Refractometric total plasma protein measurement as a cage-side indicator of hypoalbuminemia and hypoproteinemia in hospitalized dogs

Objective: To assess the relationship between total plasma protein (TPP) as measured by refractometry and serum hypoalbuminemia and hypoproteinemia in hospitalized dogs.

Design: Retrospective, observational study conducted over 6-month period between March and August 2008.

Setting: University teaching hospital.

Animals: Four hundred and three hospitalized dogs in an ICU.

Interventions: None.

Measurements and main results: TPP, serum albumin, total protein, glucose, urea, cholesterol was measured from dogs enrolled in study. TPP was evaluated as a predictor for hypoalbuminemia defined both as albumin <25 g/L (<2.5 g/dL) and albumin <20 g/L (<2.0 g/dL), and serum hypoproteinemia, defined as serum total protein <40 g/L (<4.0 g/dL), using logistic regression. Impact of glucose, urea, cholesterol, and total bilirubin on refractometric readings were also assessed. TPP predicted hypoalbuminemia at albumin concentrations of <25 g/L (<2.5 g/dL) and <20 g/L (<2.0 g/dL) (P<0.001). A TPP<60 g/L (<6.0 g/dL) predicted albumin <25 g/L (<2.5 g/dL) with 73% sensitivity and 86% specificity. A TPP<58 g/L (<5.8 g/dL) predicted a serum albumin <20 g/L (<2.0 g/dL) with 70% sensitivity and 80% specificity. For dogs with known risk factors where specificity optimization may be appropriate, refractometer TPP<50 g/L (<5.0 g/dL) and <48 g/L (<4.8 g/dL) predicted hypoalbuminemia at each level with >95% specificity, although sensitivity was poor. Refractometer TPP<58 g/L (<5.8 g/dL) predicted serum total protein of <40 g/L (<40 g/dL) with sensitivity of 82% and specificity of 84%. Hypercholesterolemia and hyperglycemia significantly affected TPP readings; an increase in serum glucose by 10 mmol/L (180 mg/dL) was associated with an average independent increase in refractometer TPP of 2.27 g/L (0.23 g/dL) (P<0.001, 95% confidence interval=1.08-3.47) and an increase in serum cholesterol of 1 mmol/L (38.6 mg/dL) was associated with an average independent increase in refractometer TPP of 1.36 g/L (0.14 g/dL) (P<0.001, 95% confidence interval=1.12-1.59).

Conclusion: Suboptimal sensitivity limits the use of refractometric TPP for prediction of hypoalbuminemia in the context of patient screening; a high proportion of false negatives may result. However, identification of a refractometric TPP<58 g/L is strongly indicative of both serum hypoalbuminemia and hypoproteinemia, with high specificity, and warrants further investigation. Refractometric readings may be falsely increased in patients with hyperglycemia or hypercholesterolemia.