Abstract
OBJECTIVE: To assess the validity of SARS-CoV-2 Antigen (Ag) detection for the diagnosis of SARS-CoV-2 infection in mildly infected or asymptomatic patients. METHODS: Observational study to evaluate diagnostic tests. Non-hospitalized patients with indication for diagnostic testing for SARS-CoV-2 infection were included. The diagnostic test to be evaluated was the determination of Ag and as a reference standard to determine the presence of viral RNA the RT-PCR was used. RESULTS: A total of 494 patients were included. Of these 71.5% (353/494) had symptoms and 28.5% (141/494) were asymptomatic (presurgery screening (35/494) and confirmed case-contact (106/494). The overall sensitivity of the Ag test was 61.1% and the specificity was 99.7%. The sensitivity and specificity in the asymptomatic group were 40% and 100% respectively, and in the symptomatic group 63.5% and 99.6% respectively. In turn, the sensitivity and specificity in the group of symptomatic patients varied according to the time of symptom evolution: in patients with recent symptoms, they were 71.4% and 99.6% respectively, while in patients with symptoms of more than 5 days of evolution, they were 26.7% and 100% respectively. In all groups studied, the presence of antigen is associated with a high viral load (Ct<30 cycles). CONCLUSIONS: The use of Ag detection test is not indicated for the diagnosis of SARS-CoV-2 infection in asymptomatic patients or with symptoms of more than 5 days of evolution, but it could be useful in patients with symptoms of 1-5 days of evolution.