Optimizing diagnostic strategy for novel coronavirus pneumonia, a multi-center study in Eastern China

Jing-Wen Ai; Hao-Cheng Zhang; Teng Xu; Jing Wu; Mengqi Zhu; Yi-Qi Yu; Han-Yue Zhang; Zhongliang Shen; Yang Li; Xian Zhou; Guo-Qing Zang; Jie Xu; Wen-Jing Chen; Yong-Jun Li; De-Sheng Xie; Ming-Zhe Zhou; Jing-Ying Sun; Jia-Zhen Chen; Wen-Hong Zhang

The copyright holder for this preprint (which was not peer-reviewed) is . https://doi.org/10.1101/2020.02.13.20022673 doi: medRxiv preprint among all patients. Meanwhile, some patients having severe symptoms had relatively low virus loads detected form nasal swab samples. For example, patient no.92 were admitted into hospital with severe pneumonia and an oxygen saturation of below 93%; however, the SARS-COV-2 viral loads of his sample is 3.1X10 3 copies/μl while mNGS detected 1,216 unique reads, relatively lower than other NCP patients (Supplement 1). A possible explanation could be that NCP patients' lesions tends to be located in the lower respiratory tract, causing the RT-PCR positivity of the upper respiratory tract sampling to be reduced. Other factors such as RNA extraction efficacy, quality of kit or potential interference of multiple primers could also affect the detection of RT-PCR assay. Possible mutation in primer or probes regions could lead to false-negative detection, albeit these regions are very reserved. Although from WGS data, we haven't found any mutation in the primer regions in these patients, however, such risk always exits. The primer design also partially relies on the chosen gene. Orf1ab and S gene are very specific genes that can effectively distinguish COVID-19 from other coronaviruses (including SARS and SARS like viruses), which have been used to design specific PCR detection primers. N and E genes of the SARS-CoV-2 may form cross-reaction with other coronaviruses.