Test-retest reliability of assessment for intraventricular dyssynchrony by tissue Doppler imaging echocardiography

The purpose of this study was to evaluate the test-retest reliability of intraventricular dyssynchrony (IVD) assessment by tissue Doppler imaging (TDI) echocardiography. Limited response rates to cardiac resynchronization may improve with TDI screening for appropriate recipients. However, the clinical applicability of TDI will depend on the reliability of the test. Repeat TDI was prospectively performed (11 +/- 11 days apart) in 15 patients with QRS intervals >120 ms and left ventricular ejection fractions <35% and 25 normal controls using the same machine, sonographer, and blinded readers for the 2 tests. There was no change in clinical status or treatment between tests. Established and clinically feasible criteria for IVD were evaluated. These were based on differences of TDI-derived activation time between 2, 4, or 12 left ventricular segments. Reliability was assessed by linear correlation and Bland-Altman analysis for quantified measures, along with percentage agreement and kappa statistics for IVD diagnosis. Despite good intrareader (r = 0.98, p <0.0001) and interreader (2 segments: r = 0.96, p <0.0001; 4 segments: r = 0.85, p <0.0001) reliability, test-retest correlations were uniformly modest for the 2-segment (r = 0.26, p = 0.11), 4-segment (r = 0.36, p = 0.021), and 12-segment (r = 0.50, p = 0.0009) measures. Test-retest agreement for IVD diagnosis by either criterion was equally limited (2 segments: 83%, kappa = 0.27; 4 segments: 75%, kappa = 0.47; 12 segments: 68%, kappa = 0.35). Bland-Altman analysis demonstrated wide confidence intervals, exceeding the diagnostic cutoff values for the respective criteria. In conclusion, the accurate assessment of IVD by TDI may be limited by poor test-retest reliability.