PubAg

Main content area

Assessment of method reliability and determination of reference intervals for rotational thromboelastometry in horses

Author:
Junge, Hannah K., Ringer, Simone K., Mayer, Nathalie, Schwarzwald, Colin C.
Source:
Journal of veterinary emergency and critical care 2016 v.26 no.5 pp. 691-703
ISSN:
1479-3261
Subject:
adults, blood, blood platelet count, blood sampling, coagulation, fibrinogen, hematocrit, horses, hospitals, observational studies, protein content, prothrombin, regression analysis, thrombin, thromboplastin
Abstract:
OBJECTIVES: To assess the measurement reliability of rotational thromboelastometry (ROTEM) measurements in horses, establish reference intervals for healthy horses, and evaluate the relationship between ROTEM variables, hematologic variables, and standard coagulation tests. DESIGN: Prospective observational study. SETTING: University teaching hospital. ANIMALS: Fifty healthy and 10 diseased adult horses. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Blood was sampled from 10 healthy and 10 diseased horses and samples were repeatedly analyzed to evaluate measurement reliability of various ROTEM variables. Four different ROTEM assays (ie, EXTEM, INTEM, FIBTEM, and APTEM) were run simultaneously under standardized conditions. The device‐related, operator‐related, and day‐to‐day variability for the majority of ROTEM variables was very low to low, as indicated by a coefficient of variation (CV) of < 15%. Most of test‐retest variability of ROTEM variables appeared to be device‐related. Blood samples from 50 clinically healthy horses were used to establish reference intervals for ROTEM variables. Multiple stepwise regression analyses identified associations of different ROTEM variables with hematocrit, total protein concentration, fibrinogen concentration, platelet count, prothrombin time, activated partial thromboplastin time, and thrombin time. CONCLUSIONS: ROTEM is a feasible method to evaluate coagulation in horses. Its measurement reliability is acceptable, but device‐related measurement variability has to be considered. Reference intervals are presented, but the influence of hematocrit, platelet count, and fibrinogen concentration may need to be taken into account when interpreting individual test results.
Agid:
5881083