质控方案设计工具在室内质量控制中的应用

Application of quality control procedure design tools in the internal quality control

目的:应用几种质控方案设计工具,根据我室现有检测方法实际性能设计理性的室内质量控制方法。方法:依据美国 CLIA’88能力验证计划的分析质量要求确定临床总允许误差(TEa),以长期室内质量控制的数据估计各方法的不精密度(CV%),以参加卫生部临床检验中心的能力对比检验(PT)成绩估计不准确度(bias%),以此计算临界误差的大小,按照方法的过程能力大小使用质控选择表格进行规则的初期选择,并用功效函数图确认质控方法的性能,确定到达特定水平质量保让时,不同规则及每批测定质控个数需检出的临界误差大小,并建立操作过程规范图。结果:我室总胆红素(TB)、肌酐(CREA)、胆固醇(CHO)的测定在低值水平处方法性能较差,不能很好地检测出误差,需要查找原因进行质量改进,其他项目可根据方法性能选择不同的规则进行控制。结论:通过质控方案设计工具的使用,建立适合本室的方法控制规则,既保证测定结果能够满足临床的质量要求,又达到节约成本的目的,同时通过积极查找并解决不满意方法存在的问题,进行持续质量改进提高我室的过程能力。

Objective To design some rational quality control protocols from the stable performance of the analytic method using quality control procedure design tools. Methods To define the clinical Total Allowable Error （TEa） according to the analytic quality requires of CLIA＇88, the size of the medically important error was firstly calculated which was from the imprecision of the method estimated by the continuing internal quality control data and from the inaccuracy of the method estimated by the difference from the target value of the PT material of National Center for Clinical Laboratories. To selecte a part of rules with the stability （f） of the method and the size of ASE, and to use the Power Function Graphs to verify the performance of the different rules, the Operating Process Specifications Charts were finally made for application in the laboratory. Results Some methods were required the Quality Improvement （QI） because of poor performance at low concentration level （TB, CREA and CHO）. The appropriate quality control procedures could be selected according to the characteristics for the rest methods. Conclusions It is important to select the appropriate control rules and the number of materials using quality control procedure design tools so as to meet the clinical quality requirements and save the costs as well . The Quality Improvement must be taken to promote the Process Capability throughout the investigation and to solve the problems of the unsatisfactory methods.