We found high heterogeneity rates among trials, probably cause by some differences in the methods, patient characteristics, and samples used. CTChest image-Yu, br / 2020 30China76-NRddPCR and RT-PCRNasal swabORF1ab and NThroat swabSputumUrineBloodZhang, br / 2020 31China14-RT-PCRNATStool-Oropharyngeal swabChest CTChest image-Zhao, br / 2020 32China173/535-RT-PCRELISAPlasmaIgM/IgG Open in a separate window em E /em , envelope protein gene; em ELISA /em , enzyme linked immunosorbent assay; em Hel /em , helicase protein gene; em LFIA /em , lateral flow immunoassay; em N /em , nucleocapsid protein gene; em NAT /em , nucleic acid tests; em NR /em , not reported; em ORF1ab /em , open reading frame 1ab gene; em RdRp /em , RNA-dependent RNA polymerase gene for SARS-CoV, SARS-CoV-2, and bat-SARS-related CoV; em RdRp-P2 /em , Rabbit Polyclonal to MT-ND5 RNA-dependent RNA polymerase specific gene for SARS-CoV-2; em RT-PCR /em , real\time reverse\transcriptase polymerase\chain reaction; em S /em , spike protein gene. ?Samples were used only for the validation of the method (no clinical application). Analytical parameters Three studies evaluated the optimization of PCR parameters for the detection of SARS-CoV-2.20 , 22 , 33 Chan et al. 20 developed and compared the performance of 3 new essays of RT-PCR of RNA-dependent RNA polymerase (RdRp)/helicase (Hel), spike (S) and nucleocapsid (N) genes from SARS-CoV-2. Corman et al. 22 assessed several SARS-related viral genomic sequences to design the best primer and probe set. Pferfferle et al. 33 investigated a set of primer and probes, targeting the E gene, for use in an automated system (Cobas 6800 System; see Table 2 ). Table 2 Analytical parameters reported by the included studies thead th align=”left” valign=”top” rowspan=”1″ colspan=”1″ Study /th th align=”center” valign=”top” rowspan=”1″ colspan=”1″ Method /th th align=”center” valign=”top” rowspan=”1″ colspan=”1″ Probe RNA /th th align=”center” valign=”top” rowspan=”1″ colspan=”1″ Gene target /th th align=”center” valign=”top” rowspan=”1″ colspan=”1″ LoD br / RNA copies/reaction (CI) /th /thead Chan, br / 2020 20RT-PCRspecific for SARS-CoVRdRp/Helicase11.2 (7.2-52.6)RT-PCRspecific for SARS-CoVSpike geneNART-PCRspecific for SARS-CoVN gene21.3 (11.6-177.0)RT-PCRspecific for SARS-CoV-2RdRp geneNACorman, 2020 22RT-PCR br / (new method)specific for SARS-CoVE gene5.2 (3.7-9.6)RT-PCR br / (new method)specific for SARS-CoVRdRp gene3.8 (2.7-7.6)RT-PCR TaqMan Fastspecific for SARS-CoVE gene3.2 (2.2-6.8)RT-PCR TaqMan Fastspecific for SARS-CoVRdRp gene3.7 (2.8-8.0)RT-PCR br / (new method)specific for SARS-CoV-2E gene3.9 (2.8-9.8)RT-PCR br / (new method)specific for SARS-CoV-2RdRp gene3.6 CFM-2 (2.7-11.2)Pfefferle, br / 2020 33RT-PCRspecific for SARS-CoV-2E gene275.72 (NR) Open in a separate window em CI /em , confidence interval 95%; em LoD /em , limit of detection; em NA /em , not applied; em NR /em , unreported. The genes E and RdRp were the most commonly used to detect the COVID-19 virus, both with high analytical sensitivity CFM-2 (technical limit of detection of 3.2 and 3.6 copies per reaction, respectively). The detection of the gene N presented lower analytical sensitivity (8.3 copies per reaction). The probe used by these studies is indicated for any SARS-CoV infection, including SARS-CoV-2. Process automation by using the open channel of the Cobas 6800 systems significantly increased the limit of detection. Diagnostic accuracy of tests Meta-analyses evaluating the parameters of accuracy (sensitivity, specificity, PLR and NLR) of the reported tests were performed (Supplementary?Table S1), results are shown in Table 3 . Table 3 Meta-analysis of the parameters of accuracy for the different diagnostic techniques thead th valign=”top” rowspan=”1″ colspan=”1″ Technique /th th align=”center” valign=”top” rowspan=”1″ colspan=”1″ Sample /th th align=”center” CFM-2 valign=”top” rowspan=”1″ colspan=”1″ No. of studies /th th align=”center” valign=”top” rowspan=”1″ colspan=”1″ Sensitivity br / (95% CI) /th th align=”center” valign=”top” rowspan=”1″ colspan=”1″ Specificity br / (95% CI) /th th align=”center” valign=”top” rowspan=”1″ colspan=”1″ PLR br / (95% CI) /th th align=”center” valign=”top” rowspan=”1″ colspan=”1″ NLR br / (95% CI) /th /thead Computed tomography-6 18,19,23,28,29,310.919 br / (0.898-0.937) br / I2??=?92.9%0.251 br / (0.210-0.295) br / I2??=?32.8%1.194 br / (0.936-1.525) br / I2??=?56.2%0.301 br / (0.043-2.124) br / I2??=?71.9%Immunological test br / (IgM and IgG)Blood, serum, plasma4 21,24,25,320.845 br / (0.822-0.866) br / I2??=?93.2%0.916 br / (0.860-0.954) br / I2??=?0.0%7.604 br / (3.903-14.817) br / I2??=?12.8%0.170 br / (0.041-0.697) br / I2??=?97.0%Immunological test br / (IgM and IgG)Blood3 21,24,320.863 (0.833-0.888) br / I2?=?96.3%0.907 (0.848-0.948) br / I2?=?0.0%8.618 (5.219-14.231) br / I2?=?0.0%0.146 (0.021-1.028) br / I2?=?99.0%Immunological test br / (IgM and IgG)Serum2 24,250.82 (0.78-0.85) br / I2?=?35.8%—Immunological test (IgM)Blood, serum, plasma5 21,24,25,27,320.770 br / (0.745-0.795) br / I2??=?89.9%0.933 br / (0.886-0.965) br / I2??=?18.5%7.295 br / (3.403-15.641) br / I2??=?96.1%0.211 br / (0.067-0.666) br / I2??=?96.1%Immunological test (IgM)Blood3 21,24,320.788 (0.754-0.819) br / I2?=?94.8%0.931 (0.882-0.964) br / I2?=?43.3%8.390 (3.367-20.905) br / I2?=?24.0%0.274 (0.072-1.043) br.