XÁC ĐỊNH GIÁ TRỊ SỬ DỤNG VÀ ĐÁNH GIÁ ĐỘ KHÔNG ĐẢM BẢO ĐO PHƯƠNG PHÁP XÉT NGHIỆM TSH BẰNG KỸ THUẬT HÓA PHÁT QUANG TRÊN MÁY MIỄN DỊCH ARCHITECT I2000SR TẠI KHOA SINH HÓA BỆNH VIỆN CHỢ RẪY

Abstract

Objective: To (1) verify the precision, bias, and linearity of the TSH chemiluminescence assay on the Architect i2000SR analyzer at Cho Ray Hospital Biochemistry Department, and (2) estimate the measurement uncertainty (MU) of this assay. Methods: A descriptive cross-sectional study was conducted. Method verification for precision and bias followed CLSI EP15-A3 guidelines, and linearity followed CLSI EP06. Measurement uncertainty was estimated using a "top-down" approach, utilizing 3 months of internal quality control (QC) data and 3 cycles of external quality assessment (EQA) data. Results: Precision: QC samples (3 levels) showed CV_WL (1.90% - 2.04%), meeting manufacturer specifications. Patient serum samples (3 levels) showed CV_WL (1.96% - 3.13%), meeting the Upper Verification Limit (UVL). Trueness: QC samples (3 levels) showed %Bias (5.01% - 6.43%), meeting manufacturer specifications (<10%). Linearity: The analytical measurement range of 0.0046 – 82.1327 mIU/L was verified, with R² = 0.9999 and a regression equation of y = 0.9983x - 0.2743. Measurement Uncertainty: The expanded MU (U) was 14.68% (k=2, 95% CI), which is within the desirable Total Allowable Error (TEa) of 23.74%. Conclusion: The TSH assay on the Architect i2000SR system at Cho Ray Hospital meets the verification requirements for precision, trueness, and linearity. The expanded MU of 14.68% is within the desirable quality level, confirming the method's reliability for clinical diagnostics.