关键词: C反应蛋白, 偏倚, 检测

Abstract: Objective  To evaluate the analysis performance of testing results between speed scattering turbidimetric method and transmitted immunoturbidimetric method for determining C reactive protein(CRP). Methods  The speed scattering turbidimetric method (Siemens Healthcare Diagnostica BN-Ⅱ specific protein analyzer) and transmitted immunoturbidimetric method (Finland Orion Diagnostica QuikRead CRP analyzer) were used to determine CRP, and the precision, linearity, interference, correlation and bias were evaluated methodologically. Results  The total coefficient of variation (CV) of the BN-Ⅱ specific protein analyzer was <6%, when the concentration of CRP was 2.0-80.0 mg/L. The total CV of the QuikRead CRP analyzer was <7%, when the concentration of CRP was 8.0-80.0 mg/L. The linearity range was 8.0-70.0 mg/L. Haemoglobin (Hb) <10 g/L and bilirubin (Bil) <300 mg/L had no significant interference (<10%) for the assay. The interference (<10%) was not significant to the BN-Ⅱ specific protein analyzer when triglyceride (TG) <20 mmoL/L. When TG >15 mmol/L, the interference of low-value CRP samples was >10%, and there was no interference in high-value CRP samples for QuikRead CRP analyzer. The correlation analysis of 50 blood samples showed that both analyzers were correlated well (r=0.98，P<0.01). There was no significant bias from linearity regression through Cusum between the 2 analyzers（P>0.05）. The values of BN-Ⅱ specific protein analyzer were slightly lower than those of QuikRead CRP analyzer. Bland-Altman curve showed that the average bias of the 2 analyzers was -2.1. 〖WTHZ〗 Conclusions  The precision, interference and linearity tests are suitable for routine CRP determination on the 2 analyzers. Although the results have bias, they meet the clinical requirements.

Key words: C reactive protein, Bias, Determination