key: cord-1021563-8cl4mt04
authors: Zobrist, Stephanie; Oliveira-Silva, Michelle; Vieira, Alexia Martines; Bansil, Pooja; Gerth-Guyette, Emily; Leader, Brandon T; Golden, Allison; Slater, Hannah; de Lucena Cruz, Catherine Duran; Garbin, Eduardo; Sagalovsky, Mariana; Pal, Sampa; Gupta, Vin; Wolansky, Leo; Vieira Dall’Acqua, Deusilene Souza; Naveca, Gomes Felipe; do Nascimento, Valdinete Alves; Villalobos Salcedo, Juan Miguel; Drain, Paul K; Tavares Costa, Alexandre Dias; Domingo, Gonzalo J; Pereira, Dhélio
title: Screening for SARS-CoV-2 in close contacts of individuals with confirmed infection: performance and operational considerations
date: 2022-05-20
journal: J Infect Dis
DOI: 10.1093/infdis/jiac204
sha: cdb46090460d9a1fb626ea408116ba0715b0f8b7
doc_id: 1021563
cord_uid: 8cl4mt04

BACKGROUND: Point-of-care and decentralized testing for SARS-CoV-2 is critical to inform public health responses. Performance evaluations in priority use cases such as contact tracing can highlight trade-offs in test selection and testing strategies. METHODS: A prospective diagnostic accuracy study was conducted among close contacts of COVID-19 cases in Brazil. Two anterior nares swabs (ANS), a nasopharyngeal swab (NPS), and saliva were collected at all visits. Vaccination history and symptoms were assessed. Household contacts were followed longitudinally. Three rapid antigen tests and one molecular method were evaluated for usability and performance against reference RT-PCR on NPS. RESULTS: Fifty index cases and 214 contacts (64 household) were enrolled. Sixty-five contacts were RT-PCR positive during at least one visit. Vaccination did not influence viral load. Gamma variants were most prevalent; Delta emerged increasingly during implementation. Overall sensitivity of evaluated tests ranged from 33%–76%. Performance was higher among symptomatic cases and cases with Ct < 34 and lower among oligo/asymptomatic cases. Assuming a 24-hour time-to-result for RT-PCR, the cumulative sensitivity of an ANS rapid antigen test was >70% and almost 90% after four days. CONCLUSIONS: The near immediate time-to-result for antigen tests significantly offsets lower analytical sensitivity in settings where RT-PCR results are delayed or unavailable.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus, which causes COVID-19, has 2 significantly burdened health systems globally, with over 22 million confirmed cases in Brazil alone as of 3 2021 [1]. A key challenge of the pandemic response is access to appropriate diagnostic testing, which is 4 critical to inform timely and targeted clinical management and public health strategies [2] . 5

The reference standard for SARS-CoV-2 testing is RT-PCR. While accurate, this method has many 6 practical limitations, including cost, laboratory infrastructure requirements, and often invasive sampling. 7

RT-PCR testing is typically centralized, which can lead to delays in reporting results to patients. Such 8 delays have important public health implications, including increased risk for transmission during the 9 period before results are available to infected individuals [3, 4] . Expanded access to decentralized and 10 point-of-care (POC) testing is essential to identify cases early and limit community transmission, 11 particularly where RT-PCR is unavailable. 12

Infected persons both with and without symptoms can transmit SARS-CoV-2 [5] [6] [7] . Due to the 13 significance of asymptomatic transmission [8] , testing these populations is often recommended, including 14 close contacts of individuals with confirmed infection as part of contact tracing, testing, and isolation 15 strategies [9, 10] . However, contact tracing can be time and resource intensive, particularly during periods 16 of high transmission, which can limit its implementation in practice. 17

Multiple platforms have been developed to enable decentralized and POC SARS-CoV-2 testing [11] . In 18 particular, rapid antigen tests have garnered interest due to their lower cost, ease of use, and rapid 19 turnaround time for results (typically under 30 minutes) [10, 12] . The World Health Organization (WHO) 20 advises that rapid antigen tests meeting minimum performance criteria can be employed in a range of use 21 cases, including for testing of asymptomatic contacts of cases [10] . Previous studies of rapid antigen test 22

performance have shown variability, with strongest performance among symptomatic individuals with 23 high viral loads in early stages of infection [11, [13] [14] [15] . Several studies have investigated test 24

performance among contacts of confirmed cases [16, 17] collected. Presence, duration, and severity of symptoms were assessed at all visits. At each visit, two 3 paired anterior nares swabs (ANS), one nasopharyngeal swab (NPS), and saliva were collected 4 (Supplementary Material B). One ANS was used to run the STANDARD Q COVID-19 Ag Nasal Test 5 during the visit. All specimens were then transferred to a laboratory where the remaining tests were 6 performed. 7

For the longitudinal study, household contacts were followed every other day for up to five visits total, or 8 until the POC screening test was positive. One additional visit was performed after this positive result, , 9 during which NPS were not collected to minimize staff exposure. Participants were considered lost to 10 follow-up after two missed visits. 11

The extracted ANS mixed with LumiraDx buffer was frozen within five hours of collection and thawed 13 before testing, no more than five days after freezing. The saliva sample was also frozen, and aliquots were 14 thawed for testing with the STANDARD Q COVID-19 Ag Saliva Test (within five days of freezing) and 15 analyze data from the EoU questionnaire, a matrix was used to rank aspects of the products' usability as 5 "satisfactory," "average," or "unsatisfactory" (Supplementary Material E) [21] . 6

Sample size and statistical analysis 7 The sample size targeted at least 50 contacts with a positive reference result, including at least 20 8 asymptomatic individuals, to meet US FDA Emergency Use Authorization requirements [31] . 9

Participants with no symptoms at the time of sampling were classified as asymptomatic. Participants were 10 considered symptomatic if they presented with cough, shortness of breath, difficulty breathing, or at least 11 two of the following symptoms at the time of sampling: fever, chills, rigor, myalgia, headache, sore 12 throat, new olfactory or taste disorder [32] . Participants who presented with one or more mild symptoms 13 but did not fit the symptomatic case definition and reported no care seeking or changes to behavior were 14 considered oligosymptomatic. 15 Sensitivity, specificity, and positive and negative predictive values were calculated using standard 16 formulas and presented with 95% CIs. Samples for which both RT-PCR and evaluated test results were 17 available were included in the analysis. Using the longitudinal dataset, trade-offs between performance 18 and utility of the evaluated tests in terms of cumulative sensitivity at specified time points were assessed 19 as a function of time-to-results. Here, we use the term 'cumulative sensitivity' to refer to the probability 20 that a rapid test will identify a SARS-CoV-2 positive individual at any point during the nine-day serial- (Table 1) The two POC ANS antigen tests demonstrated comparable performance, with overall sensitivity of 55.0% 10 for the STANDARD Q (95% CI 43.5%-66.2%) and 50.6% for LumiraDx (95% CI 39.1%-62.1%) (Table  11 2). Performance increased to >80% sensitivity for both tests among symptomatic cases but decreased to 12 <30% among oligo/asymptomatic cases. For specimens with Ct values less than 34, above which viral 13 viability is negligent and quantification is not as reliable [34,35], performance of both tests improved, 14 with sensitivities in the ranges of 90% and 60% for symptomatic and oligo/asymptomatic cases, 15 respectively. 16

The SalivaDirect PCR assay showed the highest overall performance at 75.9% sensitivity (95% CI 17 65.0%-84.9%), which increased to 88.2% (95% CI 76.1%-95.6%) among contacts with Ct<34. In all 18 scenarios, the rapid STANDARD Q Saliva Test had a sensitivity of <60%, although performance 19 increased among symptomatic positive cases at lower Ct levels. 20 Figure 3 presents the viral load of positive specimens, stratified by results of the STANDARD Q Nasal 21

and Saliva tests. Overall, specimens with low viral loads were more likely to yield negative results; 22 however, misclassification of specimens with high viral loads was more common with the saliva test. hours, >70% of contacts would have been identified by a POC test. At 48 hours, cumulative sensitivity is 13 80%, increasing to nearly 90% at four days. 14 Usability 15

In total, 12 study staff completed the usability assessment. All three POC antigen tests were considered 16 easy to use and SUS scores were acceptable (>77) (Supplementary Material J). 17

In this study, performances of three POC antigen tests (two ANS and one saliva) and one molecular assay 19 for SARS-CoV-2 in saliva were assessed among close contacts of COVID-19-positive index cases. 20

All evaluated tests demonstrated strongest performance among symptomatic cases-and particularly 21 those with Ct values <34. Performance decreased among oligo/asymptomatic cases, which is consistent 22

with results of prior studies [11, 13] and may indicate that the tests are best able to detect those most likely 23

to be infectious [34, 35] . However, there is no universal Ct value cut-off-point that corresponds to 1 infectivity, and the relationship between Ct values and viral load varies by laboratory [11] . 2

The SalivaDirect assay had the best performance, with sensitivity of up to 90% among contacts with 3

Ct<34. Although this assay uses a noninvasive sample type and a simplified procedure that minimizes 4 processing time and costs, infrastructure and training requirements still limit the feasibility of 5 implementing this test in many settings, with potential implications for time-to-results. 6

The saliva antigen test had the lowest overall performance. Other evaluations of POC saliva antigen tests 7 have also shown variable but generally sub-optimal performance [36,37]. One recent evaluation of this 8 test reported an overall sensitivity of 66.1%; however, the reference assay was conducted on saliva [38] . 9

In this study, the test was run on passively collected saliva. This may have impacted performance, as the 10 manufacturer recommends use of actively collected saliva with snorted nasal mucus. 11

The two POC ANS antigen tests-STANDARD Q Nasal and LumiraDx-demonstrated comparable 12 performance which was best among cases with Ct<34, with sensitivities in the ranges of 90% and 60% for 13 symptomatic and asymptomatic cases, respectively. Among symptomatic cases and those with Ct<34, 14 both tests met WHO performance criteria (≥ 80% sensitivity and ≥ 97% specificity) [10] . Both tests were 15 also considered easy to use; however, the LumiraDx test requires the use of an instrument. 16

Overall, the observed positivity rate among close contacts in this study ( Immediate results can impact behavior of potentially infectious individuals, encouraging earlier isolation 2 and signaling where additional testing is warranted [4] . The emergence of antiviral therapies-which are 3 more effective the sooner they are taken-further underscores the value of timely results. 4

Limitations of the study include its modest sample size, reflected in the 95% CIs reported with 6 performance indicators. Further, the STANDARD Q Nasal and LumiraDx tests are among the best-in-7 class commercial POC antigen tests. Other tests with lower performance may increase the risk of missing 8 infections against the benefit of identifying cases, to the extent that other strategies may be needed if RT-9 PCR is unavailable. Lastly, only Gamma and Delta variants were observed in this study; future research 10 should investigate implications of new variants on diagnostic performance across sample types. 11

The near immediate time-to-result of rapid antigen tests is a significant benefit that offsets reduced 13 sensitivity by decreasing diagnostic delays and onward viral transmission. Here, we demonstrate that 14 POC ANS antigen tests for SARS-CoV-2 are easy to use and perform adequately to provide prompt, 15 actionable information to both the health system and individuals. 16 The authors would like to thank all study participants as well as the clinical and laboratory staff at 25 CEPEM involved with this study. We also thank SD Biosensor and LumiraDx for facilitating the 26 availability of their tests for this study. Finally, we also thank Amanda Tsang and Christine Waresak for 27 editorial support with the manuscript. 28 

Coronavirus Disease

Considerations for diagnostic

COVID-19 tests

Impact of delays on effectiveness of contact tracing strategies for COVID-19: a modelling 9 study

Test sensitivity is secondary to frequency and 11 turnaround time for COVID-19 screening

Asymptomatic SARS-CoV-2 infection: a systematic 14 review and meta-analysis

Occurrence and transmission potential of 17 asymptomatic and presymptomatic SARS-CoV-2 infections: a living systematic review and meta-18 analysis

Estimating the extent of asymptomatic COVID-19 and 20 its potential for community transmission: Systematic review and meta-analysis

Containment of COVID-19: Simulating the impact 1 of different policies and testing capacities for contact tracing, testing, and isolation

World Health Organization. Antigen-detection in the diagnosis of SARS-CoV-2 infection

Rapid, point-of-care antigen and molecular-based tests for 8 diagnosis of SARS-CoV-2 infection

Drain P. Rapid diagnostic testing for SARS-CoV-2

Accuracy of novel antigen rapid diagnostics 13 for SARS-CoV-2: a living systematic review and meta-analysis

Performance and operational feasibility of antigen and 17 antibody rapid diagnostic tests for COVID-19 in symptomatic and asymptomatic patients in 18

Cameroon: a clinical, prospective, diagnostic accuracy study

Evaluation of a rapid antigen test 1 (Panbio™ COVID-19 Ag rapid test device) for SARS-CoV-2 detection in asymptomatic close 2 contacts of COVID-19 patients

Diagnostic accuracy of rapid antigen tests in 5 asymptomatic and presymptomatic close contacts of individuals with confirmed SARS-CoV-2 6 infection: cross sectional study

Epidemiological Surveillance Guide: Public Health Emergency of National 8 Importance due to Coronavirus Disease

LumiraDx Microfluidic Immunofluorescence Point-of-Care SARS-CoV-2 Antigen Test in 12 asymptomatic adults and children

The comparative clinical performance of four SARS-CoV-2 15

rapid antigen tests and their correlation to infectivity in vitro

Evaluation of accuracy, exclusivity, limit-of-detection and 18 ease-of-use of LumiraDx™: An antigen-detecting point-of-care device for SARS-CoV-2 19

Evaluation of an antigen-based test for hospital point

An empirical evaluation of the System Usability 1

United States Food and Drug Administration. Template for Developers of Antigen Tests

Research electronic data 11 capture (REDCap)-a metadata-driven methodology and workflow process for providing 12 translational research informatics support

Diagnostic performance of 14 rapid antigen tests (RATs) for SARS-CoV-2 and their efficacy in monitoring the infectiousness of 15 COVID-19 patients

Viral RNA load as determined by cell culture as a 17 management tool for discharge of SARS-CoV-2 patients from infectious disease wards

Performance of qualitative and quantitative antigen tests 20 for SARS-CoV-2 using saliva

Sensitivity of rapid antigen testing and RT-PCR 22