Many people who have been ill with covid-19 think that a positive antibody response can be used to draw conclusions about how to behave in different situations.

"Maybe I can visit my grandmother this Christmas anyway now that I have antibodies?" 

But according to an article published in Trends in Microbiology that summarizes the state of knowledge about antibody testing, they should not be used as a basis for this type of decision-making. 

The reason is that the tests available on the market are not of good enough quality, which leads to incorrect results.

This is a view shared by Sophia Hober, professor of molecular biotechnology at KTH. 

- Some of these quick tests are really lousy.

The tests can show that an individual has had sars-cov-2 infection, even though the person does not have it - and vice versa, according to Sophia Hober.

However, there are tests that are more reliable, says Sophia Hober, who explains that the tests she and her colleagues use measure antibodies against two parts of the virus - the nail protein and the nucleocapsid - to ensure that the result is correct. 

- Those who have antibodies to both these parts of the virus in sufficiently high amounts are those who have been ill, says Sophia Hober.

Wrong answer from the county council's test

Several of the tests that can be ordered from 1177 only measure antibodies against the nucleocapsid, says Sophia Hober.

Many people have antibodies that recognize parts of the virus even though they have not had the infection, but then not at such high levels.

To ensure that you do not report someone as positive that is not really the case, you therefore require a higher "signal" to report a response as positive.

And then there are some positives that fall under this "signal".

- We have seen several cases in our studies where people have had very clear antibodies, but which have received negative responses from methods used by the county council, says Sophia Hober. 

But the biggest risk is when a person thinks they have been ill after an antibody test, even though they do not have it.

- If the specificity of a test is 95 percent, five out of 100 will believe that they have antibodies even though they do not.

And for them, it can be dangerous.

They may meet the sick and become infected.

The antibodies fade over time

The earliest time a person should be tested for antibodies is 16 days after the illness, says Sophia Hober.

And after four months, almost 20 percent no longer have high enough levels of antibodies to be classified as positive. 

- If you take into account that the amount of antibodies decreases over time, these measurements will thus be quite uncertain. 

The researchers in the study, on the other hand, point out that antibody tests should be used by researchers and authorities to be able to see curves of deaths, increases and decreases in frequency and prevalence of infections at population level.

At the individual level, however, a positive response can rock people into a sense of false security.