Aims. The cycle threshold (Ct) value for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) nucleic acid detection is important because of the criteria for quarantine management, including release from it, which are defined in Guidelines on the Novel Coronavirus-Infected Pneumonia Diagnosis and Treatment (Provisional 9th Edition, China). As this is also currently relevant because of the recent SARS-CoV-2 epidemic in Shanghai, we discuss the SARS-CoV-2 nucleic acid detection and its problems. We focus on the gene fragments and sample types involved in nucleic acid detection and their effect on the latest criteria for release from quarantine. Methods. A total of 215 patients with SARS-CoV-2 infection were included. Pharyngeal swabs (nasopharyngeal swabs plus oropharyngeal swabs) were collected in the early stage of the disease, and pharyngeal swabs, sputum samples, and anal swabs were collected both in the middle and advanced stages of the disease. The Open reading frame 1ab (ORF lab) gene, Nucleocapsid (N) gene and Envelop (E) gene of each sample were quantitatively analyzed using fluorescence qPCR technique. Results. Exclusion of the E gene detection results had no significant effect on the interpretation of the nucleic acid Ct value of 35, with a positive concordance rate of 98.7% (95% CI 86.0%-100%) and an overall concordance rate of 99.7% (95% CI 92.9%-100%). The kappa coefficient was 0.99 (95% CI 0.92-.00). Compared with nucleic acid detection using both pharyngeal swab and sputum sample, the positive concordance rate of the detection using pharyngeal swab alone was 47.6% (95% CI 27.8%-99.3%). The kappa coefficient was 0.63 (95% CI 0.53-0.75), and the consistency was not ideal. Conclusions. Nucleic acid detection using the ORF 1ab gene and the N gene can achieve the purpose of SARS-CoV-2 detection. Nucleic acid detection using sputum samples is significant in the determination of Ct values and its significance in the development of the criteria for release from quarantine needs to be taken into account. It is suggested that to increase the accuracy of nucleic acid detection, instead of unilaterally pursuing increasing the number of target genes for amplification and improving PCR techniques, more attention should be paid to sampling and sample reliability, as well as strict quality control of the detection process.