The potential flaws in New York’s antibody testing

Andrew Cuomo Northwell Health lab
Andrew Cuomo Northwell Health lab
Kevin P. Coughlin / Office of Governor Andrew M. Cuomo
Gov. Andrew Cuomo touring the Northwell Health Core Lab in New Hyde Park .

The potential flaws in New York’s antibody testing

The data wasn't randomized, and there may be false positives.
April 26, 2020

Nearly 14% of those tested at grocery stores around New York were found to have antibodies to the new coronavirus, according to preliminary results of a statewide coronavirus antibody survey announced on Thursday. In New York City, 1 in 5 tested positive for the antibodies.

But the testing wasn’t perfect. Cuomo acknowledged that the sample wasn’t totally randomized. Since health workers performed tests on volunteers outside grocery stores, the sample consisted of people willing to leave their houses to shop and likely excluded many essential workers who were working when the tests were administered. 

What’s more, the tests might have false positives for people who simply had a cold. Tests on the market, even those with emergency authorization from the U.S. Food and Drug Administration, warn of positive results due to reactivity with non-COVID-19 antibodies, like those from other coronaviruses that cause the common cold. The state test is no exception. According to a state Department of Health fact sheet, a positive antibody test “may be due to past or present infection” with a different coronavirus strain. Yet it also says that this test, developed by the Wadsworth Center in Albany, has 93-100% specificity, so false positives are expected. On Friday, state Health Commissioner Howard Zucker said the state antibody test has “incredible sensitivity and specificity” when asked about its reliability. 

Technically, this test has not received emergency authorization from the FDA. A Department of Health spokesperson said it went through state regulatory approval with the go-ahead from the FDA for the alternate route. But formal federal authorization would have included the public disclosure of trial data to back up claims of accuracy, which does not appear to be on the Department of Health’s website where it provides information about antibody testing.

Rebecca C. Lewis
is a staff reporter at City & State.