不同检测系统对多项生化指标结果的可比性试验

目的]探讨不同检测系统之间对12项生化检测结果是否有可比性,为检验结果互认打下基础,并为临床实验室认可提供实验数据.方法]参照NCCLS的EP9-A文件要求,先检测2套检测系统的精密度和线性,用配对t检验和回归统计方法分析2套系统对12项生化指标测定结果的偏差和比对.结果]配对t检验结果表明校正前GLU1个项目差异无统计学意义;其他11个项目差异有统计学意义(P<0.05),应进行校正;将换算后的各项目的直线回归方程的斜率、截距作为仪器的校正参数,输入检测系统2中,在2个检测系统上再做批量样品的12个项目,配对t检验结果差异均无统计学意义(P>0.05).结论]当同一检验项目存在2个以上的检测系统时,应对方法和结果进行比对,并评估偏差,以保证检验结果的可比性.

COMPARARBILITY TEST OF THE RESULTS OF MUTI-BIOCHEMICAL INDEXES BY DIFFERENT DETECTING SYSTEMS

Objective] To investigate the comparability within the results of 12 biochemical detections obtained from two detecting systems through comparative analysis, by which lays down foundation for results mutual accreditation and provides experimental data for the recognition of clinical laboratory. Methods] Consulting EP9-A file of NCCLS. precision and linearity of two detecting systems were first evaluated. Afterwards, applying one system as experimental system and the other as system for comparison, 12 biochemical indexes were detected. Then. the comparison and deviation of the results between two systems were analyzed by paired t tests and statistical regression methods as well. Results] Before correction. There was no significant difference for one GLU project by using the paired t tests before adjustment, and significant difference for the other 11 projects was found. Correlation coefficients were obtained from slope coefficient, intercept of the linear regression equation and input into detecting system 2. The results of the retest show no significant difference. Conclusion] When there are more than two systems for the same test, method comparison and bias estimation are necessary to ensure the comparability.