Evaluating a Modified Sputum Preparation Method for Improved qRT-PCR Detection of COVID-19

Background: An appropriate specimen is the crucial for quantitative reverse transcriptase–polymerase chain reaction (qRT-PCR) detection of COVID-19. Sputum is suitable specimen due to its noninvasive and simple collection process. For reference method, sputum samples were prepared by adding normal saline in a 1:2 ratio, followed by mixing and vortexing until fully dissolved. However, its sticky and vicious nature can hinder the preparation process and delay results. Therefore, a modified sample preparation was required during the COVID-19 pandemic. Objective: To compare a modified sample preparation method with reference method for sputum specimens in the detection of COVID-19 by qRT-PCR. Methods: Leftover sputum specimens were included and prepared in parallel using both the modified sample preparation and reference methods. qRT-PCR for COVID-19 was performed using the Seegene Allplex SAR-CoV2 master assay. Sensitivity, specificity, and other performance parameters were statistically analyzed. Cycle threshold (Ct) values of target genes (E, RdRp/S, N, and S variants) from both methods were compared. Result: A total of 202 samples were examined. The modified sample preparation and reference methods yield positive results of 32 (15.8%) and 31 (15.3%), inconclusive results of 1 (0.5%) and 5 (2.5%), invalid results of 4 (2.0%) and 27 (13.4%), and negative results of 165 (81.7%) and 139 (68.8%), respectively. The sensitivity and specificity were 86.7% and 100%, respectively, with positive predictive value of 100% and negative predictive value of 97.1%. There were no statistically significant differences in Ct values for the E, RdRp/S, N, and S variants between the 2 methods with P values of .169, .421, .513, and .172, respectively. Conclusions: The modified sample preparation method can be used during a COVID-19 pandemic while maintaining the reliable accuracy and reducing turnaround time.