There’s more to a COVID-19 test

Dear Dr. Roach: What are the chances of a false positive or false negative with COVID-19 tests? Would it be wise to get tested with more than one brand just to make sure? — A.W.

Many laypeople tend to think of diagnostic tests as perfectly accurate. The doctor does a test for a condition, and if the test is positive, you have it, and if it’s negative, you don’t. Unfortunately, the reality is a lot more complicated.

Since very few tests are absolutely perfect, clinicians and scientists talk about probabilities. The most important number for the patient is the posttest probability. For COVID-19 testing, that’s the likelihood you have COVID-19 (for a swab test) or have had COVID-19 (for an antibody test). The post-test probability depends on both the pretest probability and the test characteristics (sensitivity and specificity) of the diagnostic test. Different manufacturers have different test characteristics, but the pretest probability has underappreciated impact.

For example, if you live in a part of the country with very little COVID-19 transmission and you have had no symptoms, your pretest probability of having the virus is low. The analysis starts with the prevalence of COVID-19 in your community (if known), and adjusts that number upward if you have had symptoms or downward if you haven’t. That pretest probability will go down if the test result is negative, and up if it is positive. However, there is never complete certainty.

In an area where there is not much COVID-19, most people who have not had symptoms but have a positive antibody test will have a false positive. Many people in a high prevalence area with a negative test will have a false negative.

Using multiple tests won’t help much. Ordering the test when it makes sense and understanding the limitations of the test are more important.