Influence of Seasonal and Operator Variations on Diagnostic Accuracy of Lateral Flow

Devices in the COVID-19 Pandemic: A Systematic Review and Meta-Analysis.

Authors:

Ashwin Krishnamoorthy1 - Ashwin.Krishnamoorthy2@uhcw.nhs.uk

Subashini Chandrapalan2 - subashini.chandrapalan@nhs.net

Ayman Bannaga 3 - aymanbannaga@yahoo.com

Gohar JalayeriNia4 - Gohar.JalayeriNia1@uhcw.nhs.uk

Yaqza Hussain5 - Yaqza.Hussain@uhcw.nhs.uk

Ian Io Lei6 – Brian.lei@uhcw.nhs.uk

Ramesh Arasaradnam7 - r.arasaradnam@warwick.ac.uk

Affiliation:
1. Department of gastroenterology research, University Hospital of Coventry &
Warwickshire NHS trust, United Kingdom (Dr. Krishnamoorthy MBBS)
2. Department of gastroenterology research, University Hospital of Coventry &
Warwickshire NHS trust, United Kingdom (Dr. Chandrapalan MD)
3. Department of gastroenterology research, University Hospital of Coventry &
Warwickshire NHS trust, United Kingdom (Dr. Bannaga PhD)
4. Department of gastroenterology research, University Hospital of Coventry &
Warwickshire NHS trust, United Kingdom (Dr. JalayeriNia MBBS)
5. Department of gastroenterology research, University Hospital of Coventry &
Warwickshire NHS trust, United Kingdom (Dr. Hussain MBBS)
6. Department of gastroenterology research, University Hospital of Coventry &
Warwickshire NHS trust, United Kingdom (Dr. Lei MBBS)
7. Department of gastroenterology research, University Hospital of Coventry &
Warwickshire NHS trust, United Kingdom (Prof. Arasaradnam PhD)

Correspondence to:
Mr Ashwin Krishnamoorthy

Department of Gastroenterology Research, University Hospitals Coventry & Warwickshire

NHS Trust, Clifford Bridge Road, Coventry, United Kingdom, CV2 2DX
02476966581

Email: ashwin.krishnamoorthy2@uhcw.nhs.uk

Author Conflicts: No competing interests declared

1
Summary

Background

Lateral flow tests (LFT) are point-of-care rapid antigen tests and allow isolation and control
of disease outbreaks through convenient, practical testing. Studies have shown significant
variation in their diagnostic accuracy. Our objective was to carry out a systematic review
into the diagnostic accuracy and to identify potential factors affecting their performance.

Methods

A systematic search of online databases was carried out to identify studies assessing
sensitivity and specificity of lateral flow tests compared to PCR. Data was extracted and
used to calculate pooled sensitivity and specificity. Meta regression analysis was carried out
to identify covariates influencing diagnostic accuracy.

Findings

76 articles with 108,820 test results were identified for analysis. Pooled sensitivity and
specificity were 72% (95% CI 68% to 76%) and 100% (95% CI 99% to 100%) respectively. Staff
operation of the lateral flow test showed a statistically significant increase in sensitivity
(p=0.04) and specificity (p=0.001) when compared to self-operation by the test subjects. The
symptomatic patient subgroup also resulted in a higher sensitivity of the test.

Interpretation

Lateral flow tests display good sensitivity and extremely good specificity for SARS-CoV-2
antigen detection; and they are even more sensitive in symptomatic patients and when
operated by trained professionals.

Funding: None.

Keywords: Coronavirus, COVID-19, SARS-CoV-2, viral antigen detection, rapid antigen test,
lateral flow device, lateral flow test, rapid antigen detection

Panel: Research in Context

Evidence before this study

2
There have been a number of systematic reviews on the topic of sensitivity and specificity of
lateral flow immunoassays in the diagnosis of SARS-CoV-2 infection; which we identified
through searching EMBASE and Medline databases with the search terms “systematic
review”, “lateral flow test”, “COVID-19”, “diagnostic accuracy”, “sensitivity” and “specificity”
between the dates of 1st January 2020 and 31st January 2022.

Added value of this study

To our knowledge this is the largest meta analysis carried out to date with the most papers
spanning the most countries and most test subjects present. We demonstrate the diagnostic
accuracy of lateral flow tests and comment on external factors influencing their accuracy
derived from our statistical analysis.

Implications of all the available evidence

The literature points towards lateral flow testing as being a potentially viable replacement
for PCR in certain situations with certain caveats. Our study reinforces that message but
reminds us that certain factors can make these tests even more accurate and so we should
proceed with them in mind.

Introduction

SARS-CoV-2 is the highly transmissible respiratory virus that emerged in late 2019 and resulted in a
massive global outbreak of viral pneumonia (COVID-19) in 2020 and 2021. Since then, various
testing methods have been developed for the efficient diagnosis of COVID-19 in order to limit
spread through isolation and quarantine of infectious patients.[1]

A lateral flow test (LFT), or lateral flow immunoassay, is a device that is intended to detect the
presence of a target substance, such as a biological antigen, as a point of care test. More recently
they have emerged as one of the most important tests for the diagnosis of SARS-COV2 infection.
The test works via binding of conjugated antibodies to a specific target antigen in a sample and this
antigen-antibody complex (positive test) is evident on the test strip through a coloured line.[2]

Although the gold standard for diagnosis of COVID-19 remains the reverse transcriptase polymerase
chain reaction (RT-PCR), it is more expensive, requires trained professionals for processing and has
a longer time for results to show.[3] Lateral flow testing, on the other hand, has a quick turnaround
time, is easily available, has relatively lower cost and requires limited training; allowing use by the
test subject.[4] It is because of these favourable characteristics that LFT has become significantly
popular worldwide and is considered a powerful tool to tackle the COVID 19 pandemic.[5]

Consequently, mass testing for COVID-19 with lateral flow test kits was rolled out in the UK and
many other countries in 2021. In the case of a positive test result, affected individual was required
to self-isolate for a period of time. This rule was meant to reduce transmission and positive
caseloads. However, the actual outcome of such rules is highly dependent on the test performance

3
and disease prevalence in the region. In fact, if the tests perform poorly, these rules could have a
negative impact on various services including healthcare and the overall economy.[6]

Several studies showed significant variation in the reported sensitivity of the different lateral
flow assays.[4] Our main objective was to conduct a systematic review and meta-analysis to
identify the sensitivity and specificity of these tests as per the literature; and also to identify
various factors that could potentially affect the performance characteristics of the lateral
flow devices with regards to the detection of COVID-19.

4
Methodology
Study design
This was a systematic review of clinical studies published in peer-reviewed journals. This
followed the guidance laid out in the Cochrane handbook for diagnostic accuracy reviews.
[7]

Search strategy
Literature search strategies were developed using medical subject headings (MeSH) and text
words related to the title. The search was performed in Medline, EMBASE, Scopus and
Cochrane, using various combinations of keywords and subject headings – “Covid 19”,
“SARS-Cov 2”, “Coronavirus”, “antigen test”, “antigen detection”, “lateral flow devices”.
To ensure maximum capture, we had scanned the reference lists of included studies, or the
relevant reviews that had been identified through the search for potential studies. We
included studies which were published until the 31 st of January 2022. Four reviewers
reviewed each paper generated from the search and excluded articles firstly based on the
abstract; and then on reviewing the full text according to the inclusion and exclusion criteria
and relevance to the topic.

Study selection
The inclusion criteria:
(1) prospective and retrospective comparative cohort studies, case-controlled studies, cross
sectional studies, and randomised controlled trials
(2) studies with both adult and paediatric populations
(3) published in English and peer reviewed and, was available as full texts in the medical
database.
(4) COVID-19 confirmed through PCR testing as the reference standard
(5) have sufficient data to calculate true positives (TP), true negatives (TN), false positives
(FP) and false negatives (FN)
(6) used commercially available antigen detection kits.

The exclusion criteria were as follows:

(1) studies which did not have a control group
(2) studies that were published as systematic reviews/meta-analyses themselves.

5
6
Quality assessment and data selection
QUADAS -2 tool was used to assess the quality of the included studies. This was undertaken
by one reviewer and double-checked by the other. The studies were assessed against four
domains - patient selection, index test, reference standard and patient flow and timing; the
risk of bias was graded as low, unclear or high. Any disagreements were resolved by
discussion. Publication bias was assessed through Deek’s regression test and funnel plot.[8]

Data synthesis and analysis

The following data was extracted from the papers: country of study and geographic region;
period of study and season in that country; antigen test kit used; specimen collected and
collected by; antigen test results - TP, TN, FP, FN, sensitivity, specificity; total population and
those confirmed to have had Covid 19; population characteristics with regard to symptoms.
The R software was used for all statistical analyses. [9]
Where studies had reported only sensitivity and specificity, the TP, TN, FP, FN were
calculated from precision estimates and their respective confidence intervals, to produce a
2x2 contingency tables. Descriptive forest plots were produced to illustrate sensitivity,
specificity and their corresponding confidence intervals using package ‘meta’ [10]
Bivariate analysis was performed using the package “mada” [11]
Summary receiver operator curve (SROC) with confidence region for sensitivity and false
positive rate (1-specificity) was produced. The area under the curve was calculated, in order
to assess the overall accuracy of lateral flow devices. The heterogeneity was assessed
through a visual distinction of forest plots and I 2 test. Meta-regression analysis was
performed to identify the potential covariates which contributed to the heterogeneity.

Results
Overall, the literature search identified 2708 articles. These were checked against the
inclusion and exclusion criteria which resulted in 76 articles. A few included studies which
evaluated more than one study cohorts - therefore the total number of study cohorts
included in our meta-analysis was 106 with a total number of 108,820 patients for analysis.
The selection process of the included studies is shown in PRISMA diagram (Figure 1).
The basic characteristics consist of geographical regions, seasons, symptomatic and self or
staff swabbing of the included studies are summarised in Table 1. There are 30 studies

7
including symptomatic patients only and 7 studies included only patients who are
asymptomatic. In terms of the swabbing methods, 48 reports (45%) did not mention how
the tests were performed. The population size of these studies varied from 56 to 15402. 86
cohorts (81%) included adult patients only.

Figure 1. PRISMA diagram demonstrating the study selection process

8
 Table 1: Basic characteristics of the studies included in the meta-analysis

Size of the
Symptomatic/ study
Geographic asymptomatic/bo Staff/ populatio
Study al region Season th self n
Not
Abdelrazik et Middle reporte
al, 2020 East / Africa Summer Both d 310
Albert et al,
2020 Europe Autumn Symptomatic Staff 412
Not
Alemany et al, reporte
2020 Europe Not reported Both d 1406
Asai et al,
2021 Far East Winter Symptomatic Staff 305
Ben
Abdelhanin et Spring
al 2021 Europe /Summer Both Staff 225
Berger et al,
2021(a) Europe Autumn Both Staff 529
Berger et al,
2021(b) Europe Autumn Both Staff 535
Not
Blairon et al, reporte
2020 Europe Spring Not reported d 56
Bulilete et al,
2020 Europe Autumn Both Staff 1362
Not
Caputo et al, reporte
2021 Europe Winter Both d 4266
Carbonell-
Sahuquillo et
al, 2021 Europe Winter Symptomatic Staff 357
Not
Cerutti et al, Spring / reporte
2020 Europe summer Both d 330
Not
Chiu et al, reporte
2021 America Spring Symptomatic d 349
Not
Coggiola et al, reporte
2021 Europe Autumn Both d 4000
Courtellemon Europe Autumn Both Staff 248

9
t et al, 2021
Drevinek et
al, 2021 Europe Autumn Both Staff 591
Drevinek et
al, 2021 Europe Autumn Both Staff 875
Escriva et al
2021 Europe Winter Both Staff 448
Not
Fenollar et al, reporte
2020 Europe Autumn Asymptomatic d 159
Not
South reporte
FIND, 2020 (a) America Summer Symptomatic d 400
Not
FIND, 2020 reporte
(b) Europe Autumn Symptomatic d 535
Not
FIND, 2020 South reporte
(c)1 America Summer Symptomatic d 400
Not
FIND, 2020 reporte
(c)2 Europe Summer Symptomatic d 1249
Not
FIND, 2020 South reporte
(d)1 America Summer Symptomatic d 453
Not
FIND, 2020 reporte
(d)2 Europe Summer Symptomatic d 676
Not
FIND, 2020 South reporte
(e)1 America Summer Symptomatic d 476
Not
FIND, 2020 South reporte
(e)2 America Summer Symptomatic d 1239
Garcia-Finana
et al, 2021 Europe Autumn Asymptomatic Staff 5504
Gili et al, 2021 Europe Autumn Both Staff 226
Gonzalez-
Donapetry et
al 2021 Europe Autumn/Winter Both Staff 440
Gremmels et
al, 2020 Europe Autumn Symptomatic Staff 1575
Gupta et al,
2021 Asia Summer Both Staff 330
Hartard et al,
2021 Europe Not reported Both Staff 378
Houston et al, Europe Winter Symptomatic Staff 728

10
2021
Ifko et al,
2021 (a) Europe Autumn Symptomatic Staff 191
Ifko et al,
2021 (b) Europe Autumn Symptomatic Staff 333
Igloi et al,
2021 Europe Spring Symptomatic Staff 970
Jegerlehner et
al, 2021 Europe Winter Both Staff 1465
Kim et al,
2022 Far East Winter / Spring Both Staff 165
Not
Kiyasu et al, reporte
2021 Far East Winter Both d 1934
Korenkov et
al, 2021 Europe Aut/Winter Both Staff 2028
Kruger et al,
2020 Europe All 4 seasons Symptomatic Staff 1108
Kruger et al, South
2022 (a)1 America All 4 seasons Both Staff 454
Kruger et al,
2022 (a)2 Europe All 4 seasons Both Staff 1239
Kruger et al, South
2022 (b)1 America All 4 seasons Both Staff 453
Kruger et al,
2022 (b)2 Europe All 4 seasons Both Staff 676
Kruger et al, South
2022 (c)1 America All 4 seasons Both Staff 400
Kruger et al,
2022 (c)2 Europe All 4 seasons Both Staff 1710
Kruger et al,
2022 (d) Europe All 4 seasons Both Staff 727
Kruger et al,
2022 (e) Europe All 4 seasons Both Staff 723
Kruger et al,
2022 (f) Europe All 4 seasons Both Staff 649
Kruger et al, South
2022 (g) America All 4 seasons Both Staff 400
Lambert- Not
Noclot et al, reporte
2020 Europe Not reported Not reported d 138
Not
Lee et al, reporte
2021 Far East All 4 seasons Not reported d 680
Not
Leiner et al, reporte
2021 Europe Winter Both d 4076

11
 Not
Leixner et al, reporte
2021 Europe Winter Symptomatic d 392
Not
Leli et al, reporte
2021 Europe Autumn/Winter Both d 792
Not
Linares et al, reporte
2020 Europe Autumn Both d 255
Not
Liotti et al, reporte
2020 Europe Not reported Not reported d 359
Not
Masia et al, reporte
2020 Europe Autumn Both d 905
Not
Menchinelli reporte
et al, 2021 Europe Winter Both d 595
Merino et al,
2020 Europe Autumn Both Staff 958
Not
Mertens et al, reporte
2020 Europe Spring Symptomatic d 328
Mitchell et al,
2021 America Not reported Both Staff 148
Murillo- Not
Zamora et al, South reporte
2021 America Winter Symptomatic d 15402
Not
Nalumansi et reporte
al, 2020 Africa Not reported Not reported d 262
Ngo Nsoga et
al, 2021 Europe Winter Both Staff 402
Oh et al, 2021 Far East Winter Both Staff 118
Pena et al, South
2021 America Summer Asymptomatic Staff 842
Not
Perez-Garcia reporte
et al, 2021 Europe Autumn Both d 356
Not
Perez-Garcia reporte
et al, 2021 Europe Autumn Both d 356
Not
Peto et al, Spring / reporte
2021 Europe summer Both d 6954
Pilarowski et America Not reported Both Not 3302
al, 2021 reporte

12
 d
Pollock et al,
2021 America Autumn Both Staff 2250
Pollreis et al,
2021 America Winter / Spring Both Staff 214
Porte et al, South
2020 (a) America Not reported Symptomatic Staff 127
Not
Porte et al, South reporte
2020 (b) America Not reported Both d 64
Not
Porte et al, South reporte
2020 (c) America Not reported Both d 64
Not
Randriamahaz Middle reporte
o et al, 2022 East/Africa Not reported Both d 256
Reichert et al,
2021 Europe Winter / Spring Asymptomatic Staff 1076
Not
Sberna et al, reporte
2021 Europe Not reported Both d 513
Not
Schildgen et reporte
al, 2020 (a) Europe Not reported Both d 73
Not
Schildgen et reporte
al, 2020 (b) Europe Not reported Both d 73
Not
Schildgen et reporte
al, 2020 (b) Europe Not reported Both d 73
Not
Scohy et al, reporte
2020 Europe Spring Both d 148
Shah et al,
2021 America Autumn Both Staff 2110
Shrestha et al,
2020 Asia Summer Asymptomatic Staff 113
Sood et al,
2021 America Winter Both Staff 766
Not
Surasi et al, reporte
2021 America Winter Both d 769
Not
Takeda et al, reporte
2020 Far East Spring Both d 162
Thirion- South Not reported Both Staff 1060
Romero et al, America

13
2021
Tinker et al,
2021 America Autumn Asymptomatic Self 1540
Torres et al,
2020 Europe Autumn Asymptomatic Staff 634
Not
Turcato et al Summer/ reporte
2021 Europe Autumn Both d 3410
Van der
Moeren et al,
2021 Europe Autumn Symptomatic Staff 351
Not
Veyrenche et reporte
al, 2020 Europe Spring Symptomatic d 65
Villaverde et
al, 2021 Europe Autumn Symptomatic Staff 1620
Wagenhauser
et al, 2021 Europe Winter Both Staff 806
Wagenhauser
et al, 2021 Europe Winter Both Staff 1029
Wagenhauser
et al, 2021 Europe Winter Both Staff 3221
Not
Weitzel et al, South reporte
2020 (a) America Spring Symptomatic d 109
Not
Weitzel et al, South reporte
2020 (b) America Spring Symptomatic d 109
Not
Weitzel et al, South reporte
2020 (c) America Spring Symptomatic d 111
Young et al,
2020 America Summer Symptomatic Staff 251
Not
Young et al, reporte
2020 America Spring Symptomatic d 251
Young et al,
2021 Europe Winter Both Staff 785

95 cohorts (90%) used nasopharyngeal swabs as a specimen for their antigen tests and the
rest of the studies had different sampling method included saliva sampling, bronchoalveolar
lavage fluids and oropharyngeal swab. The lateral flow devices, produced by 28 different
manufacturers, vary across the 106 cohorts. The geographic distribution of study

14
populations, presence or absence of symptoms and the lateral flow devices used are
demonstrated in Figure 2.

Figure 2. These 3 geographic plots demonstrate the geographic distribution of

study populations, the lateral flow devices used and the presence or absence of 15
symptoms.
The quality assessment of studies as per the QUADAS-2 tool is illustrated in Supplementary
material (Table 2). There was a relatively low risk of detection bias among the studies as per the
second domain in in Supplementary material (Table 2)– “Index Test” – meaning that most
studies carried out interpretation of the lateral flow tests independently to having knowledge of
the test subject’s PCR status. Similarly – Domain 3 – the “Reference Standard” – also had a
generally low level of observer bias – the reference test or PCR was carried out independently
without knowledge of the lateral flow result and we know PCR to have the greatest sensitivity
and specificity of any test for COVID-19 There was a slightly higher risk of selection bias noted
among studies as seen in the first domain of the table. Some studies did not clearly describe the
characteristics of participants, and this may have an impact on diagnostic accuracy when it
comes to test subjects swabbing themselves. Finally, there was also a potential for bias in terms
of unclear patient flow in the last domain – ideally studies should have carried out lateral flow
testing and PCR at the same time but if this was not described it could lead to potentially
discordant results through disease state alteration with time. [12]

Supplementary Figure 4 represents Deeks’ regression test for funnel plot asymmetry –
which showed an absence of publication bias (P>0.05) amongst the studies included in the
meta-analysis.

16
The diagnostic accuracy of lateral flow devices for the detection of SARS Cov 2 antigen.
The pooled sensitivity and specificity of an antigen test for the detection of SARS Cov 2 were
0.72 (95% CI, 0.68 to 0.76) and 1.00 (95% CI, 0.99 to 1.00). Forest plots for sensitivity and
specificity are given in Figures 3 and 4. AUC of bivariate SROC was 0.96 (Supp Figure 2).

Figure 3. Forest plot showing the sensitivities of lateral flow devices for the
detection of SARS Cov 2 and their corresponding confidence intervals. The pooled
sensitivity was 0.72 (95% CI, 0.68 to 0.76)

17
Figure 4. Forest plot showing the specificities of lateral flow devices for the detection of
SARS Cov 2 and their corresponding confidence intervals. The pooled specificity was
1.00 (95% CI, 0.99 to 1.00)

18
There was significant heterogeneity among the studies included in the meta-analysis. This
was further explored through meta-regression analysis in view of identifying the covariates
contributed to the heterogeneity. The co-variates included in the bivariate meta-regression
analysis were – presence of symptoms, manufacturer of the Covid antigen test used,
specimen type, operator of test and season of the country when the test was performed.

The z value for the regression coefficient when carried out by staff rather than test subjects
were significantly higher for sensitivity and lower for false positive rate respectively; with a p
value of 0.04 and 0.001 respectively. (see Supplementary material - Table 3)
Similarly, the sensitivity of the LFDs was found to be higher if the patients had symptoms.
However, presence of symptoms did not have any effect on the test specificity.
Therefore, our analysis implied a higher sensitivity in patients with symptoms – which has
been reported before; such as in the Cochrane review by Dinnes et al. [2]
However, our analysis also implies a potentially higher false positive rate in test subjects
undergoing the test themselves; which was not found in previous studies such as Mistry et
al. [4]
With regard to seasonal variation, specimen type, or manufacturers of the covid antigen test
used, statistical significance was not reached in any of these domains. Hence, no evidence of
any contribution from these co-variates to the heterogeneity was demonstrated.

Discussion
We evaluated the diagnostic accuracy of lateral flow devices used in the detection of COVID-
19 as per current published literature. There were 76 studies and 106 cohorts included with
a total of 108,820 patients in this analysis with minimal publication bias. The findings
showed that detection of COVID-19 using lateral flow devices had pooled sensitivity and
specificity of 0.72 (95% CI, 0.68 to 0.76) and 1.00 (95% CI, 0.99 to 1.00) respectively.
Our study includes more test subjects and spans more continents than other recent
published systematic reviews.[2,4] Furthermore our data presents a significant update
when considering the increased burden of use of lateral flow tests as opposed to earlier in
the pandemic; when PCR was the test mainly used on patients.

19
As shown in Figure 2 – the majority of the studies were conducted in Europe with much less
representation in the literature from countries in Africa or Asia and this may reflect ongoing
global health inequalities in the COVID-19 pandemic.[13] This is a positive of our study in
demonstrating the academic inequalities exposed by the global pandemic as well as the
inclusion of papers among five different continents; making our review a truly global study.

There are some limitations of our study – such as the variable thresholds used to define a
positive result depending on the device manufacturer; and the lack of reporting of exact
environmental temperatures at which an antigen test was performed amongst our studies –
which may limit our attempt to quantify seasonal variation on test accuracy.
Another limitation is the heterogeneity of studies included. However, our meta regression
analysis led to some interesting findings – mainly the increased sensitivity and reduced false
positive rate in trained staff operating the lateral flow tests – which were both statistically
significant results taking into account of the relatively low number of studies which reported
this data.

Furthermore, our study reaffirmed the increased sensitivity of lateral flow tests in
symptomatic patients; which is in keeping with other studies. [14]

Insufficient data were available to investigate effect time after symptom onset – therefore
we are not sure when the sensitivity reduces after onset of symptoms; only that we can
imply the tests are more likely to be negative once symptoms abate.

Within our study the assays shown to meet appropriate criteria, such as WHO’s priority
target product profiles for COVID-19 diagnostics (‘acceptable’ sensitivity ≥ 80% and
specificity ≥ 97%), can be considered as a replacement for laboratory-based RT-PCR when
immediate decisions about patient care must be made, or where RT-PCR cannot be
delivered in a timely manner. [15]

20
Conclusion:

This systematic review reaffirmed that rapid antigen tests have a reasonable sensitivity 0.72
and a high specificity 1.00 that meet the appropriate criteria for COVID-19 diagnostics. Our
data suggested that the test performance is better with trained staffs compared to self-
operator. In addition, the sensitivity of the Lateral flow test is also higher in patients who are
symptomatic.

Future research:
It would be beneficial to consider further studies in investigating the effect of symptom
status or timing after onset of symptom on the diagnostic yield of lateral flow tests. Further
clinical trials, comparing self-swabbing and staff-swabbing as well as monitoring the
temperature at the time of lateral flow tests being performed, are required. By addressing
these factors, this might provide methods to further optimise the precision of the lateral
flow tests.

Contributors:
(1) Substantial contribution to design of work

(2) Drafting and revising the intellectual content

(3) Final approval of version to be published

(4) Agreement to be accountable for works published

Ashwin Krishnamoorthy (1,2,3,4), Subashini Chandrapalan (1,2,3,4), Ayman Bannaga

(1,2,3,4), Gohar Jalayeri Nia (1,2,3,4), Yaqza Hussain (1,2,3,4), Ian Io Lei (2,3,4), Ramesh
Arasaradnam (1,2,3,4)

Declaration of Interest: We have no competing interests.

Data sharing:

21
Data collected for the study will be made available to others where possible. De-identified
patient data, study protocols or data dictionaries will not be available as this is not our
property but rather that of those authors of the studies within our meta analysis.
A copy of the statistical analysis plan can be made available on request to the corresponding
author and with a signed data access agreement.

Funding: None

22
 REFERENCES
[1] Mercer TR, Salit M. Testing at scale during the COVID-19 pandemic. Nature Reviews
Genetics n.d. https://doi.org/10.1038/s41576-021-00360-w.
[2] Dinnes J, Deeks JJ, Berhane S, Taylor M, Adriano A, Davenport C, et al. Rapid, point-of-
care antigen and molecular-based tests for diagnosis of SARS-CoV-2 infection. Cochrane
Database of Systematic Reviews 2021;2021.
https://doi.org/10.1002/14651858.CD013705.pub2.
[3] Mercer TR, Salit M. Testing at scale during the COVID-19 pandemic. Nature Reviews
Genetics n.d. https://doi.org/10.1038/s41576-021-00360-w.
[4] Mistry DA, Wang JY, Moeser ME, Starkey T, Lee LYW. A systematic review of the
sensitivity and specificity of lateral flow devices in the detection of SARS-CoV-2. BMC
Infectious Diseases 2021;21. https://doi.org/10.1186/s12879-021-06528-3.
[5] Andryukov BG. Six decades of lateral flow immunoassay: From determining metabolic
markers to diagnosing covid-19. AIMS Microbiology 2020;6:280–304.
https://doi.org/10.3934/microbiol.2020018.
[6] Considerations on the use of self-tests for COVID-19 in the EU/EEA Key messages. n.d.
[7] Guide to the contents of a Cochrane Diagnostic Test Accuracy Protocol. n.d.
[8] Deeks JJ, Macaskill P, Irwig L. The performance of tests of publication bias and other sample
size effects in systematic reviews of diagnostic test accuracy was assessed. J Clin Epidemiol
2005;58:882–93. https://doi.org/10.1016/J.JCLINEPI.2005.01.016.
[9] R Core Team (2020). — European Environment Agency n.d.
https://www.eea.europa.eu/data-and-maps/indicators/oxygen-consuming-substances-in-
rivers/r-development-core-team-2006 (accessed June 22, 2022).
[10] Shim SR, Kim SJ, Lee J. Diagnostic test accuracy: application and practice using R software.
Epidemiol Health 2019;41:e2019007. https://doi.org/10.4178/EPIH.E2019007.
[11] Doebler P, Holling H. Meta-Analysis of Diagnostic Accuracy with mada n.d.
[12] Stohr JJJM, Zwart VF, Goderski G, Meijer A, Nagel-Imming CRS, Kluytmans-van den
Bergh MFQ, et al. Self-testing for the detection of SARS-CoV-2 infection with rapid antigen
tests for people with suspected COVID-19 in the community. Clinical Microbiology and
Infection 2022;28:695–700. https://doi.org/10.1016/J.CMI.2021.07.039.
[13] Ivers LC, Walton DA. Editorial Novel Coronavirus Disease (COVID-19): Global Health
Equity in Pandemic Response. Am J Trop Med Hyg 2020:1–2.
https://doi.org/10.4269/ajtmh.20-0260.
[14] Lisboa Bastos M, Tavaziva G, Abidi SK, Campbell JR, Haraoui LP, Johnston JC, et al.
Diagnostic accuracy of serological tests for covid-19: Systematic review and meta-analysis.
The BMJ 2020;370. https://doi.org/10.1136/BMJ.M2516.
[15] Target product profiles for priority diagnostics to support response to the COVID-19
pandemic v.1.0 Purpose of the document 2020.

Supplementary material:

23
Full search terms for one database - Covid 19”, “SARS-Cov 2”, “Coronavirus”, “antigen test”,
“antigen detection”, “lateral flow devices

Supplementary Table 1: Different manufacturers of lateral flow device used in different studies.

Different manufacturers of the lateral flow Number of

device studies
Panbio 22
Standard Q 14
Biocredit 8
BinaxNOW 7
Standard F 6
Respi-Strip 5
Elesys 5
Lumipulse 5
Innova 4
Quidel Sofia 3
SARS Cov 2 RAT 3
Bioeasy 3
ESPLINE 2
BD veritor system 2
NowCheck 2
LumiraDx 2
Nadal 2
Huaketai New 1
Alltest 1
Liaison XL 1
AMP rapid test 1
Mologic 1
QuickNavi 1
Medsan 1
Multiple antigen tests 1
COVID 19 Gold 1
Covid-viro 1
Elecsys 1

Supplementary Table 2: Quality assessment for the studies included in the meta-analysis
using QUADAS-2

Quality assessment for the studies included in the meta-analysis using QUADAS-2 tool
PATIENT REFERENCE FLOW AND
Study SELECTION INDEX TEST STANDARD TIMING
24
 Abdelrazik et al,
2020 Low Unclear Low Unclear
Albert et al, 2020 Low Low High Unclear
Alemany et al, unclear Low High Unclear
2020
Asai et al, 2021 Low Low Low Unclear
Ben Abdelhanin unclear Unclear Low Unclear
et al 2021
Berger et al, Low Low Unclear Low
2021(a)
Berger et al, Low Low Unclear Low
2021(b)
Blairon et al, unclear Low High High
2020
Bulilete et al, Low Low Unclear High
2020
Caputo et al, Low Low Low Unclear
2021
Carbonell-
Sahuquillo et al,
2021 Low Low Low High
Cerutti et al, Low Unclear High Unclear
2020
Chiu et al, 2021 unclear Low Low Unclear
Coggiola et al, High Low High Unclear
2021
Courtellemont et High Low High Unclear
al, 2021
Drevinek et al, Low Low Low Unclear
2021
Drevinek et al, Low Low Low Unclear
2021
Escriva et al 2021 unclear Low Unclear Low
Fenollar et al, High Low High Unclear
2020
FIND, 2020 (a) Low Low Unclear Low
FIND, 2020 (b) Low Low Unclear Low
FIND, 2020 (c)1 Low Low Unclear Low
FIND, 2020 (c)2 Low Low Unclear Low
FIND, 2020 (d)1 Low Low Unclear Low
FIND, 2020 (d)2 Low Low Unclear Low
FIND, 2020 (e)1 Low Low Unclear Low
FIND, 2020 (e)2 Low Low Unclear Low
Garcia-Finana et Low Low Low Unclear
al, 2021
Gili et al, 2021 unclear Low Unclear Unclear
Gonzalez- Low Low Low Unclear

25
 Donapetry et al
2021
Gremmels et al,
2020 Low Low High Low
Gupta et al, 2021 Low Unclear High Unclear
Hartard et al,
2021 unclear Unclear Unclear Unclear
Houston et al,
2021 Low Low Low Unclear
Ifko et al, 2021
(a) Low Low Low Unclear
Ifko et al, 2021
(b) Low Low Low Unclear
Igloi et al, 2021 unclear High Unclear High
Jegerlehner et al,
2021 unclear Unclear Unclear Unclear
Kim et al, 2022 unclear Low Low Low
Kiyasu et al, 2021 High Low Unclear Unclear
Korenkov et al,
2021 Low Low Low Unclear
Kruger et al,
2020 Low Low Unclear High
Kruger et al,
2022 (a)1 Low Low Unclear High
Kruger et al,
2022 (a)2 Low Low Unclear High
Kruger et al,
2022 (b)1 Low Low Unclear High
Kruger et al,
2022 (b)2 Low Low Unclear High
Kruger et al,
2022 (c)1 Low Low Unclear High
Kruger et al,
2022 (c)2 Low Low Unclear High
Kruger et al,
2022 (d) Low Low Unclear High
Kruger et al,
2022 (e) Low Low Unclear High
Kruger et al,
2022 (f) Low Low Unclear High
Kruger et al,
2022 (g) Low Low Unclear High
Lambert-Noclot
et al, 2020 unclear Unclear High Unclear
Lee et al, 2021 Low Low Unclear Low
Leiner et al, 2021 Low Low Unclear Low
Leixner et al, unclear Low Low Low

26
 2021
Leli et al, 2021 unclear Low High Low
Linares et al,
2020 unclear Unclear Unclear Unclear
Liotti et al, 2020 Low Low Unclear High
Masia et al, 2020 Low Low Unclear Unclear
Menchinelli et al,
2021 Low Low Low Low
Merino et al,
2020 unclear Unclear Unclear Unclear
Mertens et al,
2020 Low Unclear Low Low
Mitchell et al,
2021 unclear Unclear Unclear High
Murillo-Zamora
et al, 2021 unclear Low Unclear Low
Nalumansi et al,
2020 unclear Unclear Low Low
Ngo Nsoga et al,
2021 Low Low Unclear Unclear
Oh et al, 2021 Low Low Low Low
Pena et al, 2021 unclear Low Low Unclear
Perez-Garcia et
al, 2021 Low Unclear Low Unclear
Perez-Garcia et
al, 2021 unclear Unclear Low Unclear
Peto et al, 2021 unclear Low Unclear Low
Pilarowski et al,
2021 Low Low Unclear Unclear
Pollock et al,
2021 Low Unclear Low Low
Pollreis et al,
2021 unclear Low Unclear Low
Porte et al, 2020
(a) Low Low Unclear Low
Porte et al, 2020
(b) Low Low Unclear Low
Porte et al, 2020
(c) Low Low Low Low
Randriamahazo
et al, 2022 unclear Low Unclear High
Reichert et al,
2021 Low Unclear Low Unclear
Sberna et al,
2021 Low Low Unclear High
Schildgen et al,
2020 (a) unclear Low Low Low

27
 Schildgen et al,
2020 (b) unclear Low Low Low
Schildgen et al, unclear Low Low Low
2020 (b)
Scohy et al, 2020 Low Unclear Unclear Unclear
Shah et al, 2021 Low Low Low Low
Shrestha et al, Low Unclear Low Low
2020
Sood et al, 2021 Low Low Low Unclear
Surasi et al, 2021 Low Low Low Low
Takeda et al, unclear Low Unclear Low
2020
Thirion-Romero Low Unclear Low Unclear
et al, 2021
Tinker et al, 2021 Low Low Unclear Low
Torres et al, 2020 unclear Low Low Low
Turcato et al Low Low Low Unclear
2021
Van der Moeren unclear Unclear Unclear Low
et al, 2021
Veyrenche et al, Low Low Unclear Low
2020
Villaverde et al, unclear Unclear Low Low
2021
Wagenhauser et Low Low Low Unclear
al, 2021
Wagenhauser et Low Low Low Low
al, 2021
Wagenhauser et Low Low Low Low
al, 2021
Weitzel et al, unclear Low Low Low
2020 (a)
Weitzel et al, unclear Low Low Low
2020 (b)
Weitzel et al, unclear Low Low Low
2020 (c)
Young et al, 2020 Low Low Low Low
Young et al, 2020 Low Low Low Low
Young et al, 2021 Low Unclear Low Low

Supplementary Table 3: Bivariate meta-regression analysis result table

28
 Bivariate meta-regression analysis of covariates in the detection of Covid 19

Point Standar 95% CI 95% CI Z-value p-value

estimate d error lower upper
limit limit
Fixed effects
model
Swab Sensitivity
Specimen Intercept 0.73 0.14 0.46 1.00 5.24 0.00
collected Staff 0.38 0.19 0.01 0.75 2.01 0.04
by
False positive rate
Intercept -4.06 0.22 -4.5 -3.62 -18.01 0
Staff -0.99 0.31 -1.6 -0.38 -3.18 0.001

Presence Sensitivity
of Intercept 0.03 0.36 -0.7 0.74 0.07 0.94
symptoms Symptomatic 1.1 0.4 0.31 1.9 2.72 0.006

Specificity
Intercept -5.59 0.64 -6.85 -4.34 -8.74 0
Symptomatic 0.91 0.71 -0.48 2.31 1.28 0.19

29
Supplementary Figure 1. Deeks regression test and funnel plot for the studies
included in the meta-analysis. P=0.07 suggests absence of publication bias

30
Supplementary Figure 2. Summary receiver operator curve characteristics for lateral flow
devices for the detection of Covid 19. Green circle indicates 95% confidence interval. The
study distribution is represented by triangles.

31