生物学变异允许总误差在血脂检测中的质量管理

目的探讨生物学变异允许总误差在血脂检测中的质量管理。方法对2015年参加原卫生部脂类正确度验证的胆固醇(CHOL)、三酰甘油(TG)、高密度脂蛋白(HDL-C)、低密度脂蛋白(LDL-C)进行σ分值、质量目标指数(QGI)、优先改进措施及性能评定等指标进行评价,对成绩不理想的CHOL作原因分析。结果生物学变异设定的质量规范3个层次,当允许总误差处于适当的水平时,TG、HDL-C、LDL-C的σ分值已满足6,无需改进。而CHOL的分析性能为差,需优先改进正确度。当允许总误差处于最佳水平时,4个项目中只有TG的σ分值可达到良,可进行精密度改进。当允许总误差处于最低水平时,TG、HDL-C、LDL-C的σ分值远大于6,分析性能达到优,CHOL的σ分值(2.9)已接近3σ水平,应从正确度进行纠正。结论 6σ质量理论可反映检测项目的性能,且能有效提高分析质量。

Study on the quality management of biological variation allowed total error in detection of serum lipid

Objective To study on the application of biological variation total allowable error in quality management of serum lipid detection .Methods The σ score ,quality goal index (QGI) ,priority improvement measures and performance evaluation of lipid accuracy criteria were evaluated ,including cholesterol (CHOL) ,triglyceride (TG) ,high density lipoprotein (HDL‐C) ,low density lipoprotein (LDL‐C) ,which included in the standard of accuracy of the Ministry of Health in 2015 .The reason for unsatisfactory re‐sults of cholesterol (CHOL) test were analyzed .Results Based on the three levels of quality specifications derived in biological var‐iation ,when the total allowable error was located at an appropriate level ,the σ score of TG ,HDL‐C ,LDL‐C reached 6 ,it was not re‐quired for improvement .While the score of CHOL performance analysis was poor ,accuracy was required to give priority to impro‐ving .When the total allowable error reached the best level ,only the σ score of TG achieved “good” in the four items ,improvement of precision was needed .When the total allowable error was located at the lowest level ,the σ score of TG ,HDL‐C ,LDL‐C was grea‐ter than 6 ,the score of performance analysis reached the “excellent” .The σ score of CHOL (2 .9) had been closen to 3σ ,accuracy was required to be corrected .Conclusion The biological variation derived total allowable error is easy to meet the requirements of the quality management in serum lipid determination by current technologies and methods .The theory of 6σ quality can reflect the performance of detection indexes ,and improve the quality of analysis effectively .