by Arneb » Tue Apr 07, 2020 7:52 pm
Lance anwered the question already. I'm just elaborating a bit.
PCR (polymerase chain reaction) tests have the advantage of being simple (PCR is off-the-shelf stuff nowadays), robust, and highly sensitive as well as specific. The SARS-CoV2 PCR test detects active viral replication (=production of virus genomes). The test can be positive in asymptomatic people, about 2 or 3 days before symptoms develop and some 4-8 days into the symptomatic phase for pharyngeal swabs, and longer than that in sputum or deep washings of the airways. It is also positive for almost as long in feces - but it is not clear if that represents undigested virus particles originally produced in the pharynx or active viral replication in the gut.
The antibody tests can only turn positive after the body has mounted a specific immune response. It detects antibodies in the patient's serum. The lag between infection, onset of symptoms and detectable antibodies is called the "diagnostic gap" and an obvious drawback for antibody testing. Others include the lower sensitivity and greater technical problems with test development and execution. Serological test for SARS-CoV2 exist, but they are not routine enough yet to apply en masse. They are used in people where the PCR is stubbornly negative in contrast to a clear clinical picture of Covid-19 or for people who present late in the disease and may already have lost viral replication.
At the moment, what you want is pick out infected people for isolation as soon as possible, even in (as yet) asymptomatic patients - that is precisely the problem PCR tests are ideal for. Later, when we ask ourselves who might have protective immunity and which proportion of infected people have remained asymptomatic, antibody testing will a have a greater role, hopefully in a test system that can be produced industrially for off-the-shelf test kits.
Non sunt multiplicanda entia praeter necessitatem