key: cord-1016954-nc1njugh
authors: Takeuchi, Y.; Akashi, Y.; Kato, D.; Kuwahara, M.; Muramatsu, S.; Ueda, A.; Notake, S.; Ishikawa, H.; Suzuki, H.
title: Diagnostic Performance and Characteristics of Anterior Nasal Collection for the SARS-CoV-2 Antigen Test: A Prospective Study in Japan
date: 2021-03-05
journal: nan
DOI: 10.1101/2021.03.03.21252425
sha: 34508de45e72cc4f44bbe562ffed96dd2bebd5cd
doc_id: 1016954
cord_uid: nc1njugh

We conducted a prospective study in Japan to evaluate the diagnostic performance of the antigen test QuickNavi-COVID19 Ag using anterior nasal samples and to compare the degrees of coughs or sneezes induction and the severity of pain between anterior nasal collection and nasopharyngeal collection. A total of 862 participants were included in the analysis. In comparison to the findings of reverse transcription PCR using nasopharyngeal samples, the antigen test using anterior nasal samples showed 72.5% sensitivity (95% confidence interval [CI]: 58.3%-84.1%) and 100% specificity (95% CI: 99.3%-100%). Anterior nasal collection was associated with a significantly lower degree of coughs or sneezes induction and the severity of pain in comparison to nasopharyngeal collection (p < 0.001). The antigen test using anterior nasal samples showed moderate sensitivity but was less painful and induced fewer coughs or sneezes.

The laboratory diagnosis of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-

The presence or absence of SARS-CoV-2 was defined by the results of the reference real-92 time RT-PCR test. However, if discordance existed between the reference real-time test and the in-house RT-PCR test, a re-evaluation was performed with an Xpert Xpress Examiners rated the degrees of coughs or sneezes induction from the following four 102 categories: "None", "Small, 1-2 times", "Loud, 1-2 times" and "Loud, multiple times". The 103 severity of pain was evaluated with a five-point scale (Pain score), with 1 being "no pain" 104 and 5 being "worst imaginable pain," and the participants were asked to report a number from 105 the scale. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 The comparison of SARS-CoV-2 viral loads between different sample collection sites 108 and swab types 109 We conducted an additional experiment to evaluate whether the viral loads differed 110 between sample collection sites and swab types between January 8 and 19, 2021. After 111 receiving the participants' informed consent, two anterior nasal samples were obtained from 112 the participants for whom a nasopharyngeal sample had already tested positive for SARS-113 CoV-2. Two anterior nasal swab samples were collected from each nostril using one with a 114 NP-type swab and the other with an oropharyngeal-type flocked (OP-type) swab. These 115 sample collections were performed on the same day.

The samples were diluted in 3 mL of UTM, and stored at -80°C. After several days of 117 storage, the samples were thawed, and purification and RNA extraction were performed 118 according to the above-described method. The viral concentrations in samples were 119 quantified with the following procedure. The calibration curves were generated with 5, 50, 120 and 500 copies/reaction of positive control (EDX SARS-CoV-2 Standard; Bio-Rad 121 Laboratories, Inc., Hercules, CA, USA). Quantitative RT-PCR was performed on a 122 LightCycler 96 System (Roche, Basel, Switzerland) using a THUNDERBIRD Probe One-123 step qRT-PCR Kit (TOYOBO Co. Ltd., Osaka, Japan) with a primer/probe N2 set.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 5, 2021. Table S1a and Table S1b . at the drive-through PCR center, and only 17 were obtained after hospitalization. We 152 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 5, 2021. ; https://doi.org/10.1101/2021.03.03.21252425 doi: medRxiv preprint excluded the participants who underwent duplicate tests (n=7) or for whose clinical 153 information were lacking (n=7). Finally, 862 participants were included in the analysis.

Among the 862 participants, SARS-CoV-2 was detected in 50 (5.8%) on nasopharyngeal 155 samples by the reference real-time RT-PCR test. There was one discordant sample that was 156 positive on the in-house RT-PCR test and negative on the reference real-time RT-PCR test.

The Xpert Xpress SARS-CoV-2 test provided a positive result for the sample (cycle threshold

[Ct] values for E and N2 targets: 0.0 and 39.8, respectively); thus, 51 (5.9%) participants 159 were finally considered positive for SARS-CoV-2. The discordant sample was obtained from 160 a participant who had been diagnosed with COVID-19 one month before the current 161 evaluation and who was referred to the PCR center due to refractory respiratory symptoms.

All 51 participants who were positive for SARS-CoV-2 were symptomatic (Table 1a) ; their 163 characteristics are described in Table 1b . The most common symptom was fever (80.4%), 164 followed by cough or sputum production (60.8%), sore throat (37.3%), runny nose or nasal 165 congestion (35.5%), and loss of taste or smell (27.5%).

Diagnostic performance of QuickNavi-COVID19 Ag using anterior nasal samples 168 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 5, 2021. ; https://doi.org/10. 1101 /2021 Of the 51 participants who were found to be SARS-CoV-2-positive by the RT-PCR test, 169 37 participants were found to be positive with the antigen test with anterior nasal samples 170 (Table 2) . Among the 811 SARS-CoV-2-negative participants, all participants were found to 171 be negative with the antigen test ( The pain score was obtained from 90 participants (Fig. 3) . Fifty-seven participants (63.3%) 184 reported no pain in anterior nasal collection. The median pain score of anterior nasal 185 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 5, 2021. ; https://doi.org/10.1101/2021.03.03.21252425 doi: medRxiv preprint collection and nasopharyngeal collection was 1 (IQR: 1-2) and 3 (IQR: 2-4), respectively. In 186 comparison to nasopharyngeal collection, anterior nasal collection was significantly less 187 painful (p < 0.001).

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 5, 2021. ; https://doi.org/10.1101/2021.03.03.21252425 doi: medRxiv preprint

The selection of swab type influences the uptake, extraction and recovery efficiency of the 222 collected sample (17, 18) . In this study, we compared two flocked swabs with different tip 223 sizes (NP-type and OP-type swab). There was no significant difference in the viral load of 224 the samples collected with the two types of swabs; however, the OP-type swab has a larger 225 tip and seemed to handle a larger amount of samples collected. A previous study suggested 226 that the efficiency of sample release was not associated with the absorbed volume (19), which 227 could explain the result in this study.

The present study was associated with some limitations. First, the samples used for the 229 reference real-time RT-PCR test were frozen and transported. Although the samples were 230 frozen at -80°C, the viral loads may have been decreased during the storage and transport 231 process. Nevertheless, in the case of discrepancy with in-house PCR, re-evaluation was 232 performed and did not affect the calculation of the sensitivity of the antigen test. Second, 233 asymptomatic SARS-CoV-2 positive patients were unintentionally not included in this study.

Further study is required to evaluate the clinical performance of the antigen test in those 235 patients. Finally, this study was conducted at a single center, although participants were 236 referred from multiple primary facilities. Further research should be conducted to assess the 237 generalizability of the findings.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 In conclusion, the QuickNavi-COVID19 Ag test with an anterior nasal sample showed 239 100% specificity; however, the sensitivity in the detection of SARS-CoV-2 was lower in 240 comparison to the QuickNavi-COVID19 Ag test with nasopharyngeal samples. Anterior 241 nasal collection was less invasive and induced fewer coughs or sneezes, which may be more 242 comfortable for the patient and may reduce the risk of droplet exposure to healthcare workers. severity of pain at swab insertion was assessed on a five-point scale, from 1 to 5 (Pain score).

The pain score for each collection method was obtained from the same participant (n = 90).

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted March 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 The comparison of the pain scores with the two collection procedures was performed with 347 the Wilcoxon signed-rank test.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted March 5, 2021. ; https://doi.org/10.1101/2021.03.03.21252425 doi: medRxiv preprint Sensitivity, specificity, positive predictive value, and negative predictive value are provided with 95% confidence intervals.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted March 5, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 

World Health Organization. 2020. Diagnostic testing for SARS

Another false-positive problem for a SARS