Two years ago, I collaborated with colleagues from Georgetown's School of Nursing & Health Studies on a peer-reviewed article explaining the complex interpretation and management of Zika virus test results in the Journal of the American Board of Family Medicine. After COVID-19 began spreading in the United States this spring, our team reunited virtually and added a fifth collaborator (a 4th year medical student in Georgetown's Population Health Scholar Track) to outline a practical approach to the interpretation of COVID-19 test results in clinical settings. This week, JABFM published our new paper online ahead of print in its COVID-19 subject collection.
The paper includes three clinical scenarios that illustrate the application of testing principles to hypothetical persons with a low, moderate, and high pretest probability of having COVID-19, respectively. The manuscript we initially submitted to the journal included the following scenario:Scenario 3 – High pretest probability (RT-PCR)
A 28-year-old female White House staffer with unremarkable past medical history spent two hours meeting in person with another staffer who was diagnosed the next day with COVID-19. Five days later, she developed a temperature of 101.5 degrees Fahrenheit, a nonproductive cough, and loss of taste and smell. A rapid molecular SARS-CoV-2 test is negative. She is relieved at the test result because she will be meeting with the President tomorrow.
Let’s say that her pretest probability was 80%, given her direct contact with an individual who tested positive and development of classic COVID-19 symptoms. As she is symptomatic, the timing of testing with RT-PCR is appropriate. Suppose her test has a sensitivity of 90% and specificity of 95%. The predictive value for a positive test result is 98.6% and a negative test result is 70.4%. There is a 70% chance the staffer with a negative result actually does have SARS-CoV2 infection. Suppose that she was given a test with lower sensitivity of 80%. While the positive predictive value would still be high at 98.5%, the negative predictive value would drop significantly to 54.3%. In either case, a negative test result does not rule out COVID-19. In other words, a negative molecular test result in an individual with high clinical suspicion or high pretest probability for COVID-19 should be interpreted with caution, especially when tests have low sensitivity. Despite testing negative, this individual should self-isolate for 10 days.
One of the peer reviewers asked us to rewrite this scenario in our revision because he or she felt that using a White House staffer as an example was "gratuitous," "unnecessarily provocative," and "introduces politics into discussion that we need to strive to remain apolitical." We decided to change the patient in the scenario to a college student with a fever who had recently been in close contact with a resident in her dormitory who had tested positive for COVID-19. We in no way intended for it to be prophetic, even though White House senior counselor Hope Hicks, the presumed source for the President's and First Lady Melania Trump's recently diagnosed infections, is 31 years old and tested positive after experiencing mild flulike symptoms on a return flight from a campaign rally on Wednesday evening.
News accounts thus far indicate that White House medical staff have followed CDC guidance since Ms. Hicks's diagnosis, immediately placing the President and First Lady in quarantine since they had been exposed and in isolation now that they have tested positive. They will need to self-isolate for at least 10 days or, should either develop symptoms, 24 hours after fever resolves, whatever is later. Setting aside my personal politics and the upcoming Presidential election, I wish them both a speedy and uneventful recovery. As for the rest of us: if the most powerful man in the world, who has hundreds of highly skilled protective and medical personnel working around the clock to keep him physically safe and in good health, can contract COVID-19, so can you. Wash your hands. Keep your distance. And most importantly, wear a mask.
