The reliability of SARS-CoV-2 IgG antibody testing - a pilot study in asymptomatic health care workers in a Croatian university hospital

To evaluate three fully automated serological assays in terms of reactivity to SARS-CoV-2 immunoglobulin G (IgG) and perform SARS-CoV-2 IgG antibody testing among asymptomatic health care workers (HCW) at the University Hospital Center Zagreb. Three IgG serological assays (Abbott SARS-CoV-2 IgG, Elecsys Anti-SARS-CoV-2, and MAGLUMI 2019-nCoV IgG) were initially evaluated by analyzing 42 samples from confirmed COVID-19-recovered patients and 48 negative individuals. A total of 1678 HCW (~30% of all hospital employees) were screened for SARS-CoV-2 IgG with the Abbott assay, run on Abbott Architect i2000SR. The samples exceeding the predefined cut-off (1.4 S/C) were reanalyzed with the Elecsys, MAGLUMI, and VIDAS SARS-COV-2 IgG assays. Initially, the MAGLUMI 2019-nCoV IgG produced 26.2% false negatives and the Elecsys Anti-SARS-CoV-2 produced one false positive. Among 1678 HCW, the Abbott assay showed only 10 (0.6%) positive results, with mostly mildly elevated signals. Nine of these samples were non-reactive when they were retested with the Elecsys, MAGLUMI, and VIDAS assays. As for the one remaining sample, it was positive when tested with the Elecsys assay, while the other two assays yielded negative results. SARS-CoV-2 IgG seroprevalence among asymptomatic HCW in our hospital setting was low, with different assays indicating a different number of positive samples. One of the assays yielded a large false negative rate. These findings can be attributed to differences in assay formulation but also to heterogeneity and diverse reactivity of antibodies against SARS-CoV-2 antigens.