A couple of weeks ago, my dad’s workplace, an essential business, informed everyone about a confirmed COVID-19 case within the company. My dad called me and asked about the different tests available. “Should I get the nasal swab or the antibody blood test?”
My dad wasn’t alone. Many family and friends have been asking about COVID-19 testing, particularly the antibody test. Sources like the CDC, The Atlantic, and NPR have covered the basics about this. However, we should also consider a few other questions to really understand what this fuss with antibody testing is all about.
Here, I will cover some of the commonly asked questions as a refresher and then dive a bit more into the less frequently asked questions (marked with ***).
- What does a COVID-19 antibody test do?
- ***Can a COVID-19 antibody test tell me if I had the common cold instead?
- What is the purpose of a COVID-19 antibody test?
- ***How good are the available tests?
- ***Who carries out the tests?
- What do the results mean?
- Should I get one?
What does a COVID-19 antibody test actually test for?
A COVID-19 antibody test looks at a sample of blood to see if there are antibodies made against different parts or products of the virus. Our body can fight infections by making “antibodies,” which are proteins targeted specifically to a pathogen. Once an infection begins, our immune system needs a few days to a week to start making antibodies. This lag time explains why antibody testing is better suited at looking for a previous infection rather than a current one.
Antibodies usually do one of two things: 1) bind to pathogens or cells harboring pathogens to alert the immune system, or 2) neutralize pathogens directly. All of the current FDA-authorized tests only look at binding antibodies, which do not directly kill the pathogen. This is one reason why we can’t be sure that having these antibodies mean protection against the virus in the future.
Can the COVID-19 antibody test tell if I had the common cold?
If you read the fine print for an antibody testing, you will see a similar paragraph to the one highlighted above and below.
Wait, there are other coronaviruses? Why are we not worried about them?
There are seven known types of coronaviruses, conveniently named 1) 229E, 2) NL63, 3) OC43, 4) HKU1, 5) MERS-CoV, 6) SARS-CoV, and 7) SARS-CoV-2 (CDC). Coronaviruses, especially the first four strains, have been around for decades and often cause the common cold. You may recognize strain 5, 6, and 7 as the causes of the MERS, SARS, and, now, COVID-19. Because these viruses share similar characteristics, antibodies made against one of the less harmful strains may cross react with the COVID-19 antibody test to produce a false positive result. This may decrease the strength of COVID-19 antibody testing, though more validation studies will help improve that.
What is the purpose of the COVID-19 antibody testing?
A lot of people may have COVID-19 but are asymptomatic. This “asymptomatic” population makes it very difficult to assess how wide spread (or prevalence) the infection really is. Antibody testing helps assess prevalence because it can capture a much longer time period because antibodies will stick around for a while. Additionally, a patient who had a positive PCR test can also get an antibody test to aid researchers studying how our immune system fights the virus.
Countries have played with the idea of an “immunity passport” as a way to signal you are protected from future infections. However, given that we do not know enough about the protectiveness of these antibodies, antibody testing is not currently recommended as a way of handing out “immunity passports” (World Health Organization).
How good are the currently available tests?
The FDA authorized a list of antibody tests under the Emergency Use Agreement (EUA). It is important to understand that the EUA does not mean the FDA has gone through its usual stringent and broad testing of a product. These tests are approved based on a set of criteria but there is not enough time to do the usual testing process.
To understand how good or bad these tests are, we have to learn a bit of terminology. What follows is a simplified version of FDA’s explanation.
Note: if you are well versed in biostats, please skip to the table chart down below.
Let’s begin with a scenario. If I had a confirmed COVID-19 PCR test, and now I take the antibody test three weeks later, how good is this antibody test at detecting COVID-19 antibodies?
That’s where sensitivity and specificity comes in.
- Sensitivity: the rate of true positives. If you were positive for COVID-19 and made antibodies, how good is this test at detecting the antibodies?
- Specificity: the rate of true negatives. If you were negative for COVID-19 and did not make antibodies, how good is this test at detecting zero amount of COVID-19 antibodies?
Remember that other coronavirus related antibodies may exist from prior infections, so false positives can exist. Interestingly, specificity, not sensitivity, directly tells us the rate of false positives. The true negative rate is always 100%. If the specificity of a test is 80%, it means that the test can correctly identify 80% of all the true negatives. Then, the rate of false positives (or the rate of true negatives that are falsely identified as positives) is 100% — 80% = 20%.
Sensitivity and specificity are great at validating a test, since we already know the COVID-19 status of the individuals before testing them. But the real question is: if I never got tested for COVID-19, will an antibody test tell me whether I had it?
That’s where the Positive Predictive Value (PPV) and Negative Predictive Value (NPV) comes in.
- Positive Predictive Value: if the antibody test is positive, how likely are you to have had an exposure or infection?
- Negative Predictive Value: if the antibody test is negative, how likely are you to not have had an exposure or infection?
The PPV and NPV depend on one more factor: prevalence. Prevalence is how widespread the disease is in a population. You can imagine the prevalence may change from urban NYC to the rural Midwest. If the prevalence is relatively low, then the test will have a low PPV (even if it has been validated to have good sensitivity and specificity).
That last bit certainly threw me for a loop when I first learned about it. Thankfully, to make things simple, the FDA states that they assumed a prevalence of 5% when evaluating these tests.
We are almost there.
The last piece of the puzzle is the 95% confidence interval, which essentially asks the question “how sure are we of the result?” It means the true value will fall between these two values shown for 95% of the time. For example, Abbott’s Alinity test has a PPV of 84% with a confidence interval of 46.7% to 96.3%. That means that, for 95% of the time, the true PPV will fall in the range of 46.7% to 96.3%.
The smaller that range is, the more confident we are about the PPV. For example, Roche’s Elecsys test has a PPV of 96.5% with a confidence interval of 93.9% to 98.1%. The Roche’s test is likely a better test compared to Abbott’s in this case.
With that in mind, let’s look at the currently approved antibody tests:
At first glance, these tests look similar in their sensitivity and specificity. But with the confidence interval, we can get a clearer picture of the PPV and NPV. Roche’s Elecsys test has probably the best PPV and NPV with their respective confidence intervals.
Note: our bodies make a variety (or classes) of antibodies (also known as immunoglobulins), including IgG (Immunoglobuin G) and IgM (Immunoglobulin M). These are the antibodies that the tests look for. “Pan-Ig” means the test are looking for antibodies in all classes or varieties.
Now that we understand how reliable the tests are at predicting past infection, where do we go get one? As it turns out, you can’t just walk into a clinic or your doctor’s office and ask for the Abbot or Roche test.
Who carries out the test?
You probably didn’t think you had to ask this question. The companies that carry out the testing are NOT the same companies that make the tests. After some digging around, here’s what I have found so far:
- Quest Diagnostics uses tests from Abbott Labs (Architect), Euroimmun, and Ortho-Clinical Diagnostics (VITROS Immunodiagnostic). It is hard to say which one yours will be when you are getting one.
- LabCorp does not plainly list the test(s) it uses. But, based on this press release, it looks like they are using the Roche Elecsys test.
- PlushCare uses tests from either Abbott (not sure which specific one) or Euroimmun.
- ACRPoint Labs does not specify which test(s) they use. They do state that they collect the blood and then send it to third-party labs for testing.
This gets more complicated if you are getting tested at a local clinic or hospital through your primary care doctor. Hospitals and clinics often serve as a point of collection, so they may not know which tests will be run to get you your results.
The best way to find out which test you are getting :
- Ask your healthcare provider if you should get one and, if so, where to schedule for one.
- Once you know the testing center, call and ask them if they know what specific test they will be using.
Note: you still may not find out which test you are getting after the above steps, and that is a challenge currently as not all the tests have the same PPV and NPV. Hopefully this will change as testing becomes more standarized.
Bonus: if you find out which specific test it is, go to FDA’s site, click on “Information for Recipients,” and learn more about your specific test.
What do the results mean?
If you test positive, you likely had an exposure or infection of COVID-19. “Likely” because there can be false positives. Remember how there are seven different types of coronaviruses? The antibodies to those types may trigger the antibody tests and lead to a false positive.
As a reminder, because we don’t know enough about the longitudinal protective properties of these antibodies, it is difficult to confidently say if having the antibodies may confer long term protection.
Should I get one?
First, here’s a recap of what we talked about so far:
- The COVID-19 antibody tests look for antibodies made after an exposure or infection with COVID-19. It looks for “past infection” rather than “current infection.”
- Some of the seven types of coronaviruses can cause the common cold, and your body may already have antibodies against one of them. For the currently available, FDA-authorized antibody tests, there can be false positive results.
- One purpose for COVID-19 antibody testing is to help detect the prevalence (how widespread) of the infection, especially considering lots of people can be asymptomatic.
- There are lots of antibody tests out there, and their performance can vary dramatically.
- The antibody tests are made by manufacturers, but it is the individual testing company that carries out the tests.
- If you have a positive antibody test, you likely had an exposure or infection of COVID-19. However, there can be false positives. Also, due to lack of long term studies, it is difficult to know if these antibodies will protect you from future COVID-19 infections.
Keeping those things in mind, the test accomplishes two things:
- It allows epidemiologists to study the prevalence of the infection. It also allows us to have a baseline set of data to ask the question, if someone has made antibodies to COVID-19, can they get COVID-19 again in one month, six month or five years?
- It may give you a peace of mind. I have talked to people who have had mild symptoms or no symptoms but potential exposure to COVID-19. However, they couldn’t get a COVID-19 PCR test due to shortages. Now, weeks or months later, they still grapple with the question, “did I have it?” Even though the test may not be perfect, getting tested may give at least some answers. While there are false positives, the Negative Predictive Value (NPV) of these tests are pretty good.
If you are interested in getting a test, please consider all of the above points and talk to your healthcare provider. Also keep in mind that availability and insurance coverage may vary from company to company.
COVID-19 has not only caused much tragedy around the world but also focused the spotlight on science and medicine. Yet, the barrage of daily news on these topics often leave us more confused than when we first began. I hope this has provided some clarity to your questions about COVID-19 antibody testing and that you and your loved ones stay safe during this time.
Below are some useful sources for more information on this topic:
CDC Guideline on Use of Antibody Tests: Link
FDA list of antibody tests with EUA (Emergency Use Agreement): Link
“Shoul you get an antibody test?” from The Atlantic. An excellent read with more common FAQs. Link
For a list of both PCR and antibody testing around the world, please go here.