Evaluation of two rapid antigen tests to detect SARS-CoV-2 in a hospital setting

Abstract

Successful containment strategies for the SARS-CoV-2 pandemic will depend on reliable diagnostic assays. Point-of-care antigen tests (POCT) may provide an alternative to time-consuming PCR tests to rapidly screen for acute infections on site. Here, we evaluated two SARS-CoV-2 antigen tests: the STANDARD™ F COVID-19 Ag FIA (FIA) and the SARS-CoV-2 Rapid Antigen Test (RAT). For diagnostic assessment, we used a large set of PCR-positive and PCR-negative respiratory swabs from asymptomatic and symptomatic patients and health care workers in the setting of two University Hospitals in Munich, Germany, i.e. emergency rooms, patient care units or employee test centers. For FIA, overall clinical sensitivity and specificity were 45.4% (n = 381) and 97.8% (n = 360), respectively, and for RAT, 50.3% (n = 445) and 97.7% (n = 386), respectively. For primary diagnosis of asymptomatic and symptomatic individuals, diagnostic sensitivities were 60.9% (FIA) (n = 189) and 64.5% (RAT) (n = 256). This questions these tests' utility for the reliable detection of acute SARS-CoV-2-infected individuals, in particular in high-risk settings. We support the proposal that convincing high-quality outcome data on the impact of false-negative and false-positive antigen test results need to be obtained in a POCT setting. Moreover, the efficacy of alternative testing strategies to complement PCR assays must be evaluated by independent laboratories, prior to widespread implementation in national and international test strategies.

Keywords: COVID-19 point-of-care; Diagnostic test; SARS-CoV-2 antigen test; Sensitivity; Specificity.

Fig. 1

Detection of SARS-CoV-2 in 381 PCR-positive respiratory swabs from site 1 using either the STANDARD™ F COVID-19 Ag FIA or the SARS-CoV-2 Rapid Antigen Test. Respiratory swabs were analyzed and scored “positive” or “negative” according to the manufacturers’ instructions and plotted relative to either the respective sample’s Ct/Cp value (a) or the corresponding SARS-CoV-2 RNA copy number per mL (b) determined in in RT-PCR assays. Sub-analysis of swabs taken at primary diagnosis of COVID-19 (c, d) or swabs taken at follow-up testing during hospitalization (e, f). Each symbol represents one sample. Center lines show the medians and the box limits are quartiles 1 and 3, and whiskers show maximum and minimum values

Fig. 2

Distribution of Ct/Cp values in SARS-CoV-2 RT-PCR reactions from respiratory samples taken at primary COVID-19 diagnosis at site 1. Each bar indicates the number of respiratory samples ± 1 Ct/Cp value around the Ct/Cp value given on the x-axis. The red line depicts the median of these 193 specimen

Fig. 3

Detection of SARS-CoV-2 in 66 freshly collected, PCR-positive respiratory swabs obtained at the Klinikum rechts der Isar (site 2) and analyzed at the Institute of Virology (TUM) using the SARS-CoV-2 Rapid Antigen Test (RAT). Respiratory swabs were analyzed and scored “positive” or “negative” according to the manufacturer’s instructions and plotted relative to either the respective sample’s viral load

Fig. 4

The site of swab sampling in the upper respiratory tract does not significantly affect the SARS-CoV-2 RNA load (a) (not significant (n.s.), p = 0.15, Wilcoxon rank sum test) but shows a slightly reduced antigen test reactivity for oropharyngeal swabs (b) (FIA: p = 0.029; RAT: p = 0.039; Fisher’s exact test) indicated by the asterix

Fig. 5

Prior storage of respiratory specimen at − 20 °C even slightly enhances the rate of SARS-CoV-2 antigen test reactivity for FIA