Comparative analysis and bias evaluation of the testing results of BUN,Cr,GLU and UA among three different biochemical analysis systems

Objective: To explore and evaluate the comparability and bias of the testing results of blood urea nitrogen (BUN), creatinine (Cr), glucose (GLU) and uric acid (UA) derived from three different biochemical analysis systems in three hospitals and to validate whether mutual accredit of these testing results can be performed among these hospitals. Methods: According to the document of NCCLS EP9-A, 5-7 clinic samples were detected in three kinds of different biochemical analysis systems (Beckman CX4, Olympus 2700 and Hitachi 7600) respectively everyday. The testing results of total 41 samples were recorded within 7 days and those of 40 samples were kept after one outlier value was deleted. The testing results derived from three biochemical analysis systems were analyzed comparatively and their bias was evaluated by biochemical analysis in Beckman CX4 as a control method and biochemical analyses in Olympus 2700 and Hitachi 7600 as experimental methods. Results: Both correlation (r²>0.95) and linear regression between the testing results of BUN, Cr, GLU and UA in three different biochemical analysis systems in three hospitals were good. Their anticipated biases (but with exception for Cr when its testing result was within the normal reference range) were within allowed biases. Conclusions: On the basis of good internal quality control and external quality assessment, the testing results of BUN, Cr, GLU and UA derived from three different biochemical analysis systems from the three hospitals can be mutually accredited.