Mehmet Akif BİLDİRİCİ
Journal of Turkish Clinical Biochemistry - 2026;24(2):15-20
Objective: To compare the analytical performance of 2 immunoturbidimetric C-reactive protein (CRP) reagents using the same automated analyzer and to evaluate whether an alternative reagent provides clinically acceptable results. Methods: This prospective method-comparison study used 44 residual serum samples. C-reactive protein levels were measured on a Beckman Coulter AU5800 clinical chemistry analyzer using the original Beckman Coulter reagent and an alternative Archem reagent. Method comparison was performed in accordance with CLSI EP09-A3 guidelines using percentage difference plots and Passing-Bablok regression analysis. Precision was evaluated based on CLSI EP15-A3 recommendations. Acceptable bias was defined as 23.4% according to the EFLM Biological Variation Database. Results: The mean percentage difference between the methods was -15.4% (95% CI: -18.0 to -12.9), which was within the predefined acceptable bias limits. Passing-Bablok regression analysis showed a slope of 0.78 (95% CI: 0.73-0.81) and an intercept of 0.56 (95% CI: 0.36-0.83), indicating proportional and constant differences between the methods. Both reagents demonstrated acceptable within-run and between-run precision. Within-run precision values exceeded the manufacturer's specifications, whereas between-run precision results were consistent with the reported values. Conclusion: Despite the presence of proportional and constant differences, the alternative CRP reagent demonstrated clinically acceptable analytical comparability within biological variation-based limits. These findings support its potential use in routine laboratory practice, provided that the same reagent is consistently used for patient follow-up.
