Discussion
The study results demonstrated high agreement between the rapid RT-PCR assay using the STANDARD M10 and the conventional RT-PCR using the Allplex SARS-CoV-2 for detecting SARS-CoV-2 during the Omicron surge in Korea. Discordant results between the two assays were observed in specimens having low viral loads, which were reflected with high Ct values.
Few studies have evaluated the clinical performance of STANDARD M10. Hong et al. estimated the sensitivity of STANDARD M10 among 342 positive samples proven by the conventional RT-PCR using the Allplex SARS-CoV-2 assay.8 They reported a sensitivity of 87.7% with STANDARD M10 using simulation with a virtual specimen pool comprising specimens having a Ct value of E gene ≥30 as 27.7%. In the present study, STANDARD M10 detected 118 specimens among 123 positive specimens that were positive with Allplex assay (95.9%), although the positive specimens consisted of 34.1% of specimens with a Ct value ≥30 for E gene, which demonstrates acceptable clinical performance based on the criteria by the Korean Ministry of Food and Drug Safety. These findings suggest that clinical performance would be different between the evaluation from manipulated laboratory dataset and the data acquired from consecutive patients in the real-world clinical setting.
Discordant results between STANDARD M10 and Allplex were observed in specimens with a Ct value >33 using the Allplex assay. These findings that the limit of detection of RT-PCR is associated with a viral load of the samples estimated by the Ct value are also reported by previous researchers who evaluated the performance of rapid RT-PCR. Jeong et al. found that the PPA of STANDARD M10 was 97.4% of the conventional RT-PCR assays.9 However, the PPA was decreased to 87.5% in samples with a Ct value of >30 for the E gene, similar to that noted in the present study. Recent studies compared well-known cartridge-based rapid RT-PCR assays, Xpert Xpress SARS-CoV-2 (Xpert) and ID NOW COVID-19 (Abbott, IL, USA; ID NOW) with the Roche Cobas SARS‐CoV‐2 (Roche Molecular Systems, NJ, USA) for samples with low, medium, and high SARS‐CoV‐2 viral concentrations.10,11 The two rapid RT-PCR assays showed 100% positive agreement for medium and high viral concentrations, defined as having a Ct value <30. However, for low viral concentrations defined as a Ct value >30, positive agreement for the Xpert was 97.1% (95% CI: 83.4–99.8%), whereas it was 34.3% (95% CI: 19.7–52.2%) for the ID NOW assay.11
Before the COVID-19 pandemic, the Korea Disease Control and Prevention Agency prepared a laboratory diagnostic system in response to a potential unknown disease outbreak.12 This system was successfully applied for laboratory diagnosis based on developing and implementing RT-PCR for the causative pathogen of COVID-19 in the very early phase of the COVID-19 pandemic in Korea. However, there was no reliable domestic rapid RT-PCR method in Korea, and thus, diagnosis depended on imported rapid-RT-PCR methods such as Xpert. The Xpert method yields reliable results, but it has issues with high prices and substantial undersupply of reagents in Korea. STANDARD M10 is a method similar to Xpert, a completely automated cartridge-based point-of-care testing method but less expensive than Xpert and in potentially sufficient supply. After the introduction of STANDARD M10, the emergency healthcare system of Korea could manage the Omicron surge as it can serve as a reliable method in the real-world setting.
This is the first study including consecutive data collection from a large patient cohort for clinical evaluation of STANDARD M10 in an emergency department during the Omicron surge. However, several limitations must be considered. First, because the study was performed during the Omicron surge, the performance of detecting other variants of concern could not be evaluated. Second, analytical performance was not evaluated. Third, specimens used for STANDARD M10 and Allplex in the initial test were different. Fourth, although retesting was performed with the same specimen for specimens showing discordant results, potential specimen effect could remain. Fifth, the ratio of patients with severe medical conditions (KTAS levels 1 and 2 and a portion of KTAS level 3) is lesser in this study than the real-word ratio. Because the patient groups were indicated for the Xpert assay by the Korea health insurance system, these patients underwent Xpert instead of STANDARD M10. Thus, KTAS level 1–3 patients may be underrepresented in this study.
In conclusion, the STANDARD M10 SARS-CoV-2 rapid RT-PCR assay showed reliable clinical performance compared with the conventional RT-PCR using Allplex SARS-CoV-2 in an emergency department. STANDARD M10 can be a useful method for emergency healthcare because of its advantage of cartridge-based fully automated system and short reporting time for rapidly detecting SARS-CoV-2 compared with conventional RT-PCR. The diversity of rapid-RT-PCR systems may also be useful to solve problems associated with an imbalance of resources of rapid RT-PCR technology in the world.