testing methods and the need for a larger sample, he said.
The results also suggest that it’s premature to believe herd immunity has been achieved in a way that would help slow or stop virus transmission within a population. Infections would need to be closer to 60%-75% percent, Weisfuse said, to expect it.
Jeffrey Shaman, an epidemiologist at Columbia University who is modeling the virus, in an email said the state’s antibody tests present “many opportunities for bias.”
That’s particularly true, he said, “if the specificity of the test is lower than advertised (even just a small amount); if true infection numbers are low; if lags in the system aren’t accounted for, and if not everyone develops antibodies.”
Specificity levels tell how accurate the tests are in identifying specific viral antibodies — important to make sure similar but non-COVID antibodies aren’t being wrongly included. The state says its COVID antibody test has a 93%-100% specificity level.
Still, Shaman noted that at present there are around 165,000 infections in New York City and 299,700 statewide. “If only one in 10 or one in 12 infections is confirmed, which is what we estimate from our models,” then the antibody test conclusion that there are 1.45 million to 1.75 million infections in the city “seems ballpark reasonable,” Shaman said.