Recent articles in the media have once again questioned the reliability of the serological tests bought by the government, in particular the DiaSorin test. The FAMHP is clear: on the basis of current scientific knowledge, this test meets all requirements and can be used without any problems.
The DiaSorin SARS-CoV-2 S1/S2 IgG test received a first positive validation from a Belgian laboratory (HIS-IZZ) at the end of April 2020 and was therefore added to the list of serological tests recommended by the FAMHP and Sciensano.
The criticism from a second Belgian laboratory (AZ Delta) about the performance (sensitivity) of this test was taken seriously by the FAMHP and automatically led to a thorough analysis of the raw data from both laboratories by the FAMHP inspection services. The criticisms from the second laboratory proved unfounded: some of the data from this laboratory were based on incomplete controls, incorrect device calibrations and the use of expired reagents, which resulted in unreliable data. There was therefore no reason to remove the DiaSorin test from the list of recommended tests.
In addition, a third independent Belgian laboratory (CHU Liège) carried out the validation tests again in May-June 2020 at the request of the FAMHP and confirmed the initial validation data. The DiaSorin test therefore remained on the list of recommended tests for anti-SARS-CoV-2 antibodies.
The FAMHP was mandated by the Belgian government in April 2020 to acquire a sufficient stock of serological tests. The DiaSorin test was one of the first serological tests available on the market that could be delivered in sufficient number. The test was therefore bought by the Belgian government at the end of April 2020. Tests have also been bought from other manufacturers, including Abbott and Ortho Clinical Diagnostics.
The purchase procedure for the DiaSorin test was then examined by Federal Internal Audit. The audit concluded that there were no indications of legal or deontological breaches aimed at illegally favouring an FAMHP employee or third parties, either when buying or granting the validation of the DiaSorin test. The Federal Internal Audit also underlined the exceptional circumstances of the COVID-19 crisis and the good faith of the Director General Hugues Malonne.
Reference is made in the media to scientific studies that would question the reliability of the DiaSorin test. These studies are the following:
- AZ Delta Study (Herroelen et al., 2020)
In this study, AZ Delta has reported 2 false positive results out of 56 samples analysed. However, after the FAMHP inspected the raw data, it was found that the calibration used for these measurements was not valid. A new measurement, with correct calibration this time, gave values, for these two positive samples, that were in a “grey area” and needed to be retested, which was not the case. When these 2 samples are not taken into account, the specificity is 100 %. - Danish study (unpublished)
An as yet unpublished study from Denmark, where no peer review process has yet taken place and whose results have therefore not yet been scientifically validated, has compared the performance of 16 different antibody tests in 16 different clinical laboratories. However, the samples used to determine specificity in laboratories are not always the same, which makes it difficult to make an exact comparison of the specificity values obtained. Of the 16 antibody tests, the DiaSorin test appears to have average performance, with a specificity of 97.2%. The Danish selection criterion of 99% specificity is stricter than what is taken as a consensus in other countries such as Belgium. - Dutch study (RIVM report)
This study has compared twenty-one antibody tests. Only 4 tests have reached a specificity ≥ 99%. In addition, it should be noted that serological tests with lower specificity may be used in follow-up studies. Local experts should consider this possibility in specific situations.
None of the above studies justify considering the DiaSorin test to be unreliable. Furthermore, it is intellectually incorrect not to mention the Belgian scientific studies that confirm the reliability of the DiaSorin test. The two previously reported and correctly carried out validation studies (from HIS-IZZ and the CHU of Liège) have established the specificity of the DiaSorin test at 100% (Tré-Hardy et al., 2020). Wolff et al. (LHUB-ULB) have reported a specificity of 97.9% (Wolff et al., 2020). A recent study from the Belgian reference centre for respiratory pathogens (UZ Leuven) has showed a specificity of 99.1% (Van Elslande et al., 2020). Other countries have also developed a validation process that has demonstrated the reliability of the DiaSorin test. This has also been shown in numerous international studies. This test is also used in the United States, Italy, Finland, Austria, the United Kingdom, Luxembourg, Israel and France, among others.
More and more serological tests have been placed on the market in recent months. There are more recent tests that give better results than the DiaSorin test. This makes perfect sense, because scientific knowledge is constantly expanding. But this does not mean that the DiaSorin test would thus be unreliable and can no longer be used. The fact that more and better tests are coming onto the market can only be applauded in view of the high demand during this health crisis.