Clinical efficacy of Doppler-echocardiographic indices of aortic valve stenosis:a comparative test-based analysis of outcome
Javier Bermejo, MD*,*,
Rodolfo Odreman, MD*,
Josefina Feijoo, MD*,
M. Mar Moreno, MD*,
Paz Gómez-Moreno, RDCS* and
Miguel A. García-Fernández, MD*
* Department of Cardiology, Hospital General Universitario Gregorio Marañón, Madrid, Spain
View larger version (23K):
[in a new window]
|
Figure 1 Association among hemodynamic indices of aortic stenosis. Results of the cluster analysis of variables in Group B show an initial split into two groups, flow-corrected and non–flow-corrected indices, with a fair correlation between them (rho2 = 0.37). Stroke work loss closely correlates with  P, as expected from its calculation formula. Also, there was a close association between AVA and AVR.
|
|
View larger version (23K):
[in a new window]
|
Figure 2 Follow-up data of the study population. Values are shown for patients in Group A and Group B, separated by a slash. AS = aortic valve stenosis.
|
|
View larger version (50K):
[in a new window]
|
Figure 3 Clinical efficacy of aortic stenosis indices. Prediction capability is shown for the clinical objectives of predicting: 1) symptomatic status, 2) critical aortic stenosis (AS), 3) late AS events, 4) early or late AS events in patients with depressed left ventricular function, and 5) any-cause mortality in unoperated patients. Clinical efficacy is assessed by receiver-operator characteristic analysis for populations derived from Group B (see methods section). EF = ejection fraction.
|
|
View larger version (24K):
[in a new window]
|
Figure 4 Boxplots of indices of aortic stenosis according to the symptomatic status. Distributions are shown for peak jet velocity (A), mean transvalvular pressure gradient (B), stroke-work loss (C), valve area (D), and valve resistance (E), for patients in Group B. Boxes represent the interquartile distance, whereas the white line represents the median. The shaded zone represents the 95% confidence interval for the median and the whiskers represent the limits of each distribution. Significant differences between groups are ascertained by the lack of overlap of the shaded zones. The highest difference between asymptomatic (Asympt.) and symptomatic (Sympt.) patients is observed for stroke-work loss.
|
|
View larger version (20K):
[in a new window]
|
Figure 5 Validation of clinical efficacy models in Group A. (A) Regression tree for calculating the probability of suffering critical aortic stenosis (AS) (prob) according to stroke-work loss (SWL), presence or absence of symptoms, and ejection fraction (EF). (B) Cumulative probability of long-term outcome according to SWL and symptomatic status. Patients with critical AS have been excluded. (C) Probability of early and late AS events in patients with impaired left ventricular function (EF < 0.45). p = probability of long-rank test for comparisons between categories.
|
|
View larger version (19K):
[in a new window]
|
Figure 6 Unadjusted cumulative probability of suffering aortic stenosis (AS) events (valve replacement or cardiac death). Plots are shown for peak jet velocity (Vmax) (A), P (B), stroke-work loss (SWL) (C), aortic valve area (AVA) (D), and aortic valve resistance (AVR) (E), for patients in Group B after exclusion of patients with critical AS (n = 70). There is a remarkable overlap of the outcome of the group with AVA < 0.75 cm2 and the group in the 0.75 to 1 cm2 range. Again, best discriminative power was observed for SWL.
|
|
|
