Background: Due to the high transmission rate of SARS-CoV-2, diagnostic tests have become tools for identifying patients. The key points were the virus genomes survey to design RT-LAMP primers; comparing the sensitivity and specificity of RT-LAMP and RT-qPCR; and determining the relationship among clinical symptoms, CT scan, RT-qPCR, and RT-LAMP results.
Methods: This cohort study included 444 symptomatic patients. The specificity and sensitivity of RT-LAMP were assayed. The five statistical models, simultaneously, by RapidMiner to find the best method for detecting the virus were done through the correlation between the clinical symptoms, RT-LAMP, RT-qPCR, and CT scan results. The chi-square test by SPSS 26.0 was used to calculate kappa agreement.
Results: The virus genome was detected in all the positive samples (198) by RT-qPCR and RT-LAMP. In addition, 246 samples were negative by RT-qPCR, while 88 were positive by RT-LAMP. Data mining analysis indicated that there were most associations between the RT-LAMP and CT scan data compared to RT-qPCR and CT scan data.
Conclusions: RT-LAMP could detect SARS-CoV-2 with great simplicity, speed, and cheapness. Therefore, it is logical to screen, a large number of patients by RT-LAMP, and then RT-qPCR can be used on the limited samples.
Keywords: CT scan; Data mining; RT-LAMP; RT-qPCR; SARS-CoV-2; detection.