Measurement Variation and Repeatability of Echocardiographic Variables Used to Estimate Pulmonary Artery Pressure in Dogs.

BACKGROUND: Pulmonary hypertension (PH) is an important clinical entity in dogs. Pulmonary hypertension is generally identified based on echocardiographic findings, but measurement variation of the variables most often used in this context has not been reported. OBJECTIVE: To define measurement variation of echocardiographic variables that are used to estimate pulmonary artery pressure. ANIMALS: Four dogs with subclinical tricuspid valve regurgitation were selected from a hospital population. METHODS: Prospective descriptive study - dogs were subject to repeated echocardiographic examination by 2 operators. Each dog was echocardiographically examined 12 times, once by each operator during mornings and afternoons of 3 nonconsecutive days. Measurement variation was defined by coefficients of variation and by mean square errors (MSE) from ANOVA models. Repeatability coefficients (RC) were calculated from MSE. RESULTS: Of 50 within-day, between-day, and interoperator coefficients of variation (CVs), 37 were less than 10% and 46 were less than 15%. The range of peak velocity of tricuspid valve regurgitation (VTR) VTR was 1.87-3.34 m/s. Analysis of variance revealed a significant effect (P < 0.05) of operator for 7 of 10 measurements. Within-operator/between-day/within-subject and between-operator/between-day/within-subject RCs of VTR were, respectively, 0.3 and 0.6 m/s. CONCLUSIONS AND CLINICAL IMPORTANCE: With regard to repeatability, VTR is superior to other methods used for echocardiographic estimation of systolic pulmonary artery pressure. When TR is absent, the acceleration time of pulmonary ejection and the acceleration-to-ejection time ratio are alternative measures. A change in VTR exceeding 0.3 m/s might be outside the range of expected within-operator variation.

Conventional echocardiographic assessment of the canine right heart: reference intervals and repeatability.

INTRODUCTION: We sought to define reference intervals for echocardiographic dimensions of the canine right heart. Secondarily, we intended to describe measurement variability of these dimensions. ANIMALS: Forty-five healthy adult pet dogs of diverse somatotype. MATERIALS AND METHODS: To obtain normative data used to define reference intervals, dogs underwent one echocardiographic examination by the same operator. Raw data were subject to logarithmic (log10) transformation and allometric relationships between body weight and echocardiographic variables were determined; reference intervals intended to include 95% of the population were defined. Six of the 45 dogs were randomly selected and subject to repeated echocardiographic examination by two operators. Measurement variability was quantified through calculation of coefficients of variation and repeatability coefficients. RESULTS: The strength of the linear relationships between log10 transformed echocardiographic variables and body weight varied; the range of coefficients of determination (R(2)) was 0.055-0.872. For most variables the scaling exponents were close to values expected based on presumed allometric relationships. Of the 190 within-day, between-day and inter-operator coefficients of variation generated, 165 (87%) were less than 15%. Analysis of variance revealed the operator to be a significant source of variation for 25 of 38 measurements. CONCLUSIONS: Reference intervals for echocardiographic dimensions and indices of right heart function are proposed. Repeatability of selected linear dimensions and areas obtained from two-dimensional echocardiography likely is sufficient for longitudinal clinical evaluations.