Reproducibility of Doppler echocardiographic assessment of left ventricular diastolic function in multicenter setting

Whether Doppler echocardiography allows reliable assessment of left ventricular (LV) diastolic parameters in multicenter studies is disputed.

To assess interstudy reproducibility of Doppler echocardiographic parameters of LV filling and relaxation in multicenter setting, two laboratories assessed Doppler parameters of LV filling and relaxation twice within 24 hours in 56 participants. The study sample included hospitalized patients, clinically healthy outpatients, and healthy volunteers. All studies were performed according to a standardized protocol, and read by a single reader.

In the whole study sample the reproducibility of the assessment of the ratio between the peak velocities of the early (E) and late (A) LV filling waves (E/A), of the E wave deceleration time and atrial filling fraction, measured at the mitral valve leaflet tips, was good (intraclass correlation coefficients, rho: 0.89-0.76); the reproducibility performance of the same parameters measured at the mitral anulus was less good. The reproducibility of the assessment of the isovolumic relaxation time was good (rho: 0.81), and that of the E-wave propagation rate (E-Vp), by color M-mode, was excellent (rho: 0.96). The reproducibility of the ratio E wave peak velocity/E-Vp was also fairly good (rho: 0.87). The classification of E-Vp < 45 cm/s and E-Vp > or = 55 cm/s showed the highest between-study reliability performance (agreement index kappa = 0.92 and 0.83, respectively, both P < .001), followed by E/A at mitral leaflet tips >1.5 or <0.8 (kappa = 0.78 and 0.74, respectively, both P < .01), whereas the worst reliability performance was shown by isovolumic relaxation time < 60 milliseconds (kappa = 0.05, P = not significant).

Standardized procedures for imaging and readings allowed good reliability of serial Doppler echocardiographic evaluations of LV filling and relaxation parameters in multicenter setting.