When intuition falters: repeated testing accuracy during an epidemic

Widespread, repeated testing using rapid antigen tests to proactively detect asymptomatic SARS-CoV-2 infections has been a promising yet controversial topic during the COVID-19 pandemic. Concerns have been raised over whether currently authorized lateral flow tests are sufficiently sensitive and specific to detect enough infections to impact transmission whilst minimizing unnecessary isolation of false positives. These concerns have often been illustrated using simple, textbook calculations of positivity rates and positive predictive value assuming fixed values for sensitivity, specificity and prevalence. However, we argue that evaluating repeated testing strategies requires the consideration of three additional factors: new infections continue to arise depending on the incidence rate, isolating positive individuals reduces prevalence in the tested population, and each infected individual is tested multiple times during their infection course. We provide a simple mathematical model with an online interface to illustrate how these three factors impact test positivity rates and the number of isolating individuals over time. These results highlight the potential pitfalls of using inappropriate textbook-style calculations to evaluate statistics arising from repeated testing strategies during an epidemic.