Infection Control Measures and Prevalence of SARS-CoV-2 IgG among 4,554 University Hospital Employees, Munich, Germany

Hospital staff are at high risk for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection during the coronavirus disease (COVID-19) pandemic. This cross-sectional study aimed to determine the prevalence of SARS-CoV-2 infection in hospital staff at the University Hospital rechts der Isar in Munich, Germany, and identify modulating factors. Overall seroprevalence of SARS-CoV-2-IgG in 4,554 participants was 2.4%. Staff engaged in direct patient care, including those working in COVID-19 units, had a similar probability of being seropositive as non–patient-facing staff. Increased probability of infection was observed in staff reporting interactions with SARS-CoV-2‒infected coworkers or private contacts or exposure to COVID-19 patients without appropriate personal protective equipment. Analysis of spatiotemporal trajectories identified that distinct hotspots for SARS-CoV-2‒positive staff and patients only partially overlap. Patient-facing work in a healthcare facility during the SARS-CoV-2 pandemic might be safe as long as adequate personal protective equipment is used and infection prevention practices are followed inside and outside the hospital


Calculation of Specificity and Sensitivity of the SARS-CoV-2 Antibody Tests
IgG and IgM were determined in 4,554 and 1,708 serum samples, respectively, by using a paramagnetic particle chemiluminescent immunoassay (CLIA) on an iFlash 1800 immunoassay analyzer (Shenzhen Yhlo Biotech Co., Shenzhen, China). This assay was selected as a screening assay because it detects antibodies directed against either SARS-CoV-2 S1 or N protein.
According to the manufacturer's instructions, values ≥10 AU/mL were considered positive.
To determine the sensitivity and specificity of the screening assay, confirmatory testing was performed in all serum samples that tested positive for IgM or IgG, all serum samples with IgG values between 5 and 10 AU/mL, and all serum samples from SARS-CoV-2 PCR-positive persons. For confirmation, the total antibodies against SARS-CoV-2 N protein were determined by using an electrochemiluminescent immunoassay (ECLIA) on a Cobas e411 analyzer (Roche Diagnostics, Mannheim, Germany). In all samples with incongruent results, IgG against SARS-CoV-2 S1 protein were determined by using an ELISA (Euroimmun, Luebeck, Germany), while immunoblot was used to differentiate antibodies against N, S1, and the receptor binding domain (RBD) of SARS-CoV-2 from those against seasonal coronaviruses (Mikrogen, Neuried, Germany).
Tests were considered correct if the presence of antibodies was confirmed by at least one more independent assay. Of the 108 serum samples that tested positive in the Yhlo screening assay, 93 also tested positive in the Roche IgG assay, eight were confirmed by immunoblotting (Appendix Tables 1 and 4). In one individual the screening result was considered specific due to high IgG titer and concomitant IgM positivity although no confirmatory testing could be performed (Appendix Tables 1 and 4, Sample-ID 18). In another individual testing only IgG positive, the serum amount was insufficient for confirmatory testing (Appendix Table 3, Sample-ID 114). Five IgG test results were considered false positive because screening IgG results were not confirmed by any of the other assays (Appendix Table 3). This resulted in a specificity of 99.89% for the IgG assay (4,441/4,446; Appendix Table 5).
IgM was screened in all patients until May 4 (n = 1,620). Six patients lacking prevalence of SARS-CoV-2 IgG tested positive for IgM (6/1,620). Because this could not be confirmed by the Roche ECLIA detecting IgM and IgG, these samples were considered false positive (S4 Table). If the Roche assay would have a 100% sensitivity, this would result in a specificity of 99.63% for the IgM assay. Due to the lack of a third assay, this, however, has to be considered preliminary.
To determine the sensitivity of the Yhlo IgG screening assay, 35 samples with detectable values between 5 and 10 AU/mL, i.e., below the recommended cutoff of the assay, were retested with both the Roche and the Euroimmun assay. Four samples tested positive in the Roche and Euroimmun assays, and were therefore considered false negative in the Yhlo screening assay (Appendix Tables 1 and 4). This enabled us to estimate the overall sensitivity of the IgG assay at 96.30% (104/108; Appendix Table 5).