ĐÁNH GIÁ ĐỘ TƯƠNG ĐỒNG CỦA XÉT NGHIỆM CREATININ HUYẾT THANH BẰNG PHƯƠNG PHÁP ENZYMATIC VÀ PHƯƠNG PHÁP JAFFE TRÊN MÁY SINH HÓA TỰ ĐỘNG COBAS C702 TẠI BỆNH VIỆN CHỢ RẪY

Abstract

Methods: An experimental study was conducted in accordance with established CLSI guidelines (EP15-A3, EP06, EP29, EP09-A3). The performance of both methods was evaluated for precision, trueness, linearity, and measurement uncertainty using quality control materials, patient pooled sera, and calibrators. The comparability between the two methods was analyzed using ANOVA, Passing–Bablok regression, and Bland–Altman plots. Results: Both the enzymatic and Jaffe methods successfully met the verification criteria for precision and trueness as defined by ISO 15189:2022 and manufacturer specifications. The verified linear ranges were 0.18–9.28 mg/dL for the enzymatic method and 0.27–9.05 mg/dL for the Jaffe method. The expanded measurement uncertainty (Ue) was low and clinically acceptable for both. A strong correlation was observed between the two methods with a Pearson correlation coefficient (r) of 0.996. Passing–Bablok regression analysis revealed no significant proportional or constant systematic error. However, Bland–Altman analysis demonstrated that the Jaffe method showed a consistent positive bias, particularly at lower creatinine concentrations (approximately 1 mg/dL), compared to the enzymatic method. Conclusion: Both the enzymatic and Jaffe methods are suitable for routine clinical use. However, due to its enhanced specificity and greater reliability at low concentrations, the enzymatic method is recommended for routine screening and for the assessment of early-stage renal impairment, especially in pediatric patients or individuals with subtle kidney function decline.