THURSDAY, July 2, 2020 (HealthDay News) -- Existing evidence on the diagnostic accuracy of serological tests for COVID-19 is characterized by high risks of bias and heterogeneity, according to a review and meta-analysis published online July 1 in The BMJ.
Mayara Lisboa Bastos, M.D., from the McGill University Health Centre in Montreal, and colleagues conducted a systematic review and meta-analysis to examine the diagnostic accuracy of serological tests for COVID-19 compared with a reference standard.
The researchers identified high risk of patient selection bias in 98 percent of assessments (48 of 49); 73 percent had high or unclear risk of bias from test performance or interpretation. Four studies included outpatients; only two assessed point-of-care tests. The pooled sensitivity of enzyme-linked immunosorbent assays measuring immunoglobulin (Ig) G or IgM, lateral flow immunoassays (LFIAs), and chemiluminescent immunoassays was 84.3, 66.0, and 97.8 percent, respectively. Lower pooled sensitivity was seen for LFIAs in all analyses. Pooled specificities varied from 96.6 to 99.7 percent. Eighty-three percent of the samples used for estimating specificity were from populations tested before the epidemic or not suspected of having COVID-19. Pooled sensitivity was lower for commercial versus noncommercial LFIA tests (65.0 versus 88.2 percent). Compared to testing within the first week, sensitivity was higher for testing at least three weeks after symptom onset (ranging from 69.9 to 98.9 percent versus 13.4 to 50.3 percent).
"While the scientific community should be lauded for the pace at which novel serological tests have been developed, this review underscores the need for high-quality clinical studies to evaluate these tools," the authors write.