The Quantitative assessment of beat to beat and surgically induced changes in left ventricular segmental wall motion.

Many approaches have been used to quantitate the information concerning left ventricular (LV) function which can be obtained from angiography. Such quantitative descriptors are needed to effectively evaluate changes in LV function which might result from investigative or therapeutic procedures and to generate prognostic data. One descriptor which has been widely used is segmental percent shortening. This parameter has been found sufficiently sensitive to measure differences in LV function between ventricles of normal and diseased patient populations and has been successfully employed to assess changes in LV function following several types of interventions in groups of patients. However, in spite of the extensive application of this parameter, there has been limited study of its reproducibility. Because reproducibility is so fundamental to the conclusions being drawn from observed changes in LV function following an intervention, it is the purpose of this work to test whether segmental percent shortening and several other quantitative descriptors of left ventricular function exhibit beat to beat variability. The final goal is to test the parameters which exhibit beat to beat reproducibility to see if they are sensitive enough to differentiate between segmental motion in normal and diseased ventricles and to see if these parameters can be used to measure changes in LV function following the specific therapeutic intervention of coronary artery surgery. To evaluate whether beat to beat variability was significant, a multiple factor analysis of variance (ANOVA) model was used to test each of the quantitative descriptors of LV function noted above. Segmental percent shortening was found to differ significantly (p < .05) from beat to beat. On the other hand, mean normalized regional velocity during each quarter of systole and diastole, was reproducible from beat to beat (p < .05). Also, path lengths during each quarter of systole and diastole and their ratios were found to be free of significant beat to beat variability (p < .05). None of the other quantitative descriptors which were studied were reproducible from beat to beat (p < .05). To evaluate the sensitivity of the parameters found to be reproducible from beat to beat two types of tests were employed. First, a multiple factor ANOVA model was used to test whether a significant difference in parameter values could be demonstrated between normal and diseased patients and whether these parameters changed after coronary artery surgery. Second, nonparametric tests were employed to further evaluate these differences in patient groups with specific types of anatomical disease. Mean normalized regional velocity was the parameter most sensitive to changes in LV function following surgery. In summary, segmental percent shortening was found to vary significantly from beat to beat while mean normalized regional velocity was reproducible and sensitive to the state of the ventricle for groups of patients. Segmental percent shortening may be an inappropriate descriptor of left ventricular function for use in individual patients because of this beat to beat variability even though it is useful in evaluating changes in groups of patients. Further work is required to study whether the reproducibility of mean normalized velocity evident in the beat to beat comparisons will also be found in angiogram to angiogram comparisons.