摘要
目的通过对不同检测系统进行方法学比对分析,探讨各检测系统之间测定临床常用血清酶是否具有可比性。方法依据美国临床和实验室标准化研究所EP9-A2文件要求,每天随机从临床样本中抽取8份不同浓度的血清样本,分别用两种检测系统进行相关血清酶测定,共测定5d,记录检测结果,并进行相关回归分析。根据直线方程计算在某医学决定水平处的预期偏倚(B^c)和预期偏倚的95%可信区间,并判断偏倚性是否可以接受。结果除天门冬氨酸氨基转移酶在低值处的预期偏差(绝对值)大于1/4允许总误差外,不同检测系统对常用血清酶测定结果的偏差在允许误差范围内,其检测结果具有可比性。结论当用2个以上检测系统检测同一检验项目时,应进行方法学比对和偏倚评估,判断其临床可接受性能,以保证检验结果的可比性。
Objective To investigate the comparability among different kinds of analytic systems measuring common sero-enzymes by methodical comparation of two analytic systems. Methods Based on the guide of CLSI- EP9-A2 file, involved sero-enzymes of eight serum samples with different concentrations were measured with two different detection systems for consecutive five days, then the results were recorded and subjected to correlation and regression analysis. By means of linear equation, the expectant bias at a medical decision level and 95% confidence interval were calculated to estimate acceptability of the bias. Results All of the biases of common sero-enzyme results were accepted within the allowable error range by using different analytic systems except for the predicted bias of AST（absolute value）, which was beyond a quarter of TEa at low concentration, and the results from two detection systems were comparable. Conclusion When the same test item is analyzed by over two detection systems, it is necessary to evaluate the clinical acceptability to ensure the comparability by methodical comparison and bias estimation.
出处
《检验医学与临床》
CAS
2008年第22期1366-1368,共3页
Laboratory Medicine and Clinic
关键词
酶类
实验室技术和方法
偏差
偏倚(流行病学)
质量控制
enzymes
laboratory techniques and procedures
uncertainty
bias（epidemiology）
quality control
