LAUSR

Abstract Objectives Chemiluminescence immunoassay (CLIA) automated assays (fourth-generation antigen test) for SARS-CoV-2 detection are promising because of their analytical productivity, but have lower sensitivity and specificity than rt-PCR assays. The authors of this paper evaluated a recent immunoassay implemented on Siemens Atellica IM, investigating how much this could affect the actual feasibility of this diagnostic during the pandemic. Methods From the three-day routine 134 positive and 241 negative swab samples by rt-PCR test were evaluated, selected as 1/3 positive – 2/3 negative. Results Using rt-PCR as gold standard, the specificity of immunoassay was 96.7%, while sensitivity was 68.0%. Sensitivity is inversely proportional to the viral load: 100% for cycles threshold (CT) values from 14 to 29, 95% until 30 CT, then 85, 74, 72, 68%, for 31–35 CT respectively. Conclusions Our study confirms the reliability of the fourth-generation antigen assay in recognizing negative samples. Conversely, sensitivity appears to be less reliable (68.0%) than reported in the literature. This could be due to a non-randomized study group: many swab samples were taken from patients with expected low viral load (hospitalized for COVID for more than 10–12 days or asymptomatic patients for epidemiological surveillance). The strong correlation of sensitivity and viral load could prove significant to track the infectiousness of infected people, as previous studies reported that a viral load of at least 10E6 copies of RNA/mL, corresponding to 25 CT, is the threshold of transmission of the disease.