Feasibility, repeatability, and reproducibility of contemporary diastolic parameters and classification

To evaluate feasibility, time of acquisition, retest repeatability and reproducibility of echocardiographic indexes and classification algorithms of diastolic function.

A total of 356 patients were examined before coronary artery bypass-grafting and/or aortic valve surgery. A subgroup of 50 was examined with 3 successive echocardiograms in conditions reflecting daily clinical practice. Diastolic parameters were obtained and analysed according to previous (2009) and current (2016) guidelines. Acquisition and analysis time, plus intra- and inter-observer variability were assessed.

Feasibility of diastolic parameters was between 93 and 99%, except the maximal tricuspid regurgitation velocity (TR Vmax) (65%). Mean acquisition and analysis time were highest for left atrial volumes (141 ± 24 s) in contrast to other parameters which were obtained in approximately one minute. Mean 368 and 360 s were needed to classify diastolic function according to the 2009 and 2016 algorithms, respectively (non-significant). Reproducibility was overall moderate (Pearson r = 0.62 to 0.87), with TR Vmax having the highest (r = 0.62) and mitral valve E/A ratio the lowest (r = 0.87) variation. The 2009 algorithm resulted in more indeterminate cases than the 2016 algorithm. Inter-examiner analysis resulted in reclassification of 20 vs. 8 patients using the 2009 and 2016 algorithms, respectively.

Diastolic parameters are highly feasible and moderately reproducible, except TR Vmax. The 2016 algorithm is more restrictive than the 2009 algorithm in classifying patients with advanced stages of diastolic dysfunction. Time of acquisition according to the two guidelines is not significantly different.