High-Frequency Self-Testing by Schoolteachers for Sars-Cov-2 Using a Rapid Antigen Test

Real-time polymerase chain reaction (RT-PCR) from a nasopharyngeal swab taken by a healthcare professional is the gold standard for demonstrating active SARS-CoV- 2 infection. Rapid antigen tests are also available, and provide point-of-care testing within minutes. In terms of sensitivity and specificity they are inferior to PCR. Nevertheless, their sensitivity may be sufficient to detect most symptomatic cases of SARS-CoV-2 (1), especially if testing is repeated at short intervals (high-frequency testing) (2). The test is usually simple to perform and can be done by an untrained person on a self-collected anterior nasal swab (3, 4).

Method

In the Safe School Hessen study, funded and commissioned by the Hesse Ministry of Education and the Hesse Ministry of Integration and Social Affairs, schoolteachers in two rural districts and one city in Hesse were invited to self-test with a SARS-CoV- 2 rapid antigen test (R-Biopharm, Darmstadt). The test was done on a self-collected anterior nasal swab and was carried out, analyzed, and recorded by the teachers, at home and unsupervised, every 48 h over the 7-week study period. Both video and written instructions were provided on how to carry out the test, and a study hotline was made available for queries. Samples on which the teachers reported positive or unclear results were investigated using RT-PCR at the Institute of Medical Virology, Frankfurt University Hospital. Study participants were asked whether they had had a SARS-CoV-2 infection detected by any other test outside the present study during the study period, and whether they had encountered any difficulties in carrying out the test. The study protocol was approved by the ethics committee of Frankfurt University Hospital.

Results

A total of 711 teachers from 86 schools took part in the study, 635 of whom (89.3%) submitted records, thus enabling evaluation of 11 385 rapid antigen tests. Twenty-one positive antigen tests were checked using RT-PCR. In 5 cases, SARS-CoV-2 infection was confirmed (true positive result) (table). High viral loads were present in the cycle threshold (Ct) range between 17.6 and 20.2, which is associated with in vitro infectivity (1). One teacher had a presymptomatic SARS-CoV-2 infection at the time of testing, and four teachers already had symptoms, many of which were nonspecific (table). No asymptomatic infections were detected. In 16 cases, SARS-CoV-2 infection was ruled out by RT-PCR (false positive antigen test). Frequently, the very first antigen test carried out was positive (10 out of 16 false positive antigen tests, 62.5%). This could indicate the presence of a persistent host factor in the nasal mucosa of study participants. Protein A from Staphylococcus aureus or from other coagulase-positive Staphylococcus spp. can bind the Fc region of IgG, thus leading to false reactive findings in a lateral flow system (5). Seven false reactive samples were tested for Staphylococcus aureus by PCR (target: nuc and MREJ) and culture, and S. aureus was detected in 6 of 7 samples (85.7%). In four cases, PCR-confirmed SARS-CoV-2 infection was reported from a source external to the study during the study period but according to the participants was not detected by their antigen testing (presumed false negative antigen tests). An invalid test result (control line did not show) was recorded in 49 of 11 385 tests (0.43%). As to technical difficulties in carrying out the test, 92.0% of participants reported never having any, 7.0% rarely, 1.0% occasionally, and 0.0% frequently.

Table. Characterization of true positive, false positive, and false negative antigen test results from the SAFE School Hessen study.

No.	SW	Ag test result*1	RT-PCR test result*2	Ct value	Ag test classification	7-day incidence *4	Symptom classification at the time of Ag test	SymptomsbeforeAg test
1	1	+	Negative (1)	n. a.	False positive	17.71	Asymptomatic	n. a.
2	1	+	Negative (2)	n. a.	False positive	17.71	Symptomatic – sore throat	0 days
3	1	++	Negative (3)	n. a.	False positive	17.71	Asymptomatic	n. a.
4	1	++	Negative (3)	n. a.	False positive	14.58	Asymptomatic	n. a.
5	1	+	Negative (1)	n. a.	False positive	17.71	Asymptomatic	n. a.
6	2	+	Negative (3)	n. a.	False positive	11.75	Asymptomatic	n. a.
7	2	+++	Negative (3)	n. a.	False positive	24.45	Asymptomatic	n. a.
8	2	++	Negative (3)	n. a.	False positive	24.45	Asymptomatic	n. a.
9	2	+++	Negative (1)	n. a.	False positive	31.47	Asymptomatic	n. a.
10	3	+++	Negative (3)	n. a.	False positive	28.10	Asymptomatic	n. a.
11	3	+	Negative (3)	n. a.	False positive	39.91	Symptomatic – rhinitis	4 days
12	3	+	Negative (2)	n. a.	False positive	28.10	Asymptomatic	n. a.
13	3	++	Negative (3)	n. a.	False positive	39.91	Asymptomatic	n. a.
14	4	+++	Positive (3)	18.9*3	True positive	108.76	Symptomatic – fever, cough, headache, aching limbs	3 days
15	5	+++	Positive (2)	20	True positive	267.12	Symptomatic – headache, fatigue	2 days
16	5	+	Negative (2)	n. a.	False positive	211.35	Symptomatic – sore throat	0 days
17	6	++	Negative (3)	n. a.	False positive	213.60	Symptomatic – sore throat	3 days
18	6	Negative	Positive (4)	n.r.	False negative	213.60	Presymptomatic	0 days
19	6	+++	Positive (3)	17.6*3	True positive	275.56	Symptomatic – fatigue, aching limbs	0 days
20	6	+	Positive (3)	19.7*3	True positive	139.45	Symptomatic – cough, nasal congestion, loss of smell and taste, headache	2 days
21	7	++	Positive (3)	20.2*3	True positive	252.66	Presymptomatic	n. a.
22	7	Negative	Positive (4)	n.r.	False negative	252.66	Symptomatic – nasal congestion	2 days
23	7	+	Negative (3)	n. a.	False positive	171.51	Asymptomatic	n. a.
24	7	Negative	Positive (4)	n.r.	False negative	171.51	Symptomatic – nausea	2 days
25	n. a.	Negative	Positive (4)	n.r.	False negative	n. a.	n. a.	n. a.

Ag, antigen; Ct, cycle threshold; n. a., not applicable; RT-PCR, real-time polymerase chain reaction; SW, study week; n.r., not reported(PCR carried out elsewhere)

*¹ Antigen tests carried out and results interpreted by study participants: + positive: test line fainter than control line; ++ positive: test line and control line equally clear; +++ positive: test line clearer than control line

*² RT-PCR tests used: (1) Cepheid Xpert Xpress SARS-CoV-2; (2) Abbott Alinity M SARS-CoV-2; (3) Roche Cobas SARS-CoV-2 on the Cobas 6800; (4) PCR carried out elsewhere

*³ PCR target: ORF region

*⁴ Local incidence

Discussion

Self-testing by schoolteachers allowed early detection of cases of SARS-CoV- 2 infection, potentially preventing transmission in the school or home environment. Of all the positive antigen tests, 76.2% were false positives. This means that positive antigen tests need to be reviewed early to minimize potential confusion and unnecessary isolation measures. An antigen test with optimized specificity combined with high sensitivity should be used. In the study, testing was particularly effective in the presence of symptoms that can be nonspecific and prevalent in the general population, and when local incidence was high. Because false negative results can also occur, the use of antigen testing should not justify relaxation of locally existing precautions. We suspect that low viral load at the time of testing led to the false negative antigen test results, but errors in carrying out the test are also possible. Users should be aware of the possibility of false positive and false negative results.

The results of the study are also available as a pre-print: www.medrxiv.org/content/10.1101/2020.11.02.20223859v1