Author + information
- Ramin Farzaneh-Far, MD and
- Nelson B. Schiller, MD, FACC⁎ ()
- ↵⁎Box 0214, Department of Medicine, University of California, San Francisco, California 94143-0214
We read with interest the recent study by Bangalore et al. (1) evaluating the prognostic significance of left atrial (LA) enlargement in patients undergoing stress echocardiography. The authors demonstrate that left atrial dimension provides incremental prognostic significance in patients referred for stress echocardiography, irrespective of the presence of inducible ischemia. Although we applaud the authors for highlighting the importance of left atrial enlargement in this patient population, there are several methodologic concerns which profoundly limit the applicability of their findings to clinical practice.
First, the use of M-mode linear dimension to measure left atrial volume is inaccurate and varies widely among individual readers (2). We have previously shown that left atrial volume is most accurately estimated by 2-dimensional (2D) echocardiography using biplane methods (area–length or method of discs) (3); this approach is recommended by the American Society of Echocardiography for use in clinical trials (4). Left atrial dimension consistently underestimates left atrial volume and is therefore a specific but insensitive test for left atrial enlargement (5). As such, a significant proportion of patients with true left atrial enlargement were likely categorized as normal. The authors’ qualification that “unidimensional measurement is still the most common method worldwide to quantify LA size,” although true, is not adequate justification for the investigative use of an inferior measurement. A suitable analogy would be the use of urinary dipstick testing rather than serum glucose measurement to determine the prevalence of diabetes in a study population.
Second, the resting ejection fraction used in the study analysis was based on visual estimation or “eyeballing.” This technique is not only inaccurate in determining ejection fraction compared with 2D measurement techniques, but also suffers from wide interobserver variability and poor reproducibility (6). Moreover, the mean ejection fraction in the patients with dilated left atria was 48%, suggesting established systolic dysfunction at baseline rather than “relatively preserved [left ventricular] ejection fraction,” as the authors contend.
Third, the authors did not report or correct for Doppler indices of diastolic dysfunction in the multivariate analysis. If, as they suggest, left atrial size is a marker of the severity and duration of diastolic function, then the independent prognostic value of left atrial enlargement cannot be established without taking diastolic dysfunction into account.
Given these methodologic concerns, the authors’ conclusion that left atrial size should be routinely incorporated in the prognostic interpretation of stress testing is not justified and furthermore would be unlikely to impact clinical decision making. For example, in the presence of a positive stress echocardiogram it is doubtful that coronary angiography would be averted because of the single measurement of a small left atrial dimension. Conversely an enlarged left atrial dimension in the setting of a normal stress echocardiogram is of unclear significance and would not, on its own, merit further invasive workup.
- American College of Cardiology Foundation
- Bangalore S.,
- Yao S.,
- Chaudhry F.
- Abhayaratna W.P.,
- Seward J.B.,
- Appleton C.P.,
- et al.
- Tsang T.S.M.,
- Abhayaratna W.P.,
- Barnes M.E.,
- et al.