The diagnostic accuracy of Rapid Antigen Tests (RAT) has been widely studied in various applications and in diverse populations.
Sensitivity, in the order of 75% in pooled estimates, is significantly influenced by the presence or absence of symptoms, viral load, and the timing of sampling relative to the onset of symptoms.
Specificity, in the order of 99% in pooled estimates, is consistently high across tests, populations, and sampling methods.
Post-test probability of being an infectious case following a positive test is highest in individuals with a high pre-test probability (population prevalence > 5%), such as those with COVID-19 symptoms, and those in settings with a high level of community transmission. Here, the positive predictive value is in the order of 95%. However, when used in settings with a lower pre-test probability (population prevalence < 0.5%), as in screening asymptomatic individuals, the positive predictive value is considerably reduced, as low as 25%.
Post-test probability of being an infectious case following a negative test is less than 1% (negative predictive value > 99%) in all settings except those with the highest levels of community transmission.
Badea, A; Reeder, B; Groot, G; Muhajarine, N; Minion, J; Miller, L; Howell-Spooner, B. In real world settings, what is the validity of Rapid Antigen Tests (RATs) in identifying SARS-CoV-2 and how well do they predict disease? 2022 Jan 12, Document no.: EOC211201 RR. In: COVID-19 Rapid Evidence Reviews [Internet]. SK: SK COVID Evidence Support Team, c2022. 15 p. (CEST rapid review report).