关键词: 脂蛋白（a）, 检测系统, 一致性

Abstract:

Objective To evaluate the consistency of different systems for determining lipoprotein（a） [Lp（a）],and to investigate whether calibrators can improve the consistency or not. Methods A total of 6 different kinds of Lp（a） reagents（Roche,BSBE,Diasys,Yunda,Medicalsystem and Maccura）were chosen,and they were combined with Hitachi 7600 automatic analyzer separately to set up different systems. A total of 51 samples were collected,and each sample was determined by 6 different systems. Roche system was as a reference system,and the samples were retested by the other 5 systems after being calibrated by Roche system. The results of same sample that was determined by different systems before and after calibration were compared. Results There was difference for the determination of Lp（a） among different systems [the mean of coefficient of variation （CV） was 29.8%,and the median of CV was 28.9%]. However,the difference could be decreased after calibration （the mean of CV was 17.6%,and the median of CV was 13.3%]. The intercept of regression equation of each system to the reference system reduced after calibration. The results still can not be exchanged by conversion factor. The difference between the results from Roche system and the other systems was decreased after calibration. Compared with the reference system,the proportions of bias < ±30% and < ±15% were increased. Conclusions The results from different systems have difference. When calibrating systems by using the same calibrator,the difference of results can be decreased,but it still can not be completely eliminated.

Key words: Lipoprotein（a）, Determination system, Consistency