Intra- and interobserver reproducibility of left ventricular mechanical dyssynchrony assessment by real time...

Real time three-dimensional (3D) echocardiography allows the assessment of left ventricular (LV) mechanical dyssynchrony and may be useful in predicting response to cardiac resynchronization therapy. However, reproducibility of 3D dyssynchrony in past reports varied widely. We evaluated intra- and interobserver reproducibility of parameters of LV mechanical dyssynchrony by 3D echocardiography and explored the impact of image quality as a possible source of variability.

LV volumes and ejection fraction (LV EF) were measured by 3D echocardiography in 90 subjects. LV mechanical dyssynchrony was calculated as the standard deviation of the time-to-minimum segmental volume (Tmsv) over 16 (SDI16) and 12 (SDI12) segments and as the maximum time-difference in Tmsv over 16 (Dif16), 12 (Dif12), and 2 (DifS-L) segments. Opposing wall delay in peak myocardial systolic velocity (S-L delay) by tissue Doppler imaging (TDI) was also evaluated.

Feasibility of 3D measurements was 88.9%. Intra- and interobserver intraclass correlation coefficients were excellent for LV volumes, LV EF, and S-L delay (all ≥ 0.90), very good for SDI16 (0.83 and 0.85), moderate to good for SDI12, Dif16, Dif12, and DifS-L (0.51-0.81). No systematic bias was present between readings for 3D dyssynchrony, but limits of agreement resulted fairly large for most parameters with the exception of SDI16, which showed reproducibility slightly lower than TDI. Suboptimal image quality was a significant source of variability in 3D-dyssynchrony assessment.

The 3D assessment of LV volumes and dyssynchrony was feasible in our community-based cohort. SDI16 was the most reproducible among 3D-dyssynchrony parameters and the least affected by image quality.