不同检测系统17项常规生化结果的比对和偏倚评估  

The contrast and bias estimation on the result of routine biochemistry items in different biochemical testing systems

作　　者：张秀明 李炜煊[2] 郑松柏 陈滨鸿[2] 曹永坚 孙蕾 林莲英 柯培锋 马骥 

机构地区：[1]广州中医药大学第二附属医院检验科,广东广州510120 [2]佛山市第一人民医院检验科,广东佛山528000

出　　处：《检验医学》2007年第2期166-170,共5页Laboratory Medicine

基　　金：国家"863"课题资助项目(2004AA2Z3762);广东省医学科研基金立项课题资助项目(A2006252)

摘　　要：目的探讨不同检测系统间常规生化结果的可比性,为不同实验室检验结果的互认和实验室认可提供依据。方法参考美国临床实验室标准化委员会(NCCLS)的EP9-A2文件,以Modular PPI检测系统为目标,用患者新鲜血清对不同自建检测系统的17项常规生化结果与目标系统进行比对,计算自建检测系统(Y)和目标检测系统(X)之间的相对偏差,以美国临床医学检验部门修正法规(CLIA′88)允许总误差的1/2为标准,判断检验结果的可比性。结果自建检测系统1的总胆红素(TBil)、直接胆红素(DBil)、γ-谷氨酰基转移酶(GGT)、三酰甘油(TG)、总钙(TCa)、无机磷(IP)与目标系统不具可比性,自建检测系统2的白蛋白(Alb)、TBil、DBil、TG、IP与目标系统不具可比性,自建检测系统3的所有17个项目与目标系统均具可比性。结论自建检测系统与目标系统的常规生化结果依项目不同存在不同程度的偏差。当用2个以上的检测系统检测同一检验项目时,应进行方法比对和偏倚评估,以实现检验结果的可比性。Objective To study the results comparability of different biochemical testing systems by bias estimation and method comparison. Methods According to National Committee for Clinical laboratory Standards（NCCLS） document EP9-A2, 17 biochemistry items in different self-buih biochemical testing system and aim system were tested respectively to obtain the correlation coefficient and the linear equation. Then the system bias between self-buih testing system and aim system could be evaluated by obtained equation. The comparability of different investigated system was judged according to the half of clinical laboratory improvement amendment 88 （CLIA＇88） allowed total error. Results The results of total bilirubin（TBil）, direct bilirubin（DBil）, γ-glutamy itransferase（GGT）, triglycerides（TG）, total calcium （TCa^2＋ ） and inorganic phosphate（IP） tested in system 1 and aim system showed no comparability. The results of albumin（ Alb）, TBil, DBil, TG and IP tested in system 2 and aim system showed no comparability. There was comparability between system 3 and aim system in all items. Conclusions Routine biochemistry results had different bias between self-buih testing system and Roche Modular PPI system. If the same item is tested in more than two systems, it is necessary to judge clinical acceptability by method comparison and bias estimation so as to insure the comparability.

关 键 词：偏差 比对研究 实验室 质量管理 

分 类 号：R195.1[医药卫生—卫生统计学]