key: cord-0925412-saf76msy
authors: Kolwijck, E.; Brouwers-Boers, M.; Broertjes, J.; van Heeswijk, K.; Runderkamp, N.; Meijer, A.; Hermans, M.H.A.; Leenders, S. C.
title: Validation and implementation of the Panbio COVID-19 Ag rapid test for the diagnosis of SARS-CoV-2 infection in symptomatic hospital healthcare workers
date: 2021-04-16
journal: Infection prevention in practice
DOI: 10.1016/j.infpip.2021.100142
sha: 530d3a523de729a287468d3eec62bd71d338b966
doc_id: 925412
cord_uid: saf76msy

Background RT-PCR is the current recommended laboratory method to diagnose SARS-CoV-2 in healthcare workers (HCW). As RT-PCR is not widely available and is time-consuming, it limits decision making on removal from and return to work of possibly contagious HCW. Aim In this study we evaluated the Panbio™ COVID-19 Ag rapid test (PanbioCAgRT) in 825 hospital HCW. Methods and findings: This study consisted of two phases. In the validation phase, we tested hospital HCW with mild symptoms (three days or less) in parallel using the PanbioCAgRT and the RT-qPCR test. The PanbioCAgRT demonstrated 86.7% sensitivity, 100% specificity, 100% PPV and 98.5% NPV with regard to RT-qPCR. For HCW with PanbioCAgRT-/RT-qPCR+, the median Ct value was 30.9, whereas for the HCW with PanbioCAgRT+/RT-qPCR+ the median Ct value was 19.3 (p<0.001). In the second phase, we implemented an on-site antigen test-based strategy for symptomatic hospital HCW: HCW that tested positive with the PanbioCAgRT on-site were considered SARS-CoV-2 positive and were sent home. HCW that tested negative with the PanbioCAgRT on-site were allowed to work with PPE pending RT-qPCR test results from the laboratory. Sensitivity of the antigen test-based strategy was 72.5% and NPV was 97%. For HCW with PanbioCAgRT-/RT-qPCR+ median Ct values were 27.8. Conclusion The PanbioCAgRTt validated in this study showed a high sensitivity and specificity in samples obtained from HCW with high viral loads. The antigen-based testing strategy proposed in this study seems to be effective, safe and easy to implement in a wide range of occupational healthcare settings.

Background 28 RT-PCR is the current recommended laboratory method to diagnose SARS-CoV-2 in healthcare 29 workers (HCW). As RT-PCR is not widely available and is time-consuming, it limits decision making on 30 removal from and return to work of possibly contagious HCW. 31

Aim 32

In this study we evaluated the Panbio™ COVID-19 Ag rapid test (PanbioCAgRT) in 825 hospital HCW. 33

This study consisted of two phases. In the validation phase, we tested hospital HCW with mild 35 symptoms (three days or less) in parallel using the PanbioCAgRT and the RT-qPCR test. The 36

PanbioCAgRT demonstrated 86.7% sensitivity, 100% specificity, 100% PPV and 98.5% NPV with 37 regard to RT-qPCR. For HCW with PanbioCAgRT-/RT-qPCR+, the median Ct value was 30.9, whereas 38 for the HCW with PanbioCAgRT+/RT-qPCR+ the median Ct value was 19.3 (p<0.001). In the second 39 phase, we implemented an on-site antigen test-based strategy for symptomatic hospital HCW: HCW 40 that tested positive with the PanbioCAgRT on-site were considered SARS-CoV-2 positive and were 41 sent home. HCW that tested negative with the PanbioCAgRT on-site were allowed to work with PPE 42 pending RT-qPCR test results from the laboratory. Sensitivity of the antigen test-based strategy was 43 72.5% and NPV was 97%. For HCW with PanbioCAgRT-/RT-qPCR+ median Ct values were 27.8. 44

Introduction 1 Access to rapid, reliable testing for SARS-CoV-2 infection in and possible contagiousness of 2 healthcare workers (HCW) is key to maintaining an adequate work force. Currently, reverse 3 transcriptase real time polymerase chain reaction (RT-PCR) is the recommended test for detection of 4 SARS-CoV-2 infection in symptomatic HCW. [1, 2] In the Netherlands, high-throughput testing centres 5 for symptomatic healthcare employees using RT-PCR are widely available in hospitals, with test 6 results offered within six to ten hours. Most HCW in long-term care facilities and primary care 7 centres can make use of these hospital-based testing centres as well, but logistic difficulties may 8 prolong time to results. 9

In HCW crucial to the healthcare system, rapid test results can enable employees who tested 10 negative for SARS-CoV-2 to return to work quickly (wearing personal protective equipment (PPE) 11 until their symptoms have resolved). HCW crucial to the healthcare system that test positive might 12 be removed sooner from their work environment, therewith lowering the risk of further spreading 13 the virus to colleagues. 14 Antigen tests are generally less sensitive than RT-PCR for SARS-CoV-2 detection, but are easy to 15 perform, relatively inexpensive, have short turnaround times, and may facilitate decentralized 16 testing. In addition, early data suggest that antigen tests may play a key role in rapidly identifying 17 those at highest risk for transmitting disease. [3] [4] [5] [6] [7] [8] Abbott Panbio TM COVID-19 Ag rapid test 18 (PanbioCAgRT) is a lateral flow immunoassay that is used to detect the nucleocapsid protein of SARS-19

CoV-2 in approximately 15 minutes. Antigenic tests have been shown to be most valid in the days 20 around the onset of symptoms, when the viral load in the nasopharynx is highest. [3, [9] [10] [11] [12] [13] ] 21

To our knowledge, this is the first study that was initiated to evaluate the use of rapid antigen tests in 22 hospital HCW. The primary objective of the study was to determine the clinical specificity and 23 sensitivity of the PanbioCAgRT for detection of SARS-CoV-2 compared to the RT-qPCR in hospital 24 HCW being mildly symptomatic for three days or less (validation phase). Secondary objective was to 25 determine the clinical sensitivity of the PanbioCAgRT for different cycle threshold (Ct) value groups.

After the validation phase, an antigen test-based strategy for HCW was implemented in our hospital 1 and described in this study. 2 3

Ethics 5

The medical research ethics committee (MREC) Brabant decided it was not necessary to subject the 6 study to the Medical Research Involving Human Subjects Act (WMO) and did not require full review 7 by an accredited MREC (NWO2021-10). All participants have provided informed consent. 8 9

The study was performed at a single peripheral teaching hospital in the Netherlands, with 700 beds 11 offering general and complex medical care. Currently, the hospital has over 4000 employees. 12

The validation phase of the study was conducted between October 5th and October 30 th 2020. The 13 implementation phase was conducted between November 5 th and November 29 th 2020. The overall 14 incidence of SARS-CoV-2 infection (confirmed by RT-PCR) among all routinely tested hospital HCW 15 (also HCW that were not included in the study) varied during the course of the study and was 9% 16 during the first two weeks of the study (validation phase), 16% during week three to four (validation 17 phase), 11% during week five to six (implementation phase), and 10% in week seven to eight 18 (implementation phase). 19 20

Hospital HCW employed at the Jeroen Bosch hospital with mild symptoms suggestive of COVID- 19 22 (according to the national COVID-19 protocol of the National Institute for Public Health and the 23 Environment (RIVM)), presenting to the hospital testing centre were informed about the study and 24 enrolled if they consented. Only participants that were 16 years or older and were symptomatic for 25 three days or less were included. Asymptomatic HCW with recent exposure to a SARS-CoV-2 infected 1 person were excluded. 2

In the validation phase, HCW were tested by both PanbioCAgRT and RT-qPCR. During this phase, the 3 participants and occupational health personnel were unaware of the PanbioCAgRT results. 4

In the implementation phase, a rapid antigen-based test strategy for symptomatic HCW (aged ≥16 5 years) in our hospital was realised. The excellent specificity of the PanbioCAgRT (see results) led us to 6 decide to extend the test period by one day (from three days or less from symptom onset to four 7 days or less from symptom onset) to create more flexibility for our HCW to schedule. With the rapid 8 antigen-based strategy, hospital HCW that tested positive with the PanbioCAgRT were considered 9 SARS-CoV-2 positive, (due to excellent specificity during the validation phase (see results)), did not 10 undergo further testing and were directly sent home after test results became available. HCW that 11 tested negative with the PanbioCAgRT received a confirmation RT-qPCR test and could, if considered 12 crucial to healthcare, return to work pending RT-qPCR results, using PPE at all times. 13 14

From each symptomatic HCW, two nasopharyngeal samples and one oropharyngeal sample were 16 collected. One nasopharyngeal and one oropharyngeal swab were taken for routine viral genome 17 detection by RT-qPCR and placed in a single 3 mL viral transport medium (VTM). A second 18 nasopharyngeal swab, provided in the PanbioCAgRT test kit, was taken for antigen testing as 19 recommended by the manufacturer. The second nasopharyngeal swab was obtained through the 20 contralateral nostril. All swabs from individual participants were taken by the same trained health 21 care professional. 22

In the validation phase, both the VTM tube and PanbioCAgRT tube were, as recommended by the 23 manufacturer, transported within two hours to the hospital's medical microbiological laboratory. The 24

PanbioCAgRT was immediately performed by an experienced laboratory technician and the VTM 25 tube was tested by RT-qPCR on the same day.

In the implementation phase, the PanbioCAgRT was performed immediately at the sample collection 1 site by a trained nurse. When the PanbioCAgRT was positive, the VTM tube of the HCW was 2 discarded. When the PanbioCAgRT was negative, the VTM tube was transported to the hospital's 3 medical microbiological laboratory and was tested using RT-qPCR on the same day. 4

Adequate PPE was used while collecting the swabs and performing the antigen tests. Primary outcome was the PanbioCAgRT clinical sensitivity and specificity with 95% confidence 6 interval (95% CI) compared to qRT-PCR, which was considered gold standard. The secondary 7 outcome was clinical sensitivity with 95% CI compared to qRT-PCR stratified by Ct value category. 8

Continuous variables were shown as median and interquartile range (IQR). Mann-Whitney U-test was 9 used to compare differences between Ct values of PanbioCAgRT+/PCR+ and PanbioCAgRT-/PCR+ 10 subjects. A P value of ≤0.05 was considered statistically significant. All data was analyzed using 11 None of the samples that tested positive by PanbioCAgRT tested negative by RT-qPCR. Therefore, the 19 overall sensitivity of the PanbioCAgRT was 86.7% (95% CI:72.5-94.5), specificity was 100% (95% CI: 20 98.8-100), positive predictive value was 100% (95% CI: 88.8-100) and negative predictive value was 21 98.5% (95% CI: 96.6-99.4) ( Table I) In the validation phase, we found an overall clinical specificity of 100% (95% CI: 98.8%-100%) and 15 sensitivity of 86.7% (95% CI: 72.5%-94.5%) of the PanbioCAgRT compared to RT-PCR. The sensitivity 16 in our validation study was slightly higher than found in most previous reports, [3, [9] [10] [11] 14] and 17 comparable to other studies. [12, 15] In line with other studies, SARS-CoV-2 RNA load was significantly 18 higher in antigen test+/RT-PCR+ samples than in antigen test-/RT-PCR+ samples. [3,9- In two recent studies, SARS-CoV-2 could not be cultured from antigen test-/RT-qPCR+ samples. [3, 7] study was close to previous published results on virus culture (around 10 6 copies/ml or Ct 25), 1 suggesting that patients with RT-PCR proven COVID-19 but negative PanbioCAgRT were unlikely to be 2 contagious. [3] A recent Dutch study showed that 97.3% of mildly symptomatic infectious individuals 3 (patients with positive SARS-CoV-2 cultures) were detected with PanbioCAgRT. [6] In three other 4 recent reports the sensitivity of the antigen test for detection of SARS-CoV-2 infectivity based on 5 virus culture were 96.4%, 84%, and 86.4% respectively.[4,5,8] Although more studies on this topic 6 are needed, antigen tests may play a key role in rapidly identifying HCW at highest risk for 7 transmitting disease. 8

In a healthcare setting, the best test is not necessarily one that determines whether a person has any 9 evidence of SARS-CoV-2, but may be the one that quickly and accurately identifies individuals 10 capable of transmitting the infection to others. Laboratory-independent antigen tests could be key to 11 detecting contagious HCW to prevent further transmission, but also keeping non-contagious HCW at 12 work during this challenging pandemic. [3] With our antigen-based strategy, highly contagious HCW 13 (with a positive antigen test), could be rapidly removed from the work environment and stopped 14 being a risk for colleagues and patients. Less or non-contagious HCW (with negative antigen test), 15 could return to work wearing PPE pending RT-PCR results. Moreover, if staff shortages continue 16 despite mitigation strategies, it might be considered that (SARS-CoV-2 infected but) antigen test 17 negative HCW, might return to work wearing PPE until symptoms have resolved without performing 18 PCR as secondary confirmation. [1] However, in this case one should consider to repeat antigen 19 testing the following day(s) to ensure the viral load remains low. 20

The rapid antigen-based rapid strategy that we implemented in our hospital was easy to perform and 21 had a short turnaround time of approximately 20 minutes as the PanbioCAgRT was performed on-22

site. Remarkably, the sensitivity of the PanbioCAgRT in the implementation phase (72.5%) was lower 23 than the sensitivity in the validation phase (86.7%). In addition, in the implementation phase, one 24 sample with high viral load (Ct 21.8) was not detected by the antigen test. A possible explanation the antigen test was performed by experienced laboratory technicians and in the implementation 1 phase, the antigen test was performed by trained nurses. Though all antigen tests that were 2 performed on-site were photographed and reassessed at a later time point by our laboratory 3 technicians without any inconsistencies. Another difference between the two study phases was the 4 prolonged inclusion time in the implementation phase of four days or less after symptom onset, 5 compared to three days or less after symptom onset in the validation phase. However, the sensitivity 6 in the implementation phase did not change when we excluded HCW with symptoms for four days 7 from our analysis (data not shown). Another difference between the two study phases was the time 8 from collection of the nasopharyngeal swab and preparation of the extraction tube with the swab 9

and buffer (at the testing site), to the dispense of the fluid into the specimen well on the test device. 10

In the validation phase, this time was at least 30 minutes (maximum two hours), as it required 11 transport of the extraction tubes to the laboratory. In the implementation phase, this time was only 12 a few minutes, as the antigen test was performed at the sampling site. In the validation phase, the 13 prolonged incubation of the swab with the extraction buffer could have had a positive effect on the 14 sensitivity, but this hypothesis needs to be investigated in a technical validation study. This 15 somewhat longer incubation time certainly did not influence the specificity of the test, as all positive 16

PanbioCAgRT were confirmed by RT-PCR as mentioned. Other factors such as the prevalence of 17 than RT-PCR and might therefore be implemented in a wide range of occupational healthcare 1 settings. 2

This research did not receive any specific grant from funding agencies in the public, commercial, or 15 not-for-profit sectors. 

Strategies to Mitigate Healthcare Personnel 19

Staffing Shortages

European Commission recommendation on the use of rapid antigen tests for the 21 diagnosis of SARS-CoV-2 infection

Coronavirus 2 Viral Culture

CoV-2 rapid antigen test: Potential to help reduce community spread?

From more testing to 6 smart testing: data-guided SARS-CoV-2 testing choices

Urgent need of rapid tests 8

for SARS CoV-2 antigen detection: Evaluation of the SD-Biosensor antigen test for SARS-CoV-2

Rapid Antigen Test for SARS-CoV-2 in Municipal Health Service Testing 12

Real-life 14 validation of the Panbio COVID-19 Antigen Rapid Test (Abbott) in community-dwelling 15 subjects with symptoms of potential SARS-CoV-2 infection

Panbio Covid-19 rapid antigen detection test device for the screening of patients with Covid-18

Panbio TM rapid antigen test 20 for SARS-CoV-2 has acceptable accuracy in symptomatic patients in primary health care

Diagnostic accuracy of two commercial SARS-CoV-2 Antigen-detecting rapid tests at the point 24 of care in community-based testing centers

26 the onset of symptoms

Performance evaluation of a SARS-CoV-2 rapid antigentest: test performance in the 4 community in the Netherlands

Multicenter evaluation of the Panbio TM COVID-19 rapid antigen-detection test for the 7 diagnosis of SARS-CoV-2 infection

Predicting Infectious Severe 9

Acute Respiratory Syndrome Coronavirus 2 From Diagnostic Samples

Viral RNA load as determined by 12 cell culture as a management tool for discharge of SARS-CoV-2 patients from infectious 13 disease wards

SARS-CoV-2 informs infectivity and safe de-isolation assessments during COVID-19

Molecular Testing: Correlation of SARS-CoV-2 Culture with Molecular Assays and Cycle 19 Thresholds

Viral cultures for COVID-19 infectious 21 potential assessment -a systematic review

We thank all laboratory and occupational healthcare personnel working at Jeroen Bosch hospital for 5 their commitment to complete this project. We thank Jeroen Schellekens for his help with the RT-6 PCR analysis, Ray van Gaal for his help with collection of the samples for the antigen test, and Melissa 7 van Beuningen for her assistance in the organization of the rapid antigen-based testing strategy in 8 our hospital. 9