A prospective clinical study.
A cardiac surgery operating room in a single cardiovascular center.
Forty-five patients undergoing elective CRT lead implantation.
CIFT and CI2D were determined simultaneously before and after CRT using CI3D as the reference method.
BNP was measured before CRT and 6 months after CRT. Areas under the receiver operator characteristic curves (AUCs) were calculated for each method of measurement to predict BNP decrease. AUC was largest for CI3D (AUC = 0.735, p = 0.017). Bland-Altman analysis revealed that the percentage error was 58% for CIFT and 28% for CI2D. A polar plot analysis showed that the mean angular bias was -7.26° and 0.64°, the radial limits of agreement were 70° and 29.4°, and the concordance rate was 67.7% and 93.8% for CIFT and CI2D, respectively.
CI significantly increased after CRT in patients whose BNP level decreased 6 months after CRT. However, only CI3D could predict decreases in BNP 6 months after CRT. Although CI2D was acceptable compared with CI3D, the tracking ability of CI changes was just below acceptable. CIFT has a wide limit of agreement with CI3D, with a poor tracking ability.