Many COVID-19 cases are currently diagnosed outside of the health care setting, typically with use of at-home rapid antigen tests. Multiple studies have confirmed that although antigen tests are less sensitive than “gold standard” polymerase chain reaction tests, they have good operating characteristics among acutely symptomatic persons, with an extremely low false-positive rate (Lee, November 2021; Chu, April 2022; Soni, July 2023). What is less clear, however, is whether these tests remain highly accurate in the hands of outpatients in the “real world” versus when used by health professionals.
This question was the subject of a large study at a single community testing site in Baltimore, where 953 patients presenting for COVID-19 testing underwent clinician-collected swabs for both rapid (BinaxNOW) and PCR testing. After clinician tests were collected, patients then performed and interpreted their own rapid test. The sensitivity of clinician and self-collected rapid testing was very good and relatively similar at 88.2% (clinician) vs. 83.9% (patient); notably, sensitivity was much lower in asymptomatic cases (63% [clinician] vs. 52% [self]) as compared to symptomatic cases (90% [clinician] vs. 88% [self]). Overall, very high and concordant specificities were observed: 99.6% (clinician) vs. 99.8% (patient). Interestingly, ~5% of positive self-collected rapid tests were read by the patient as negative yet were deemed to be positive results by the clinician team when re-read.
Conclusion: This study supports that self-collected rapid tests show similar sensitivity and specificity to those performed by clinicians. In particular, false-positive testing is exceedingly uncommon, though false negatives occur among asymptomatic persons — such cases tend to have high cycle threshold values, suggesting lower viral loads. Inaccurate interpretation of positive antigen tests by some patients indicates a role for targeted education to improve their use in nonclinical settings.