Effects of Patient Characteristics on Diagnostic Performance of Self-Collected Samples for SARS-CoV-2 Testing

Smith-Jeffcoat, Sarah E.; Koh, Mitsuki; Hoffman, Adam; Rebolledo, Paulina A.; Schechter, Marcos C.; Miller, Halie K.; Sleweon, Sadia; Rossetti, Rebecca; Kasinathan, Vyjayanti; Shragai, Talya; O’Laughlin, Kevin; Espinosa, Catherine C.; Khalil, George M.; Adeyemo, AdeSubomi O.; Moorman, Anne; Bauman, Brenda L.; Joseph, Kahaliah; O’Hegarty, Michelle; Kamal, Nazia; Atallah, Hany; Moore, Brooks L.; Bohannon, Caitlin D.; Bankamp, Bettina; Hartloge, Claire; Bowen, Michael D.; Paulick, Ashley; Gargis, Amy S.; Elkins, Christopher; Stewart, Rebekah J.; da Silva, Juliana; Biedron, Caitlin; Tate, Jacqueline E.; Wang, Yun F.; Kirking, Hannah L.

We evaluated the performance of self-collected anterior nasal swab (ANS) and saliva samples compared with healthcare worker–collected nasopharyngeal swab specimens used to test for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). We used the same PCR diagnostic panel to test all self-collected and healthcare worker–collected samples from participants at a public hospital in Atlanta, Georgia, USA. Among 1,076 participants, 51.9% were men, 57.1% were >50 years of age, 81.2% were Black (non-Hispanic), and 74.9% reported >1 chronic medical condition. In total, 8.0% tested positive for SARS-CoV-2. Compared with nasopharyngeal swab samples, ANS samples had a sensitivity of 59% and saliva samples a sensitivity of 68%. Among participants tested 3–7 days after symptom onset, ANS samples had a sensitivity of 80% and saliva samples a sensitivity of 85%. Sensitivity varied by specimen type and patient characteristics. These findings can help physicians interpret PCR results for SARS-CoV-2.