To improve the performance of nucleic acid amplification assay (RT-PCR)and obtain high quality screening results of SARS-CoV-2 detection through retrospective study of the data generated from confirmatory test.

RT-PCR kit A from our laboratory and the kit B provided by Bozhou Center for Disease Control and Prevention (CDC) were both utilized to analyze fifty throat swabs specimen in a double blind experimental manner. In the follow-up routine detection with reagent A, 1 354 of the 37 250 samples were found to have abnormal test results, among which 1 286 were failed test amplification samples, and 68 were abnormal test results of suspicious samples with end jump but no complete amplification. 1 354 samples were reexamined with reagent A, and the results of the two tests were statistically analyzed.

The coincidence rate of 50 samples with known results was 96.0% (48/50), indicating the reliability of reagent A. The results of 1 286 throat swabs specimen with failed amplification retested with the same A kits showed that there are 1 114 specimen of negative results, 15 specimens of abnormal ORF 1ab gene signal, 10 specimen of abnormal N gene signal, and 147 specimens were remained failure of detection with no data. The re-tested result from 57 specimens with abnormal single gene signal at the first test by using kit A showed high consistence with 94.70% coincidence rate and with 10.53% (6/57) positive rate. The re-tested results from analyzing 11 specimens with double abnormal genes signal at the first test by kit A also showed complete consistence results (100% coincidence rate) with 81.82% positive rate (9/11).

The Kit A meets laboratory requirements for rapid early screening of viruses. Meanwhile,the specimen with any abnormal detective signal during the RT-PCR reaction should be retested to ensure reliable results. At the same time, the precise operating in laboratory with a high quality of specimen are the most important criteria for obtaining the high quality data.