Current best evidence for clinical care (more info)
Serologic testing for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) promises to assist in assessing exposure to and confirming the diagnosis of coronavirus disease 2019 (COVID-19), and to provide a roadmap for reopening countries worldwide. Considering this, a proper understanding of serologic-based diagnostic testing characteristics is critical. The aim of this study was to perform a structured systematic review and meta-analysis to evaluate the diagnostic characteristics of serological-based COVID-19 testing. Electronic searches were performed using Medline (PubMed), EMBASE, and Cochrane Library. Full-text observational studies that reported IgG or IgM diagnostic yield and used nucleic acid amplification tests (NAATs) of respiratory tract specimens, as a the reference standard in English language were included. A bivariate model was used to compute pooled sensitivity, specificity, positive/negative likelihood ratio (LR), diagnostic odds ratio (OR), and summary receiver operating characteristic curve (SROC) with corresponding 95% confidence intervals (CIs). Five studies (n=1,166 individual tests) met inclusion criteria. The pooled sensitivity, specificity, and diagnostic accuracy for IgG was 81% [(95% CI, 61-92);I2=95.28], 97% [(95% CI, 78-100);I2=97.80], and 93% (95% CI, 91-95), respectively. The sensitivity, specificity, and accuracy for IgM antibodies was 80% [(95% CI, 57-92);I2=94.63], 96% [(95% CI, 81-99);I2=92.96] and 95% (95% CI, 92-96). This meta-analysis demonstrates suboptimal sensitivity and specificity of serologic-based diagnostic testing for SARS-CoV-2 and suggests that antibody testing alone, in its current form, is unlikely to be an adequate solution to the difficulties posed by COVID-19 and in guiding future policy decisions regarding social distancing and reopening of the economy worldwide.
Discipline / Specialty Area | Score |
---|---|
Occupational and Environmental Health | |
Hospital Doctor/Hospitalists | |
Internal Medicine | |
Infectious Disease | |
Respirology/Pulmonology | |
Family Medicine (FM)/General Practice (GP) | |
General Internal Medicine-Primary Care(US) | |
Emergency Medicine | |
Public Health | |
Most practitioners are already well aware of the limitations of serology testing due to the high heterogeneity of serology test kits offered, as shown in this meta-analysis.
This is a good summary of fairly well-disseminated current knowledge. There is not much new, however.
Unlike the Cochrane review (PMID 32584464), this review did not report accuracy based on days since exposure. Perhaps as a result, the current review found high heterogeneity of results.
This is a formal analysis of what is already known to any physician who is familiar with these tests.
There is significant heterogeneity, but the likelihood ratios are quite good IgG +ve 27, -ve 0.2; IgM +ve 20; -ve 0.21. It emphasizes my concern with using PPV and NPV. This test performs well to rule in with a medium to high pre-test probability, and will be not bad for ruling out the disease. Eg pre-test prob 20% will yield a +ve post test of 87% and a -ve post test 4.8%. for IgG. Not bad at all.
The study does not take into account the more than 150 serological tests available.
This is on COVID-19 antibody testing; the pooled analysis supports conclusions to date. Antibody testing is unreliable. Until better methods are developed, its use is limited.