Observer-dependent variability of quantitative clinical endpoints: the example of canine echocardiography


Herve P. Lefebvre, UMR 181 Physiopathologie et Toxicologie Expérimentales, Ecole Nationale Vétérinaire de Toulouse, 23 chemin des Capelles, 31076 Toulouse cedex 03, France. E-mail:


The objective of this study was to determine the intra- and inter-observer variability of echocardiographic measurements in dogs. Four observers with different levels of experience in echocardiography performed 192 echocardiographic examinations of six dogs on four different days. The lowest within- and between-day coefficients of variation (CV) (%) were 13.8 and 5.2 for the right ventricle in diastole, 8.9 and 4.5 for the interventricular septal thickness in diastole (6.3 and 7.0 in systole), 7.7 and 9.4 for the left ventricular free-wall thickness in diastole (8.1 and 5.2 in systole), 3.1 and 5.0 for the left ventricular end-diastolic diameter (6.2 and 7.0 for end-systolic diameter), 10.2 and 10.8 for the left ventricular shortening fraction, and 8.2 and 9.8 for the left atrium/aorta ratio, respectively. Most of these lowest CVs were observed by the two most experienced observers. Conversely, all maximum values were obtained with the two less experienced observers. These differences in observer-dependent variability may considerably influence the minimum number of animals required to detect a treatment-associated change in echocardiographic variables.