Comparative accuracy of oropharyngeal and nasopharyngeal swabs for diagnosis of COVID-19

March 26, 2020

Catherine Carver, Nick Jones
Edinburgh / Oxford

On behalf of the Oxford COVID-19 Evidence Service Team
Centre for Evidence-Based Medicine, Nuffield Department of Primary Care Health Sciences
University of Oxford

VERDICT
The only current COVID-19 specific data comparing OP with NP comes from two low quality, non-peer-reviewed studies and should be viewed with caution. It is not possible to accurately assess sensitivity from the existing data and there are no data to assess the diagnostic impact of combining both tests.

We found two COVID-19 studies, one with 213 patients and 205 OP and 490 NP samples (1) and another with nine patients and an unknown number of samples (2). The smaller study by Wolfel found no difference in detection rates between OP and NP but Yang, the larger of the two studies, reported OP swabs detected the COVID-19 virus less frequently than NP swabs and should not be used in place of NP swabs. This difference was most notable at days 8+ after illness onset, with about a 20-percentage point minimum difference in positive rates between OP and NP swabs. The difference was less at 0-7 days, with about 60% COVID-19 infected patients positive on OP vs about 70% on NP.

Both studies recruited hospitalised patients and it is unclear whether the primer used in the RT-PCR was the same as used in other countries. Applicability to other healthcare settings is therefore uncertain. Overall, the data are not robust but we would caution against relying on OP alone over NP.

BACKGROUND
SARS-CoV-2 is the virus being detected for tests for the illness COVID-19, for ease of communication this review will refer to COVID-19 virus testing. COVID-19 virus testing techniques are constantly evolving as the drive for faster tests, using less complex technology builds. Current NHS guidance for COVID-19 virus testing states that the preferred screening/testing technique is molecular diagnosis using real-time RT-PCR (RdRp gene) assay.

One issue is how to collect samples for testing, as the mechanism affects the viability of expanding to mass community testing by distributing home tests or using lay screeners. Current Public Health England guidance advises samples from the upper respiratory tract should be sought as EITHER:

•         individual nose and throat swabs in separate collection tubes
•         combined nose and throat swab in one collection tube containing universal transport medium
•         single swab used for throat then nose
•         nasopharyngeal aspirate

They also recommend collecting a lower respiratory tract sample in the form of sputum if obtainable.

The US CDC is currently recommending collecting only nasopharyngeal swabs (NP), not NP and oropharyngeal swabs (OP). However, OP would be easier to do without training than NP, so one question is whether there is a significant difference in sensitivity between the two methods or using them in combination.

SEARCH STRATEGY
Search terms: COVID-19, SARS-CoV-2, wuhan coronavirus, 2019-nCoV, 2019nCoV, nasopharyngeal, oropharyngeal, swab, diagnosis, diagnostic, test, sensitivity

Databases: PubMed, Embase, Web of Science, Scopus and Google Scholar. Google searches were also used to provide access to government, international institute and industry information. Searches were limited to 2020 because the first test was developed in January 2020.

Hits: 527 unique citations

Useful papers: 2.

In addition, one potentially relevant Chinese review by Ye and colleagues (3) was found but excluded because it was in Chinese and from the abstract seemed to be based on data from MERS/SARS/H1N1. Another paper by Wang et al (4) was excluded because on contacting the authors ‘pharyngeal’ was found to include either OP and NP whilst ‘nasal’ meant nostril.

SUMMARY OF FINDINGS FROM MOST USEFUL PAPERS 

COVID-19 virus testing: OP vs NP

We found two papers that compared OP vs NP, both were low quality, non-peer-reviewed preprints and should therefore be viewed with caution. Yang 2020 (1) examined 205 throat swabs and 490 nasal swabs (confirmed with author as OP and NP respectively) in 213 hospitalised COVID 19 patients in China. They reported positive test rates for these different swabs in both mild and severe cases at different time points after illness onset. They concluded that OP swabs should not be used for diagnosis. This was based on the following positive rates:

Day 0-7

Mild: OP 61.3% (46/75) NP 72.1% (147/204)

Severe: OP 60% (12/20) NP 73.3% (11/15)

Day 8-14

Mild: OP 29.6% (8/27) NP 53.6% (96/179)

Severe: OP 50% (18/36) NP 72.3% (34/47)

Day 15+

Mild: OP 11.1% (1/9) NP 54.5% (6/11)

Severe: OP 36.8% (14/38) NP 50% (17/34)

No significance calculations were performed.

This contrasts with a smaller study by Wolfel 2020 (2) which described OP and NP in nine COVID-19 patients in Germany. They reported all swabs (unknown number) from all patients taken on days 1-5 of symptoms tested positive. They also stated that swabs taken after day 5 had a detection rate of 45.95% and found no discernible differences in viral loads or detection rates when comparing NP and OP.

CRITICAL APPRAISAL CHECKLIST
The CASP checklist for diagnostic tests was used for critical appraisal. Both papers were found to be of low quality.

First, it is not possible to calculate sensitivity because both papers assume that all positives are true positives and that all negatives are true negatives. There are little public data on the false positive and false negative rates of the various RT-PCR based tests but this has been raised as a potential issue (see Ai 2020 (5) and Zhuang 2020 (6)– for the latter only abstract is in English). Test sensitivities may be affected by the primer used in the RT-PCR, which is worth noting when comparing papers where the primer is not stated, such as Yang, to UK data. For instance, Chan et al (7) described how the COVID-19-RdRp/Hel assay was positive for an additional 42 RdRd-P2-negative specimens [119/273 (43.6%) vs 77/273 (28.2%), P<0.001. It is unclear whether different primers would alter the relative sensitivities of different sample types.

Second, there is a lack of clarity over the reference standard the tests are being compared to. In the case of Yang it is presumably having a positive RT-PCR from any of the samples taken (OP, NP, sputum, BALF) but this is unclear. If this is the case, whilst understandably pragmatic in the circumstances, it means the reference standard has a fluid sensitivity and specificity, and that the reference standard and the test in question may be the same thing. This is also a problem in Wolfel where it appears the reference standard was a positive NP or OP.

Finally, neither paper provided a detailed description of sampling technique, which might be important in terms of interpreting the accuracy of each approach.

OP vs NP in other respiratory viruses

Given the minimal data on COVID-19 we also briefly looked at OP vs NP in other respiratory viruses. However previous research on OP vs NP in respiratory viruses is equivocal. For instance, in a study of 2,331 paired specimens 2011 Kim reported (8) “The sensitivities were variable among the eight viruses tested; neither specimen was consistently more effective than the other.” Their study did not include a coronavirus. Equally, in a smaller study in 2009 Lieberman (9) examined 12 respiratory viruses and found a significantly higher sensitivity for NP over OP, although this was not statistically significant for coronaviruses (the study was underpowered to detect differences within virus groups). With regards combining sampling methods, a 2019 review of 14 studies (10) examining influenza viruses also found NP typically had a higher sensitivity than OP, but “a combination of two less-invasive swabbing methods, such as nasal and oropharyngeal swabs, had about the same sensitivity as did nasopharyngeal specimens”. However, these results are from influenza and Kim’s data suggests virus specific variation in site sensitivity, moreover this was a descriptive review because the lack of homogeneity amongst the data precluded meta-analysis.

CONCLUSION
In short, whilst there is data that detection rates vary by sample site (see Wang 2020) the only current COVID-19 specific data comes from two low quality, non-peer-reviewed studies and should be viewed with caution. It is not possible to accurately assess sensitivity from the existing data and there are no data to assess the diagnostic impact of combining both tests. However, the larger of the two studies we found reported oropharyngeal swabs detected the COVID-19 virus less frequently than nasopharyngeal swabs and should not be used in place of nasopharyngeal swabs, particularly from day 8+ of symptom onset. Notably, both studies were based on hospitalised patients (though Wolfel included some tests from ultimately hospitalised patients who were also tested in the community) so applicability to primary care may be limited. Overall, the data are not robust but we would caution against relying on OP over NP.

Disclaimer: The article has not been peer-reviewed; it should not replace individual clinical judgement and the sources cited should be checked. The views expressed in this commentary represent the views of the authors and not necessarily those of the host institution, the NHS, the NIHR, or the Department of Health and Social Care. The views are not a substitute for professional medical advice.

REFERENCES

1.     Yang Y, Yang M, Shen C, Wang F, Yuan J, Li J, et al. Evaluating the accuracy of different respiratory specimens in the laboratory diagnosis and monitoring the viral shedding of 2019-nCoV infections. medRxiv. 2020 Feb 17;2020.02.11.20021493.
2.     Wolfel R, Corman VM, Guggemos W, Seilmaier M, Zange S, Mueller MA, et al. Clinical presentation and virological assessment of hospitalized cases of coronavirus disease 2019 in a travel-associated transmission cluster [Internet]. Infectious Diseases (except HIV/AIDS); 2020 Mar [cited 2020 Mar 20]. Available from: http://medrxiv.org/lookup/doi/10.1101/2020.03.05.20030502
3.     Ye B, Fan C, Pan Y, Ding R, Hu HX, Xiang ML. [Which sampling method for the upper respiratory tract specimen should be taken to  diagnose patient with COVID-19?]. Zhonghua Er Bi Yan Hou Tou Jing Wai Ke Za Zhi. 2020 Mar 13;55(0):E003.
4.     Wang W, Xu Y, Gao R, Lu R, Han K, Wu G, et al. Detection of SARS-CoV-2 in Different Types of Clinical Specimens. JAMA [Internet]. 2020 Mar 11 [cited 2020 Mar 20]; Available from: http://jamanetwork.com/journals/jama/fullarticle/2762997
5.     Ai T, Yang Z, Hou H, Zhan C, Chen C, Lv W, et al. Correlation of chest CT and RT-PCR testing in coronavirus disease 2019 (COVID-19) in China: a report of 1014 cases. Radiology. 2020;200642.
6.     Chan JF-W, Yip CC-Y, To KK-W, Tang TH-C, Wong SC-Y, Leung K-H, et al. Improved molecular diagnosis of COVID-19 by the novel, highly sensitive and specific COVID-19-RdRp/Hel real-time reverse transcription-polymerase chain reaction assay validated in vitro and with clinical specimens. J Clin Microbiol. 2020;
7.     Kim C, Ahmed JA, Eidex RB, Nyoka R, Waiboci LW, Erdman D, et al. Comparison of Nasopharyngeal and Oropharyngeal Swabs for the Diagnosis of Eight Respiratory Viruses by Real-Time Reverse Transcription-PCR Assays. PLoS ONE [Internet]. 2011 Jun 30 [cited 2020 Mar 25];6(6). Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3128075/
8.     Lieberman D, Lieberman D, Shimoni A, Keren-Naus A, Steinberg R, Shemer-Avni Y. Identification of Respiratory Viruses in Adults: Nasopharyngeal versus Oropharyngeal Sampling. J Clin Microbiol. 2009 Nov 1;47(11):3439–43.
9.   Spencer S, Thompson MG, Flannery B, Fry A. Comparison of Respiratory Specimen Collection Methods for Detection of Influenza Virus Infection by Reverse Transcription-PCR: a Literature Review. J Clin Microbiol [Internet]. 2019 Aug 26 [cited 2020 Mar 25];57(9). Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6711916/