COVID-19: the HAS defines criteria to assess the reliability of serological tests

To date, the only tests recommended to diagnose COVID-19 during the acute phase are molecular tests by RT-PCR. These rely on the detection of the virus genome drawn from a nasopharyngeal sample.

Another category of tests is serological testing. These require a blood sample and allow for the detection of antibodies produced by a patient’s immune cells in response to a foreign pathogen. To detect antibodies against the SARS-CoV-2 virus, serological tests can be used approximately one week following the onset of symptoms.

The presence of antibodies establishes a patient's immune system to the virus; and thus, confirms the occurrence of the COVID-19 infection.

The first available studies provide valuable pathophysiological information on the production of anti-SARS-CoV-2 antibodies, but the data is still insufficient to precisely define the role of these tests in COVID-19 management.

Prior to establishing serological testing in France, the HAS developed specifications that detail the quality and requirement criteria to facilitate both the development and the evaluation of these tests.

A critical step in defining the screening strategy in France

Serological tests can be carried out by various methods. Automated tests, of the ELISA type for example, can only be carried out in a biomedical analysis lab. On the other hand, unit tests can in principle be used as a rapid diagnostic test (biomedical analysis performed in a lab), a rapid diagnostic orientation test (performed outside a biomedical lab, by a health professional) or as a self-testing method (carried out by the patient, without the reporting of results).

Following a Ministry of Health request, the HAS developed and validated in 5 days the specifications allowing to validate the clinical performance of all these serological tests in association with the National Reference Centers (CNR), drawing from the latest scientific and medical data available to date (and subject to daily scanning for updates). Among the evaluation criteria, the HAS retains for example: the presence of CE marking, the minimum threshold value for clinical sensitivity estimated to be between 90% and 95% depending on test usage (value represents the probability of obtaining a positive test in the presence of infection, i.e. confirm the presence of anti-SARS-CoV-2 antibodies in the blood of the COVID-19 infected patient), and even the minimum specificity threshold value clinical estimated at 98% (value represents the probability of obtaining a negative test in non-infected patients, i.e. confirm whether the test is indeed negative for absence of anti-SARS-CoV-2 antibodies in the blood of non-infected donors).

On one hand, these specifications should allow manufacturers of serological tests to have access to French requirements in terms of reliability, facilitating their development. On the other, they should allow the CNR to assess reliability according to equal technical requirements. Additional work will be carried out between the end of April and the beginning of May which will take into account the results of studies currently underway. This data will be essential for HAS to recommend the role of these tests in the management and fight against COVID-19.