National Measurement Laboratory, LGC, Teddington, Middlesex, UK.
Department of Microbial Sciences, School of Biosciences & Medicine, Faculty of Health & Medical Science, University of Surrey, Guildford, UK.
Clin Chem. 2021 Dec 30;68(1):153-162. doi: 10.1093/clinchem/hvab219.
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) RNA quantities, measured by reverse transcription quantitative PCR (RT-qPCR), have been proposed to stratify clinical risk or determine analytical performance targets. We investigated reproducibility and how setting diagnostic cutoffs altered the clinical sensitivity of coronavirus disease 2019 (COVID-19) testing.
Quantitative SARS-CoV-2 RNA distributions [quantification cycle (Cq) and copies/mL] from more than 6000 patients from 3 clinical laboratories in United Kingdom, Belgium, and the Republic of Korea were analyzed. Impact of Cq cutoffs on clinical sensitivity was assessed. The June/July 2020 INSTAND external quality assessment scheme SARS-CoV-2 materials were used to estimate laboratory reported copies/mL and to estimate the variation in copies/mL for a given Cq.
When the WHO-suggested Cq cutoff of 25 was applied, the clinical sensitivity dropped to about 16%. Clinical sensitivity also dropped to about 27% when a simulated limit of detection of 106 copies/mL was applied. The interlaboratory variation for a given Cq value was >1000 fold in copies/mL (99% CI).
While RT-qPCR has been instrumental in the response to COVID-19, we recommend Cq (cycle threshold or crossing point) values not be used to set clinical cutoffs or diagnostic performance targets due to poor interlaboratory reproducibility; calibrated copy-based units (used elsewhere in virology) offer more reproducible alternatives. We also report a phenomenon where diagnostic performance may change relative to the effective reproduction number. Our findings indicate that the disparities between patient populations across time are an important consideration when evaluating or deploying diagnostic tests. This is especially relevant to the emergency situation of an evolving pandemic.
严重急性呼吸综合征冠状病毒 2（SARS-CoV-2）的 RNA 量，通过逆转录定量 PCR（RT-qPCR）测量，被提议用于分层临床风险或确定分析性能目标。我们研究了重复性以及设置诊断截止值如何改变 2019 年冠状病毒病（COVID-19）检测的临床灵敏度。
对来自英国、比利时和大韩民国的 3 个临床实验室的 6000 多名患者的定量 SARS-CoV-2 RNA 分布[定量循环（Cq）和拷贝/mL]进行了分析。评估了 Cq 截止值对临床灵敏度的影响。使用 2020 年 6/7 月 INSTAND 外部质量评估计划 SARS-CoV-2 材料来估计实验室报告的拷贝/mL，并估计给定 Cq 的拷贝/mL变化。
当应用世界卫生组织建议的 25 的 Cq 截止值时，临床灵敏度下降到约 16%。当应用模拟的 106 拷贝/mL 检测限时，临床灵敏度也下降到约 27%。对于给定的 Cq 值，实验室间的变异超过 1000 倍（99%CI）。
虽然 RT-qPCR 在应对 COVID-19 方面发挥了重要作用，但由于实验室间的重复性差，我们建议不要使用 Cq（循环阈值或交叉点）值来设置临床截止值或诊断性能目标；校准的基于拷贝的单位（在病毒学的其他地方使用）提供了更具重复性的替代方案。我们还报告了一种现象，即诊断性能可能相对于有效繁殖数而改变。我们的研究结果表明，在评估或部署诊断测试时，随时间推移患者人群之间的差异是一个重要的考虑因素。这在不断演变的大流行的紧急情况下尤其重要。
