Widespread use of antigen tests? Nice idea, questionable effect

To detect the SARS-CoV-2 virus infection, antigen as well as PCR testing is increasingly used, often on a population base. This idea is a priori not bad, but especially in case of a population based testing (e.g. in a population of schoolchildren) the benefits of such a testing are at least questionable. It is not easy to determine sensitivity of the antigen test as we must take into account the fact that the test must be calibrated on a viral load being specified in advance which does not happen. Interpretation of the test result then also differs depending on whether we are testing a patient being indicated on the basis of clinical symptoms or whether we are performing a population based screening. Interpretation of both positive and negative test result differs significantly in those two situations as it is necessary to take into account also so called pre-test probability of infection. When considering the population based testing, it is quite important to use a test with almost 100% specificity as this parameter being even mildly lower than 100% may lead to considerable problems. At the end of the paper, we identify situations when the use of the antigen testing makes sense also outside the scope of testing clinically indicated individuals.