Table 1

Characteristics of included studies reporting diagnostic accuracy outcomes for virus tests

ReferenceStudy designPopulation/samplesTest (supplier) (target); sample site
Kim et al (2020)35 Systematic review and meta-analysis. 68 studies were included: 19 studies (n=1502) reported on RT-PCR. Initial search in Medline and Embase from 1 December 2019 to 16 March 2020. The search was updated to 3 April 2020Studies on COVID-19 that reported the diagnostic sensitivity and/or specificity of chest CT scans and/or RT-PCR assaysIndex tests: initial RT-PCR test (target varied among studies); nasopharyngeal swab, throat swab or sputum. Reference standard: repeated RT-PCR tests. RT-PCR results were extracted within 14 days of symptom onset
Baek et al (2020)36 Design/validation study (samples used collected retrospectively). Korea, number of centres not clearPatients with COVID-19. n=14. No demographic details reportedRT-LAMP assay (developed in-house); (N primer); nasal swabs
Fang et al (2020)68 Retrospective case series. Single centre (China). 19 January 2020 to 4 February 2020People with eventual confirmed diagnosis of COVID-19 infection who had an RT-PCR test and CT scan within 3 days or less. Eventual confirmed diagnosis is defined as through repeated RT-PCR testing of negative patients, until a positive test is received. n=51. Median age 45 years (IQR 39 to 55 years). 29 men:22 womenIndex test: initial RT-PCR (Shanghai ZJ Bio-Tech) (primer not specified); throat or sputum samples. Reference standard: eventual confirmed diagnosis through RT-PCR
Fang et al (2020)43 Retrospective case series. Single centre (China). January 2020 to February 2020 (specific dates not specified)People with COVID-19. n=32 (8 ICU patients; 24 non-ICU patients). Age range 35 to 54 years. Sex not reportedRT-PCR (NR) (primer not specified); from nasal swabs, blood, faecal, urine, saliva and tear samples
Harrington et al (2020)37 Prospective case series. Five centres (USA)Symptomatic patients meeting current criteria for diagnosis of COVID-19. n=524. Demographic details not reportedIsothermal amplification (ID NOW COVID-19 assay (Abbott)) (RdRp); nasal swabs. Reference standard was RT-PCR (Abbott RealTime SARS-CoV-2 assay performed on the Abbott m2000 system)
He et al (2020)69 Retrospective case series. Single centre (China). 10 January 2020 to 28 February 2020Hospitalised patients with suspected COVID-19 who underwent high-resolution chest CT and real-time RT-PCR. n=82. Median age 52 years (range 8 to 74 years). 49 malesRT-PCR (BGI Genomics); (primer NR); nasopharyngeal swab, oropharyngeal swab, endotracheal aspirate or bronchoalveolar lavage. Reference standard: eventual confirmed diagnosis through RT-PCR
Lee et al (2020)70 Retrospective case series. Single centre (Singapore). Up to 29 February 2020 (start date not reported)Patients admitted to hospital with suspected COVID-19 infection. n=70. Demographics not reportedRT-PCR (in-house or A*STAR Fortitude Kit (Accelerate Technologies)); (N, Orf1ab); nasopharyngeal swabs
Lu et al (2020)38 Design/validation study. China. Dates of sampling not reportedPatients with suspected COVID-19 admitted to hospital and quarantined. n=56. Demographics not reportedIndex test: RT-LAMP (in-house assay) (N); throat swabs. Reference standard: RT-PCR (LifeRiver Bio)
Shen et al (2020)62 Retrospective case series. Single centre (China).
22 January to 18 February 2020
Subjects judged at high risk of SARS-CoV-2 infection. n=5630. Median age 51 years (IQR 36–63). Male 2631 (46.7%)RT-PCR (Shanghai Huirui Biotechnology); (Orf1ab, N); throat swabs
Yan et al (2020)39 Development/validation study. Centre NR. Dates NRPatients with pneumonia and suspected SARS-CoV-2 infection. n=130 specimens. Characteristics NRRT-LAMP (Loopamp RNA amplification kit; Loopamp Real-time Turbidimeter, both Eiken Chemical, Tokyo, Japan, used to perform and monitor the RT-LAMP reaction) (Orf1ab and spike). Reference standard: RT-PCR. Sampling from swabs (not specified) and bronchoalveolar lavage fluid
Zhen et al (2020)40 Prospective comparative evaluation. Single centre (USA)Symptomatic patients suspected of having COVID-19. n=108 samples (unclear if from unique patients). Patient characteristics not reportedThree index tests, all ‘sample to answer’ platforms: (1) RT-PCR (Cepheid Xpert Xpress SARS-CoV-2 assay, performed on GeneXpert instrument system); (N2, E); (2) isothermal amplification (Abbott ID NOW COVID-19 assay); (RdRp); (3) DNA hybridisation and electrochemical detection (GenMark ePlex SARS-CoV-2 assay) (target NR). All compared with RT-PCR (Hologic Panther Fusion SARS-CoV-2 assay); (Orf1ab). Nasopharyngeal specimens
Zhang et al (2020)71 Retrospective case series. China, two centres. Collection from 29 December 2019 to 16 February 2020People with laboratory-confirmed COVID-19 (via RT-PCR). n=290. Median age 57 years (22–88 years). 155 (53.4%) maleRT-PCR (Shanghai Bio-germ Medical Technology) (Orf1ab, N primers); pharyngeal swab samples
  • NR, details not reported; RT-LAMP, reverse transcription loop-mediated isothermal amplification.