Feasibility and reproducibility of left ventricular rotation by speckle tracking echocardiography in elderly individuals and the impact of different software.

Abstract

BackgroundChanges in ventricular rotation measured by two-dimensional speckle tracking echocardiography (2DSTE) are early indicators of cardiac disease. Data on the clinical feasibility of this important measure are scarce and there is no information on the comparability of different software versions. We assessed the feasibility, reproducibility and within patient temporal variability of 2DSTE in a large community based sample of older adults. We additionally compared 2DSTE results to those generated by 3DSTE.Methods and results1408 participants underwent transthoracic echocardiography. Using Philips Qlab 8.1 peak LV rotation at either the base or the apex was analysable in 432 (31%) participants. Peak twist measurements were achieved in 274 (20%) participants. 66 participants were randomly selected for the reproducibility study. 20 additional participants had scans 4-6 weeks apart for temporal variability and 3D echocardiography to assess the agreement between 2DSTE and 3DSTE. Reproducibility was evaluated using the intraclass coefficient of correlation (ICC). Better reproducibility for rotation and twist were obtained when measured at the endocardium, and when using more recent software versions, Peak twist and rotation were significantly different using two versions of the same software. Agreement with 3DSTE was better using newer software.ConclusionFeasibility of 2DSTE is low in this cohort of elderly individuals severely limiting its utility in clinical settings. However if high quality images can be acquired assessment of ventricular rotation by 2DSTE is reproducible. Caution should be taken when comparing measurements of ventricular rotation by software from different vendors or different versions of software from the same vendor.