Testing the accuracy of coronavirus antibody tests

If we want to safely ease social distancing requirements, eventually we're going to need to know who has had Corona virus and whether they're immune to getting it again, that's where antibody tests come in. But currently there are some questions about their accuracy. [MUSIC] Also known as serological tests, antibody tests can help identify people who have developed an immune response to Coronavirus By detecting antibodies in their blood. More than 100 tests have been brought to market since the FDA ease its restrictions to help meet the global demand, but recent data suggest that some of these tests are not as accurate as they need to be. Inaccurate antibody tests can lead to deadly missteps in treatment, information gathering, and policy decisions. That's why the COVID-19 testing project, a collaboration between researchers from the University of California Chan Zuckerberg bio hub has set out to test the accuracy of antibody tests in head to head matchups to find out which ones are the most reliable. So the way that we determine the accuracy of the test, you usually break that out into two metrics, sensitivity and specificity. So sensitivity is what percent of the time do I expected positive results and how often do I actually receive one. The other major metric is specificity, when you look on samples where you're not expected to have a positive result, to look at how often you stop False positives. And here we saw high variability in the specificity of the tests to be compared 10 different lateral flow assays or rapid serology tests, which look a lot like pregnancy tests with two different Eliza's, which are common laboratory based tests. For antibody detection, and we compare them all head to head against this common set of patient blood.>> The study looked at blood samples taken from before the Coronavirus outbreak began as well as samples from people who had tested positive for Corona virus infections and found that the specificity or how often the tests produced false positives vary greatly depending on the test Some of them were very non-specific they had false-positive rates that were over ten percent. So if you tested a hundred people it would report ten of them being anti-body positive incorrectly. On the other hand, we found three of the chest were over ninety-eight percent specific. Showing out they might potentially actually be useful. Some of these tests are already causing problems around the world. The UK reportedly spent $20 million on antibody tests, only to have Oxford researchers discovered that they weren't accurate enough to be of much use. So what's the best way to ensure that you're getting a trustworthy antibody test? Don't buy tests online, don't get them from a neighbor or a friend or a friend of a friend. Talk to your doctor, get a prescription, do your homework, understand what antibody test you're getting, how it works. The FDA recently announced that it will wrapping up oversight of antibody tests, in response to criticism and antibody test study, such as the COVID-19 testing project. And even if you get the most accurate antibody test there is it won't be much good if you don't understand what the results of that test really mean.>> A negative result does not necessarily mean that you haven't been exposed to the virus. A positive result does not necessarily mean that you have protective immunity. Just because you have antibodies doesn't mean. Necessarily that you have the right of neutralizing antibodies that might be expected to protect cells or tissues against infection by virus. And so further experiments are needed to do that. And this is a subject of really active and intense research. And we and many other scientists around the world are beginning to connect those dots between what it means to have a positive antibody results In a test and what it means to have actual protective immunity. Just because antibodies don't guarantee immunity doesn't mean the information we get from antibody tests isn't useful. antibody tests can help public health officials keep track of where the virus is being spread, especially considering a lot of people who are infected and contagious, Don't even show symptoms, and therefore likely didn't get tested during infection. People with antibodies could also help scientists understand the relationship between antibodies and immunity to coronavirus, laying the groundwork for the eventual reopening of society. This is an unprecedented time where developing the tests And understanding the underlying biology of the disease and the virus are really happening at the same time. We're going to get the answers, but we need time to get there. The COVID-19 testing project has submitted their preliminary findings for peer review, but their work is far from over. One of the things that we'd like to do is to evaluate more tests. We've also stood up in the company website at covidtesting project.org so that we can actually update in an ongoing way. The results of more tests and the performance as we evaluate them. I just want to thank all these scientists and researchers who worked with us on this study and are working hard to combat this and we're in lab or in hospital all over the world. Everyone can help everyone else by social distancing and doing what's proven to work. As always, thanks so much for watching. I'm your host, Jesse Oral. Stay safe out there, everybody.