Diagnostic Accuracy of SARS-CoV-2 Antigen Tests for Community Transmission Screening: A Systematic Review and Meta-Analysis

Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2) caused the global pandemic of coronavirus disease 2019 (COVID-19). Rapid identification and isolation of infectious patients are critical methods to block COVID-19 transmission. Antigen tests can contribute to prompt identification of infectious individuals. This meta-analysis aims to evaluate the diagnostic accuracy of antigen tests for SARS-CoV-2. We conducted a literature search in PubMed, Embase, the Cochrane Library, and Biomed Central databases. Studies evaluating the diagnostic accuracy of antigen tests for SARS-CoV-2 in community participants were included. Only English-language articles were reviewed. We included eligible studies that provided available data to construct a 2 × 2 table on a per-patient basis. Overall sensitivity and specificity for antigen tests were generated using a bivariate random-effects model. Eighteen studies with 34,865 participants were retrieved. The meta-analysis for SARS-CoV-2 antigen tests generated a pooled sensitivity of 0.82 and a pooled specificity of 1.00. A subgroup analysis of ten studies that reported outcomes for 5629 symptomatic participants generated a pooled sensitivity of 0.87 and a pooled specificity of 1.00. Antigen tests might have higher sensitivity in detecting SARS-CoV-2 in symptomatic patients in the community and may be an effective tool to identify patients to be quarantined to prevent further SARS-CoV-2 transmission.


Introduction
Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2) caused the global pandemic of coronavirus disease 2019 . Asymptomatic cases make COVID-19 difficult to monitor and prevent. It is estimated that at least 50% of COVID-19 patients contract the virus from asymptomatic people [1]. To break the transmission chains of SARS-CoV-2, testing infected individuals and tracing and quarantining their contacts have been used as major nonpharmaceutical interventions [2]. Rapid identification and isolation of infectious patients with SARS-CoV-2 are critical methods to block COVID-19 community The study was reported according to "Preferred Reporting Items for a Systematic Review and Meta-Analysis of Diagnostic Test Accuracy Studies: The PRISMA-DTA Statement" [10].
We conducted a literature search for relevant studies in PubMed, Embase, the Cochrane Library, and Biomed Central. A literature search was conducted using multiple search terms, including (COVID-19 OR severe acute respiratory syndrome coronavirus 2 OR SARS-CoV-2) AND (antigen test OR SARS-CoV-2 antigens OR Mass Screening OR Community Participation) AND (RT-PCR OR Reverse Transcriptase Polymerase Chain Reaction OR COVID-19 Nucleic Acid Testing) AND (sensitivity OR specificity). A combination of free text and MeSH terms was used to identify relevant studies. We limited our search results to studies performed with human participants. Detailed search strategies are presented in Table S1.

Inclusion and Exclusion Criteria
Studies evaluating the diagnostic accuracy of antigen tests for SARS-CoV-2 with reference standards in participants with suspected SARS-CoV-2 infection in the community were included, but review articles were excluded. Respiratory specimens were collected from symptomatic or asymptomatic individuals. Studies that defined RT-PCR technology as the reference standard were included. Only English-language articles were reviewed. The literature search was conducted with no time restrictions. Studies that provided sufficient data to construct a 2 × 2 table on a per-patient basis were included. We excluded case reports, case series, proposals, protocols, conference abstracts, in-house tests, and preprint articles. The last literature search was performed on 1 August 2021. One reviewer initially screened titles and abstracts for potentially eligible studies. After eliminating irrelevant studies, two reviewers independently examined full-text articles that met the inclusion criteria. Disagreements between the reviewers were resolved through joint discussions.

Quality Assessment
The quality of the included studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool [11]. Antigen tests for the SARS-CoV-2 virus were the index tests and RT-PCR test results for SARS-CoV-2 were the reference standards. The QUADAS-2 tool consists of the following four domains: patient selection, index test, reference standard, and flow and timing. Each domain includes questions that allow an assessment of the risk of bias. The quality assessment of the diagnostic test comprises the risk of bias and the applicability for individual studies. A study is considered high-quality if each domain in the study exhibits a low risk of bias.

Statistical Analysis
We extracted data on true positives, true negatives, false positives, and false negatives from each included study to construct 2 × 2 tables for calculating values of the pooled sensitivity, pooled specificity. If 2 × 2 tables could not be extracted from the main text, we searched the supplementary material of the study for additional information. The sensitivity of a test is the proportion of those with the target condition correctly identified as having the condition, whereas the specificity of a test is the proportion of those without the target condition correctly identified as not having the condition [12].
We conducted a meta-analysis using a bivariate random-effects model to generate a summary of sensitivity, specificity on a per-patient basis. We also graphed the summary receiver operating characteristic (SROC) curve to determine the overall diagnostic performance of the index tests. The closer the curve approaches the upper-left corner, the higher the overall performance is [13]. Possible causes of heterogeneity between studies were explored through pre-specified subgroup analysis, which included the following: days after symptom onset, asymptomatic participants, and symptomatic individuals. Summary estimates, including pooled sensitivity, specificity, and DOR, were generated with associated 95% confidence intervals (CIs). All analyses were performed using MetaDiSc version 1.4 (Universidad Complutense, Madrid, Spain) and MetaDTA software (National Institute for Health Research Complex Review Support Unit, Glasgow, UK) [14,15]. A value of p < 0.05 was considered statistically significant.

Quality Assessment
We applied the QUADAS-2, which has four domains to evaluate the quality of studies, in our meta-analysis. Regarding patient selection, eight studies enrolled patients randomly or consecutively. All studies avoided a case-control study design, which might have overestimated the diagnostic accuracy. Based on the rules in this domain, eight studies were judged to have a low risk of bias in the patient selection domain [16][17][18][19]28,30,31,33]. Regarding index tests, all studies recorded that index tests were interpreted without knowledge of the results of the reference standard. All studies in the meta-analysis were judged to have a low risk of bias in the index domain. Regarding the reference standard, all studies indicated that the reference standard likely correctly classified the target condition. Regarding the flow and timing domain, 17 studies demonstrated that all patients received a reference standard [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30]32,33]. Seven studies indicated that all patients were included in the analysis [24][25][26][27][28]30,33]. Five articles were judged as having a low risk of bias in the flow and timing domain [24,[26][27][28]33]. With regard to applicability, index tests and reference standards of studies in our meta-analysis matched our review title. Table 3 presents the quality of studies. Figure 3 demonstrates the risk of bias of individual studies in the meta-analysis.        The first quadrant of a circle represents patient selection, the second quadrant represents the index test, the third quadrant represents the reference standard, and the fourth quadrant represents flow and timing. Circles on the SROC plot are colored depending on their quality assessment score: green for low, red for high, and gray for unclear risk of bias.

Discussion
Our major findings indicated that antigen tests had high sensitivity and excellent specificity in detecting SARS-CoV-2 in individuals in the community. If a test (in this case, an antigen test) has high specificity and yields a positive result, a clinician can be nearly certain that the disease (in this case, COVID-19) is present [35]. Antibody testing also plays a crucial role in understanding the seroprevalence of COVID-19 in the community and identifying individuals who are immunoreactive against SARS-CoV-2 [34]. RT-PCR is the standard diagnostic tool for SARS-CoV-2 detection. A previous study reported that RT-PCR positivity may persist over 3 weeks after illness onset, although most mild cases will yield a negative result. However, a positive RT-PCR result reveals only SARS-CoV-2 RNA and does not necessarily indicate the presence of a replicating virus [36].
Based on the subgroup analysis of our meta-analysis, antigen tests might have higher sensitivity in detecting SARS-CoV-2 in symptomatic individuals in the community, which indicates that antigen tests might be reliable for SARS-CoV-2 detection among the contagious population. In another subgroup analysis of studies involving asymptomatic participants, antigen tests had insufficient sensitivity in detecting SARS-CoV-2 in the asymptomatic population in the community. Surveillance testing regimens that can sever enough transmission chains to reduce community spread should complement current clinical diagnostic tests. Antigen tests could be used to enable true community-wide surveillance regimens for SARS-CoV-2 [37]. The current meta-analysis provided evidence of high sensitivity of antigen tests in identifying symptomatic individuals in the community. The antigen test is a valuable nonpharmaceutical intervention strategy to contain SARS-CoV-2. Recent research has suggested that when antigen tests are used, a pretest quarantine period of 5 days is not inferior to a quarantine of 10 days for travelers and a postexposure quarantine period of 10 days is not inferior to a quarantine of 14 days [38]. Gradual release of nonpharmaceutical interventions coupled with a high-efficacy vaccine strategy might prevent subsequent waves of SARS-CoV-2 transmission. Although vaccination can allow for some relaxation of nonpharmaceutical control measures, such relaxation should be performed gradually to avoid large-scale public health consequences [39].
Frequent use of antigen tests might help to identify infected individuals and reduce COVID-19 transmission [6] The benefits of administering antigen tests in suspected cases are the rapid diagnosis for clinical treatment and management (including protection of first-line staff) and the ability to quarantine infected individuals. Contact tracing becomes feasible so that positive cases can be isolated to minimize SARS-CoV-2 spread [40]. Diagnostic testing plays a key role in COVID-19 outbreak control. To end the pandemic, the accurate application of high-volume diagnostic testing and the rapid use of the results may help with the timely implementation of appropriate therapies and prevention of further spread [41]. Antigen tests could increase overall COVID-19 testing capacity and have the advantages of shorter turnaround times and reduced costs [42]. Antigen tests are most likely to have high performance in patients with high viral loads (Ct values ≤ 25), which usually appear in the presymptomatic (1-3 days before symptom onset) and early symptomatic (within the first 5-7 days of illness) phases of COVID-19 [43].
A study reported that a requirement to quarantine until an RT-PCR or antigen test on Day 7 after exposure (with early release if negative) might prevent as much transmission as the standard 14-day quarantine period [44]. Testing of asymptomatic health care workers has been suggested to reduce nosocomial transmission of COVID-19 [45]. Therefore, antigen tests can be used for screening and serial testing (every 2-3 days) of residents and staff in health care, home care, and long-term care facilities in areas where there is ongoing community transmission. When a first case is confirmed in a resident or staff member of a closed setting, a comprehensive testing strategy of all residents and staff should be considered [42].
Different testing strategies, including focused symptomatic testing, focused asymptomatic testing, mass testing, and systematic meaningful asymptomatic repeated testing, are adopted to prevent transmission. Many of these strategies use antigen tests [46]. Widespread community transmission has become entrenched in many countries and has required the testing of populations to identify and isolate infected individuals. Although the effects of mass antigen testing are difficult to distinguish from those of concurrent interventions, an obvious reduction in SARS-CoV-2 infections was observed after mass antigen testing in Slovakia. However, this reduction was restricted to regions with high SARS-CoV-2 prevalence, and testing had little effect in areas with lower viral prevalence [47]. Due to an increasing prevalence of new SARS-CoV2 variants with possible clinical implications, monitoring and detecting the spread of different variants in the general population in a timely method is critical [48]. Antigen testing is evolving. Sensitivity of SARS-the CoV-2 antigen test with a self-collected nasal swab is comparable with that of a professionalcollected nasopharyngeal swab. Patients suspected of COVID-19 may be able to perform the antigen test and test by themselves [18]. Based on the outcomes of the meta-analysis, antigen tests might have higher sensitivity in symptomatic patients. Hence, we suggested that RT-PCR could be performed after negative antigen test results in symptomatic patients and positive antigen test results in asymptomatic patients in the community transmission screening algorithm for SARS-CoV-2 [8].
Although this meta-analysis demonstrated that antigen tests may be sensitive in detecting SARS-CoV-2 in the community, our study had limitations. The Ct cutoff values of the studies in the meta-analysis were limited. Statistical data of antigen tests stratified by Ct cutoff value are limited. Statistical data of antigen tests for patients under 18 years of age are limited. The majority of studies in this meta-analysis did not report the days after symptom onset of participants. No study in the meta-analysis reported SARS-CoV-2 variants.

Conclusions
Our major findings indicated that antigen tests had high sensitivity in detecting the SARS-CoV-2 virus in symptomatic patients in the community. Antigen tests might have a higher sensitivity in detecting SARS-CoV-2 within 7 days after symptom onset.
Antigen tests are sensitive in detecting SARS-CoV-2 in patients with a Ct value less than or equal to 35. Therefore, antigen tests might be an effective tool in the effort to block SARS-CoV-2 transmission.