LAUSR

OBJECTIVE To assess the agreement between measurements of total protein (TP) concentrations in canine serum samples between a commercially available veterinary digital refractometer (DR), an analog handheld refractometer (AR), and a laboratory-based chemistry analyzer (LAB). An additional objective was to assess the effects of various potential interferents (ie, hyperbilirubinemia, increased BUN, hyperglycemia, hemolysis, and lipemia) on DR measurements. SAMPLE 108 canine serum samples. PROCEDURES Serum samples were measured in duplicate on the DR, which reported TP concentration, assessed via optical reflectance and critical angle measurement. These serum samples were also assessed on the AR and LAB for comparison. Serum samples with grossly visible lipemia, hemolysis, and icterus were noted. Medical records were retrospectively assessed to determine concentrations of BUN, glucose, and bilirubin. RESULTS Method comparisons among the various data generated by the analyzers were completed using linear regression, Bland Altman, and calculation of intraclass coefficients. Mean bias between DRTP and LABTP in samples without potential interferents was 0.54 g/dL with 95% limits of agreement of -0.17 to 1.27 g/dL. One-third of DRTP samples without potential interferents had > 10% difference from their LABTP comparison. Interferents, particularly marked hyperglycemia, can result in inaccurate measurements on the DR. CLINICAL RELEVANCE There was a statistically significant difference between DRTP and LABTP measurements. TP measurements in samples with any potential interferent, particularly hyperglycemia, should be assessed cautiously on DR and AR.