Performance evaluation of six homocysteine cycle enzymatic detection systems

Objective To evaluate the performance of six homocysteine (Hcy) analysis systems. Methods This is a methodological evaluation study. We analyzed six cycle enzymatic systems, and their correlation and deviation compared with chemiluminescence microparticle immunoassay (CMIA) from Abbott Architect plus i2000 system. Precision, accuracy, anti-interference and analytical measuring range (AMR) were evaluated, according to the CLSI EP5-A2, EP15-A2, EP7-A2, EP6-A, EP9-A2 guidelines. To assess the accuracy, we used the reference material SRM 1955 from National Institute of Standards and Technology (NIST), and EQA samples from CAP and National Center of Clinical Laboratory. Regression analysis was conducted using Passing-Bablok method. Linear regression measurement was performed using Cusum method, with a statistical significance level set at P＜0.05. Correlation analysis was conducted using Pearson coefficient, with P<0.05 indicating significant difference. Deviation analysis was performed using Bland-Altman method. Results In the six systems (A1-D2) except A1, the within-run CVs were all 0.975 (P 40). Compared with CMIA system, the Bland-Altman results showed that the maximum average absolute deviation was 2.9 μmol/L. Conclusions The cycle enzymatic method used to measure homocysteine has good precision, linear range, and anti-interference ability. But it is noticeable that the results of cycle enzymatic was higher than those of CMIA. Meanwhile, the six systems do not apply to measuring urine samples.（Chin J Lab Med，2014,37：173-178） Key words: Homocysteine; Clinical enzyme tests; Immunoassay