Overestimation of canine albumin concentration with the bromcresol green method in heparinized plasma samples.

Albumin concentrations are routinely measured in dogs with bromcresol green (BCG)-binding assays on automated chemistry analyzers. Several variables affect this assay, including the length of reaction time, sample type, and lack of specificity of BCG for albumin. We observed that albumin concentrations measured with BCG appeared higher in heparinized plasma samples in sick dogs. The objective of this study was to determine the effect of anticoagulant and assay procedure on BCG albumin concentrations in clinically ill dogs. We hypothesized that albumin concentrations would be overestimated in heparinized plasma compared with serum because of the combination of heparin and fibrinogen. Furthermore, we hypothesized that the overestimation would be influenced by assay parameters. Blood was collected from 32 clinically ill dogs into tubes containing heparin, citrate, or no anticoagulant. Citrate was chosen to assess the effect of fibrinogen in the absence of heparin. Albumin concentration was measured in all 3 sample types from each dog using 2 different BCG procedures on an automated chemistry analyzer. The BCG procedures (standard and modified) differed in the wavelengths used for absorbance readings (standard, 600/700; modified, 570/505) and the time point at which absorbance was measured (standard, 100 seconds; modified, 40 seconds). In addition, the modified method incorporated a sample blank. Globulin fractions, fibrinogen concentration, and indices of lipemia, hemolysis, and icterus were evaluated for their contribution to the overestimation of albumin concentration in heparinized plasma compared with serum samples. Albumin concentrations were significantly higher (P </=.05) in heparinized plasma (mean +/- SE, 3.8 +/- 0.1 g/dL) than in serum (3.6 +/- 0.2 g/dL) or citrated plasma (3.2 +/- 0.1 g/dL). Overestimation was evident only with the standard BCG procedure. Multiple linear regression analysis indicated that fibrinogen was largely responsible for the higher albumin concentration in heparinized plasma with the standard method. Based on these results, heparin is not recommended as an anticoagulant for albumin measurement in dogs when using a BCG method unless the assay uses a sample blank and a reaction time of <1 minute.