In times of high numbers of infections, many people rely on rapid antigen tests. But the differences in quality are large. In an interview, Gunnar Schwan from Stiftung Warentest explains what consumers should pay attention to – and warns against useless tests.
Mr. Schwan, the number of cases with the corona virus is currently increasing again. There are many rapid tests on the market. How do consumers recognize which are good?
The Paul-Ehrlich-Institut (PEI) has tested many of these tests and determined how high the measurement accuracy is – i.e. how reliably the rapid tests detect an infection. The Federal Institute for Drugs and Medical Devices (BfArM) has supplemented this data. However, the lists were difficult for consumers to read. We at Stiftung Warentest have therefore prepared the results so that they can be filtered according to various criteria. It contains the results for 61 corona self-tests. 18 tests have yet to be checked by the PEI, the other 43 are classified as “suitable”. They meet the minimum criterion required by the PEI: a measurement accuracy of at least 75 percent with a high viral load.
What does this value mean?
The number results from a comparison of the rapid test with the PCR analysis. PCR is the best known method to detect infection. The PEI examiners compared the results of these tests with those of the rapid tests and examined whether the rapid tests had a similar effect. The measurement accuracy is also referred to as sensitivity. A value of 75 percent means that a rapid test detects three out of four infected people with a high viral load who are also positive according to the PCR. In this case, an infected person would fall through the cracks.
Three out of four? That sounds bad.
Yes, the rate is quite weak. However, this is only the minimum criterion. There are many tests that score higher.
The database contains various search criteria. Which setting makes the most sense?
We recommend the filter “high measurement accuracy across all virus concentrations”. After all, consumers do not know how high their viral load is at the time of the test. Eleven self-tests remain. These can then be further sorted, for example according to the place where the sample is taken, i.e. in the nose or in the mouth or throat.
Do you have any other advice?
After setting the filters, I would always use the test with the greatest measurement accuracy. The best test in the database achieves a sensitivity of 90 percent based on the measurement accuracy across all virus concentrations.
In other words, nine out of ten infections are detected. There’s room for improvement there too, isn’t there?
Absolutely. A single negative result should therefore not be taken as a free pass.
How can security be increased?
Through a PCR test or through repeated testing, for example the following day. This increases the chance that during the course of the virus infection the point may be reached at which the viral load is high enough to detect the infection with a self-test.
Assuming the tests are any good.
A test should at least meet the minimum criterion of 75 percent with a high viral load. Numerous tests in the PEI evaluation did not make it. Some even had a measurement accuracy of zero percent and were therefore unable to detect the virus at all. Anyone who gets to such a quick test has bad cards. Repeated testing doesn’t help either. Even a coin toss would have more significance. There is at least a 50/50 chance.
Are such tests still sold?
That’s to be expected, yes. Although the providers have to vouch for the quality, it usually takes quite a while for the tests to really disappear from the market – even if it is only due to a simple market shakeout because they are no longer bought.
Can rapid tests also detect the omicron variant of the coronavirus?
Many rapid tests detect the corona virus via the so-called nucleo protein (N), not via the spike protein, which is heavily modified in omicron. That is why many rapid tests are basically also able to detect Omicron. Alone: How well they actually do this is currently being investigated by the Paul Ehrlich Institute.
We hear again and again that rapid tests, the sample of which is taken in the throat instead of in the nose, become positive more quickly. An infection can therefore be indicated earlier. Is that true?
In principle, consumers should proceed as described in the instructions for use when testing, because incorrect use can affect the result of the rapid test. Anyone who gets along better with a throat swab can of course use a test that is to be used accordingly. There doesn’t appear to be much qualitative difference between tests sampled from the nose or throat. Both test variants are represented in our list with high measurement accuracy for all virus concentrations.
What should I do if I have symptoms of a respiratory disease?
In order to avoid contagion, it is best to stay at home and cure the infection – especially if you do not know for sure whether you are dealing with an infection with the corona virus. A doctor should be consulted if symptoms are severe or persistent.
To the freely accessible self-test database of Stiftung Warentest
Source: Stern
