Antigen tests for Covid-19: are they reliable?

Updated:24/09/2020 16: 41h


There is no perfect test. They all have their advantages and disadvantages. We have already explained what the three types of tests consisted of: PCR, antibodies and antigens. The latter have become fashionable after the Community of Madrid announced their massive use.

The tests that allow diagnosing the infection are those based on PCR and on antigens because they directly detect the virus (the genome or its proteins). That they test positive does not always imply that the virus is active and infective: we can detect its genome or its proteins, but that the virus is not complete, only remains.


On the other hand, antibody tests detect molecules produced by our body when we are infected, so they are used to evaluate the disease. Testing positive for these does not always mean that we are infectious or that we have active disease.

The diagnosis of a disease is not based only on a microbiological test. These help, but the doctor also takes into account other clinical aspects, symptoms, and other tests.

How an antigen test works

Antigen tests confirm the presence of the virus by detect their proteins (antigens). With some difference, more or less all have the same foundation.

Specific antibodies are fixed on a support that will react against some protein of the virus. The protein from the surface of the envelope (protein S), which projects outwards, is usually used. If there are viral particles in the sample, they will be attached to the antibody. It is as if the virus or its proteins have been captured by the antibody.

Next, a second antibody against the virus is added so that a sandwich is formed: antibody-virus-antibody. This second antibody will be labeled or flagged in some way to show the reaction. If the reaction is positive, it shows that there were virus proteins. That is, the person was infected.

The advantages of this test are its speed and simplicity. It does not require expensive reagents, machines, or qualified technical personnel. They are a lot cheaper than PCR. They are usually manufactured as a pregnancy test: a sample is taken from the nose with a swab or saliva, a few drops of a reagent that extracts the virus antigens are added, it is placed in the device and waiting less than 30 minutes for the corresponding reactive bands to appear.

Its specificity (the probability that a healthy person will test negative) is similar to that of PCR. This means that the number of false positives is low. But its sensitivity (the probability that an infected person will test positive) is less than that of PCR.

This means that can give more false negatives than PCR. PCR is much more sensitive than antigen detection: while with the PCR technique we can detect one viral RNA molecule per microliter, with antigen tests we need thousands or tens of thousands of virus proteins per microliter for the result is positive.

So why do we say that this type of test can be a good diagnostic tool?

As they have a lower sensitivity than PCR, antigen tests are positive at higher concentrations of the virus and this can have its advantage. Although we do not know what viral load means that a person is infectious or not, we can assume that the higher the viral load, the greater the probability that one is contagious.

Antigen tests can be very useful early in the infection, when the viral load is highest: a few days before symptoms appear and a week after. The problem with PCR is that it is so sensitive that can remain positive for several weeks after the appearance of symptoms because it detects even remnants of the non-active, non-infectious viral genome.

We can do antigen tests much more frequently: a test (cheap and simple) that we can do twice a week, for example, is better than another (more expensive and complex such as PCR) that we do every two weeks.

The status of the infection should always be correlated with the clinical history and other diagnostic information. The interpretation of a test must always be done within a clinical context.

For example, if the antigen test is negative but the person has any symptoms, it could be combined with PCR, which is much more sensitive. Antigen testing can be a tool very useful in primary care. As they can be repeated much more easily than PCR, they can be a good alternative to monitor and follow-up in certain groups or collectives: nursing home, health centers, schools.

What I am not so clear about is whether these tests are useful for a massive screening of asymptomatic patients.

Another issue to keep in mind is that there are several companies that sell antigen tests. Although the rationale is similar, the results do not have to be the same. The tests can vary in the type of antibodies they use, the protein of the virus that they detect or the way to reveal the reaction. Sensitivity and specificity may be different between them and should first be evaluated. Let’s remember the fiasco of the famous Chinese rapid tests.

Ignacio López-Goñi. Professor of Microbiology, University of Navarra.

This article was published in The Conversation.

See them


Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.