The SARS-CoV-2 coronavirus epidemic has triggered a major development of molecular methods that identify active infection through viral RNA detection and serological methods that detect specific antibodies, usually IgG and IgM directed, most commonly against viral spike or nucleocapsid protein. With knowledge of the kinetics of the immune response to SARS-CoV-2 virus infection, serological methods can be used as an appendix to molecular methods, in determining seroprevalence, in knowledge of the virus, and the development of potential vaccines. Despite the need for rapid implementation of methods into routine practice, a verification step is required before the use, in order to evaluate the diagnostic and analytical properties. Based on the results, we can determine whether the method in the laboratory gives quality and reliable results and fits its purpose.
In order to use it as quickly as possible, we verified and compared three qualitative serological assays that determine antibodies against SARS-CoV-2 nucleocapsid protein in the serum. We selected two automated methods, the Roche Elecsys Anti-SARS-CoV-2 and Abbott SARS-CoV-2 IgG and the point of care test SARS-CoV-2 IgM / IgG Antibody Assay Kit by Colloidal Gold Method. By comparison with the RT-PCR method, the diagnostic properties of the assays were evaluated on 91 RT-PCR negative and 60 RT-PCR positive samples. We found that the point-of-care test has poorer diagnostic characteristics, as it gives the most false results (8,61 %). The least false positive results (2,65 %) were given by the Roche Elecsys Anti-SARS-CoV-2 assay with the best diagnostic sensitivity of 98,33 % (91,06–99,96). We concluded that the method gives such good results due to the automated detection of both classes of antibodies. Among the methods, we determined the level of agreement using the Cohen's kappa coefficient (к). An almost perfect agreement with к = 0,90 (0,83–0,97) was proven between the automated assays. Repeatability, intermediate and interlaboratory precision were assessed to automated methods. We found that both meet the manufacturer’s criteria. We can conclude that automated methods give better results than the point-of-care test, especially the one that determines both classes of antibodies. Their agreement between the results is also better than agreement between the results gained by the point-of-care test. In addition to better diagnostic properties, both automated methods meet the manufacturer's criteria in terms of accuracy, which means that the results can be trusted and interpreted correctly.
|
