Analytical quality assessment and method comparison of immunoassays for the measurement of serum cobalamin and folate in dogs and cats

Abstract

Serum cobalamin and folate are often measured in cats and dogs as part of laboratory testing for intestinal disease, small intestinal dysbiosis, or exocrine pancreatic deficiency. We performed an analytical validation of human immunoassays for cobalamin and folate measurement (AIA-900 analyzer, Tosoh Bioscience) and compared results with those obtained using chemiluminescence assays (Immulite 2000 analyzer, Siemens Medical Solutions Diagnostics). Accuracy, precision, total observable error (TEobs%), and σ values were calculated for the immunoassays. Correlation and agreement were evaluated with Deming regression, Passing–Bablok regression, and Bland–Altman analysis. Cobalamin intra-assay and inter-assay CVs were 1.8–9.3% and 2.6–6.8%, respectively. Folate intra-assay and inter-assay CVs were 1.5–9.1% and 3.4–8.1%, respectively. TEobs (%) were ≤19 and ≤31 for cobalamin and folate, respectively. Sigma values were 3.60–11.50 for cobalamin and 2.90–7.50 for folate. Regression analysis demonstrated very high or high correlations for cobalamin [ r = 0.98 (dogs), 0.97 (cats)] and folate [ r = 0.88 (dogs), 0.92 (cats)] but Bland–Altman analysis revealed poor agreement for both. The immunoassays had good analytical performance for measuring cobalamin and folate in both species. Results obtained by the 2 analyzers cannot be used interchangeably and should be interpreted using instrument-specific reference intervals. Further studies are required to establish immunoassay-specific reference intervals and to evaluate the diagnostic performance and clinical utility of the analyzer for these analytes.