Estimates of biological variation in routinely measured biochemical analytes in clinically healthy dogs.

ABSTRACT

BACKGROUND: The degree of biological variation in routinely measured concentrations and activities of biochemical analytes has not been well defined in client-owned pet dogs. OBJECTIVES: The aim of this study was to define indices of biological variation and, where appropriate, indices of individuality and critical change values for routinely measured serum biochemical analytes in a group of clinically healthy dogs owned and housed privately. METHODS: A prospective cohort study was conducted. Serum samples obtained from clinically healthy adult dogs at varying intervals over a 12-week period were analyzed. For each sample, a panel of 14 analytes was measured. Three levels of outlier analyses were applied (analytical, intra-individual, and inter-individual), followed by nested ANOVA to calculate intra-individual, inter-individual, and analytical coefficients of variation (CV(I), CV(G), and CV(A), respectively). RESULTS: Specimens from 11 dogs were analyzed. Individuality indices ranged from 0.9 for glucose and total triglyceride concentrations to 3.4 for ALT activity. Analytical variation (CV(A) ) was > ½ CV(I) for 9/14 analytes, failing to meet criteria for acceptable analytical variation when defining critical change values. Where analyzer performance was acceptable, critical change values ranged from 26.4% for glucose concentration to 84.0% for total triglyceride concentration. CONCLUSIONS: Many frequently measured analytes included in routine serum biochemical panels have high individuality. Thus, use of standard reference intervals to monitor changes over time in an individual is likely to miss meaningful biological change.