Assessment of three automated assays for C-reactive protein determination in dogs

OBJECTIVE: To determine the characteristics of an automated canine C-reactive protein (CRP) assay and evaluate 2 human CRP assays for use in dogs. Animals-56 client-owned dogs with pyometra and 11 healthy control dogs. PROCEDURES: Samples from 11 dogs with high (> 100 mg/L) or low (< 10 mg/L) CRP concentrations (determined by use of a canine ELISA) were evaluated by use of the automated canine CRP assay. Intra- and interassay imprecision was determined (by use of those 2 plasma pools), and assay inaccuracy was assessed by use of logistic regression analysis of results obtained via ELISA and the automated canine CRP assay. Two automated human CRP assays were used to measure plasma CRP concentration in 10 dogs. RESULTS: By use of the ELISA, mean +/- SD plasma CRP concentration was 96.1 +/- 38.5 mg/L and 10.1 +/- 23.2 mg/L in dogs with pyometra and control dogs, respectively. The automated canine assay had intra-assay coefficients of variation (CVs) of 7.8% and 7.9%, respectively, and interassay CVs of 11.1% and 13.1%, respectively. Results from the automated assay were highly correlated with results obtained via ELISA. The human assay results did not exceed 0.4 mg/L in any dog. CONCLUSIONS AND CLINICAL RELEVANCE: The automated canine CRP assay had less interassay imprecision, compared with the ELISA. The 2 human CRP assays were not suitable for analysis of canine plasma samples. The automated canine CRP assay was more precise than the ELISA for serial evaluations of plasma CRP concentration in dogs.