Table 2

Characteristics of included studies reporting diagnostic accuracy outcomes for antibody tests

ReferenceStudy designPopulation/samplesTest (supplier) (target); sample site
Cassaniti et al (2020)17 Cohort study. Single centre (Italy). Collection date NR3 cohorts: (1) healthy volunteers with negative RT-PCR for COVID-19; (2) hospitalised patients with positive COVID-19 RT-PCR; (3) patients with suspected COVID-19 at their first access at emergency room. n=110 (30 healthy volunteers; 30 patients with COVID-19; 50 patients with suspected COVID-19). Baseline characteristics reported separately for each cohortVivaDiag COVID-19‐19 IgM/IgG rapid point-of-care lateral flow immunoassay (VivaChek) (target NR); serum or blood samples. Serum samples were obtained at median 7 days (IQR 4–11 days) after positive result for hospitalised patients. Reference/comparator: RT-PCR (RdRp and E primers); respiratory samples
Dohla et al (2020)18 Single centre (Germany). Dates NRPeople within a community setting (high-prevalence area), presenting with COVID-19 symptoms (n=39) and people diagnosed with COVID-19 (n=10). Median age 46 years (IQR 28–72 years). 29/49 female (49.0%)IgG/IgM point-of-care test (NR) (target NR); fingertip prick blood or serum. Reference standard: repeated RT-PCR (Altona Diagnostics) (target NR); throat swabs. Serum samples for previously diagnosed individuals were also analysed using the antibody test
Hoffman et al (2020)19 Validation study. Centre NR. Study dates NRPatients with confirmed COVID-19 or convalescents (n=29). Controls: healthy volunteers without any known history of COVID-19 (n=24); blood donor sera from healthy adults (n=80) and babies (n=20) collected during 2018IgG/IgM Rapid Test Cassette (Zhejiang Orient Gene Biotech Company) (target NR); blood/serum samples. Reference standard: RT-PCR
Jin et al (2020)20 Retrospective study. Single centre (China)People with a laboratory-confirmed SARS-CoV-2 infection in hospital, and at least one viral serological test (n=43). Median age 47.0 years (IQR 34.0–59.0 years). 39.5% male. Control group: patients with suspected SARS-CoV-2 infection who were excluded and quarantined at home (n=33). Median age 31.0 years (IQR 25.5–37.5 years). 66.7% male. Suspected infected patients were discharged from hospital when they received two negative PCRs, performed in a 24-hour intervalIgM and IgG chemiluminescence assay (CLIA)(Shenzhen YHLO Biotech) (targets N protein and spike protein). Reference standard: confirmed diagnosis from RT-PCR (target not specified); sampling not clearly reported but includes oral swabs, anal swabs and sputum. Duration between first symptoms and serological test (CLIA) was 18 days (IQR 11–23 days) in the COVID-19 group, 3.0 days (2.0–8.0 days)
Li et al (2020)21 Prospective development study. Single centre (China). 12 February 2020 to 20 February 2020People with suspected (RT-PCR negative) or confirmed (RT-PCR positive) COVID-19. n=278 (89 confirmed; 189 probable). n=273 controls were included. Baseline characteristics NRIgM and IgG colloidal gold assay (NR) (targets serum antibodies against N protein); serum specimens. RT-PCR assumed to be the reference standard (described as a ‘control’ by the authors); primer/target and sampling methods not known
Li et al (2020)22 Prospective development study. 8 centres (China). Dates NRPeople with suspected COVID-19. n=525 specimens (397 clinical positive; 128 clinical negative). Characteristics NRIgM/IgG rapid point-of-care lateral flow immunoassay (Jiangsu Medomics Medical Technologies) (targets antibodies against spike protein); blood (including serum and plasma). Reference standard: RT-PCR; respiratory specimens
Liu et al (2020)10 Prospective study. Single centre (China). 18 January to 26 February 2020Hospitalised patients diagnosed with COVID-19. All patients were laboratory confirmed by RT-PCR. n=314 (214 patients; 100 healthy controls). Baseline characteristics NRIgM ELISA IgG ELISA (NR) (targets antibodies against N and spike); serum. Median time of sample collection was 15 days (range 0 to 55)
Shen et al (2020)23 Prospective cohort study. Single centre (China). 20 January and 2 February 2020People with suspected COVID-19. Suspected COVID-19 was defined as a pneumonia that had related epidemiological history and fulfilled two of the following: fever and/or respiratory symptoms; imaging indicative of pneumonia; low/normal white cell count or low lymphocyte count. n=150. Median age: PCR-positive group 38 years (IQR 46–56 years); PCR-negative group 32 years (IQR 20–42.5). Sex: PCR-positive group 60.8% male; PCR-negative group 56.6% male. Control: healthy donors (n=26)IgM/IgG colloidal gold immunochromatography antibody kit (Shanghai Outdo Biotech Company) (target NR); blood. Reference standard: RT-PCR (target NR); nasopharyngeal and oropharyngeal swabs. At least two different samples were obtained from each patient for RT-PCR. If the result was inconclusive, repeated sample collection was required. A patient with at least one positive RT-PCR was confirmed as positive. Patients with two consecutive negative results were defined as PCR negative, but would only be diagnosed as non-COVID-19 if the symptoms could be explained by another condition or infection resolved following the corresponding treatments. Any other PCR-negative result was treated as inconclusive
Spicuzza et al (2020)24 People with confirmed COVID-19 (n=23) or suspected COVID-19 (n=7). Confirmed COVID-19 was defined as consistent radiological/clinical findings, with positive RT-PCR. Suspected COVID-19 was defined as suggestive radiological/clinical findings but negative RT-PCR. Control: asymptomatic controls with negative RT-PCR (n=7). n=37. Mean age: confirmed COVID-19 57±17 years; suspected COVID-19 67±15 years. Controls age NR. Sex characteristics NRIgG/IgM POC Antibody Rapid Test Kit (Beijing Diagreat Biotechnologies Company) (spike); blood/serum/plasma. Reference standard: RT-PCR (NR) (target NR); nasopharyngeal swab or bronchial aspirate
Xiang et al (2020)11 Single centre (China). 19 January to 2 March 2020People with suspected (n=24) or confirmed (n=85) COVID-19. Diagnosis of laboratory-confirmed COVID-19 was defined as positive nucleic acid tests for SARS-CoV-2 by RT-PCR. Diagnosis of suspected COVID-19 was based on negative RT-PCR, but satisfying 1 of the epidemiological history criteria and 2 of the clinical criteria. Control: samples from healthy blood donors or from hospitalised patients with other diseases (n=60)IgM/IgG ELISA (Livzon) (NR); serum. Serum samples were obtained at different time periods after symptom onset. Reference standard: RT-PCR (ORF1ab and N); nasopharyngeal and/or oropharyngeal swabs
Xu et al (2020)33 Retrospective study. Single centre (China). 20 January 2020 to 17 February 2020Patients with suspected COVID-19. n=284 participants: 186 COVID-19 patients with RT-PCR–positive result; 19 COVID-19 cases diagnosed by clinical symptoms; 79 controls with other diseases (negative RT-PCR). Baseline characteristics NRIgM and IgG fully automated assay (NR) (target NR); serum samples. Comparator: RT-PCR. Reference standard: Diagnosis through positive RT-PCR or clinical symptoms
Zhao et al (2020)12 Retrospective study. Single centre (China). 11 January 2020 to 9 February 2020People with COVID-19. All enrolled cases were confirmed to be infected with SARS-CoV-2 by RT-PCR. n=173 patients (535 samples). Median age 48 years (IQR 35 to 61). 51.4% femaleIndex tests: IgM ELISA (Beijing Wantai Biological Pharmacy Enterprise) (spike protein). IgG ELISA (N). Total antibody (Ab) ELISA (spike protein); plasma samples. Comparator: RT-PCR result. Reference standard: confirmed COVID-19 through positive RT-PCR
  • NR, details not reported.