Gudbjartsson 2020 [A].

Study characteristics
Patient Sampling	Purpose: Diagnosis of current acute‐phase infection, current convalescent‐phase infection, and prior infection Design: Multi‐group study to estimate sensitivity and specificity [1] Confirmed COVID cases (1237 patients; 2102 samples; possible overlap of patients between [1a] and [1b]) [1a] Hospitalised (48 patients; 249 samples) [1b] Recovered (1215 patients; 1853 samples) [2] PCR‐ or not tested [2a] Pre‐pandemic: 2017 (n = 472) [2b] Early 2020 (n = 470) [2c] Health Care (n = 18,609) [2d] Reykjavik (n = 4843) [2e] Vestmannaeyjar (n = 663) [2f] Quarantine (n = 4222) Only groups [1b] and [2a] were eligible for our review. Recruitment: [1] From February 28 to May 1, 1797 patients were found to be SARS‐CoV‐2 positive by qPCR. We collected samples from a group of hospitalised qPCR‐positive persons and invited all qPCR‐positive persons who had recovered from infection to donate samples, both shortly after recovery and again approximately 3 months after recovery (a total of 2102 samples from 1237 persons). [1a] 48 out of 101 (48%) hospitalised Icelandic COVID‐19 patients during their hospital admission The most common reason for missing samples was that the patient had been discharged before commencement of the study, followed by the patient not consenting to participate in the study. [1b] We invited all qPCR‐positive persons to give a blood sample after recovery (defined as at least two weeks from qPCR diagnosis and one week after end of symptoms) and again on July 1, on average 100 days after diagnosis with qPCR. Non‐participation was because of refusal or inability to participate because of health or geographic constraints. [2a] Persons participating in the deCODE health study in the year 2017 [2b] Persons participating in the deCODE health study from February 18 through March 9 2020 [2c], [2d], [2e] Persons who had neither tested qPCR‐positive nor been quarantined to evaluate seroprevalence outside quarantine and the spread of the virus in Iceland (the Health Care, Reykjavik, and Vestmannaeyjar sample groups) [2f] Samples from quarantined persons who had not tested qPCR‐positive Prospective or retrospective: [1a], [1b], [2c], [2d], [2e], [2f] Prospective [2a], [2b] Retrospective Sample size: 30,576 (1237) people of which 2325 (1853) samples from 1687 (1215) patients were eligible for our review. Further detail: Inclusion criteria: either that the person had not been tested positive with qPCR (Neg/NA) (group [2]) or that the person had been positive with qPCR assay (positive) (group 1])
Patient characteristics and setting	Setting: [1b] Convalescent/community Location: Former inpatients or outpatients of Landspitali ‐ The National University Hospital of Iceland (LUH), Reykjavik. Country: Iceland Dates: [1b] 3 April to 8 July 2020 Symptoms and severity: [1b] Not stated (1215 of 1797 PCR+ COVID patients included in [1b]: Of the 1797 confirmed COVID patients, 1746 (97.2%) were treated as outpatients while the remaining 51 (2.8%) patients were admitted to hospital at the time of diagnosis. Now all recovered with at least 1 week without symptoms) Demographics: [1b] 48% male Age: Mean 43 (SD 16) years Exposure history: Not stated Non‐Covid group 1: [2a] Pre‐pandemic Source: Persons participating in the deCODE health study in the year 2017 (2 January to 4 December 2017) Characteristics: 41% male Age: mean 57 (SD 16) years
Index tests	Test name: Name not stated [A] Roche Elecsys chemiluminescence assay [B] Wantai ELISA [C] EDI ELISA [D] EDI ELISA [E] Euroimmun ELISA [F] Euroimmun ELISA Manufacturer: [A] Roche International, Basel, Switzerland [B] Wantai/Nordic BioSite, Täby, Sweden [C] EDI/Eagle Biosciences, Amherst, NH, United States [D] EDI/Eagle Biosciences, Amherst, NH, United States [E] Euroimmun AG, Luebeck, Germany [F] Euroimmun AG, Luebeck, Germany Antibody: [A] Total antibodies [B] Total antibodies [C] IgG [D] IgM [E] IgG [F] IgA Antigen target: [A] Nucleocapsid (anti‐N) [B] Spike 1 RBD (anti‐S1‐RBD) [C] Nucleocapsid (anti‐N) [D] Nucleocapsid (anti‐N) [E] Spike subunit 1 (anti‐S1) [F] Spike subunit 1 (anti‐S1) Evaluation setting: [A]‐[F] Lab tests performed in lab Test method: [A] ECLIA [B]‐[F] ELISA Timing of samples: [1b] at least two weeks from qPCR diagnosis and one week after end of symptoms; (text and Fig 2 stated "25 days after diagnosis" for the earliest time point) and again on July 1, on average 100 days after diagnosis with qPCR (487/1215 recovered patients with at least 2 samples at least 30 days apart); up to 4 months after PCR+ Samples used: Serum samples were frozen in aliquots at ‐80°C. Test operator: Not stated Definition of test positivity: All measurements were done according to manufacturer ‘s instructions. The ELISA results are expressed as optical density (OD) and the ECLIA results as log light emission. [C] and [D] For the IgG and IgM anti‐N assays, we ran four negative controls per 96 well plate and subtracted the mean OD of the negative controls from the OD. After subtraction of the negative controls, the manufacturer recommended OD thresholds for positive results were 0.198 for the IgG anti‐N assay [C] and 0.11 for the IgM anti‐N assay [D]. [A] and [B] The manufacturer recommended OD thresholds for positive results were 1 (0 for log(OD)) for the pan‐Ig anti‐N assay [A] and 0.19 for the pan‐Ig anti‐S1‐RBD assay [B]. [E] and [F] For the IgG and IgA anti‐S1 assays, the manufacturer recommended using two negative controls and two calibrator samples per plate and declaring samples positive if they have greater OD than the difference of the mean OD for the calibrator samples minus the mean OD for the negative control samples. The mean threshold was 0.33 for the IgG anti‐S1 assay [E] and 0.36 for the IgA anti‐S1 assay [F]. Blinding reported: Not stated Threshold predefined: yes, thresholds for positivity were supplied by the assay manufacturers.
Target condition and reference standard(s)	Reference standard: Testing for SARS‐CoV‐2 was performed either at Landspitali – The National University Hospital of Iceland (LUH) or deCODE using similar qPCR methods. LUH: WHO recommended screening method: single probe pan‐screening assay for betacoronaviruses, followed by confirmatory measurements for all positive samples using an nCoV‐2019 specific assay The broad betacoronavirus assay is based on probes for a conserved region of the E‐gene, whereas confirmatory testing assays were done using either nCoV2019 specific probes for the RdRp gene or the TaqMan™ Fast Virus 1‐step Master Mix, 2019‐nCoV Assay kits v1 from Thermo Fisher. Samples in the E‐gene screening assay with Ct < 35 were considered strong positive and went for confirmatory testing using RdRp, whereas samples with Ct values between 35‐37 were considered weak positive and were confirmed using the TaqMan™ Fast Virus method. Samples with Ct values from 37‐40 were classified as inconclusive and were tested again to confirm their status. deCODE: SARS‐CoV‐2 screening was performed using qPCR assays in either a singleplex (Method 1) or a multiplex (method 2) format, respectively. Method 1 uses the three probe TaqMan™ Fast Virus 1‐step Master Mix, 2019‐nCoV Assay kits v1 and 2019‐nCov control kit from Thermo Fisher. Method 2 uses the TaqPath™ COVID‐19 CE‐IVD RT‐PCR kit from Thermo Fisher. Results criteria for methods 1 and 2: Samples with FAM™ dye Ct 7 values < 37 in at least two of three assays were classified as positive. Samples with FAM™ dye Ct values between 37 and 40 were classified as inconclusive and their testing repeated. If repeated testing gave the same result with at least two probes the sample was classified as positive. If repeated testing gave positive results for only one probe the test was considered inconclusive and a new sample from the subject was requested. Samples with undetected FAM™ dye Ct values or values equal to 40 in all three assays were classified as negative if the human RNaseP assay was positive (VIC™ dye Ct < 40). Samples used: Not stated Timing of reference standard: Not stated Blinded to index test: yes, prior index test Incorporated index test: no Definition of non‐COVID cases: [2a] Pre‐pandemic Samples used: [2a] Pre‐pandemic Timing of reference standard: [2a] Pre‐pandemic Blinded to index test: [2a] Pre‐pandemic, prior index test Incorporated index test: no
Flow and timing	Time interval between index and reference tests: [1b] at least two weeks from qPCR diagnosis [2] Not stated All patients received same reference standard: No Missing data: yes (see Table S3: only 1134/1215 and 437/472 samples tested with test [C]; only 1145/1215 and 434/472 samples tested with test [D]), results for tests [E] and [F] not reported, groups [1a], [2b], [2c], [2d]. [2e] and [2f] excluded from review Uninterpretable results: Not stated Indeterminate results: No intermediate results as per manufacturer's instructions Unit of analysis: [1b] Samples, but for persons with multiple samples, only the results for the most recently obtained sample were used. [2a] Patients
Comparative
Notes	Funding: Not stated Publication status: Published paper Source: New England Journal of Medicine Author COI: Not stated
Methodological quality
Item	Authors' judgement	Risk of bias	Applicability concerns
DOMAIN 1: Patient Selection
Was a consecutive or random sample of patients enrolled?	Unclear
Was a case‐control design avoided?	No
Did the study avoid inappropriate exclusions?	Unclear
Did the study avoid inappropriate inclusions?	Unclear
Could the selection of patients have introduced bias?		High risk
Are there concerns that the included patients and setting do not match the review question?			High
DOMAIN 2: Index Test (All tests)
DOMAIN 2: Index Test (Antibody tests)
Were the index test results interpreted without knowledge of the results of the reference standard?	Unclear
If a threshold was used, was it pre‐specified?	Yes
Could the conduct or interpretation of the index test have introduced bias?		Unclear risk
Are there concerns that the index test, its conduct, or interpretation differ from the review question?			Unclear
DOMAIN 3: Reference Standard
Is the reference standards likely to correctly classify the target condition?	Yes
Were the reference standard results interpreted without knowledge of the results of the index tests?	Yes
The reference standard does not incorporate the index test	Yes
Could the reference standard, its conduct, or its interpretation have introduced bias?		Low risk
Are there concerns that the target condition as defined by the reference standard does not match the question?			High
DOMAIN 4: Flow and Timing
Was there an appropriate interval between index test and reference standard?	Unclear
Did all patients receive the same reference standard?	No
Were all patients included in the analysis?	No
Did all participants receive a reference standard?	Yes
Were results presented per patient?	Yes
Could the patient flow have introduced bias?		High risk