Estimation of left ventricular chamber and stroke volume by limited M-mode echocardiography and validation by two-dimensional and Doppler echocardiography.

This study has been designed to improve estimation of stroke volume from linear left ventricular (LV) dimensions measured by M-mode echocardiography, in symmetrically contracting ventricles. In experimental studies, the ratio of LV epicardial long/short axes "Z" is about 1.3. We measured systolic and diastolic epicardial long and short axes by 2-dimensional echocardiography in 115 adults with widely varying LV short-axis dimensions (LV end-diastolic dimension = 3.95 to 8.3 cm). In a learning series of 23 normotensive and 27 hypertensive subjects, Z(diastole) was 1.3 +/- 0.1 and Z(systole) = 1.2 +/- 0.1, similar to findings in experimental animals. Regression equations were developed by comparing LV volumes by M-mode and 2-dimensional echocardiography. In a test series (65 subjects), LV volumes were calculated using separate regression equations for end-diastolic volume ([LV end-diastolic dimension] 4.765 - 0.288 x posterior wall thickness]) and for end-systolic volume ([LV end-systolic dimension] [4.136 - 0.288 x posterior wall thickness]). Because the term 0.288 x wall thickness was only about 8% of the first term between brackets, the average wall thickness in the learning series was substituted in the Z-volume formulas applied to the test series: end-diastolic volume = (4.5 x [LV end-diastolic dimensions]2) and end-systolic volume = (3.72 x [LV end-diastolic dimension]2). The mean relative error produced with this simplified method was 0.9%. in diastole and 1.4% in systole. Compared with Teichholz' M-mode volume method, Z-derived end-diastolic volume in the test series was equally well related to 2-dimensional volumes (both r = 0.88), with a better intercept (1.5 vs -23 ml, p <0.001) and a slope closer to the identity line (1.1 vs 1.4). Similar results were found for systolic volumes. In a second test series of 1,721 American Indian participants in the Strong Heart Study without mitral regurgitation or segmental LV wall motion abnormalities, Doppler-derived LV stroke volume (70 +/- 14 ml/beat) was similarly predicted by the Z-derived method (r = 0.65, 70 +/- 11 ml/beat) and Teichholz formulas (r = 0.64, 72 +/- 13 ml/beat), but Z-derived volumes had a regression line significantly closer to the identity line (p <0.005). Thus, LV chamber and stroke volumes can be determined from M-mode LV diameters over a wide range of LV sizes and in epidemiologic as well as clinical populations. The performance of this new method appears better than that obtained using the Teichholz formula, with a formula that is easy to handle and makes calculation of LV volumes by pocket calculator possible, even from limited echocardiographic studies.