Search

Background: The incremental role of bedside tissue Doppler echocardiography and B-type natriuretic peptide (BNP) over the clinical judgment has been recently reported in the emergency diagnosis of congestive heart failure with a normal left ventricular ejection fraction (HFNEF). However, how well does this diagnostic strategy be applicable in the setting of atrial fibrillation is unknown., Objective: To investigate the usefulness of bedside tissue Doppler echocardiography and BNP in the emergency diagnosis of HFNEF in elderly patients with permanent, nonvalvular atrial fibrillation., Methods: Forty-one consecutive elderly patients with an ejection fraction > or =50% (mean age 84 years; 22 with HFNEF and 19 with noncardiac cause), hospitalized for acute dyspnea at rest, were prospectively enrolled; bedside septal E/Ea and BNP were obtained at admission., Results: By multivariable logistic regression analysis including the clinical judgment of heart failure, E/Ea and BNP, E/Ea (P = 0.014) and BNP (P = 0.018) provided independent diagnostic information. Optimal cutoffs were 13 for E/Ea (area under the ROC curve of 0.846, P < 0.0001; sensitivity 81.8%, specificity 89.5%) and 253 pg/ml for BNP (area under the ROC curve of 0.928, P < 0.0001; sensitivity 86.4%, specificity 89.5%). The concordance between the clinical judgment and BNP concentration at the cutoff of 253 pg/ml correctly classified 24 of 25 patients; E/Ea at the cutoff of 13 correctly classified 14 of the 16 patients with discrepancy., Conclusion: Bedside tissue Doppler echocardiography and BNP provide useful additional diagnostic information over the clinical judgment for the emergency diagnosis of HFNEF in elderly patients with permanent, nonvalvular atrial fibrillation.

Background: Based on the hypothesis that it reflects left ventricular (LV) diastolic pressures, B-type natriuretic peptide (BNP) is largely utilized as first-line diagnostic complement in the emergency diagnosis of congestive heart failure (HF). The incremental diagnostic value of tissue Doppler echocardiography, a reliable noninvasive estimate of LV filling pressures, has been reported in patients with preserved LV ejection fraction and discrepancy between BNP levels and the clinical judgment, however, its clinical validity in such patients in the presence of BNP concentrations in the midrange, which may reflect intermediate, nondiagnostic levels of LV filling pressures, is unknown., Methods: 34 patients without history of HF, presenting with acute dyspnea at rest, BNP levels of 100-400 pg/ml and normal LV ejection fraction were prospectively enrolled (17 with congestive HF and 17 with noncardiac cause). Tissue Doppler echocardiography was performed within 3 hours after admission., Results: Unlike BNP (P = 0.78), Boston criteria (P = 0.0129), radiographic pulmonary edema (P = 0.0036) and average E/Ea ratio (P = 0.0032) were predictive of congestive HF by logistic regression analysis. In this clinical setting, radiographic pulmonary edema had a positive predictive value of 80% in the diagnosis of congestive HF. In patients without evidence of radiographic pulmonary edema, average E/Ea > 10 was a powerful predictor of congestive HF (area under the ROC curve of 0.886, P < 0.001, sensitivity 100% and specificity 78.6%)., Conclusion: By better reflecting LV filling pressures, bedside tissue Doppler echocardiography accurately differentiates congestive HF from noncardiac cause in dyspneic patients with intermediate, nondiagnostic BNP levels and normal LV ejection fraction.

The accuracy of the tissue Doppler E/Ea and color M-mode Doppler E/Vp indexes in diagnosing congestive heart failure (HF) was investigated in 20 chronic hypertensive patients with acute dyspnea and normal left ventricular ejection fractions who met Vasan's criteria for definite diastolic HF, compared with 20 gender- and age-matched hypertensive patients with noncardiac cause of acute dyspnea. The E/Ea ratio appeared to be more reproducible (variability 4% to 9% vs 6 to 14%) and more precise (sensitivity 77.8%, specificity 100%, and accuracy 89.5% for the optimal cutoff of 11 vs sensitivity 73.7%, specificity 75%, and accuracy 74.3% for the optimal cutoff of 1.5) than the E/Vp ratio in this clinical setting.