Self-collected oral, nasal and saliva samples yield sensitivity comparable to professional-collected oro-nasopharyngeal swabs in SARS-CoV-2 diagnosis
Maximilian Gertler,Eva Krause,Welmoed van Loon,Niklas Krug,Franka Kausch,Chiara Rohardt,Heike Rössig,Janine Michel,Andreas Nitsche,Marcus A. Mall,Olga Nikolai,Franziska Hommes,Susen Burock,Andreas K. Lindner,Frank P. Mockenhaupt,Ulrich Pison,Joachim Seybold
DOI: https://doi.org/10.1101/2021.04.13.21255345
2021-04-20
Abstract:Summary/Abstract Introduction Containment of the COVID-19 pandemic requires broad-scale testing. Laboratory capacities for real-time-PCR were increased, and are complemented by Ag-tests. However, sample-collection still requires qualified personnel and protective equipement, may produce transmission to others during conduct and travel, and is perceived uncomfortable. We tested sensitivity of three simplified self-sampling techniques compared to professional-collected combined oro-nasopharyngeal samples (cOP/NP). Methods From 62 symptomatic COVID-19 outpatients, we obtained simultaneously three self- and one professional-collected sample after initial confirmation in a testing centre: (i) combination swab (tongue, cheek, both nasal vestibula, MS, (ii) saliva sponge combined with both nasal vestibula, SN, and (iii) gargled tap water, GW, (iv) professionally-collected cOP/NP (standard). We compared the results of SARS-CoV-2 PCR-assays detecting E-gene and ORF1ab for the different sample types and performed bivariate statistical analysis to determine the variables reducing sensitivity of the self-collecting procedures. Results SARS-CoV-2 RNA was detected in all 62 professionally-collected cOP/NP. MS and SN samples showed a sensitivity of 95.2% (95%CI 86.5-99.0) and GW samples of 88.7% (78.1-95.3). Compared to the median ct -values of cOP/NP samples for E-gene (20.7) and ORF1ab (20.2) these were higher for MS (22.6 and 21.8), SN (23.3 and 22.3), and for GW (30.3 and 29.8). For MS and SN samples but not for GW specimens, false negativity in bivariate analysis was associated with non-German mother-tongue, number of sampling errors, and with symptom duration. For symptom duration of ≤8 days, test sensitivity for SN samples was 98.2% (95%CI 90.4-100.0) and for MS 96.4% (95%CI 87.7-99.6) and drops after day 8 below 90%. Discussion The study is limited to sensitivity of self-collection in symptomatic patients. Still, in this group, self-collected oral/nasal/saliva samples are reliable alternatives to professional-collected cOP/NP samples, if symptom duration does not exceed eight days and operational errors are minimized. Self-sampling could contribute to up-scaling of safe and efficient testing.