Comparative study of two automated high-sensitivity C-reactive protein methods in a large population

Abstract

Increased serum C-reactive protein (CRP) is associated with future risk of coronary heart disease in apparently healthy individuals. Numerous high-sensitivity CRP (hs-CRP) methods are available but their comparability in large populations has not been assessed. This study aimed to evaluate the performance of two CRP methods in a large Asian population. We compared the Tina-quant CRP immunoturbidimetric assay (Roche COBAS INTEGRA) to the N high-sensitivity latex-enhanced immunonephelometric (BN 100 nephelometer, Dade Behring) assay using 4118 serum samples from the International Collaborative Study on Atherosclerosis and Stroke in Asia (Inter ASIA). The median hs-CRP value for the N high-sensitivity CRP method (1.23 mg/L) was significantly lower than that for the Tina-quant method (1.50 mg/L), P < 0.001. The two methods were highly associated (r = 0.9916). Deming regression analysis gave a slope of 0.958 [95% confidence interval (CI): 0.954-0.962] with an intercept of 0.280 [95% confidence interval (CI): 0.268-0.292]. The mean of the method differences was 0.19 mg/L and the limits of agreement (LOA), which encompass 95% of results, were -0.36-0.74 mg/L. We found the percentages of low, average, and high-risk results were 42.9, 33.8, and 23.3 for the N high-sensitivity CRP and 33.2, 41.1, and 25.7 for the Tina-quant method. The percentage of samples concordant by both methods was 87.4%. The Tina-quant CRP classified more subjects into the high-risk group. The two hs-CRP methods were highly associated and are suitable for screening large populations. © 2004 The Canadian Society of Clinical Chemists. All rights reserved.