A modified myocardial performance (Tei) index based on the use of valve clicks improves reproducibility of fetal left cardiac function assessment

Abstract

Objective: To determine whether a modified myocardial performance index (Mod-MPI) involving assessment of the movements (clicks) of the mitral valve (MV) and aortic valve (AV), improves intra- and interobserver agreement as compared to the previously reported method for MPI estimation. Methods: The Mod-MPI was recorded by two experienced operators in the left cardiac chambers of 25 normally grown fetuses using pulsed Doppler ultrasonography. The isovolumetric contraction time (ICT) was measured from the closure of the MV to the opening of the AV, the ejection time (ET) from the opening to the closure of the AV, and the isovolumetric relaxation time (IRT) from the closure of the AV to the opening of the MV. The Mod-MPI was calculated as (ICT + IRT)/ET. In addition, the MPI was estimated without using the valve clicks (F-MPI) as previously described. Intra- and interobserver agreement were then analyzed for both modalities. Results: There was a significantly lower intra- and interobserver variability in the estimation of all time periods with the Mod-MPI than with the F-MPI (ICT: intraobserver, 9.9% vs. 13.9%; interobserver 9.9% vs. 15.6%; IRT: intraobserver, 9.9% vs. 14.8%; interobserver 10.4% vs. 18.3%; and ET: intraobserver, 4.5% vs. 6.1%; interobserver 2.8% vs. 5.2%, respectively). Intraclass correlation coefficient (IntraCC) for the Mod-MPI was 0.8 (95% confidence interval (95% CI), 0.56-0.9) and for the F-MPI, the IntraCC was 0.62 (95% CI, 0.26-0.84); P = 0.01. Agreement between observers using the Mod-MPI showed a mean difference of 0.0 with 95% limits of agreement (LA) -0.09 (95% CI, -0.1 to -0.075) to 0.09 (95% CI, 0.075-0.1) and for the F-MPI the mean difference was -0.01 with 95% LA -0.26 (95% CI, -0.3 to -0.22) to 0.25 (95% CI, 0.21-0.29). Conclusion: Calculation of the Mod-MPI based on Doppler echoes of the MV and AV clicks is associated with a lower variation and better inter- and intraobserver agreement than the previously used method for fetal cardiac evaluation. Copyright © 2005 ISUOG. Published by John Wiley & Sons, Ltd.