Quantitative estimation of left atrial linear dimension on a transthoracic echocardiogram using an...

A proportionate and incremental association between left atrial (LA) dilatation on a transthoracic echocardiogram (TTE) and increasing severity in abnormal atrial depolarization can be described by the regression formula: LA dimension in parasternal long-axis view (in millimeters) = 2.47 + 0.29 [P-wave duration (in milliseconds)]. However, prospective testing of this formula for LA quantification with comparison to other TTE indexes is lacking. We prospectively obtained 12-lead electrocardiograms for 72 consecutive patients before individual, nonemergent TTE assessment. P waves were assessed independently to the nearest 10 ms for application of the formula with a Bland-Altman plot. P-wave durations were obtained specifically from lead II and also from any lead that yielded the widest measurement. There was a strong formulaic correlation with LA measurement by TTE (p <0.00000001; r = 0.662), irrespective of electrocardiographic lead used. However, as the measurement difference increased between that in any lead and lead II, correlation strength increased. Correlation was not significantly affected by commonly measured electrocardiographic and TTE indexes, such as, PR interval, QRS complex, and left ventricular end-diastolic as well as systolic dimensions, after adjustment for these variables. Moreover, when the P-wave axis remained within the normal range, the correlation strength increased. The Bland-Altman plot also showed good agreement of LA dimension assessment between formulaic estimation and TTE measurement. In conclusion, there is good agreement and correlation between formulaic estimation and that of TTE for measurement of LA linear dimension. The LA regression formula is an indirect asset that could perhaps supplement LA quantification on TTE in certain circumstances.