Echocardiographic assessment of simultaneously measured left ventricular filling pressures in patients with normal left ventricular ejection fraction

2020 
BACKGROUND Assessment of left ventricular (LV) diastolic function is part of routine echocardiographic examinations. Accuracy of the 2016 ASE/EACVI algorithm to detect LV diastolic dysfunction in patients with a normal LV ejection fraction (LVEF) has been examined but simultaneous measurements of LV pressures and echocardiographic parameters of diastolic function are sparse. METHODS The accuracy of multiple echo parameters of diastolic dynamics and the 2016 guidelines were assessed by simultaneous transthoracic echocardiography and LV pressure recordings in 120 patients (derivation cohort) and 60 patients (validation cohort) with suspected coronary artery disease and normal LVEF. Receiver-operating characteristic (ROC) curves defined optimal cut points for each echocardiographic parameter. A new algorithm was proposed to estimate LV diastolic pressures using 5 parameters based on ROC data: tricuspid regurgitation velocity >280cm/s, average e' 13, velocity of pulmonary vein A-wave reversal >32 cm/s, and left atrial volume index >32 mL/m2 . Performances of the 2016 guidelines and a proposed algorithm were examined for detecting a LV pre-A >12 and LV end-diastolic pressure (LVEDP) >15 mm Hg. RESULTS In the derivation cohort, the 2016 algorithm identified an elevated LVEDP >15 mm Hg with an accuracy of 74.2% (63.8-82.9); the modified algorithm improved accuracy to 86.0% (77.6-92.1), P < .05. In the validation cohort, the proposed algorithm improved sensitivities with accuracies remaining like the 2016 algorithm. CONCLUSIONS LV diastolic pressures in patients with normal LVEF were reliably assessed by the 2016 guidelines. The proposed algorithm improved sensitivities and may improve the accuracies for detecting abnormal LV filling pressures.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    15
    References
    2
    Citations
    NaN
    KQI
    []