I have had a number of questions recently expressing interest in antibody testing for COVID-19 and wondering when it will be available at Santa Monica Primary Care. This is not yet a modality we are offering in the office not only because of lack of availability but also insufficient validation data. Today, the American Public Health Lab Association provided support for this approach referring to “crappy” (their word!) tests that are flooding the market. Let me explain further and try to quantify “crappy.”

First – what is an antibody test? Unlike a throat or nasal swab which indicates the presence of active infection, an antibody test can distinguish between someone who has been exposed the virus from someone who lacks immunity. Two antibody classes are typically checked: IgM which is the first antibody to appear in the response to initial exposure and IgG which is a secondary immune response that ensures long term immunity.

There are pitfalls to this testing however. For instance, an antibody test can yield a falsely negative result in an infected patient who has not yet developed an adequate response to the virus. Similarly a test may be incorrectly positive if antibody to a coronavirus other than the pandemic novel strain is present (note that there are four human coronaviruses which an estimated 15-30% of seasonal “common” colds).

While antibody tests by themselves are of limited value in immediate diagnosis, using this type of test will undoubtedly help us better understand how the immune response against the SARS-CoV-2 virus develops. We can also begin to estimate how many people in the overall population may have been infected. Lastly, these test results can aid in determining who may donate a part of their blood called convalescent plasma, which may serve as a possible treatment for those who are seriously ill from COVID-19.

Currently available antibody testing in the US are limited by the fact that they are qualitative (Yes/No answer) as opposed to quantitative (How much?) testing. Tests on the market now are akin to home pregnancy testing which are “+/-” yet do not give any information about how far along one is in a pregnancy. There is currently significant debate about the degree of immunity conferred by coronavirus exposure and how we might deem one ‘sufficiently’ immune. One massive potential pitfall of Yes/No testing is providing patients with misplaced confidence about their own level of protection – leaving them susceptible to community spread infection.

Throughout my COVID-19 posts I have repeated a theme that the US population should demand better in our pandemic response. Validation series of these new tests, pushed to the market under Emergency Use Authorization guidelines, are lacking and we should demand better. Some tests on the market have only been double checked against 50 or 60 known cases and controls. One commonly marketed test (linked below) reports a sensitivity of 88.7% (identifying a known positive case as positive) and a specificity of 90.6% (identifying a known negative case as negative). These numbers sound impressive – but compare them to the sensitivity of HIV testing which has a sensitivity of about 99.9%. The specificity of these tests is slightly lower, about 99.5%

As we shelter in place for the next month, there will no doubt be rapid improvement in antibody and serologic testing. I do not have an expectation that testing will approach the operating characteristics of HIV testing, but we should continue to demand more accurate testing if such will be employed in the population at large. For the moment, it is best to insist on something better than a ‘crappy’ test which can be sold because the FDA has done away with a requirement to submit validation data.

There is no rush – our advice for the next month will remain the same. Stay at home, avoid contact with our most susceptible populations, wear a mask when performing essential activities and wash your hands. Lets wait for and insist on better.