Rapid direct nucleic acid amplification test without RNA extraction for SARS-CoV-2 using a portable PCR thermocycler

Soon Keong Wee; Suppiah Paramalingam Sivalingam; Eric Peng Huat Yap

We have observed, even in spiked water samples, that amplification efficiencies and the Cq values obtained at the limit of detection are dependent on factors such as the reagent mastermix, the model of thermocycler, the reaction volumes and the cycling protocols used (Table 1 ). This variability would be even more significant in a clinical setting, with the added confounding factors such as type of sample, sampling method, dilution in transport buffers, duration and conditions of storage, and the reagents and protocols for RNA extraction, that are difficult to standardize across healthcare settings and laboratories. While removing the RNA extraction step could help to reduce this variability in RNA yield and quality, variability in sample type and sampling procedure still renders the entire assay nonquantitative. Hence, we caution over-interpretation of Cq values, especially in the absence of in-house quality control and calibration data for each combination of protocol, reagents and thermocycler. These technical factors may also partly explain some of the false negatives and temporal variability in viral RNA shedding reported in various studies, that use fixed criteria Cq < 37 for positive, Cq between 37 to 40 as inconclusive for repeat testing and Cq > 40 as negative. In fact, our data demonstrates that as viral RNA concentrations approach the threshold of detection, our assay LoD has Cq mean value of > 37 and may exceed 40 (Table 1 ). Since negative controls in these probe-based assays remain consistently negative, it should be possible to report Cq > 37 as positive. Hence to achieve better standardization of quantitative results, we propose that instead of reporting Cq, laboratories may report the equivalent viral copy number in per unit sample volume, derived from running serial dilutions of viral or RNA controls in-house.