不同生化分析系统间检测结果的偏差评估

Method comparison of different biochemical analysis systems

目的通过对同一临床实验室不同生化分析系统进行方法比对和预期偏差评估,探讨不同生化分析系统间检测结果是否一致或检测结果的偏差是否在允许范围内。方法依据EP9-A文件方案,以Olympus Au2700(日本)全自动生化分析仪为对比方法,Vitros250(美国强生)干化学分析仪为试验方法,用患者血清对尿素、白蛋白、葡萄糖等8个项目进行检测,计算相关系数和直线回归方程,对两分析系统之间的预期偏差进行评估。结果在检测的8个项目中,直接胆红素(DBIL)两系统检测结果的相关系数r<0.975,其余项目测定结果的相关系数r<0.975。结论当实验室内同一检测项目存在两套以上分析系统时,应对其进行方法对比和偏差估计,以保证检测结果的准确稳定。

Objective To explore the results between different biochemical analysis systems in a same laboratory the same or the deviation of the test results in a allowed range, through method comparison of the different biochemical analysis systems and the assessment of expected deviation. Methods Based on EP9-A file, and use Olympus Au 2700 （Japan） automatic biochemical analyzer for the comparison method. Vitros250 （Johnson of United States） dried chemical analyzer for testing method, test urea, albumin, glucose, etc. 8 items with serum of patients, then calculate the correlation coefficient and linear regression equation to assess the expected deviations between the two analysis systems. Results In the eight items, the correlation coefficient of the two systems in the test of direct bilirubin （DBIL） was less than 0. 975, arid the correlation coefficient of the remaining items was less than 0. 975. Conclusions If there are more than two analysis systems of a test item, we should make methods comparison and bias estimation in order to ensure the accuracy and stability of the test results.