RESUMEN
BACKGROUND: Cardiac structure and function in dogs are commonly assessed using echocardiography. A variety of linear, area, and flow-based measurements can be used to calculate left ventricular (LV) total stroke volume (TSV) and forward stroke volume (FSV), but the reproducibility of many of these measurements has not been fully studied. We hypothesized that survey of echocardiographic variables would identify those with high reproducibility and inform future investigation of different methods to measure LV TSV and FSV. METHODS: The reproducibility of 25 geometric and flow-based echocardiographic measurements was prospectively evaluated in 23 healthy dogs by two experienced observers. Reproducibility (i.e., interobserver agreement) was described using intraclass correlation coefficients. The reproducibility of various methods to calculate LV TSV and FSV was explored. RESULTS: Reproducibility was generally good to excellent. Variables of LV width, length, and area and aortic and sinotubular junction diameter and velocity time integral were among measures with the highest reproducibility. Measurements of mitral annular diameter and mitral inflow velocity time integral possessed lower reproducibility. Calculation of LV TSV using measurements involved in the cube and bullet formulas demonstrated higher reproducibility than the Simpson's method of disks or mitral inflow methods. Calculation of LV FSV using LV outflow tract and aortic diameters from the right parasternal view generally demonstrated higher reproducibility compared with the left-sided view. CONCLUSIONS: The reproducibility of many simple geometric and flow-based echocardiographic measurements is high. Comparison of the reliability of different measurement informs future investigation of echocardiographic methods to determine LV TSV and FSV in dogs.