Table 3

Claims of test accuracy from websites (selected verbatim) and evidence identified from the manufacturers’ instructions for use sheet (IFU), published papers and pre-prints

Website information			Published information
What do they say about accuracy?	Sensitivity (%)	Specificity (%)	Study details	Sensitivity (%)	Specificity (%)
Abbott’s antibody test (Abbott Architect SARS-CoV-2 IgG)
London Medical Laboratory 33 Sensitivity: This test has proven to be 100% accurate in identifying antibodies to SARS-CoV-2 coronavirus at 14 days after onset of COVID-19 symptoms. Specificity: It is 99.63% specific. Or put another way, only 0.37% of over 1000 people tested who could not have been exposed to SARS-CoV-2 showed a false positive result.	100 (n=88)	99.63 (n=1070)	Manufacturer’s clinical performance of test 41 Positive samples tested: 122 serum and plasma specimens were collected at different times from 31 subjects who tested positive for SARS-CoV-2 by a polymerase chain reaction (PCR) method and who also presented with COVID-19 symptoms. Negative samples tested: 1070 specimens, 997 specimens were collected prior to September 2019 (pre-COVID-19 outbreak). An additional 73 specimens were collected in 2020 from subjects who were exhibiting signs of respiratory illness but tested negative for SARS-CoV-2 by a PCR method (unclear how many participants) Reference standard: PCR method	100 (n=88)	99.63 (n=1070)
MyHealthcare Clinic 29 The manufacturer of the antibody test (Abbott Laboratories) reports that an independent clinical performance evaluation of the test performed in the UK confirmed the following accuracy levels: Sensitivity (ability to identify positive cases) of 99.7%. Specificity (ability to identify negative cases) of 100%.	99.7 (n=NR)	100 (n=NR)	Bryan et al 2020 55 Positive samples tested: 125 patients who tested RT-PCR positive for SARS-CoV-2 for which 689 excess serum specimens were available (unclear how many at each time point). Negative samples tested: 1020 serum specimens collected prior to SARS-CoV-2 circulation in the USA Reference standard: PCR method and pre-COVID-19 samples	53.1 (at 7 days) (n=NR) 82.4 (at 10 days) (n=NR) 96.9 (at 14 days) (n=NR) 100.0 (at 17 days—data driven choice) (n=NR)	99.9 (n=1020)
The Online Clinic (Online Clinic (UK) Limited) 26 Tests have a sensitivity of 100%. When using a patient-collected sample with one of our home sampling kits, the sensitivity of this test has been shown to reduce slightly to 97.5. Recent studies suggest a specificity of 99.9% and 99.8% respectively.	97.5–100 (n=NR)	99.8–99.9 (n=NR)	Phipps et al 2020 56 Positive samples tested: Only six patients with samples 14 days post-symptom onset, the point at which the highest sensitivity was recorded. 173 suspected COVID-19 cases with 76 were confirmed positive by PCR methods. Negative samples tested: Plasma samples from healthy donors (2019 blood donations and 2020 blood donations from healthy donors without recent illness) Reference standard: PCR method for suspected COVID-19 cases to confirm positives; for negatives apparent healthy donors	38 (all days) (n=76) 7 (<3 days) (n=15) 30 (3–7 days) (n=27) 33 (5–15 days) (n=15) 83 (after 14 days) (n=6)	100 (n=656)
Atruchecks Limited 31 Abbott claim their test is 100% sensitive (88 samples) and 99.6% specific (1070 samples). Our UK Accredited Partner Lab have run their own internal verification of these claims and achieved a sensitivity of 98.5% (132 samples) and a specificity of 99.5%. (186 samples).	98.5 (n=132) 100 (n=88)	99.5 (n=186) 99.6 (n=1070)	Public Health England (PHE) evaluation of the Abbott antibody test 57 Positive samples tested: 96 samples defined by a positive PCR from a swab sample for that patient Negative samples tested: 759 negative samples were included in the evaluation (351 samples that are rheumatoid factor, CMV, EBV, or VZV positive; 11 seasonal coronavirus positive samples; and 395 historic samples, two samples unclear). These samples have been chosen based on their collection before mid-2019 to ensure they are SARS-CoV-2 antibody negative, but will contain samples containing antibodies to other seasonal coronaviruses to provide an additional screen for the assay Reference standard: PCR method and pre-COVID-19 samples	92.7 (all days) (n=96) 93.4 (≥14 days) (n=82) 93.9 (≥21 days) (n=76)	100 (n=759)
Epitope Diagnostics inc. (EDI) Antibody test (EDI Novel Coronavirus COVID-19 IgG ELISA Kit)
Summerfield Healthcare 16 We have a trusted product which is specific to COVID-19 and sensitive. As with all of these kits they undergo regular testing to ensure accuracy and reliability which on the last occasion were 100% accurate for both positive and negative samples. Antibody test is very specific for COVID-19 (some inferior tests can mistake other infections for COVID-19 and wrongly reassure you). It is also very sensitive for the specific IgG antibody.	NR (n=NR)	NR (n=NR)	Manufacturer’s clinical performance of test 58 Positive samples tested: RT-PCR confirmed positive patients. Negative samples tested: Normal healthy patients with samples collected prior to the COVID-19 outbreak. Reference standard: PCR method and pre-COVID-19 samples	98.4 (n=187)	99.8 (n=624)
	NR (n=NR)	NR (n=NR)	Krüttgen et al 2020 59 Positive samples tested: The sera of the 31 patients with positive SARS-CoV-2 PCR were collected 11.9 days (±5.0 days) post-onset of symptoms. 22 sera were considered SARS-CoV-2 IgG positive (positive on at least two assays). Negative samples tested: 53 sera were regarded as IgG negative Reference standard: A serum was regarded as SARS-CoV-2 IgG negative if at least three of the four assays compared here (for the Euroimmun assay, the EDI assay, the Mikrogen assay, and the Viramed assay) had a negative test result applying the manufacturer’s interpretation criteria. A serum was regarded as SARS-CoV-2 IgG positive if at least two of the four assays had a positive test result (comparator tests are also part of reference standard).	100 (n=22)	88.7 (n=53)
Randox PCR antigen test (Randox COVID-19 Home Testing Kit)
Randox 23 This is a PCR-based test, utilising Randox Biochip Technology, to provide an accurate diagnosis for COVID-19.	NR (n=NR)	NR (n=NR)	Manufacturer’s clinical performance of test 42 Positive samples tested: NR Negative samples tested: NR Reference standard: NR	NR (n=NR)	NR (n=NR)
	NR (n=NR)	NR (n=NR)	Public Health England (PHE) evaluation of the Randox antigen test 60 Positive samples tested: None Negative samples tested: The assessment sample-panel totalled 195 specimens, including upper or lower respiratory clinical specimens negative for SARS-CoV-2 as determined by the validated in-house PHE PCR assay and dilutions of SARS-CoV-2 Reference standard: PHE PCR assay	NR (n=0)	100 (n=195)
Primerdesign Ltd antigen test (Coronavirus (COVID-19) genesig Real-Time PCR assay)
Rightangled Healthcare 22 Studies confirm Primerdesign COVID-19 assays are highly specific for the detection of SARS-CoV-2 virus and detection of coronavirus COVID-19 disease. Independent Clinical Performance Evaluation of Primerdesign COVID-19 assay by the National Infection Service, Public Health England, Colindale confirmed the specificity of this assay using upper or lower respiratory clinical samples from patients and known SARS-CoV-2 positive material. PHE confirmed the assay showed >98% specificity to SARS-CoV-2 virus in clinical samples. An Independent Clinical Performance Evaluation by an NHS Clinical Pathology Laboratory using patient samples with respiratory symptoms confirmed the assay was 100% specific when tested against known positive and negative SARS-CoV-2 clinical samples.	98 (n=50)	100 (n=50)	Manufacturer’s clinical performance of test 43 Positive samples tested: Contrived oropharyngeal swabs (50 positive) Negative samples tested: Contrived oropharyngeal swabs (50 negative) Reference standard: 50 swabs were contrived with SARS-CoV-2 whole viral genomic RNA	94.7 (1–2 x LoD) (n=38) 100 (3x LoD) (n=7) 100 (4–5 x LoD) (n=5)	100 (n=50)
			van Kasteren et al 2020 61 Positive samples tested: Clinical samples previously submitted for routine SARS-CoV-2 diagnostics for which the presence of various amounts of SARS-CoV-2 RNA had been confirmed using in-house PCR. Negative samples tested: Clinical samples with confirmed respiratory viruses (influenza virus type A (n=2), rhinovirus (n=2), RSV-A and -B) Reference standard: SARS-CoV-2 RNA had been confirmed using in-house PCR	62.5 (n=16)	100 (n=6)
			Public Health England (PHE) evaluation of Primerdesign antigen test 47 Positive samples tested: None Negative samples tested: : The assessment sample-panel totalled 195 specimens, including upper or lower respiratory clinical specimens negative for SARS-CoV-2 as determined by the validated in-house PHE PCR assay, and dilutions of SARS-CoV-2 Reference standard: PHE PCR assay	NR (n=0)	100 (n=195)
LabCorp antigen test (COVID-19 RT-PCR test with Pixel by LabCorp COVID-19 test home collection kit)
LabCorp 35 Customers provided with link to FDA Emergency Use Authorisation Summary.	100 (NP swabs) (n=40) 100 (BALs) (n=40)	100 (NP swabs) (n=50) 100 (BALs) (n=50)	Manufacturer’s clinical performance of test 45 Positive samples tested: Positive samples were comprised of 40 NP swabs and 40 BALs spiked with quantitated live SARS-CoV-2. 10 samples each were spiked at 8 x, 4 x, 2 x, and 1x LoD. Negative samples tested: Negative samples include 50 NP swabs and 50 BALs. Reference standard: Contrived samples	100 (NP swabs) (n=40) 100 (BALs) (n=40)	100 (NP swabs) (n=50) 100 (BALs) (n=50)
Rutgers Clinical Genomics antigen test (Rutgers Clinical Genomics Laboratory TaqPath SARS-CoV-2 Assay)
Hims 36 Customers provided with link to FDA Emergency Use Authorisation Summary.	100 (n=30)	100 (n=30)	Manufacturer’s clinical performance of test 44 Positive samples tested: 30 contrived positive samples were tested Negative samples tested: 30 contrived negative samples were tested Reference standard: Contrived samples	100 (n=30)	100 (n=30)