核酸单反应性质量控制多规则的建立
Establishment of a multi-rule quality control mechanism for blood nucleic acid screening results based on Poisson distribution
谢君谋 1蓝岚茵 1杜荣松 1黄爱琼 1刘旭映 1廖芬芳 1王淏1
作者信息
- 1. 广州血液中心 广州市医学重点实验室 血液安全重点实验室,广东 广州 510095
- 折叠
摘要
目的 探讨建立 1 个基于Poisson分布的监控核酸筛查实验室污染情况的质控机制.方法 收集 2022-2023 年基立福Panther核酸检测系统的核酸单反应性率和鉴别试验反应性率相关数据,采用Poisson分布概率法计算每日核酸检测单反应性的概率,利用单反应性率和鉴别反应性率及ROC曲线分析法进一步验证该质控模型的灵敏度和特异性,并提出相应的多规则化的质控策略.结果 仅以P=0.05 为失控限,Poisson分布概率质控模型判定P<0.05 的失控点为 40 个,结合鉴别反应性率进行人工验证后发现失控点为 20 个,其判定失控的灵敏度为 60.0%,特异性为 95.6%,失控点较多,结合鉴别反应性率和运用多规则化质控方法进行质量监控后,2022-2023 年核酸检测单反应性警告点为 18 个,失控点为 2 个.结论 基于Poisson分布概率的多规则核酸单反应性监控机制,更适用于核酸实验室监测预警实验室污染,有利于提升核酸实验室管理能力和检测质量.
Abstract
Objective To establish a quality control mechanism based on the Poisson distribution for monitoring con-tamination in nucleic acid testing(NAT)laboratories.Methods The study collected NAT single-reactivity rates and dis-criminatory test reactivity rates from the Grifols Panther NAT system from 2022 to 2023.The Poisson distribution probability method was used to calculate the daily probability of nucleic acid single reactivity.Sensitivity and specificity of the quality control model were further validated using single-reactivity rates,discriminatory reactivity rates and ROC curve analysis,fol-lowed by proposing corresponding multi-rule quality control strategies.Results Using P=0.05 as the threshold for out-of-control,the Poisson distribution probability quality control model identified 40 out-of-control points with P<0.05.After man-ual verification using discriminatory reactivity rates,20 out-of-control points were confirmed.The sensitivity of determining out-of-control was 60.0%,and the specificity was 95.6%.Due to the high number of out-of-control points,it is recommen-ded to combine discriminatory reactivity rates and employ multi-rule quality control methods for quality monitoring.Under the multi-rule quality control,there were 18 warning points and 2 out-of-control points for NAT single reactivity from 2022 to 2023.Conclusion The multi-rule NAT single-reactivity monitoring mechanism based on the Poisson distribution probability is more suitable for nucleic acid laboratories to monitor and warn of laboratory contamination,thereby enhancing the manage-ment capabilities and detection quality of nucleic acid laboratories.
关键词
Poisson分布/核酸检测/实验室污染Key words
Poisson distribution/nucleic acid testing(NAT)/laboratory contamination引用本文复制引用
基金项目
广州市卫生健康科技项目(20231A011074)
广州市科技计划项目(2023A03J0550)
出版年
2024
国家科技期刊平台
NSTL
万方数据