Evaluation of an automated CRISPR-based diagnostic tool for rapid detection of COVID-19

The performance of an automated commercial CRISPR/Cas based technology was evaluated and compared with routine RT-PCR testing to diagnose COVID-19. Suspected and discharged COVID-19 cases were included and tested with CRISPR-based SARS-CoV-2 test and RT-PCR assay using throat swab and sputum specimens. The diagnostic yield was calculated and compared using the McNemar test. A total of 437 patients were included for analysis, including COVID-19 (n = 171), discharged cases (n = 155), and others (n = 111). For the diagnosis of COVID-19, the CRISPR-SARS-CoV-2 test had a sensitivity and specificity of 98.2% (168/171) and 100.0% (266/266), respectively; the RT-PCR test had a sensitivity and specificity of 100.0% (171/171) and 100.0% (266/266), respectively. No significant difference was found in the sensitivity of CRISPR-SARS-CoV-2 and RT-PCR. In conclusion, the CRISPR-SARS-CoV-2 test had a comparable performance with RT-PCR and showed several advantages, such as short assay time, low cost, and no requirement for expensive equipment.