Diagnostic Accuracy of an Integrated Echocardiographic Algorithm to Estimate Left Ventricular Filling Pressure

The purpose of this study was to investigate the reliability of an algorithmic classification system using invasively measured left atrial pressure (LAP) in a large-scale cohort. Although estimation of left ventricular filling pressure (LVFP) using an integrated echocardiographic algorithm is recommended, the usefulness of this algorithm has not been fully validated. The authors enrolled 1,967 patients (age 68 ± 10 years) whose LAP was directly measured within the left atrium during catheter ablation for atrial fibrillation. Patients were classified into 3 groups based on the echocardiographic algorithm: normal (group N, n = 1,282), undetermined (group U, n = 160), and elevated (group E, n = 346) LAP groups. Invasively measured LAP and echocardiographic parameters estimating LVFP were compared among the groups. The median LAP was 12.6 ± 5.7 mm Hg in the entire cohort. LAP was significantly higher in group E than that in the other groups (groups E vs U vs N, 14.2 ± 6.3 mm Hg vs 13.5 ± 5.9 mm Hg vs 12.0 ± 5.5 mm Hg; P < 0.001). Among group E patients, 43.1% had elevated LAP (≥15 mm Hg), whereas 56.9% had normal LAP (<15 mm Hg). Of the patients in group N, 69.0% had normal LAP, whereas 31% had elevated LAP. Although the correlation between invasively measured LAP and E/e′, peak tricuspid regurgitant velocity, and left atrial volume index was modest, the number of abnormal values correlated significantly with elevated LAP ( P < 0.001). The classification using combined echocardiographic parameters in the recommendations may be useful for detecting patients with normal LVFP but may be limited for detecting elevated LVFP.