Reproducibility of the 2016 American Society of Echocardiography-European Association of Cardiovascular Imaging Algorithm for Estimation of Left Ventricular Filling Pattern: Not Perfect but Good Enough.

Abstract

The 2016 American Society of Echocardiography (ASE) and European Association of Cardiovascular Imaging (EACVI) guidelines for the evaluation of left ventricular diastolic function reported a new algorithm to assess diastolic function and to estimate left ventricular filling pressure (LVFP). At least five to six different parameters were necessary to conclude, each of them with their own inter-observer variability. This article examines the reproducibility of each parameter of the algorithm and its influence on the final decision of the clinician. Echocardiographic exams of 12 non-selected patients without any known cardiac disease or follow-up but addressed to the hospital for symptoms were analyzed by two readers (one junior and one senior) in five French cardiologic tertiary centers. Inter-observer reproducibility at each step of the algorithm and final decision were analyzed. There was mild agreement on the final decision. The main reasons of discrepancy were disagreement on the significance of mitral annular calcifications and measured values that are just around the cut-off (despite good reproducibility, a slight variation could lead to misclassification of a dichotomous choice between a normal measure and a pathologic measure). Without considering performance, this multicentric French study puts forward limits to the actual algorithm recommended for LVFP pattern assessment. Agreement is excellent in caricatural (easy) cases (left ventricular pressure clearly normal or clearly elevated) but a great discordance exists in the gray zone. Improvement in the algorithm and in the method for LVFP determination is proposed.