Validation of COVID-19 serologic tests and large scale screening of asymptomatic healthcare workers

Abstract

Objectives: Serologic testing for SARS-CoV-2 is an important element in the fight to slow the COVID-19 pandemic. This study aimed to validate two serologic tests for total (IgM, IgG, IgA) SARS-CoV-2 antibodies, (i) the Ortho-Clinical Diagnostics Anti-SARS-CoV-2 Total Antibody assay for the Vitros 5600 analyzers and (ii) a manual laboratory developed ELISA (FDA EUA pending), for use in parallel orthogonal testing of asymptomatic healthcare workers and affiliates of the University of Maryland Medical System. Design and Methods: Validation and verification of the two tests was performed using samples from hospitalized patients that were found to be PCR positive for SARS-CoV-2, samples pre-COVID-19, and samples from individuals with current/previous infections with other viruses. Healthcare workers and affiliates from across the University of Maryland Health System were provided testing free of charge and their results were reported as reactive or non-reactive if the two tests were concordance, or indeterminate if the results were discordant. Results: Validation testing found the Ortho Vitros test to be 100% (73/73) sensitive, and 99.3% (152/153) specific, while the UMMC ELISA was found to be 97.6% (204/209) sensitive and 100% (288/288) specific. Real world testing among 8399 healthcare workers found that 2.9% (247/8399) of healthcare workers were positive for anti- SARS-CoV-2 antibodies by both tests. An indeterminate rate of 1.1% (91/8399), in which one test reported reactive results, and one as non-reactive was also seen. Conclusions: Parallel orthogonal testing improves the positive and negative predictive value of serologic testing in populations with low prevalence. The use of an indeterminate result from parallel orthogonal testing allows for the follow-up and re-testing, which helps resolve discrepancies between assays.