Biological variation of thromboelastrography variables in 10 clinically healthy horses.

Abstract

Objective: To assess the utility of population-based reference intervals (PRIs) for interpreting thromboelastography (TEG) variables in horses using biological variation data.

Design: Prospective cohort biologic variation study conducted over a 5-week period.

Setting: Veterinary teaching hospital and research facility.

Animals: Ten clinically healthy horses randomly selected from a veterinary school research and teaching herd.

Interventions: Horse health was determined using physical examination, CBC, and biochemical and coagulation profiles prior to the start of the study. Subsequently, once weekly blood sampling for TEG testing was performed for 5 weeks.

Measurements and main results: The 4 TEG variables reaction time (R), clot formation time (K), angle, and maximum amplitude (MA) were measured, and coefficient of variation representing within- and between-horse biological variation (CVi and CVg , respectively) and coefficient of variation representing analytical variation (CVa ) were calculated using a nested ANOVA after removing outlier data. The CVi , CVg , and CVa for R were 26.8%, 5.2%, and 5.9%; for K were 31.0%, 0.0%, and 5.9%; for angle were 9.4%, 6.2%, and 21.7%; and for MA were 3.4%, 4.1%, and 4.4%, respectively. Index of individuality (IOI) was then calculated for each variable using the formula {( CVi² + CVa²/CVg²)}¹/². IOI for R was 5.3, for angle was 3.8, and for MA was 1.4; IOI was not assessed for K.

Conclusions: PRIs are appropriate for TEG variables, R, angle, and MA when interpreting results from individual horses based on calculated IOI values equal to or greater than 1.4. PRIs are likely appropriate when interpreting K, but IOI could not be calculated for this variable.