Comparison of High Sensitivity C-Reactive Protein Assay with a Wide Assay Range

Kang, So Young; Suh, Jin Tae; Kim, Jeong Hun; Lee, Woo In; Lee, Hee Joo

Abstract

BACKGROUND
C-reactive protein (CRP) concentrations are measured by two different assays depending on the clinical concern: the conventional CRP assay for estimating the extent of inflammation and the high sensitivity-CRP (hs-CRP) assay for assessing the risk of cardiovascular diseases. To integrate the conventional CRP test detecting acute phase response and the hs-CRP assay with a lower detection limit, we evaluated the performance characteristics of hs-CRP assay methods with a wide range of concentrations. METHODS: Immunonephelometric assay (BNII) and two turbidoimmunometric assays (TIA), the Hitachi 7600 with Daiichi reagent (Daiichi-Hitachi) and the Roche Cobas Integra 700 (Integra), were evaluated for the precision with 8 levels of pooled sera ranging from 0.3 to 120 mg/L and the agreement of TIA with the BNII assay using regression analysis and Bland-Altman analysis with 89 patient samples.
RESULTS
The within-day coefficients of variation (CVs) of BNII were less than 10% in all levels of pooled sera. The CVs of Daiichi-Hitachi and Integra exceeded 10% in pooled sera below 1.0 mg/L and 0.5 mg/L, respectively. Both Daiichi-Hitachi and Integra appeared to be linear over the entire range of CRP concentrations used and were comparable with the results of BNII (Daiichi-Hitachi: y=0.98x+0.13, r=0.97, Integra; y=1.02x+0.22, r=0.99). However, at the concentrations over 100 mg/L TIA and BNII showed discrepant results. CONCLUSIONS: Both Daiichi-Hitachi and Integra showed a good precision over a wide range of CRP. However, the discrepant results found at very high concentrations require standardization among different assay methods or instruments.