Diagnostic accuracy of a SARS-CoV-2 rapid antigen test among military and civilian personnel of an Air Force airport in central Italy

BackgroundMost SARS-CoV-2 rapid antigen detection tests (RADTs) validation studies have been performed on specimens from COVID-19 patients and negative controls or from mostly symptomatic individuals. Herein we evaluated the diagnostic accuracy of AFIAS COVID-19 Ag, hereinafter denominated as AFIAS, during a COVID-19 screening program surveillance testing conducted among personnel of an Italian military airport.MethodsNasopharyngeal swabs (NPSs) were collected from study participants and were analysed by both AFIAS and RT-PCR assay. A questionnaire collecting demographic and exposure data were administered to all participants. AFIAS accuracy parameters including Cohen's kappa (K) were determined.ResultsOverall, from November 2020 to April 2021, 1294 (NPSs) were collected from 1183 participants (88.6% males, 11.4% females; mean age were 41.3, median age 42). Forty-nine NPSs (3.78%) were positive by RT-PCR, while 54 NPSs were positive by AFIAS. Overall baseline sensitivity, specificity, positive and negative predictive values were 0.633, 0.981, 0.574, 0.985, respectively and K was 0.585 (moderate). AFIAS sensitivity tended to be higher for NPSs with higher viral load. A higher sensitivity (0.944) compared to the overall baseline sensitivity (0.633) was also found for NPSs from participants with COVID-19 compatible symptoms, for which K was 0.891 (almost perfect). Instead, AFIAS sensitivity was quite poor for NPSs from asymptomatic participants. Most false negative NPSs in this group had moderate viral load.ConclusionOverall, AFIAS showed high specificity but only moderate sensitivity, mainly because of the high proportion of asymptomatic participants. However, AFIAS showed good sensitivity for NPSs with high viral load and nearly optimal accuracy parameters for NPSs from participants with COVID-19 compatible symptoms. Thus, taking into consideration its performance features, this test can be useful for COVID-19 case identification and management as well as for infection control.