Coronavirus Testing 101
What you need to know about the RNA and antibody tests: what they’re for, how they work, and why they’re not the answer to everything

By Amy Rogers, MD, PhD
Coronavirus testing has been top of the news for weeks. As more and faster “coronavirus tests” roll out across America, another type of laboratory test related to coronavirus is entering the conversation. In this post I explain what these two types of tests are, how they are performed, their limitations, and how they can be used.
When people say “coronavirus test”, they generally are talking about a diagnostic test to detect the presence or absence of SARS-CoV-2 (the virus) in a person. The purpose of this test is to give a simple yes or no answer to the question, am I infected? Is the virus present in my body?
The other test is for antibodies against the coronavirus. The purpose of this test is to answer the question, was I recently infected by SARS-CoV-2, whether I was aware of the infection or not?
The first test identifies people who are potentially infectious. The second test identifies people who are potentially immune. Both tests will be needed for our society to resume normal activities in a safe manner.
The basic coronavirus test is a miracle of modern medical technology, developed within an astonishingly brief period of time after SARS-CoV-2 was first discovered in December. It is not a single test but many similar tests independently developed by different countries, private testing labs, pharmaceutical companies, public labs, universities, etc. It’s important to understand that these tests have a range of accuracy and none are perfect. Given the urgent circumstances under which the tests have been developed, they have been rushed through regulatory approval, or exempted from approval under FDA Emergency Use Authorization. The tests are not 100% reliable; no medical test is. So if you’re anxious because you want a test but can’t get one, keep in mind that a coronavirus test result might not matter because it could very well be wrong. If you have a fever and cough, you should behave as if you have COVID-19 regardless of whether a test comes back negative or positive. A test result should never override common sense. As an old medical dictum says, treat the patient, not the test.
The CDC’s coronavirus test and nearly all others use the same underlying technology, called RT-PCR (reverse transcription-polymerase chain reaction) to detect viral RNA. RNA is an information-carrying molecule, similar to DNA. RNA and DNA both encode information in a sequence of four chemical building blocks, much like a language written using an alphabet with four letters. Humans store their permanent genetic library in DNA. Coronaviruses, on the other hand, use RNA. The test looks for two or three RNA sequences that are unique to SARS-CoV-2 and not found in any other coronaviruses.
RNA is less stable and harder to work with than DNA. So the first step in the coronavirus test is to convert any viral RNA in the sample into a DNA mirror image. This is called reverse transcription. In the subsequent steps, molecular probes search for the viral sequence(s) of interest. In order to detect even the tiniest amount of virus in a sample, the test then amplifies or copies the viral sequence over and over until enough material is present to detect. This amplification series is called polymerase chain reaction or PCR.
Like all medical tests, coronavirus tests give the correct answer most of the time, but not all of the time. Some people who test negative are actually infected with the virus. Some people who test positive actually are not. I’ve read many anecdotal tales of people with COVID-19 symptoms whose test came back negative, but days later as their disease worsened a re-test came back positive. One limitation of the coronavirus test is not the fault of the RT-PCR, but rather the sample we give it. Because COVID-19 is primarily an infection of the respiratory system, testing is done by swabbing organic material from a person’s nasal passages. But SARS-CoV-2 can also cause gastrointestinal sickness (diarrhea, vomiting). What if a person is carrying the virus in their gut but not in their nose? Their coronavirus swab will come back negative. Any particular swab could, at random, miss the virus.
A positive test means coronavirus RNA was present in the specimen. But how do we interpret that result? We don’t know whether that means the person is necessarily infectious. It is safest to assume that they are. We know that you can give the virus to others without having symptoms yourself. At this time we don’t know how long an asymptomatic person can do that, or how long they will continue to shed virus after recovering from COVID-19. I have seen a report of a person still testing positive thirty days after recovery but most patients seem to clear the virus sooner once their symptoms are gone. The test cannot answer these epidemiologic questions. The test cannot tell us what its answer means or what we should do with the information for a particular person.
I’ve raised these complications to emphasize that while coronavirus testing can generate extremely important information about the progression of the pandemic in a community, on an individual basis, a single test result is not the only, or even the most important, determinant of what actions you should take.
The Blood Test for Antibodies
As we grapple with the challenge of ending quarantines and resuming group interactions, a powerful tool will be serologic tests for coronavirus antibodies.
An antibody test, or serology / serologic test, does not look for the virus. Instead, it looks at a person’s immune system and asks whether they have done battle against the coronavirus. The evidence is in the immune system’s weapons: antibodies.
Parts of the human immune system learn from experience. In particular, when a new virus infects the body, B lymphocytes (a kind of white blood cell) manufacture antibodies. Antibodies are a class of proteins with one end that sticks, or binds, to a bit of the virus (the antigen). The other end of the antibody activates warrior-cells of the immune system. A staggering number of different antibodies with different antigen-binding ends are made. I’ll skip the details, but over the course of battle, the system figures out which antibodies are most effective at binding the virus. Then it remembers and keeps those particular antibodies around for the next time.
Antibodies exist in your blood, specifically in the liquid component called serum. Serum can therefore be used as medicine. If a person has lots of antibodies against a virus in their serum, you can inject that serum (the antibodies) into a person who is sick. The antibodies will bind to the virus and recruit the sick person’s immune system, possibly helping them to recover.
Antiviral antibodies in a person’s serum tell a history of what viruses the person has encountered at some point. They can even hint at how long ago the infection occurred, because one class of antibodies (IgM) is produced early, and another class (IgG) is produced later.
Convalescent antibody tests, serology tests, coronavirus antibody tests — whatever you want to call them — are being developed and deployed right now in many formats, in many places, to detect antibodies that bind to the new coronavirus. The purpose of these tests is to identify people who have recovered from a SARS-CoV-2 infection, whether they were aware of it or not.
This information is extremely useful for two things.
First, as epidemiologic data. One giant question hanging over this pandemic is, how many asymptomatic or otherwise undiagnosed infections have there been? How prevalent is the virus in the community? Knowing this would help us pin down the mortality rate. The more silent infections there are, the less deadly the virus is. Also, this provides information about herd immunity and how close we are to being able to drop all restrictions.
Second, antibody tests could identify people who are potentially safe to go back to work or into social groups. This would allow us to reactivate ordinary life while minimizing the threat of another disease spike.
But there are some major caveats to this strategy.
- Previous infection does not guarantee protective immunity.
For a variety of reasons, infection by some viruses only happens once in your life. From then on, your immune system is able to fight it off. But other viruses can attack you more than once. Your immune memory isn’t adequate to fully protect you from re-infection. Unfortunately, the “common cold” is an example of a viral infection that you can get again and again. Many “common colds” are caused by coronaviruses. We simply do not know yet whether SARS-CoV-2 is a once-and-done, or an over-and-over kind of infection. A person could have antibodies against the coronavirus but remain susceptible to infection.
- Even if natural infection does lead to protective immunity against SARS-CoV-2, we don’t know exactly what to test for.
Pretend the virus is a box of crayons. Each crayon is a potential antigen (a target for an antibody). Let’s say the virus needs red and blue to make you sick. Neutralizing the yellow, green, orange, and white crayons does not fully disable the virus. Now imagine you have a test that can detect antibodies against all the colors in the virus. You test a person’s blood and find antibodies against red, orange, and white. Clearly this person has encountered the crayon-virus. But is this person protected? In order to interpret the test result, you would have to know which colors confer protective immunity, and which ones are not good enough.
With SARS-CoV-2, we can make educated guesses but we don’t yet know for sure which antigens are important for immunity.
This is also the challenge with creating a vaccine, by the way. Which antigens (colors) should be used?
Another interesting angle is legal liability. If a serology test is used to determine whether an individual person may engage in a riskier behavior (such as, working in a hospital or grocery store), then the manufacturer of the test could conceivably be liable if the person subsequently gets COVID-19. Companies developing such tests need to be given protection from getting sued, because with the current state of knowledge, even a well-designed antibody test cannot guarantee that a person is immune to SARS-CoV-2.
Even with these elements of uncertainty, I’m confident that antibody testing will become part of our strategy in the coming weeks. Scientists are studying the serum antibodies of people who have recovered from COVID-19 to try to get the answers we need. Even if a minor bout of COVID-19 does not confer strong, long-lasting immunity, it probably provides at least some protection in the short term.