COVID-19 – antibody testing and immunity
SciLine reaches out to our network of scientific experts and poses commonly asked questions about newsworthy topics. Reporters can use these responses in news stories, with attribution to the expert.
What are Quotes from Experts?
May 6-7, 2020
What is the reliability of COVID-19 antibody tests?
“Some of the tests can measure antibodies in the blood accurately, although only four of the tests on the market have been through the FDA emergency use authorization (EUA) process. An EUA-approved test has demonstrated to the FDA that it performs to a specified standard, so an EUA test would be preferred. It takes a couple of weeks for people to make antibodies, and therefore these tests can’t be used to diagnose COVID-19. The tests can show if a person has antibodies, which may mean they have been infected. However, it’s important to remember that even an excellent lab test, with high sensitivity and specificity, has a low rate of false positives. For example, if a COVID-19 antibody test is 95% specific but the infection rate in a population is only 5%, then most of the positive results will actually be false positives. So the results have to be interpreted in the context of what we call ‘Pretest likelihood:’ how likely is it this person was infected? Were they exposed to a case, or did they have a consistent illness?” (Posted May 6, 2020)
“We remain in the infancy of antibody assay development; there’s a lot to be excited about, but the performance of the assays isn’t where they need to be for widespread roll out.” (Posted May 6, 2020)
“Reliability is unknown at this time as most tests have not been validated. I suspect there is great variation in the reliability of the various tests depending on how they are done, their manufacturer, etc.” (Posted May 6, 2020)
“We have antibody tests right now in our hospital. And we’re considering using these on all our health care workers and saying, OK, who has a positive test? The trouble is, depending on the test, there’s a whole range of reliability. We haven’t even determined how high the levels of false positivity or false negativity are. … And so we don’t know what to tell people when we test them. If it turns out to be positive, we have to say, well, it could also be false positive. Or and if they turn out to be negative, we have to say, well, it could still be false negative. And so we don’t know what to say in terms of people’s immunities. The other part is even if they are positive, we don’t yet know what immunity means.… So we can’t guarantee, given our current science, that when we give them a positive test [result], that actually means the test is actually positive and that they’re actually immune. And so even if they’re positive, we have this caveat.” (Posted May 7, 2020)
Does testing positive for COVID-19 antibodies mean you are immune to the virus?
“The tests don’t prove that a person is immune and protected. We assume that antibodies will provide protection for at least weeks to months, but that hasn’t been proven yet. Nonetheless, the tests will be valuable to know who has already been infected and is probably at lower risk.” (Posted May 6, 2020)
“There is so much about COVID-19 that we are still learning. If past is prologue, then we should expect that those who recover from COVID-19 should have a more refined immune response the next time they see the virus. The challenge with any respiratory virus is a) the virus can change frequently, resulting in versions of the virus that escape our immunity (for instance, influenza) and b) over time, the amount of antibody circulating in our bloodstream and at the surface of our nose and throat goes down – this makes us vulnerable to these types of infections at multiple times in our lives. What we don’t yet know about COVID-19 is how long immunity lasts and whether subsequent infections follow the typical pattern of being milder than the first one.” (Posted May 6, 2020)
“Not necessarily. Having antibodies does not mean one is protected. However, in the case of COVID-19 there is evidence that most people are making neutralizing antibodies, which in other diseases has been a good marker of protection. For COVID-19, we just don’t know if these make people immune as we have known this virus for only 5-6 months.” (Posted May 6, 2020)
“The biggest question is, if you have these antibodies, if I give you the virus again, will you get infected? … If I give it to you now, will you get infected? If I give it to you six months from now, will you get infected? A year from now, will you get infected? And that is the basis, the question that drives everything else. And you can’t do that study in humans because, clearly, what you don’t want to do is try to infect people again. And so the data that we currently have is actually data from animal studies, a small study that was done in four animals that showed – in nonhuman primates – that when you took these survivors, these animals who would survive the disease, and you gave them the virus again, they did not get sick. Again, a very small number, but it does give us hope that if they have these levels of antibodies that, potentially, we may see immunity.” (Posted May 7, 2020)
What are the pros and cons of using antibody status as a factor in deciding who can return to work?
“The pro would be that people who have been infected are likely to be immune for some period of time, though again this is not yet proven. In addition, immunity sometimes prevents disease symptoms, but not infection and the ability to transmit virus. The cons are that we don’t know how long they are protected, and if they are protected against infection and transmission, or just protected against disease? This last point is critical for people working in high-risk settings (hospitals, nursing homes, etc.) where asymptomatic transmission could be deadly.” (Posted May 6, 2020)
“The challenge right now is that none of the antibody assays that are commercially scalable can tell us what we need to know. They can tell us if we have antibodies to some part of the virus, but they cannot give us absolute certainty. Moreover, returning to work shouldn’t necessarily be benchmarked by who has antibody, but who has stopped shedding the virus that infected them in the first place. Therefore, testing for the actual virus is much more likely to be an important aspect of return to work plans than testing for the immune response.” (Posted May 6, 2020)
“A pro is that antibody status is something one could easily test for. A con is that the interpretation of antibody test results is not straightforward, and any answer is nuanced.” (Posted May 6, 2020)
What is herd immunity and what portion of the population needs COVID-19 antibodies in order to achieve it?
“Herd immunity refers to having enough immune individuals that a virus cannot spread rapidly and infect large numbers of people. The necessary threshold for effective herd immunity depends on the contagiousness of the virus: for example, measles is extremely contagious, and thus at least 90% of people need to be vaccinated to prevent outbreaks. Experts have estimated a herd immunity of 60-70% would probably provide benefit against COVID-19. The United States is a long way from this; even with 1 million proven cases, if those tested account for only 10% of all cases, the total number of individuals with demonstrated immunity would still be less than 10% of the US population.” (Posted May 6, 2020)
“Population immunity—or herd protection—describes the phenomenon of enough people in a community becoming immune to a germ so that when the germ is introduced, say from someone traveling from ‘outside the herd,’ the germ doesn’t take hold but meets a dead end because of immunity in most of the people. We have no way of knowing yet how many in a population have to be immune or partially immune to the virus to achieve some type of protection, but we do know that as more and more people become immune, the number of individual people that are infected for every one case of COVID-19 will go down considerably. Take chickenpox as an example. Imagine a class of 4th graders. There are 15 students in the class, all of whom have been vaccinated for chickenpox. If a new student comes to the school and happens to have chickenpox, the virus has nowhere to go because of the immunity of the classroom. No outbreak occurs, and all is well. But what If only 12 students have been vaccinated? Now that one original case will likely become 4 cases. This is why population immunity is so important – it’s the most surefire way to stop germs in their tracks.” (Posted May 6, 2020)
“This depends, but usually more than 70% of the population needs antibodies to achieve herd immunity.” (Posted May 6, 2020)
“We’re trying to obtain something called herd immunity – the concept that there are enough people in a group of people that are immune to a disease. Those who haven’t yet had the disease could keep themselves from getting infected because everybody around them has already had it and are no longer going to be infectious. And so there’s calculations around how this is measured that has to do with the transmission of the disease itself… Generally, for this disease – for COVID-19 – we’d probably want to see between 60 to 70% of the population being immune before we can say we’re comfortable. I think others may give you other numbers, but that’s the general sense.” (Posted May 7, 2020)
How long might immunity last?
“Immunity against other common respiratory viruses (human metapneumovirus (HMPV), RSV, influenza, etc.) is not lifelong, and people are infected with these viruses repeatedly throughout life. Any of us who are parents have gotten HMPV, RSV, and others from our kids. However, immunity lessens the severity of repeat infections: infants with HMPV or RSV may be sicker or even hospitalized because it’s their first infection, but older kids and young adults would just have a cold. When people are higher risk, such as those older than 65 or the immunocompromised, they are again susceptible to severe disease from these repeat infections.” (Posted May 6, 2020)
“If we extrapolate from other coronaviruses, we have some degree of immunity for 1-2 years; however, immunity is a spectrum. If someone were to say there’s going to be the global emergence of a new virus that makes you have a runny nose for 2-3 days, maybe with a low-grade fever, we would not be taking the extraordinary measures we currently are. The challenge with COVID-19 is that the case fatality rate is alarmingly high. Therefore, we are hopeful that even if it turns out that absolute immunity does not last very long, that amount of protection from severe disease would still be a cause for celebration.” (Posted May 6, 2020)
“This is unknown at this time.” (Posted May 6, 2020)
The text and video on this page are licensed as Creative Commons CC BY-SA 4.0. Journalists are free to use any text or video on this page with or without attribution to SciLine.
John Williams, MD, Professor of Pediatrics, University of Pittsburgh; Director, i4Kids; Chief, Pediatric Infectious Diseases, UPMC Children’s Hospital of Pittsburgh
I serve on the Scientific Advisory Board of Quidel, and an Independent Data Monitoring Committee for GlaxoSmithKline, neither activity involved with this topic.
C. Buddy Creech, MD, MPH, Director, Vanderbilt Vaccine Research Program, Associate Professor, Pediatric Infectious Diseases, Vanderbilt University School of Medicine and Medical Center
None that are relevant.
Arturo Casadevall MD, PhD, Chair of Molecular Microbiology & Immunology, Bloomberg Distinguished Professor, Johns Hopkins School of Public Health
Nahid Bhadelia, MD, MA, Medical Director, Special Pathogens Unit, Boston University School of Medicine