Tissue Doppler Imaging Measurement of Left Ventricular Systolic Function in Children: Mitral Annular Displacement Index Is Superior to Peak Velocity

Background Doppler tissue imaging (DTI)–derived mitral annular systolic peak S-wave velocity (S′) correlates with left ventricular (LV) ejection fraction (EF). The authors hypothesized that DTI mitral annular displacement, which is equal to the velocity-time integral of the DTI S′ wave, might be superior to S′ to analyze LV systolic function. Because S′ varies with age, it was expressed as S z , the z -score variance from normal S′ for each subject. Because displacement varies with heart size, it was expressed as a displacement index, or the DTI S′-wave velocity-time integral divided by the end-diastolic distance from the mitral annulus to the LV apex. The aims of this study were to (1) measure the accuracy, sensitivity, specificity, and positive and negative predictive values of displacement index compared with S z to detect systolic dysfunction; (2) compare displacement index with other quantitative parameters of longitudinal systolic function, including color DTI–derived strain and two-dimensional speckle-tracking echocardiography (2D)–derived mitral annular displacement and strain; and (3) determine the effects of age, heart rate (HR), and body surface area (BSA) on displacement index. Methods Displacement index and S z results were compared with EF and with each other using statistical tests, including independent t tests, linear regression, receiver operating characteristic curve analysis, and 2 × 2 probability tables. Displacement index was also compared with other parameters of longitudinal systolic function, age, HR, and BSA using regression analysis. Results Forty-six patients had normal (EF ≥ 55%) and 34 abnormal (EF < 55%) LV function. Groups were statistically equivalent ( P > .05) for age, HR, and BSA and statistically different ( P < .001) for all measured parameters of systolic function. Displacement index and EF were linearly related. Receiver operating characteristic curve analysis showed the sensitivity of displacement index to be greater than that of S z throughout the study range. Probability table analysis demonstrated that for predicting EF < 55%, the sensitivity, accuracy, and negative predictive value were greater for displacement index than for S z . Displacement index was linearly correlated with 2D mitral annular longitudinal displacement, 2D LV basal segment longitudinal strain, and color DTI LV basal segment longitudinal strain. Displacement index was not affected by age, HR, or BSA. Conclusions Displacement index is linearly