Analysis of consistency in aspartate aminotransferase detection results and bias correction using reference materials as calibrators.

OBJECTIVE: To compare the results of two mainstream aspartate aminotransferase (AST) reagent systems and evaluate the bias correction effect of recalibration with reference material, thereby exploring feasible approaches to improve AST measurement consistency. METHODS: The AST detection results of Roche and Beckman reagents from the SCCL Enzymes Trueness Verification program (2022-2024) were analyzed. Five distinct measurement procedures were established and compared using 50 clinical samples: three based on Beckman instruments, utilizing reagents without pyridoxal-5-phosphate (PLP) but with different calibrators (Beckman-assigned or secondary reference materials with/without PLP assignment), and two based on Roche instruments, using reagents with or without PLP. Method comparative analysis was performed using Passing-Bablok regression and Bland-Altman plots. RESULTS: In the trueness verification program, the Beckman measurement procedure demonstrated higher median bias exceeding the acceptable threshold of 7.5%, whereas the Roche measurement procedure met this standard. In our comparative study, two procedure comparisons showed acceptable mean bias. Four groups without PLP supplementation showed correlation coefficients (r) of 1.00 and low residual standard deviations. In contrast, the three PLP-supplemented groups displayed r values of approximately 0.95 with substantially larger residual standard deviations. CONCLUSION: The Beckman AST measurement procedure without PLP exhibited significant deviations from other procedures. The use of reference materials effectively reduced inter-system differences. However, the PLP-containing IFCC reference procedure cannot be applied to the traceability of PLP-free reagents, and no PLP-free reference procedure is available internationally-this constitutes the fundamental obstacle to standardization. While unified calibration may improve comparability, true standardization remains a long-term goal.