Comparative evaluation of SARS-CoV-2 serological tests shows significant variability in performance across different years of infection and between the tests.

Introduction: While the global COVID-19 pandemic is slowly coming under control, current efforts are focused on understanding the epidemiology of endemic SARS-CoV-2. The tool of choice for doing so remains serological tests that detect SARS-CoV-2 induced antibodies. However, the performance of these tests should be evaluated to ensure they comply with the specific performance criteria desired by each country that they are used in. Methods: Here, we use pre-COVID-19 plasma and plasma from SARS-CoV-2-infected individuals collected in 2020, 2021 and 2022 to evaluate the performance of two commercial Rapid Lateral Flow (RLF) tests (the PANBIO™ COVID-19 IgG/IgM rapid test and the LABNOVATION™ COVID-19 (SARS-CoV-2) IgG/IgM rapid test) and one commercial ELISA test (the PLATELIA™ SARS-CoV-2 total Ab). Results: We find that whereas the specificity of the two RLF tests is ≥ 95%, it was 91% for the ELISA tests. However, at 14 days post-COVID-19 date of diagnosis (DoD), only the ELISA test constantly achieved a sensitivity of ≥80% over all the three years. In addition, the rate of detection of the two RLF tests varied across the years with a sensitivity ranging from <80% in 2021 to >80% in 2022. More importantly the capacity of these two RLF tests to detect IgG antibodies decreased with time. On the contrary, the sensitivity of the ELISA test was still above 80% more than six months post DoD. Conclusion: We recommend that sero-epidemiological surveys focused on testing antibodies should not rely on performances reported by the assay manufacturers. They should include a formal evaluation of the selected assays to ensure its limitations and strengths conform with the data-accuracy requirements of the surveys.