Reproducibility of right ventricular function by longitudinal strain and other echocardiographic parameters in the ELSA-Brasil study.

Abstract

Right ventricular (RV) free wall longitudinal strain (RVFWLS), and four-chamber longitudinal strain (RV4CLS) using speckle tracking echocardiography have demonstrated increased accuracy and discrimination to measure right ventricular function in different clinical conditions. Reproducibility data of these measures are scarce and mainly tested in small or reference populations. The main objective of this study was to investigate their reproducibility, and of other traditional RV parameters, from unselected participants of a large cohort study. RV strain reproducibility was analyzed using echocardiographic images of 50 participants from a randomly selected sample from The ELSA-Brasil Cohort. Images were acquired and analyzed following the study protocols. The mean RVFWLS was -26.9 ± 2.6% and the mean RV4CLS was -24.4 ± 1.9%. The intra-observer reproducibility parameters of RVFWLS demonstrated a coefficient of variation (CV) of 5.1% and an intraclass correlation coefficient (ICC [95%CI] 0.78[0.67-0.89]), and for RV4CLS were CV = 5.1% and ICC = 0.78[0.67-0.89]. Reproducibility for RV fractional area change was CV = 12.1%; ICC = 0.66 [0.50-0.81] and for RV basal diameter was CV = 6.3%; ICC = 0.82 [0.73-0.91]. The inter-observer reproducibility for RVFWLS was CV = 8.3%; ICC 0.54[0.34-0.74] and for RV4CLS, CV = 6.3%; ICC = 0.53[0.34-0.73], following the same pattern among conventional RV parameters. We found adequate reproducibility of RV longitudinal strain parameters. This information is relevant for the long-term follow-up of cohort participants and reinforces the utility of RV longitudinal strain as a tool to monitor subclinical changes in RV systolic function.