Visual estimation of bedside echocardiographic ejection fraction by emergency physicians

The objective of this study was to determine whether bedside visual estimates of left ventricular systolic function (LVSF) by emergency physicians (EP) would agree with quantitative measurement of LVSF by the modified Simpson's method (MSM), as recommended by the American Society of Echocardiography.

After limited focused training, 2 trained EPs performed bedside echocardiography (BECH) procedures s between January and June 2012 to prospectively evaluate patients presenting to the emergency department (ED) with dyspnea. EPs categorized their visually estimated ejection fractions (VEF) as either low or normal. Formal echocardiography were ordered and performed by an experienced cardiologist using the MSM and accepted as the criterion standard. We compared BECH results for each EP using chi-squared testing and performed correlation analysis by Pearson correlation coefficient.

Of the 146 enrolled patients with dyspnea, 13 were excluded and 133 were included in the study. Comparison of EPs vs. cardiologist's estimate of ejection fraction yielded a Pearson's correlation coefficient of 0.77 (R, p<0.0001) and 0.78 (R, p<0.0001). Calculated biserial correlations using point-biserial correlation and z-scores were 1 (rb, p<0.0001) for both EPs. The agreement between EPs and the cardiologist was 0.861 and 0.876, respectively. The sensitivity, specificity, positive predictive value, negative predictive value, and the positive and negative likelihood ratios for each physician were 98.7-98.7%, 86.2-87.9%, 0.902-0.914, 0.980-0.981, 7.153-8.175, 0.015-0.015, respectively.

EPs with a focused training in limited BECH can assess LVSF accurately in the ED by visual estimation.