Estimation of left ventricular systolic function by nonvolumetric echocardiographic analysis in subjects with poor left ventricular visualization: a pilot study.

Background: Accurate assessment of left ventricular (LV) systolic function is an essential requirement in clinical cardiology. Several echocardiographic methods provide quantitative analysis of LV volumes and ejection fraction (EF) based on the precise tracing of endocardial borders. Often, however, technically limited studies prohibit such direct analysis, and alternative techniques must be applied.

Hypothesis: Nonvolumetric echocardiographic methods which do not require endocardial edge definition and tracing may accurately provide quantitative LV systolic function data.

Methods: A pilot study was conducted to validate and compare two recently described indirect echocardiographic methods of LV systolic function analysis, with LVEF by radionuclear cardiac angiography (RNCA). Thirty-two consecutive patients undergoing RNCA for clinical indications also underwent echocardiography within 24 h, with LV analysis performed by the techniques of (1) atrioventricular plane displacement (AVPD) and (2) mitral valve leaflet coaptation point to interventricular septum distance at end-systole (MVC-IVS).

Results: Thirteen patients had an echocardiogram with poor two-dimensional visualization of LV endocardial borders. One patient could not be evaluated by the MVC-IVS method and two others by the AVPD method because of technical limitations. Chi-square analysis to compare how each method could discriminate between an RNCA LVEF of < or > or = 50% demonstrated high correlations for the AVPD method (r = 0.6530, p < 0.0005) and the MVC-IVS method (r = -0.7029, p < 0.0001). Sensitivity, specificity, positive and negative predictive values, and test accuracy for the AVPD and MVC-IVS methods were 85 and 80%, 88 and 94%, 85 and 92%, 82 and 83%, and 83 and 87%, respectively.

Conclusion: This pilot study demonstrates that both alternative echocardiographic methods may be useful in the assessment of LV systolic performance, even in the setting of poor LV endocardial border visualization. A larger study is warranted to apply and contrast these methods in different patient subsets.