In order to generate independent performance data regarding accuracy of COVID-19 antigen-based rapid diagnostic tests (Ag-RDTs), prospective diagnostic evaluation studies across multiple sites are required to evaluate their performance in different clinical settings. This report describes the clinical evaluation the GENEDIA W COVID-19 Ag Device (Green Cross Medical Science Corp., Chungbuk, Korea) and the ActiveXpress+ COVID-19 Complete Testing Kit (Edinburgh Genetics Ltd, UK), in two testing sites Peru and the United Kingdom. Nasopharyngeal swabs collected from 456 symptomatic patients at primary points of care in Lima, Peru and 610 symptomatic participants at a COVID-19 Drive-Through testing site in Liverpool, England were analyzed by Ag-RDT and compared to RT-PCR. Analytical evaluation of both Ag-RDTs was assessed using serial dilutions of direct culture supernatant of a clinical SARS-CoV-2 isolate from the B.1.1.7 lineage. For GENEDIA brand, the values of overall sensitivity and specificity were 60.4% [95% CI 52.4-67.9%], and 99.2% [95% CI 97.6-99.7%] respectively; and for Active Xpress+ the overall values of sensitivity and specificity were 66.2% [95% CI 54.0-76.5%], and 99.6% [95% CI 97.9-99.9%] respectively. The analytical limit of detection was determined at 5.0 x 102 pfu/ml what equals to approximately 1.0 x 104 gcn/ml for both Ag-RDTs. The UK cohort had lower median Ct values compared to that of Peru during both evaluations. When split by Ct, both Ag-RDTs had optimum sensitivities at Ct<20 (in Peru; 95% [95% CI 76.4-99.1%] and 100.0% [95% CI 74.1-100.0%] and in the UK; 59.2% [95% CI 44.2-73.0%] and 100.0% [95% CI 15.8-100.0%], for the GENDIA and the ActiveXpress+, respectively). Whilst the overall clinical sensitivity of the Genedia did not meet WHO minimum performance requirements for rapid immunoassays in either cohort, the ActiveXpress+ did so for the small UK cohort. This study illustrates comparative performance of Ag-RDTs across two global settings and considers the different approaches in evaluation methods.
Read full abstract