Assessment of Sensitivity and Specificity of Nasopharyngeal and Throat Swabs in Detection of COVID-19 Infection Among Admitted Patients: A Scientific Perspective

Abstract

Background: Early and accurate detection of respiratory viruses (RV) is important for patient management. We have previously shown that self-collected nasal swabs (NS) are feasible and as sensitive as clinician-collected nasal washes for detection of RV, but the additive benefit of self-collected throat swabs is unknown. Objectives: To test the rise in auto sufficient nasal yields to the throat swabs in patients with upper respiratory (URTI) symptoms for PCR identification of RV. Study Design: Patients with URTI symptoms self-collected paired polyurethane foam NS and nylon flocked throat swabs and completed a symptom survey. Swabs were tested for 12 RV by real-time reverse transcription (RT)-PCR. Descriptive, McNemar's, and Wilcoxon signed rank statistical tests were used. Results: The sample was made up of 115 paired swab nasals and throat, with at least 1 specimen being positive for RV (71/115 (62 percent), including 51 positive for both specimens, 17 positive for NS only and 3 favorable for RV only with throat swab. NS was 96 percent sensitive (95 percent CI: 88-99) compared with 76 per cent in throat swabs, p<0.001 (95 percent CI: 65-85). The median PCR period threshold (Ct) of 51 concordant samples was lower in NS (25.1) than in swabs of the throat (32.0). Conclusion: Self-collection of NS was significantly more sensitive than self-collection of throat swabs for detection of RV by RT-PCR. The addition of throat sampling does not appear to increase the diagnostic load in the self-testing setting. © 2018, EBX. All rights reserved.