Chat with us, powered by LiveChat Health care | Abc Paper
+1(978)310-4246 credencewriters@gmail.com
  

RESEARCH ARTICLE

The accuracy of healthcare worker versus self

collected (2-in-1) Oropharyngeal and Bilateral

Mid-Turbinate (OPMT) swabs and saliva

samples for SARS-CoV-2

Seow Yen TanID
1☯*, Hong Liang Tey2☯, Ernest Tian Hong Lim3☯, Song Tar Toh4☯, Yiong

Huak Chan
5
, Pei Ting Tan

6
, Sing Ai Lee

7
, Cheryl Xiaotong Tan

8
, Gerald Choon Huat Koh

9‡
,

Thean Yen Tan
10‡

, Chuin Siau
11‡

1 Department of Infectious Diseases, Changi General Hospital, Singapore, Singapore, 2 Department of

Dermatology, National Skin Centre, Singapore, Singapore, 3 Emergency Department, Woodlands Health

Campus, Singapore, Singapore, 4 Department of Otorhinolaryngology- Head and Neck Surgery, Singapore

General Hospital, Singapore, Singapore, 5 Biostatistics Unit, Yong Loo Lin School of Medicine, Singapore,

Singapore, 6 Clinical Trials and Research Unit, Changi General Hospital, Singapore, Singapore, 7 Sheares

Healthcare Group Pte Ltd, Singapore, Singapore, 8 Temasek International Pte Ltd, Singapore, Singapore,

9 MOH Office for Healthcare Transformation, Singapore, Singapore, 10 Department of Laboratory Medicine,

Changi General Hospital, Singapore, Singapore, 11 Department of Respiratory & Critical Care Medicine,

Changi General Hospital, Singapore, Singapore

☯ These authors contributed equally to this work.
‡ These authors are joint senior authors on this work.

* [email protected]

Abstract

Background

Self-sampling for SARS-CoV-2 would significantly raise testing capacity and reduce healthcare

worker (HCW) exposure to infectious droplets personal, and protective equipment (PPE) use.

Methods

We conducted a diagnostic accuracy study where subjects with a confirmed diagnosis of

COVID-19 (n = 401) and healthy volunteers (n = 100) were asked to self-swab from their

oropharynx and mid-turbinate (OPMT), and self-collect saliva. The results of these samples

were compared to an OPMT performed by a HCW in the same patient at the same session.

Results

In subjects confirmed to have COVID-19, the sensitivities of the HCW-swab, self-swab,

saliva, and combined self-swab plus saliva samples were 82.8%, 75.1%, 74.3% and 86.5%

respectively. All samples obtained from healthy volunteers were tested negative. Compared

to HCW-swab, the sensitivities of a self-swab sample and saliva sample were inferior by

8.7% (95%CI: 2.4% to 15.0%, p = 0.006) and 9.5% (95%CI: 3.1% to 15.8%, p = 0.003)

respectively. The combined detection rate of self-swab and saliva had a sensitivity of 2.7%

(95%CI: -2.6% to 8.0%, p = 0.321). The sensitivity of both the self-collection methods are

higher when the Ct value of the HCW swab is less than 30. The specificity of both the self-

swab and saliva testing was 100% (95% CI 96.4% to 100%).

PLOS ONE

PLOS ONE | https://doi.org/10.1371/journal.pone.0244417 December 16, 2020 1 / 11

a1111111111

a1111111111

a1111111111

a1111111111

a1111111111

OPEN ACCESS

Citation: Tan SY, Tey HL, Lim ETH, Toh ST, Chan

YH, Tan PT, et al. (2020) The accuracy of

healthcare worker versus self collected (2-in-1)

Oropharyngeal and Bilateral Mid-Turbinate (OPMT)

swabs and saliva samples for SARS-CoV-2. PLoS

ONE 15(12): e0244417. https://doi.org/10.1371/

journal.pone.0244417

Editor: Dong-Yan Jin, University of Hong Kong,

HONG KONG

Received: September 10, 2020

Accepted: December 9, 2020

Published: December 16, 2020

Peer Review History: PLOS recognizes the

benefits of transparency in the peer review

process; therefore, we enable the publication of

all of the content of peer review and author

responses alongside final, published articles. The

editorial history of this article is available here:

https://doi.org/10.1371/journal.pone.0244417

Copyright: © 2020 Tan et al. This is an open access
article distributed under the terms of the Creative

Commons Attribution License, which permits

unrestricted use, distribution, and reproduction in

any medium, provided the original author and

source are credited.

Data Availability Statement: All relevant data are

within the manuscript and its Supporting

Information files

https://orcid.org/0000-0001-7870-2879

https://doi.org/10.1371/journal.pone.0244417

http://crossmark.crossref.org/dialog/?doi=10.1371/journal.pone.0244417&domain=pdf&date_stamp=2020-12-16

http://crossmark.crossref.org/dialog/?doi=10.1371/journal.pone.0244417&domain=pdf&date_stamp=2020-12-16

http://crossmark.crossref.org/dialog/?doi=10.1371/journal.pone.0244417&domain=pdf&date_stamp=2020-12-16

http://crossmark.crossref.org/dialog/?doi=10.1371/journal.pone.0244417&domain=pdf&date_stamp=2020-12-16

http://crossmark.crossref.org/dialog/?doi=10.1371/journal.pone.0244417&domain=pdf&date_stamp=2020-12-16

http://crossmark.crossref.org/dialog/?doi=10.1371/journal.pone.0244417&domain=pdf&date_stamp=2020-12-16

https://doi.org/10.1371/journal.pone.0244417

https://doi.org/10.1371/journal.pone.0244417

https://doi.org/10.1371/journal.pone.0244417

http://creativecommons.org/licenses/by/4.0/

http://creativecommons.org/licenses/by/4.0/

Conclusion

Our study provides evidence that sensitivities of self-collected OPMT swab and saliva sam-

ples were inferior to a HCW swab, but they could still be useful testing tools in the appropri-

ate clinical settings.

Introduction

The current “gold standard” for testing for SARS-CoV-2 requires health care workers to collect

a nasopharyngeal (NP) sample from a patient. NP sampling is very uncomfortable for the

patient and requires deployment of trained personnel and use of personal protective equip-

ment (PPE) which are in limited supply.

A prior study has shown that a combination of oropharyngeal and anterior nares swabs is

equivalent in sensitivity to an NP swab in 190 ambulatory symptomatic patients [1]. In another

study on 236 ambulatory subjects, the performance of self-collected nasal and throat swabs is

at least equivalent to that of health worker collected swabs for the detection of SARS-CoV-2

and other respiratory viruses [2].

The international community is actively searching for an even less invasive means of sample

collection: saliva. In a recent study by Yale University on 29 subjects [3], it was suggested that a

large volume sample of saliva collected from COVID-19 inpatients can be more sensitive than

NP swabs for SARS-CoV-2 detection, and saliva samples had significantly higher COVID-19

viral titres than NP swabs (p = 0.001). Furthermore, the same study showed that sensitivity of

COVID-19 in saliva was more consistent throughout extended hospitalization compared to

NP swabs.

In addition, there are a number of studies done on saliva testing for COVID-19 which have

shown promising results, reporting 91.7%, and 100% positivity in saliva samples of COVID-19

patients [4, 5]. Iwasaki et al found an overall concordance rate of 97.4% for COVID-19 detec-

tion with a strong concordance between NP swabs and saliva sampling (κ = 0.874) among 66
COVID-19 negative and 10 COVID-19 positive subjects [6]. Furthermore, a study done by To

et al. showed that viral RNA could still be detected in saliva samples in a third of their twenty-

three patients 20 days or longer after symptoms onset despite the development of COVID-19

antibodies [7]. A meta-analysis conducted on 26 saliva studies also showed a positive detection

rate of 91%, comparable to the detection rate of 98% from nasopharyngeal swabs [8]. All these

studies had small sample sizes (all <30 COVID-positive subjects) and only one study also sam-

pled COVID-negative subjects.

It is still currently unknown whether a self-collected combined Oropharyngeal and Bilateral

Mid-Turbinate (OPMT) sample, or a self-collected saliva sample is equivalent to a swab done

by a health care worker (HCW). If the self-collection of samples is proven to be a reliable alter-

native to a HCW swab, it would reduce the reliance of trained personnel to collect samples and

enable a rapid increase in testing capacity. It would also reduce greatly the biosafety risk that is

posed to HCWs and help with PPE conservation efforts.

Materials and methods

Study design and trial oversight

This was a prospective study involving 401 subjects who were previously tested positive for

COVID-19 by RT-PCR, and 100 healthy volunteers. This study was approved by the

PLOS ONE Accuracy of self-testing for COVID-19

PLOS ONE | https://doi.org/10.1371/journal.pone.0244417 December 16, 2020 2 / 11

Funding: This study was funded by Sheares

Healthcare Group Pte Ltd. The funder provided

support in the form of salaries for authors SAL and

CXT, but did not have any additional role in the

study design, data collection and analysis, decision

to publish, or preparation of the manuscript. The

specific roles of these authors are articulated in the

‘author contributions’ section. Besides that, author

CXT is employed by Temasek International Pte Ltd,

and was acting on behalf of Sheares Healthcare

Group Pte Ltd for the study. Temasek International

Pte Ltd did not have any additional role in the study

design, data collection and analysis, decision to

publish, or preparation of the manuscript.

Competing interests: Author CXT is employed by

Temasek International Pte Ltd, and was acting on

behalf of Sheares Healthcare Group Pte Ltd for the

study. This commercial affiliation does not alter our

adherence to PLOS ONE policies on sharing data

and materials.

https://doi.org/10.1371/journal.pone.0244417

SingHealth Centralised Institutional Review Board. Written informed consent was obtained

from the subjects.

Participants

The first group consisted of patients who were confirmed to have COVID-19, and who were

cared for in either a hospital (Changi General Hospital), or a community care facility (Com-

munity Care Facility @ EXPO). Diagnosis of COVID-19 was confirmed via a positive RT-PCR

from a nasopharyngeal swab. The subjects in this group were recruited within 3 days of admis-

sion to the study site and they were recruited from 31 May 2020 to 10 June 2020. The patients

who were eligible were approached directly at the study site, and were invited to participate in

the study, and the study procedures were carried out on the same day. Recruitment was carried

out until the target sample size was achieved. At the time of the study, the majority of COVID-

19 cases belong to the migrant worker population, which primarily consisted of healthy young

male adults, mainly from Bangladesh and India. Hence, this group of subjects is not represen-

tative of the general population in Singapore.

Inclusion criteria applicable to this group include:

• Male and female patients, � 21 years-old

• Tested positive for COVID-19

• Admitted to study site within the previous 3 days

• Ability to provide informed consent

• Compliance with all aspects of study protocol, methods and provision of samples

• Ability to read and understand English

Exclusion criteria applicable to this group include:

• Nosebleeds in past 24 hours

• Previous nasal surgery in past 4 weeks

• Acute facial trauma within 8 weeks

• Unable to demonstrate understanding of study and instructions

• Experienced severe adverse reactions on prior nose and/or throat swabs

• Not willing to have all 3 samples collected

The second group comprised 100 healthy volunteers who were asymptomatic and well on

the day of the study, with no recent COVID-19 exposure. This was done on 18 and 19 July

2020. The study subjects were recruited via an open advertisement.

Inclusion Criteria for this group include:

• Males and females, � 21 years-old

• Ability to provide informed consent

• Capable of understanding and complying with the requirements of the study

• Ability to read and understand English

Exclusion Criteria applicable to this group were:

• Displaying symptoms of an acute respiratory infection

PLOS ONE Accuracy of self-testing for COVID-19

PLOS ONE | https://doi.org/10.1371/journal.pone.0244417 December 16, 2020 3 / 11

https://doi.org/10.1371/journal.pone.0244417

• Known close contact with an individual diagnosed with COVID-19 within the last 3 months

• Previously diagnosed with COVID-19

• Nosebleeds in past 24 hours

• Previous nasal surgery in past 4 weeks

• Acute facial trauma within 8 weeks

• Unable to demonstrate understanding of study and instructions

• Experienced severe adverse reactions on prior nose and/or throat swabs

• Not willing to have all 3 samples collected

Test procedures

Study subjects underwent three sequential test sample collection procedures within one study

visit in the following order:

1. Each subject self-collected a sample combining OP and bilateral MT swabs using a single

swab stick;

2. A trained healthcare worker then collected a combined OP and bilateral MT swab using

another single swab stick;

3. The subject then self-collected a saliva sample.

Study subjects were shown instructional videos for both the OPMT self-swab and saliva col-

lection prior to commencing the test procedures. Study team members were present on site to

observe and supervise the self-collection process. Posterior oropharyngeal saliva, commonly

described as deep throat saliva was collected for this study. Synthetic fibre swabs were used for

collection of the OP and MT samples by both subject and healthcare worker, and immediately

placed in universal transport medium (UTM), while saliva samples were collected using the

SAFER-Sample™ (by Lucence Diagnostics). All samples were double bagged and stored at air-
conditioned room temperature in a chiller bag and transported to assigned laboratory on the

same day. Upon arrival in the laboratory, they were stored at 2˚C to 8˚C. All samples were pro-

cessed with 24 hours of sample collection.

Nucleic acid extraction was performed using PerkinElmer Nucleic Acid Extraction Kits

(KN0212) on the Pre-Nat II Automated Workstation (PerkinElmer1, United States), Extrac-

tion of swab samples followed the indicated protocol for oropharyngeal swabs, while extraction

of saliva samples followed a protocol consisting of pre-liquefaction with dithiothreitol (protocol

attached in S1 File). Reverse transcription polymerase chain reaction (RT-PCR) was performed

on the Quantstudio
TM

5 Real Time PCR system (Thermo Fisher, United Kingdom) using the

PerkinElmer1 SARS-CoV-2 Real-time RT-PCR Assay. The targets were the ‘N’ gene and

‘ORF1ab’ gene. There is an internal control target that is present in every RT-PCR reaction. The

cycle threshold (Ct) values of the ‘N’ gene were used in the analysis involving Ct values.

Outcomes

The primary objective of the study was to evaluate the accuracy of self-collected (2-in-1) OPMT

swabs and self-collected saliva samples for SARS-CoV-2 versus that of HCW-collected (2-in-1)

MT and OP swabs. The secondary objective was to evaluate the correlation of PCR Ct values of

self-collected saliva samples and swabs with comparator healthcare worker-collected swabs.

PLOS ONE Accuracy of self-testing for COVID-19

PLOS ONE | https://doi.org/10.1371/journal.pone.0244417 December 16, 2020 4 / 11

https://doi.org/10.1371/journal.pone.0244417

Sample size

Firstly, we postulated that OPMT self-swabbing was as accurate as HCW-obtained swabs. Pos-

tulating a 100% accuracy, 400 subjects will be required to achieve a lower 95% confidence

interval 99.1% (which gives a less than 1% error rate). With the computed sample size of 400

subjects, a non-inferiority could be achieved with at most a 7% difference for OPMT self-swab-

bing compared to the HCW-obtained swabs. If the study included only subjects who were

diagnosed with COVID-19, all positive results would be regarded as true positives. Hence, to

address that gap in the form of specificity of self-swabs and saliva testing in the diagnosis of

COVID-19, a further study on 100 healthy subjects was conducted. The hypothesis was that

with 100% accuracy, the error rate for a false negative was 3.6%.

Statistical analysis

All analyses were performed using SPSS 25.0 with statistical significance set at p < 0.05.

The estimates for the positivity results of the 3 methods were presented as numbers and

percentages. The differences with 95% confidence interval (CI) between self-collection meth-

ods and HCW-obtained swabs to assess for non-inferiority was calculated. Sensitivity and

specificity of the two self-collection methods were compared with HCW-obtained swabs and

results were stratified by Ct values. Spearman’s test was used to assess the correlation of the

PCR Ct values across the 3 groups.

Results

A total of 401 COVID-19 positive and 100 COVID-19 negative subjects were recruited. Of the

401 COVID-19 positive subjects, 23 were recruited from Changi General Hospital, and 378

were recruited from the community care facility @ Expo. The symptomatic COVID-19 posi-

tive subjects that were recruited were well patients, whose clinical presentation was that of an

upper respiratory tract infection. None of the subjects required oxygen supplementation.

Only the demographic data of subjects from Changi General Hospital was known. The full

demographic data of the subjects that were admitted to the community care facility could not

be made available to us due to prevailing regulations of the study site during the period when

the study was conducted, hence we do not have the data of the age of the subjects that were

admitted to the community care facility. However, we were able to surmise that the age range

of patients admitted to the community care facility was 21 to 45, due to the admission criteria

to the facility, and the inclusion criteria for the study. A summary of the profile of recruited

subjects are listed in Tables 1 and 2 below.

All subjects went through the test procedures—500 participants (400 COVID-19 positive,

100 COVID-19 negative) were able to provide all 3 samples, and one subject was unable to

provide a saliva sample despite a prolonged attempt. All participants tolerated the test proce-

dures well and did not experience any adverse events.

In the group of subjects who were COVID-19 positive, twenty-seven (6.7%) patients were

tested negative across all 3 samples. This may be explained by the fact that they are recovering

and viral shedding may have ceased at point of testing. Forty-two (10.5%) subjects reported

�1 symptom of acute respiratory infection (ARI) (e.g. fever, cough, rhinorrhoea, sore throat,

malaise) on the day of study recruitment while 371 (92.5%) subjects reported being within 7

days from onset of COVID-19 illness.

The detection rates of the HCW swab, self-swab, saliva, and combined self-swab plus saliva

samples were 82.8%, 75.1%, 74.3% and 86.5% respectively (Table 3). Compared to HCW-

swabs, the detection rate was lower for self-swab by 8.7% (95% confidence interval, CI = 2.4%

to 15.0%, p = 0.006) and for saliva samples by 9.5% (95%CI = 3.1% to 15.8%, p = 0.003). When

PLOS ONE Accuracy of self-testing for COVID-19

PLOS ONE | https://doi.org/10.1371/journal.pone.0244417 December 16, 2020 5 / 11

https://doi.org/10.1371/journal.pone.0244417

the results of both the self-swab and saliva testing were combined, the detection rate was

higher by 2.7% (95%CI = -2.6% to 8.0%, p = 0.321) but this was not statistically significant.

The sensitivities of the self-swab, saliva and combined self-swab plus saliva testing, when

compared to the HCW swab were 83.6%, 80.6% and 92.3% respectively. Table 4 shows the con-

tingency tables comparing the HCW swab vs self-swab; HCW swab vs saliva, and HCW vs

combined self-swab plus saliva respectively.

Using the Ct values (‘N’ gene) of HCW swabs as reference, 3 categories of Ct values (i.e. 30) were studied. It was observed that the sensitivity of self-swab (Table 5) and

saliva testing (Table 6) performed better at the lower Ct values, suggesting that the sensitivity of

self-collection methods approaches to that of HCW swab, when the viral load was higher.

There was a good correlation of PCR Ct values between self-swab and HCW swab (r = 0.825,

p<0.001) but moderate correlation between saliva samples and HCW swab (r = 0.528, p<0.001).

The self-swab has a better agreement with the HCW swab. Using Wilcoxon Signed Rank Test,

the difference in CT values between self-swab and HCW swab is statistically significant, where

p = 0.026. Similarly for the saliva and HCW swab, where p = 30

(i.e. when viral load is low), and the SARS-CoV-2 virus often cannot be isolated or cultured

after day 11 of illness [10]. Thus, the results of this study support the use of self-testing

methods as a replacement for a HCW swab in the early phase of COVID-19 illness when

viral loads are high, and the sensitivities of the self-swab and saliva are similar to that of the

HCW swab.

The strength of our study is the large number of subjects confirmed to have COVID-

19. Besides that, the study also included a high proportion of asymptomatic individuals

who were picked up because of Singapore’s proactive mass screening policy. The combi-

nation of self-swab and saliva sampling performed well in these asymptomatic subjects,

implying that the strategy of combined self-testing, has the ability diagnose COVID-19 in

asymptomatic individuals with a sensitivity equivalent to that of a swab by a HCW. The

study results from the healthy volunteers indicate a low false positive rate with self-collec-

tion methods.

These findings, indicate that self-collection methods may be a useful tool for COVID-19

surveillance in the asymptomatic individuals, and in situations where testing capacity needs to

be scaled up rapidly, without a need for large increase of manpower, and without increased

infectious exposure to the swabbing staff. Testing strategies can be tailored based on the target

population and the intended use of the various tests on its own or in combination.

Table 3. Detection rates of various modalities in all subjects.

HCW Swab Self-Swab Saliva Self-Swab + Saliva

Count 336 301 297 347

Percentage 83.8% 75.1% 74.3% 86.5%

95% CI 79.8% – 87.3% 70.1% – 79.2% 69.7% – 78.5% 82.8% – 89.7%

https://doi.org/10.1371/journal.pone.0244417.t003

Table 4. Comparison between HCW swabs and the self-swab/saliva.

HCW Swab

Not detected Detected p value�

Self-swab

Not detected 45 55 <0.001

Detected 20 281 (83.6)

Saliva

Not detected 37 65 <0.001

Detected 27 270

Self-swab plus saliva

Not detected 27 26 0.207

Detected 37 310

� p value was obtained from McNemar test

https://doi.org/10.1371/journal.pone.0244417.t004

PLOS ONE Accuracy of self-testing for COVID-19

PLOS ONE | https://doi.org/10.1371/journal.pone.0244417 December 16, 2020 7 / 11

https://doi.org/10.1371/journal.pone.0244417.t003

https://doi.org/10.1371/journal.pone.0244417.t004

https://doi.org/10.1371/journal.pone.0244417

The way the study findings were presented are unlike most studies involving saliva testing

for COVID-19. This is probably due to the fact that our study is carried out on subjects who

are already known to have COVID-19, unlike most studies which are done in testing centres

where the potential subjects’ results are still unknown. This also meant that the sampling was

done later in the subjects’ trajectory of illness, as they were first tested positive for COVID-19,

then enrolled into the study. The later sampling possibly had a negative impact on the sensitiv-

ity of the saliva [11].

Another key study limitation, is that the demographics of the COVID-19-positive popula-

tion was skewed, consisting solely of male migrant workers, the worst affected group of the

pandemic in Singapore, at the time this study was conducted. Hence the results from this

study might not be applicable to the general population, without the inclusion of paediatric

and elderly population segments. The migrant worker population in this study, which consist

of generally young and healthy males, is also not representative of the demographics of

Singapore.

The addition of the stabilising solution to the deep throat saliva sample, could have also

decreased the yield of the saliva testing. Studies utilising saliva test kits that do not require the

addition of stabilising fluid generally report equivalent sensitivities of the saliva test when com-

pared to a HCW swab [3, 12]. Hence the use of stabilising solution is a key consideration in

future design of saliva test kits.

The study team members observed that, despite clear instructions, many subjects still

needed guidance with the self-collection methods. For the self-swab, the most commonly

encountered scenario was that, the subjects needed guidance in breaking the swab stick. The

saliva collection presented a greater challenge to the subjects. The flow of saliva from the fun-

nel into the collection container was not smooth, and the additional step of adding the stabilis-

ing fluid required prompting. These necessitated the presence of a trained staff to troubleshoot

and ensure that the correct steps are carried out. We believe that these observations are useful

in the re-design of collection containers to enhance results and end users’ acceptability. Both

the self-swab and saliva collection require dexterity and this would limit its applicability in seg-

ments of the population who are not able to do so.

We caution against widespread, unsupervised implementation of self-collection methods.

The reliability and effectiveness of self-collection methods may also be dependent on social

and economic drivers, hence potentially influencing the test performance. For example, indi-

viduals who face a potential loss of income or unemployment if tested positive or travellers

having a test done at immigration clearance may deliberately do a suboptimal self-test to influ-

ence the test outcome.

Table 5. Sensitivity of self-swab, stratified by Ct values of HCW swab.

HCW Swab Ct Number of subjects Sensitivity

30 195 73.3% (66.5–79.4)

https://doi.org/10.1371/journal.pone.0244417.t005

Table 6. Sensitivity of saliva, stratified by Ct values of HCW swab.

HCW Swab Ct Number of subjects Sensitivity

30 194 70.6% (63.7–76.9)

https://doi.org/10.1371/journal.pone.0244417.t006

PLOS ONE Accuracy of self-testing for COVID-19

PLOS ONE | https://doi.org/10.1371/journal.pone.0244417 December 16, 2020 8 / 11

https://doi.org/10.1371/journal.pone.0244417.t005

https://doi.org/10.1371/journal.pone.0244417.t006

https://doi.org/10.1371/journal.pone.0244417

Hence, it is important to have designated personnel to supervise the self-collection process,

ensuring that the correct test procedures are carried out. These personnel need not be a HCW

and the supervision process will have a lower exposure risk (supervisor can be >1m away from

subject), compared to the HCW-swabbing process where a HCW is <1m away and face-to-

face with the subject.

Conclusion

This study demonstrates that while self-collection methods have a sensitivity of approximately

75%, it is inferior to the rate obtained by the health care worker administered swab (83.8%). The

sensitivity of the self-collection methods is, however, higher and correlates better when Ct values

of the HCW swabs are less than 30. The combined results of the saliva and self-swab test achieve a

sensitivity equivalent to that of a health care worker administered swab. The specificity of the self-

collection methods is 100%. Together with high specificity, we postulate that …

Research Letter | Infectious Diseases

Comparison of Unsupervised Home Self-collected Midnasal Swabs
With Clinician-Collected Nasopharyngeal Swabs for Detection
of SARS-CoV-2 Infection
Denise J. McCulloch, MD, MPH; Ashley E. Kim, BS; Naomi C. Wilcox, MPH; Jennifer K. Logue, BS; Alex L. Greninger, MD, PhD; Janet A. Englund, MD; Helen Y. Chu, MD, MPH

Introduction

Increased diagnostics are urgently needed to contain the spread of coronavirus disease 2019
(COVID-19). Home self-collected swabs may increase testing access while minimizing exposure risk
to health care workers and depletion of personal protective equipment, allowing for early community
detection of COVID-19. A comparison of unsupervised home self-collected swabs with clinician-
collected nasopharyngeal swabs for COVID-19 diagnosis has not been well described.

Methods

This cross-sectional study was approved by the University of Washington institutional review board
and follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)
reporting guideline. Participants provided electronic informed consent. Study participants were
recruited from symptomatic outpatients testing severe acute respiratory syndrome coronavirus 2
(SARS-CoV-2)–positive and symptomatic health care workers presenting to drive-through clinics (eFigure
and eAppendix in the Supplement). Participants were provided test kits for unsupervised home self-
collection of a midnasal swab. Home swab performance was compared with clinician-collected
nasopharyngeal swabs, which were collected by medical assistants and nurses. Cycle thresholds (Ct)
are a semiquantitative measure of viral load. Positive test results for SARS-CoV-2 by both approaches
were defined as true positives. Results with a positive clinician swab and negative home swab were
defined as false negatives. Sensitivity was defined as true positives divided by the sum of true positives
and false negatives. Cohen κ was calculated for agreement between the 2 qualitative test results. The
threshold for statistical significance was set at 2-tailed P < .05.

Results

Of 185 total participants, 158 (85%) enrolled at drive-through clinics, and 27 (15%) enrolled after a
positive SARS-CoV-2 test. Among the 185 participants, 41 (22.2%) yielded SARS-CoV-2 positive test
results via clinician-collected nasopharyngeal swab, home self-collected midnasal swab, or both. One
hundred fifty-eight participants (85%) were health care workers, of whom 14 (9%) tested positive.
Among participants with COVID-19, common symptoms included myalgia (33 participants [80.5%]),
cough (28 participants [68.3%]), and fever (26 participants [63.4%]). Compared with clinician
swabs, sensitivity and specificity of home swabs was 80.0% (95% CI, 63%-91%) and 97.9% (95% CI,
94%-99.5%), respectively (Table). Cohen κ statistic was 0.81 (95% CI, 0.70-0.93), suggesting
substantial agreement.

Cycle thresholds of home swabs were positively correlated with clinician swabs (correlation
coefficient, 0.81; P 200
countries, resulting in over 2.97 million confirmed patients
and 206  000 deaths as of April 27, 2020.1 Since its outbreak
last year, research groups used whole genome/RNA sequenc-
ing and identified viral cause of COVID-19, which possesses
genetic sequence with ~80% similarity to the genome of the
severe acute respiratory syndrome virus coronavirus 2 (SARS-
CoV).2–4 The novel coronavirus was hence named SARS-CoV-2.
Currently, the most likely transmission route is direct contact
and/or air droplet spread,5,6 which is backed up by the find-
ings that SARS-CoV-2 can be isolated in aerosol (<5 µm) for

at least up to 3 hours.7 Unfortunately, US Food and Drug
Administration (FDA) has yet approved any vaccines or thera-
peutics in clinical use for SARS-CoV-2, and most countries that
successfully limit the spread of COVID-19, including Taiwan,
primarily rely on rapid case screening, identification, quaran-
tine, and contact tracing. As the symptomatic signs (44%-89%
fever, 68% cough, 38% fatigue, 34% sputum production, and
19% short of breath) and computed tomographic scans are
non-specific,8 molecular techniques become the gold standard
for COVID-19 diagnosis.1

2. CURRENT FIRST-LINE DIAGNOSTIC TEST
FOR COVID-19
The reverse transcription polymerase chain reaction (RT-PCR)
to detect SARS-CoV-2 is thus far the primary method for diag-
nosis of COVID-19.1 The clinical specimens for RT-PCR can be
obtained from upper respiratory tract by nasopharyngeal swabs,
washes, aspirates, or oropharyngeal swabs, or from lower respir-
atory tract by sputum collection, bronchoalveolar lavage (BAL),
or tracheal aspirates. The specific region serving as the targets
for the PCR include the RdRP (RNA-dependent RNA polymer-
ase) gene, the E (envelop protein) gene, or the N (nucleocapsid)
gene.9,10 Meanwhile, the serology tests that examine the produc-
tion of specific IgM and IgG antibodies against SARS-CoV-2 in
response to infection is also useful for surveillance and of value
to complement certain limitations of PCR as a sole diagnostic
tool. According to US FDA, the SARS-CoV-2 antibodies can

*Address correspondence. Dr. Cheng-Hsien Wu and Dr. Shou-Yen Kao,
Department of Stomatology, Taipei Veterans General Hospital, 201, Section 2,
Shi-Pai Road, Taipei 112, Taiwan, ROC. E-mail address: [email protected]
(S.-Y. Kao). (C.-H. Wu)

Author Contributions: Dr. Kai-Feng Hung and Dr. Yi-Chen Sun contributed equally
to this work.

Conflicts of interest: The authors declare that they have no conflicts of interest
related to the subject matter or materials discussed in this article.

Journal of Chinese Medical Association. (2020) 83: 891-894.

Received April 27, 2020; accepted April 27, 2020.

doi: 10.1097/JCMA.0000000000000396.
Copyright © 2020, the Chinese Medical Association. This is an open access
article under the CC BY-NC-ND license (http://creativecommons.org/licenses/
by-nc-nd/4.0/)

Abstract: As of April 15, 2020, the US Food and Drug Administration has granted emergency use authorization to a first
saliva test for diagnosis of severe acute respiratory syndrome coronavirus 2 infection, the device developed by RUCDR Infinite
Biologics laboratory, Rutgers University. A key feature that distinguishes the saliva-based test from nasopharyngeal or oropharyn-
geal (throat) swabs is that this kit allows self-collection and can spare healthcare professionals to be at risk during collecting
nasopharyngeal or oropharyngeal samples, thereby preserving personal protective equipment for use in patient care rather than
sampling and testing. Consequently, broader testing than the current methods of nasal or throat swabs will significantly increase
the number of people screening, leading to more effective control of the spread of COVID-19. Nonetheless, a comparison of
saliva-based assay with current swab test is needed to understand what and how we can benefit from this newly developed assay.
Therefore, in this mini-review article, we aimed to summarize the current and emerging tools, focusing on diagnostic power of dif-
ferent clinical sampling and specimens.

Keywords: Nasopharynx; Oropharynx; Saliva, Severe acute respiratory syndrome coronavirus 2

mailto:[email protected]

892 www.ejcma.org

Hung et al J Chin Med Assoc

be detected several days after initial infection and can still be
detectable afterward, thus providing a long period of window
for indirectly detecting SARS-CoV-2 for both active and recent
past infections.11 However, as serological assays are currently in
development and several challenges remain (such as the cross-
reactivity with other virus, as well as the undetermined kinetics
of immune response), the RT-PCR still play a pivotal role in the
identification of SARS-CoV-2 infection.

3. CLINICALLY RELEVANT ISSUES OF COVID-19
TESTING
Just like the concerns from public health experts for any of the
pandemic, two issues of diagnostic testing worth further con-
sideration. In addition to the criteria of who needs to be tested,
an important issue relates to the diagnostics itself. Specifically,
for RT-PCR, while a positive test result certainly identifies the
presence of virus, a negative result may not necessarily rule
out SARS-CoV-2 infection. The potential false-negative result
could be caused by low virus loads, improper sampling sites and
timings, poor technique, and even mutations of viral genome.
About the clinical sampling, the US Centers for Disease Control
and Prevention (CDC) guideline recommends collecting upper
respiratory specimen for asymptomatic patients. For patients
who develop a productive cough, sputum can be used for SARS-
CoV-2 testing, although the induction of sputum is not recom-
mended. However, nasopharyngeal swab sampling is technically
challenging, requires healthcare professionals, and may impose
risk for aerosol generation. These drawbacks thus necessitate
the implementation of additional diagnostic approach.

4. DESCRIPTION AND PRINCIPLE OF SALIVA-
BASED COVID-19 TESTING
The newly approved saliva-based COVID-19 testing kit is built
on the existing TaqPath SARS-CoV Assay, developed by the
Rutgers Clinical Genomics Laboratory, to qualitatively iden-
tify RNA from virus. This assay employs primers and probes
validated by the emergency use authorization (EUA) for respira-
tory, nasopharyngeal, and oropharyngeal specimens. To enable
testing saliva specimen, the collection protocols and nucleic
acid extraction buffers are modified. Saliva specimens can be
transported and stored at ambient temperature but have to be
processed within 48 hours of collection. The recommended sys-
tem for RNA extraction is the PerkinElmer Chemagic 360 with
Chemagic Viral DNA/RNA 300 Kit H96. The RT-PCR can be
performed using Applied Biosystems TaqPath Combo Kit on the
ThermoFisher Applied Biosystems QuantStudio 5 Real-Time
PCR System or the Applied Biosystems ViiA7 Real-Time PCR
System. The logistics and details can be found at https://www.
fda.gov/media/136875/download.

5. COMPARISON OF DIFFERENT TYPES OF
CLINICAL SAMPLES AND SPECIMENS FOR SARS-
COV-2 DETECTION
As mentioned previously, an accurate identification of respira-
tory viruses is critically affected by the source of clinical speci-
mens. While several studies on up to 15 common respiratory
viruses suggest that the use of nasopharyngeal swabs provides
a higher sensitivity than that of nasopharyngeal washings or
oropharyngeal swabs,12,13 this is not necessarily the case for
SARS-CoV-2, as the infectivity and the predilection for trans-
mission may differ significantly between viruses. In addition,
even if a given type of clinical specimen offers a relatively higher
accuracy in diagnosis, it remains an open question whether the

technique-demanding test is the most needed during a pandemic
with global shortage of medical supplies as of today.

Currently, the available data comparing the sensitivity for
SARS-CoV-2 detection using nasal, pharyngeal, or oral swab
are very limited. One study from a Chinese group examined
213 hospitalized SARS-CoV-2 patients with a total of 205
oropharyngeal and 490 nasopharyngeal swabs at various time
points of disease course. They found that nasopharyngeal
swabs have overall higher positive rates (53.6%-73.3%) than
oropharyngeal (throat) swabs (11.1%-61.3%), regardless of
whether the patients were in mild or severe disease conditions.
Notably, this study showed highest positive rate using sputum
specimens, which is generally regarded as a type of lower res-
piratory tract sample.14 Separately, a study examining the sen-
sitivity of SARS-CoV-2 detection with different clinical samples
from 205 patients in China showed that BAL fluid has highest
positive rate (93%), followed by sputum (72%), nasal swabs
(63%), brush biopsy (46%), and pharyngeal swabs (32%).15
In contrast to the findings from these studies, another study
examining nine hospitalized COVID-19 patients in Germany
showed that there are no discernible differences in virus loads
or positive rates between nasopharyngeal versus oropharyngeal
swabs, with an overall detection rate of 45.95% being reported,
although the numbers of nasopharyngeal and oropharyngeal
swabs taken were not described. Notably, this study found that
only two among nine patients have higher virus load (>3 in
threshold cycle [Ct] value) in sputum samples than swabs, thus
leading to the conclusion that simple throat swabs will provide
sufficient sensitivity for screening.16

In addition to nasopharyngeal and oropharyngeal swabs, a
few groups also examined the potential of saliva as the clini-
cal specimen for SARS-CoV-2 detection. In this regard, a study
of 12 patients confirmed by PCR-detection of virus RNA using
nasopharyngeal or sputum specimens found that the coughed-
out saliva from 11 patients were positive for SARS-CoV-2.17
Importantly, virus RNA was not detected in saliva samples col-
lected from another 33 patients whose nasopharyngeal speci-
mens were tested negative for SARS-CoV-2. Consistent results
were obtained by the same group examining a different set
of patients, showing that SARS-CoV-2 was detectable in self-
collected saliva of 20 of 23 confirmed patients.17 These studies
revealed that salivary virus loads corresponded to the severity
of disease and declined after treatment, although the differ-
ences were not statistically significant possibly because of the
small sample size. Similar to those studies with nasal or throat
samples, these reports showed that the SARS-CoV-2 can still be
detected in saliva among a third of patients 20 days or longer
after initial diagnosis, thus supporting the idea that saliva may
represent as an appropriate specimen for screening patients ever
with SARS-CoV-2 infection.

Another approach to collect saliva sample is mediated by oral
swabs, which is easily applicable even for non-professional indi-
viduals. In two studies examining saliva sample collected by oral
swabs, 15 out of 39 (50%) and 25 out of 25 (100%) patients were
tested SARS-CoV-2 positive, respectively.18,19 Although it remains
premature to reach any conclusion, these studies indeed imply that
the saliva, either collected by coughing out or oral swabs, is a legit-
imate clinical specimen for SARS-CoV-2 detection.

6. SALIVA PROVEN TO BE A VALUABLE CLINICAL
SPECIMEN FOR DETECTION OF SARS-COV:
LESSONS FROM 17 YEARS AGO
As both of the COVID-19 and SARS are caused by coronaviruses
and can be transmitted through respiratory droplets, the studies
on SARS-CoV may provide a hint as we continue to navigate

https://www.fda.gov/media/136875/download

https://www.fda.gov/media/136875/download

www.ejcma.org 893

Mini-Review. (2020) 83:10 J Chin Med Assoc

and unravel COVID-19. Concerning the value of saliva as the
clinical specimen for coronavirus detection, a study of 17 SARS
cases in Taiwan showed that a substantial amount of SARS-CoV
RNA were detected in saliva (7.1 x 103 to 6.4 x 108 copies/mL)
and throat wash (9.6  x  102 to 5.9  x  106 copies/mL) from all
patients. Importantly, the highest detection rate of saliva/throat
wash samples appeared as early as 4 days after disease onset,
thus implying that these clinical specimens can be used for virus
detection.20 Another previous study examined the SARS-CoV
loads in different clinical samples and found that the virus RNA
could be detected in saliva (5.2  x  102 copies/mL), although its
level was relatively lower than that in throat swabs (5.5 x 102
copies/mL), sputum (1.2  x  106 copies/mL), and endotracheal
aspirates (2.8  x  106 copies/mL).21 It is noted that the amounts
of SARS-CoV virus RNA detected in these two studies differ by
a significant amount, which can be possibly associated with the
timing of sampling: The samples in the first study were taken
between day 2 and day 9, whereas the samples in the second
study were taken after a median duration of 12 days (2 to 54
days) after the onset of symptoms. Indeed, the observation that
SARS-CoV and other coronavirus peaks at around 10 days
after onset of disease was commonly shared between studies of
various clinical samples.22–24 Collectively, saliva is of potential
diagnostic value for and should play a role in detection of SARS-
CoV-2 infection.

7. GENERAL CONSIDERATION TO USE SALIVA AS A
DIAGNOSTIC FLUID FOR VIRUS DETECTION
Since saliva is easily collected and clinically informative for dis-
ease detection, the consideration that maximizes the benefit of
using saliva as a diagnostic fluid deserves more attention. Thus
far, the approach and protocol for collection of saliva sample
has yet officially standardized; however, it is likely that the diag-
nostic value of saliva is closely related to how saliva sample is
obtained. This concept is supported by a study examining saliva
specimen collected directly from the opening of salivary glands
of 31 confirmed cases, showing that only four patients (12.9%)
were tested positive for SARS-CoV-2 detection,25 which is sig-
nificantly lower than the positive rate derived from examina-
tion of coughed-out saliva. It is possible that the advantage of
using the spit saliva is partly attributed to the potential avail-
ability for multiple targets, such as desquamated oropharyngeal
mucous epithelial cells and respiratory secretions with shedding
viruses. This concept is supported by a previous study, which
found replicating SARS-CoV in the cells collected by throat
wash from SARS patients. This characteristic of benefit stands
for the sputum and saliva. Indeed, it has been estimated that the
SARS-CoV-2 load of sputum is 106 to 1011 particles/mL, whereas
the virus load of saliva is 108 to 109 particles/mL;26–28 however,
unlike the sputum comprising a large amount of mucus that
hampers RNA extraction, saliva (~70% to 90% water) is sup-
posed to give at least a comparable load of viral RNA. As to the
sampling protocols, a 0.5-hour or up to overnight fasting before
saliva collection has been shown in multiple studies to increase
the concentration of RNA.29–32 It is also recommended to have
the subject rinse their mouth with water but not disinfectant
mouthwash. The same guidelines should be used for both of the
spitting/coughing out and oral swab approaches.

In conclusion, the diagnostic testing is crucial for control-
ling the COVID-19 pandemic. Any implementation of clinical
sampling for diagnosis should take into considerations of the
sensitivity of assays, risks to healthcare professionals, and global
shortage of equipment. Many studies showed that sputum is
superior to nasopharyngeal swabs in detection of SARS-CoV-2
infection. However, while the virus is often reliably detected
in sputum, this clinical specimen is not always obtainable for

patients without productive coughs and induction of cough may
even enhance the spread of virus. On the other hand, several
preliminary reports showed that the viral load in saliva is com-
parable with that in sputum. Moreover, the collection of saliva is
minimally invasive and can be self-administrated. Accordingly,
the saliva-based SARS-CoV-2 diagnostics seems to be poten-
tially promising and appealing. Notably, this is a rapidly mov-
ing research topic and the current evidence is not peer-reviewed
and, therefore, is still far from leading to a solid conclusion.
Nevertheless, it is reasonable to incorporate the saliva-based
SARS-CoV-2 assay into a part of multiple lines of diagnostics,
which believably may further facilitate the identification of
COVID-19 patients.

REFERENCES
1. Centers for Disease Control and Prevention. Interim guidelines for col-

lecting, handling, and testing clinical specimens from persons for coro-
navirus disease 2019 (COVID-19). 2020.

2. 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. Doi:
10.1148/radiol.2020200642.

3. Zhou P, Yang XL, Wang XG, Hu B, Zhang L, Zhang W, et al. A pneumo-
nia outbreak associated with a new coronavirus of probable bat origin.
Nature 2020;579:270–3.

4. Lu R, Zhao X, Li J, Niu P, Yang B, Wu H et al. Genomic characterisa-
tion and epidemiology of 2019 novel coronavirus: implications for virus
origins and receptor binding. Lancet 2020; 395: 565-74.

5. Centers for Disease Control and Prevention. How COVID-19 Spreads.
2020.

6. Li X, Zai J, Wang X, Li Y. Potential of large “first generation” human-to-
human transmission of 2019-nCoV. J Med Virol 2020;92:448–54.

7. van Doremalen N, Bushmaker T, Morris DH, Holbrook MG, Gamble
A, Williamson BN, et al. Aerosol and surface stability of SARS-CoV-2 as
compared with SARS-CoV-1. N Engl J Med 2020;382:1564–7.

8. Guan WJ, Ni ZY, Hu Y, Liang WH, Ou CQ, He JX et al. Clinical char-
acteristics of coronavirus disease 2019 in China. N Engl J Med 2020.
Doi:10.1056/NEJMoa2002032.

9. Wu A, Peng Y, Huang B, Ding X, Wang X, Niu P, et al. Genome composi-
tion and divergence of the novel coronavirus (2019-nCoV) originating
in China. Cell Host Microbe 2020;27:325–8.

10. Sexton NR, Smith EC, Blanc H, Vignuzzi M, Peersen OB, Denison MR.
Homology-based identification of a mutation in the coronavirus RNA-
dependent RNA polymerase that confers resistance to multiple muta-
gens. J Virol 2016;90:7415–28.

11. Haveri A, Smura T, Kuivanen S, Osterlund P, Hepojoki J, Ikonen N et
al. Serological and molecular findings during SARS-CoV-2 infection:
the first case study in Finland, January to February 2020. Euro Surveill
2020; 25. Doi: 10.2807/1560-7917.ES.2020.25.11.2000266

12. 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;47:3439–43.

13. Li L, Chen QY, Li YY, Wang YF, Yang ZF, Zhong NS. Comparison
among nasopharyngeal swab, nasal wash, and oropharyngeal swab for
respiratory virus detection in adults with acute pharyngitis. BMC Infect
Dis 2013;13:281.

14. Yang YYM, Shen C, Wang F, Yuan J, Li J, et al Evaluating the accu-
racy of different respiratory specimens in the laboratory diagnosis and
monitoring the viral shedding of 2019-nCoV infections. medRxiv 2020;
02.11.20021493.

15. 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. 2020.
Doi:10.1001/jama.2020.3786.

16. Roman Wölfel VMC, Wolfgang G, Michael S, Sabine Z, Marcel A. Müller
D et al. Virological assessment of hospitalized cases of coronavirus dis-
ease 2019. medRxiv 2020. https://doi.org/10.1038/s41586-020-2196-x.

17. To KK, Tsang OT, Chik-Yan YC, Chan KH, Wu TC, Chan JMC, et al.
Consistent detection of 2019 novel coronavirus in saliva. Clin Infect Dis
2020. Doi:10.1093/cid/ciaa149.

894 www.ejcma.org

Hung et al J Chin Med Assoc

18. Azzi L, Carcano G, Gianfagna F, Grossi P, Gasperina D, Genoni A,
et al. Saliva is a reliable tool to detect SARS-CoV-2. J Infect 2020.
Doi:10.1016/j.jinf.2020.04.005.

19. Zhang W, Du RH, Li B, Zheng XS, Yang XL, Hu B, et al. Molecular and
serological investigation of 2019-nCoV infected patients: implication of
multiple shedding routes. Emerg Microbes Infect 2020;9:386–9.

20. Wang WK, Chen SY, Liu IJ, Chen YC, Chen HL, Yang CF, et al. SARS
Research Group of the National Taiwan University/National Taiwan
University Hospital. Detection of SARS-associated coronavirus in throat
wash and saliva in early diagnosis. Emerg Infect Dis 2004;10:1213–9.

21. Drosten C, Chiu LL, Panning M, Leong HN, Preiser W, Tam JS, et al.
Evaluation of advanced reverse transcription-PCR assays and an alter-
native PCR target region for detection of severe acute respiratory syn-
drome-associated coronavirus. J Clin Microbiol 2004;42:2043–7.

22. Oh MD, Park WB, Choe PG, Choi SJ, Kim JI, Chae J, et al. Viral load
kinetics of MERS coronavirus infection. N Engl J Med 2016;375:1303–5.

23. Perlman S, Netland J. Coronaviruses post-SARS: update on replication
and pathogenesis. Nat Rev Microbiol 2009;7:439–50.

24. Peiris JS, Chu CM, Cheng VC, Chan KS, Hung IF, Poon LL, et al. HKU/
UCH SARS Study Group. Clinical progression and viral load in a com-
munity outbreak of coronavirus-associated SARS pneumonia: a pro-
spective study. Lancet 2003;361:1767–72.

25. Chen L. Detection of 2019-nCoV in saliva and characteriza-
tion of oral symptoms in COVID-19 patients. Lancet 2020.
Doi:10.2139/ssrn.3556665.

26. Roman C, Wolfgang G, Michael S, Sabine Z, Marcel AM, Daniela N, et
al. Clinical presentation and virological assessment of hospitalized cases
of coronavirus disease 2019 in a travel-associated transmission cluster.
medRxiv 2020. Doi:10.1101/2020.3.05.20030502.

27. Pan Y, Zhang D, Yang P, Poon LLM, Wang Q. Viral load of SARS-CoV-2
in clinical samples. Lancet Infect Dis 2020;20:411–2.

28. Yinon F, Rob P, Ron M. SARS-CoV-2 (COVID-19) by the numbers.
medRxiv 2020. Doi:10.7554/eLife.57309.

29. Ishikawa S, Sugimoto M, Kitabatake K, Sugano A, Nakamura M,
Kaneko M, et al. Identification of salivary metabolomic biomarkers for
oral cancer screening. Sci Rep 2016;6:31520.

30. Goode MR, Cheong SY, Li N, Ray WC, Bartlett CW. Collection and
extraction of saliva DNA for next generation sequencing. J Vis Exp
2014; 4:253.

31. Laidi F, Bouziane A, Lakhdar A, Khabouze S, Amrani M, Rhrab B, et al.
Significant correlation between salivary and serum Ca 15-3 in healthy
women and breast cancer patients. Asian Pac J Cancer Prev 2014;15:
4659–62.

32. Mandel ID. The diagnostic uses of saliva. J Oral Pathol Med 1990;19:119–25.

diagnostics

Article

Self-Collected versus Healthcare Worker-Collected
Swabs in the Diagnosis of Severe Acute Respiratory
Syndrome Coronavirus 2

Johan H. Therchilsen 1,*, Christian von Buchwald 1, Anders Koch 2, Susanne Dam Nielsen 2,
Daniel B. Rasmussen 2, Rebekka Faber Thudium 2, Nikolai S. Kirkby 3,
Daniel E. T. Raaschou-Pedersen 4, Johan S. Bundgaard 4, Kasper Iversen 5, Henning Bundgaard 4

and Tobias Todsen 1,6

1 Department of Otorhinolaryngology, Head and Neck Surgery and Audiology, Rigshospitalet,
Copenhagen University Hospital, 2100 Copenhagen, Denmark; [email protected] (C.v.B.);
[email protected] (T.T.)

2 Department of Infectious Diseases, Rigshospitalet, Copenhagen University Hospital,
2100 Copenhagen, Denmark; [email protected] (A.K.); [email protected] (S.D.N.);
[email protected] (D.B.R.); [email protected] (R.F.T.)

3 Department of Clinical Microbiology, Rigshospitalet, Copenhagen University Hospital,
2100 Copenhagen, Denmark; [email protected]

4 Department of Cardiology, Rigshospitalet, Copenhagen University Hospital, 2100 Copenhagen, Denmark;
[email protected] (D.E.T.R.-P.); [email protected] (J.S.B.);
[email protected] (H.B.)

5 Department of Cardiology, Copenhagen University Hospital Herlev, 2730 Herlev, Denmark;
[email protected]

6 Copenhagen Academy for Medical Education and Simulation, Rigshospitalet,
Copenhagen University Hospital, 2100 Copenhagen, Denmark

* Correspondence: [email protected]; Tel.: +45-26250191

Received: 23 August 2020; Accepted: 7 September 2020; Published: 9 September 2020
����������
�������

Abstract: The aim of this study was to compare the sensitivity of self-collected versus healthcare worker
(HCW)-collected swabs for Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) testing.
Symptomatic individuals referred for SARS-CoV-2 testing were invited to provide mobile-phone
video-instructed self-collected oropharyngeal and nasal samples followed by a HCW-collected
oropharyngeal sample. All samples were sent for analysis to the same microbiology laboratory,
and the number of SARS-CoV-2-positive participants in the two tests was compared. A total of
109 participants were included, and 19 participants had SARS-CoV-2-positive results. The diagnostic
sensitivity of the self-collected and HCW-collected swabs was 84.2% and 89.5%, respectively, with an
acceptable agreement, Cohens kappa 0.82, p 0.80 was considered an acceptable intertest
agreement [13]. The sensitivity of each test was calculated as the number of positive test/total number
of true positive patients for both the self- and the HCW-collected samples. Differences in group
characteristics were compared with the chi-square test for categorical variables and the Student’s
t-test for continuous variables; p-values were Bonferroni-corrected because of multiple statistical
testing. The statistical analysis was performed using a statistical software package (PASW, version 26.0;
SPSS Inc, Chicago, IL, USA), and two-sided significance levels of 0.05 were used for all analyses.

2.5. Ethics and Data Management

Ethical approval was granted in the form of an exemption letter from the regional ethical committee
of the Capital Region of Denmark (protocol no. H-20027981, approved on 24 April 2020), and the
Danish Data Protection Agency approved the management of patient-sensitive information during the
study (record number: P-2020-467). All participants gave verbal and written informed consent prior
to enrolment.

3. Results

A total of 109 participants were included in the study. Demographics of the participants and their
self-recorded symptoms are shown in Table 1.

Diagnostics 2020, 10, 678 5 of 10

Table 1. Clinical characteristics and self-recorded symptoms of the study participants.

All Participants
SARS-CoV-2

Negative
SARS-CoV-2

Positive
p Value

(Chi-Square)

Clinical characteristics

Number of participants 109 90 19

Sex, female–n (%) 76 (70%) 65 (72%) 11 (58%) 0.22

Mean age—years mean/median
(SD)

39 (13) 40.4 (13) 32.6 (5.9) 0.33

Healthcare education n (%) 26 (25%) 21 (23%) 6 (32%) 0.45

Median days since first symptoms
mean/median—(Range) IQR

3.0 (0–65) 3.0 (0–65) 7.0 (2–25) 0.10

Self-recorded Symptoms, n (%)

Fever 41 (38) 32 (36) 9 (47) 0.35

Cough 54 (50) 44 (49) 10 (53) 0.80

Lethargy 55 (50) 43 (48) 12 (63) 0.24

Throat pain 59 (54) 51 (57) 8 (42) 0.23

Headache 62 (57) 48 (53) 14 (74) 0.11

Respiratory problems 14 (13) 12 (13) 2 (11) 0.11

Loss of smell 18 (17) 8 (9) 10 (53) <0.001 *

Diarrhea 14 (13) 11 (12) 3 (16) 0.67

* p-value after Bonferroni correction.

Among the 109 participants, 19 patients had SARS-CoV-2-positive results from self-collected
samples, HCW-collected samples, or both. The proportion of SARS-CoV-2-positive samples was
16/109 (14.7%) for the self-collected samples in comparison to 17/109 (15.6%) for the HCW-collected
samples, as shown in Table 2. Acceptable agreement between self-collected and HCW-collected swabs
was found, Cohens kappa 0.82, p < 0.001, without any significant difference in diagnostic sensitivity for
the self-collected and HCW-collected samples, corresponding to 84.2% and 89.5%, respectively, p = 0.81.
However, of the 19 positive samples, only 14 (74%) were found positive by both tests. Combining the
self-collected samples to the HCW-collected samples added two more SARS-CoV-2-positive cases out
of the 109 included participants, who would otherwise have been tested as false negative using only
the HCW-collected samples.

As to the preference of the swab technique, 47/109 (43.1%) of the participants preferred sample
self-collection, 29/109 (26.6%) preferred collection by HCWs, and 33/109 (30.3%) did not have any
preference. An error in the sampling technique by a participant was registered for 16/109 (14.7%) of the
self-collected oropharyngeal samples, 18/109 (16.5%) for nasal samples; 4/109 (3.7%) of the participants
made a double error. Six sampling errors were observed in the 19 SARS-CoV-2-positive patients,
including a double sampling error, and in 1/3 of the false negative participants, a sampling error for
the middle turbinate sample was observed. Discordant results favoring HCW collection regarded
primarily patients 20–25 days post-symptom onset. If inclusion of patients were limited to two weeks
post-symptom onset, then 14/14 self-collected vs. 13/14 HCW-collected samples would have been
positive for SARS-CoV-2. A subgroup analysis found that samples from participants with a healthcare
education detected SARS-CoV-2 in 5/6 (83%) of the true positive cases; a similar detection rate was
found for the participants without a healthcare education 11/13 (85%).

Loss of smell was the only self-recorded symptom that differed significantly between
SARS-CoV-2-positive and SARS-CoV-2-negative participants (53% versus 9%, p < 0.001) (Table 1).

Diagnostics 2020, 10, 678 6 of 10

Table 2. Samples positive for SARS-CoV-2 by sampling method.

Participant
(SARS-CoV-2
True Positive)

Time from
Self-Reported

Symptom Onset

Healthcare
Worker-Collected

Oropharyngeal Swab

Self-Collected Middle
Turbinate/Oropharyngeal Swab

1 8 Positive Positive
2 25 Positive Negative
3 12 Positive Positive
4 3 Positive Positive
5 2 Positive Positive
6 7 Positive Positive
7 9 Positive Positive
8 2 Positive Positive
9 5 Positive Positive

10 5 Negative Positive
11 25 Negative Positive
12 4 Positive Positive
13 3 Positive Positive
14 10 Positive Positive
15 17 Positive Positive
16 6 Positive Positive
17 20 Positive Negative
18 6 Positive Positive
19 23 Positive Negative

4. Discussion

We found that self-collection of sample swabs for SARS-CoV-2 testing is a well-tolerated diagnostic
method with a sensitivity almost equivalent to that of collection of oropharyngeal samples by HCWs.
A very good and significant intertest correlation was found between the diagnostic outcomes of
self-collected and HCW-collected samples, although we found a slightly lower sensitivity for the
self-collected samples compared with the HCW-collected ones. The sensitivity of 84.2% of the
self-collected samples is comparable to values found in other studies [9–11] and slightly higher than
that determined in a study from Washington, USA [12]. However, in the latter study, there was a delay
between the diagnostic tests and the shipping of the self-collected samples at room temperature from
participants’ homes, which could explain the observed difference [12]. A strength of our study is that
the self- and HCW-collected samples were obtained at the same time, which decreased the risk of
change of viral load between the tests. Further, all of the 218 (109 × 2) samples were stored in the
same type of transport medium and at the same temperature and analyzed at the same laboratory,
which increases the internal validity of our test results. We used a simple self-collection sampling
technique in this study, whereby all participants were guided by a mobile phone video-instruction,
and no HCW were needed for guidance. This could therefore increase the applicability of the method
using a single swab and a vial for self-testing.

Interestingly, the loss of smell was the only self-reported symptom for which a difference
was seen between SARS-CoV-2-positive and -negative participants. Approximately half of the
SARS-COV-2-positive participants reported a loss of smell compared to 9% of the SARS-COV-2-negative
participants, a result that is highly significant even after a Bonferroni correction for multiple statistical
testing. These findings are comparable with those of other studies describing olfactory loss associated
with SARS-CoV-2 infection [14,15]. Therefore, we recommend heightened awareness when patients
describe a sudden loss of smell, and it should be considered whether to offer them a test for SARS-CoV-2.

The findings of the present study should be interpreted in accordance to the following limitations.
We do not know the true SARS-CoV-2 infection status in our participants, as our results are

dependent on a positive HCW-collected and/or self-collected swab. We can therefore expect that some
of the participants will be false-negative, and a lower sensitivity of both tests may be expected. Further,
the HCW-collected and self-collected swabs were not performed in the same way, as the self-collection

Diagnostics 2020, 10, 678 7 of 10

technique also included the acquisition of a nasal sample in addition to the oropharyngeal sample.
However, as the study aim was to compare a new self-testing technique with the recommended
standard technique for SARS-CoV-2 testing in Denmark (HCW-collected oropharyngeal sample),
we chose this method. The small sample size of our study was without sufficient power to conduct
a non-inferiority statistical comparison of the sensitivity between HCW-collected and self-collected
samples for SARS-CoV-2 testing. A priori, we did not expect self-collected samples to be as sensitive as
HCW-collected, and we were surprised by the almost equivalent sensitivity of the self-collection- and
HCW-collection-based SARS-CoV-2 testing methods found in our study. These findings may be the
result of combined oropharyngeal and nasal sampling for the self-collected samples that were compared
to HCW-collected oropharyngeal swabs. In our study, we found two additional SARS-CoV-2-positive
patients—corresponding to an increase in sensitivity of 10%—by adding self-collected oropharyngeal
and nasal samples to HCW-collected oropharyngeal swabs. Assuming a similar sensitivity for the
self- and the HCW-collected samples, this emphasizes the importance of nasal samples in SARS-CoV-2
testing. These findings are also in accordance with a recent study that found a higher concentration of
SARS-CoV-2 RNA in nasopharyngeal samples than in oropharyngeal ones in SARS-CoV-2-positive
patients [16]. Another limitation is the presence of a high proportion of health professionals among
the participants, which may impact on the generalizability of our results. A lower sensitivity of the
self-collected swabs could be expected when the technique is introduced in the general community,
with older and fewer healthcare-educated patients. However, we found no difference in the proportion
of false-negative test results between participants with or without a healthcare education. Another
factor which might impact the sensitivity of the test method were it to be performed without a HCW’s
supervision is the patients’ personal situation. Some patients might have a social or economic interest
or a fear of a positive result.

In conclusion, we found an acceptable sensitivity for swab samples collected from the nasal and
oropharyngeal cavities for SARS-CoV-2 testing by patients solely guided by a written pictorial guide
and Supplementary Video S1 instructions on their mobile phone. The SARS-CoV-2 self-test described
here is therefore a reliable testing method with a low false-negative rate compared to the technique
based on HCW-collected swabs and might be used in community testing or settings where HCW
time and personal protective equipment need to be economized. Future studies should explore the
diagnostic accuracy and cost-effectiveness of this method when implemented in a larger and more
heterogenous cohort of patients tested for COVID-19.

Supplementary Materials: The instruction Supplementary Video S1 was constructed specifically for this study.
https://drive.google.com/file/d/1ch4At6pvYThB5qKz32PFUJ69aOUur2Wg/view?usp=drive_web.

Author Contributions: Conceptualization, T.T., H.B., and C.v.B.; methodology, T.T., H.B., J.H.T., and C.v.B.;
software, T.T., D.B.R., and J.S.B.; validation, J.H.T. and T.T.; formal analysis, J.H.T. and T.T.; investigation, J.H.T.,
D.E.T.R.-P., and R.F.T.; data curation, J.H.T.; writing—original draft preparation, J.H.T. and T.T.; writing—review
and editing, J.H.T., C.v.B., A.K., S.D.N., D.B.R., R.F.T., N.S.K., D.E.T.R.-P., J.S.B., K.I., H.B., and T.T.; project
administration, J.H.T.; funding acquisition, T.T. and J.H.T. All authors have read and agreed to the published
version of the manuscript.

Funding: This research received no external funding.

Conflicts of Interest: Swabs from Zymo Collection Swab, Zymo Research, Irvine, CA, USA were given free of
charge by Nordic Biosite, 1301 Copenhagen, Denmark. The funders had no role in the design of the study; in
the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish
the results.

https://drive.google.com/file/d/1ch4At6pvYThB5qKz32PFUJ69aOUur2Wg/view?usp=drive_web

Diagnostics 2020, 10, 678 8 of 10

Appendix A

Diagnostics 2020, 10, x FOR PEER REVIEW 9 of 11

Appendix A

Figure A1. Written Instructions.

Start by blowing your nose and examine

which nostril has the highest air flow.
Select this nostril for the nasal swab. If
the air flow is similar, choose the right
nostril.

Find a good light source, optionally your

phone. Light into your mouth while
looking at yourself in the mirror. Say
“ahh” to optimize your view.

Unscrew the lid on the transparent tube

and place it in a cup so the liquid doesn't
run out. Unpack the bag containing the
swab stick.

You will now be able to view your palate
arches with your uvula hanging in the

middle. Furthest back is your pharynx
(*) and to both sides are your tonsils (*)
– where the swab shall be taken from.

(both * + * )

Now you are ready to the self-collecting
swab procedure.

Move the swab to the back wall of the
pharynx and avoid hitting the tongue or

the cheek on the way.

With the swab against the back wall of
the pharynx rotate the swab stick.

Repeat this procedure on one of your
tonsils with the same swab stick.

The same swab stick should now also be
used to take a sample from your nose.

Place two fingers 3-4 cm below the tip
of the swab stick in the cotton wool end.

Insert the swab stick into your nostril
until your fingers touch your nostril. The

swab stick is now 3-4 cm inside your
nose. Turn the swab stick around a
couple of times and take it out.

Put the cotton end of the swab into the
transparent tube and with the swab

stick in the tube break the swab stick
where it is indicated so you can screw
the lid on. Tighten the lid.

G uide to self-collected sw ab for C O VID -19

3)

4) 5) 6)

7) 8) 9)

S can and
view video

2) 1)

Figure A1. Written Instructions.

Diagnostics 2020, 10, 678 9 of 10
Diagnostics 2020, 10, x FOR PEER REVIEW 10 of 11

Figure A2. Questionnaire to the participants.

References

1. WHO Laboratory Testing for 2019 Novel Coronavirus (2019-nCoV) in Suspected Human Cases. Interim
Guidance. Available online: https://www.who.int/publications-detail/laboratory-testing-for-2019-novel-
coronavirus-in-suspected-human-cases-20200117 (accessed on 17 January 2020).

2. Mohammadi, A.; Esmaeilzadeh, E.; Li, Y.; Bosch, R.J.; Li, J.Z. SARS-CoV-2 detection in different respiratory
sites: A systematic review and meta-analysis. EbioMedicine 2020, 102903, doi:10.1101/2020.05.14.20102038.

3. Dhiman, N.; Miller, R.M.; Finley, J.L.; Sztajnkrycer, M.D.; Nestler, D.M.; Boggust, A.J.; Jenkins, S.M.; Smith,
T.F.; Wilson, J.W.; Cockerill, F.R., III; et al. Effectiveness of patient-collected swabs for influenza testing.
Mayo Clin. Proc. 2012, 87, 548–554.

4. Seaman, C.P.; Tran, L.T.T.; Cowling, B.J.; Sullivan, S.G. Self-collected compared with professional-collected
swabbing in the diagnosis of influenza in symptomatic individuals: A meta-analysis and assessment of
validity. J. Clin. Virol. 2019, 118, 28–35.

5. Elliot, A.J.; Bermingham, A.; Charlett, A.; Lackenby, A.; Ellis, J.; Sadler, C.; Sebastianpillai, P.; Powers, C.;
Foord, D.; Povey, E.; et al. Self-Sampling for community respiratory illness: A new tool for national
virological surveillance. Eurosurveillance 2015, 20, 1–5.

6. Fisher, C.E.; Boeckh, M.; Jerome, K.R.; Englund, J.; Kuypers, J. Evaluating addition of self-collected throat
swabs to nasal swabs for respiratory virus detection. J. Clin. Virol. 2019, 115, 43–46.

7. Available online: https://www.gov.uk/guidance/coronavirus-covid-19-getting-tested#the-testing-process
(accessed on 28 July 2020).

8. Available online: https://www.cdc.gov/coronavirus/2019-ncov/lab/guidelines-clinical-specimens.html
(accessed on 28 July 2020).

9. Wehrhahn, M.C.; Robson, J.; Brown, S.; Bursle, E.; Byrne, S.; New, D.; Chong, S.; Newcombe, J.P.; Siversten,
T.; Hadlow, N. Self-collection: An appropriate alternative during the SARS-CoV-2 pandemic. J. Clin. Virol.
2020, 128, 104417.

10. Altamirano, J.; Govindarajan, P.; Blomkalns, A.L.; Kushner, L.E.; Stevens, B.A.; Pinsky, B.A.; Maldonado,
Y. Assessment of Sensitivity and Specificity of Patient-Collected Lower Nasal Specimens for Sudden Acute
Respiratory Syndrome Coronavirus 2 Testing. JAMA Netw. Open 2020, 3, e2012005.

Figure A2. …

ASSIGNMENT

1. Critically read each paper submitted by your group members (5-6). Post a paragraph synthesizing what you think were the main/important results of the studies. Include what results/conclusions were similar across all studies and what were the differences.

2. Respond to one of your group member’s posts.

Can you help with this assignment?
In this assignment, I have to read about four research articles posted by each member in my group on our discussion board. After reading the all the four articles, I have to post a paragraph synthesizing what I think were the main or important results of the studies. Include what results or conclusions that were similar across all studies and what were the differences.

I will upload the four different articles so you can best help me with the assignment.

error: Content is protected !!