As we enter the second quarter of the COVID-19 pandemic, with testing for severe acute respiratory syndrome coronavirus 2 (SARS–CoV-2) increasingly available (though still limited and/or slow in some areas), we are faced with new questions and challenges regarding this novel virus. When to test? Whom to test? What to test? How often to test? And, what to do with test results? Since SARS–CoV-2 is a new virus, there is little evidence to fall back on for test utilization and diagnostic stewardship (1). Several points need to be considered to begin answering of these questions; specifically, what types of tests are available and under which circumstances are they useful? This understanding can help guide the use of testing at the local, regional, state, and national levels and inform those assessing the supply chain to ensure that needed testing is and continues to be available. Here, we explain the types of tests available and how they might be useful in the face of a rapidly changing and never-before-experienced situation. There are two broad categories of SARS–CoV-2 tests: those that detect the virus itself and those that detect the host’s response to the virus. Each will be considered separately.

We must recognize that we are dealing with (i) a new virus, (ii) an unprecedented pandemic in modern times, and (iii) uncharted territory. With this in mind, in the absence of either proven effective therapy or a vaccine, diagnostic testing, which we have, becomes an especially important tool, informing patient management and potentially helping to save lives by limiting the spread of SARS–CoV-2. What is the most appropriate test, and for whom and when?

Hypothetically, if the entire world’s population could be tested all at once, with a test providing 100% specificity and sensitivity (unrealistic, obviously), we might be able to identify all infected individuals and sort people into those who at that moment in time were asymptomatic, minimally/moderately symptomatic, and severely symptomatic. The asymptomatic and minimally/moderately symptomatic could be quarantined to avoid the spread of the virus, with the severely symptomatic managed and isolated in health care settings. Contract tracing could be carried out to find those at risk of being in the incubation period by virtue of their exposure. Alternatively, testing for a host response, if, again, the test were hypothetically 100% sensitive and specific, could identify those previously exposed to the virus and (if we knew this to be true, which we do not) label those who are immune to the virus, who could be tapped to work in settings where potentially infected individuals (e.g., sick patients in hospitals) might otherwise pose a risk. Unfortunately, these hypothetical scenarios are not reality. However, with this ideal situation as a guide, what we do have available as tests today should be carefully considered in terms of how they can be leveraged to move the current crisis closer to the ideal situation, especially in the absence of therapeutics or vaccines.

Although the virus can be cultured, this is dangerous and not routinely done in clinical laboratories. While detection of viral antigens is theoretically possible, this approach has not, to date, been a primary one, but one that those participating in the summit considered to deserve further research.