The sky has its limits in covid-19 testing

At the time of writing, in July 2020, the COVID-19 pandemic has already inflicted dramatic international restrictions, including airports closing and limiting international travel. It has been suggested that re-opening of airports should involve and even rely on testing travelers for COVID-19. This paper discusses the methodology of estimating the detection and diagnostic accuracy of COVID-19 tests. It explains the clear distinction between the technical characteristics of the tests, the detection measures, and the diagnostic measures that have clinical and public health implications. It demonstrates the importance of the prevalence of COVID-19 in terms of determining the ability of a test to yield a diagnosis. We explain the methodology of evaluating diagnostic tests, using the predictive summary index (PSI), and the minimum number of tests that need to be performed in order to correctly diagnose one person, which is estimated by 1/PSI. In a population with low prevalence, even a high-sensitivity test may lead to a high percentage of false positive diagnoses, resulting in the need for multiple high-cost tests to achieve a correct diagnosis. Thus, basing a policy for opening airports on diagnostic testing, even with the best test for COVID-19, has some limits.