Background: Currently, more than 500 different AgPOCTs for SARS-CoV-2 diagnostics are on sale, for many of which no data about sensitivity other than self-acclaimed values by the manufacturers are available. In many cases these do not reflect real-life diagnostic sensitivities. Therefore, manufacturer-independent quality checks of available AgPOCTs are needed, given the potential implications of false-negative results. Objective: The objective of this study was to develop a scalable approach for direct comparison of the analytical sensitivities of commercially available SARS-CoV-2 antigen point-of-care tests (AgPOCTs) in order to rapidly identify poor performing products. Methods: We present a methodology for quick assessment of the sensitivity of SARS-CoV-2 lateral flow test stripes suitable for quality evaluation of many different products. We established reference samples with high, medium and low SARS-CoV-2 viral loads along with a SARS-CoV-2 negative control sample. Test samples were used to semi-quantitatively assess the analytical sensitivities of 32 different commercial AgPOCTs in a head-to-head comparison. Results: Among 32 SARS-CoV-2 AgPOCTs tested, we observe sensitivity differences across a broad range of viral loads (~7.0*10⁸ to ~1.7*10⁵ SARS-CoV-2 genome copies per ml). 23 AgPOCTs detected the Ct25 test sample (~1.4*10⁶ copies/ ml), while only five tests detected the Ct28 test sample (~1.7*10⁵ copies/ ml). In the low range of analytical sensitivity we found three saliva spit tests only delivering positive results for the Ct21 sample (~2.2*10⁷ copies/ ml). Comparison with published data support our AgPOCT ranking. Importantly, we identified an AgPOCT offered in many local drugstores and supermarkets, which did not reliably recognize the sample with highest viral load (Ct16 test sample with ~7.0*10⁸ copies/ ml) leading to serious doubts in its usefulness in SARS-CoV-2 diagnostics. Conclusion: The rapid sensitivity assessment procedure presented here provides useful estimations on the analytical sensitivities of 32 AgPOCTs and identified a widely-spread AgPOCT with concerningly low sensitivity.

No abstract