Biological variation of routine haematology and biochemistry measurands in the horse

BACKGROUND: Clinical pathology results are typically interpreted by referring to population-based reference intervals. The use of individualised (subject-based) reference intervals is more appropriate for measurands with a high degree of variation between individuals. OBJECTIVES: To determine the biological variation of routinely analysed equine haematology and biochemistry measurands and calculate indices of individuality and reference change values which enable production of individualised reference intervals, in a group of healthy, privately owned horses. STUDY DESIGN: In a prospective cohort study, thirty-nine privately owned horses were sampled by jugular venipuncture for analysis of haematology and biochemistry measurands at weekly intervals for 6 weeks. METHODS: Haematology was analysed on the day of collection. Serum was frozen and biochemistry analyses performed on thawed samples. Duplicate results were obtained and the coefficient of variation was calculated for analytical variation, within-subject variation and between-subject variation. The index of individuality and reference change value were derived for each measurand. RESULTS: Haematology (red blood cell count, mean corpuscular haemoglobin and mean cell volume) and biochemistry measurands (total protein, globulins, albumin, gamma-glutamyl transferase, aspartate aminotransferase) demonstrated high individuality, indicating that individualised reference intervals are more appropriate for evaluation of these measurands. Two haematology (mean corpuscular haemoglobin concentration and platelets) and three biochemistry measurands (chloride, glucose and sodium) had low individuality, indicating that the use of traditional population-based reference intervals is appropriate for these measurands. Remaining measurands had intermediate individuality suggesting interpretation of the reference change value should occur with consideration of the population-based reference interval. MAIN LIMITATIONS: The use of privately owned horses, variable management and environmental factors. CONCLUSIONS: The use of individualised reference intervals is justified for many measurands in horses, supporting the use of serial sampling, consideration of biological variation and application of reference change values for improved clinical decision making and patient management in equine practice.