Coronary sinus diameter to estimate congestion and predict survival

Figures

• 

  Fig. 1. Maximum diameter of CS and IVC before and after hemodialysis (HD).Panel A: Correlation plot between pre-dialysis (●) and post-dialysis (○) maximum diameters of CS and IVC (pre/post-HD CSmax and IVCmax, respectively; Spearman ρ = 0.86 for the aggregate, p < 0.0001). Since the regression slopes do not differ between pre/post HD conditions (p = 0.08) a single regression line with 95 %CI is plotted for all data. Median values and IQR for pre and post-dialysis of CS and IVC are shown in red and blue, respectively. Panel B: distribution across quartiles of HD-extracted fluid volume. Data presented as median, interquartile range. Pre-dialytic max diameters of both CS (left, red) and IVC (right, blue) increased across quartiles of fluid volume extraction (1 to 4); the distribution of post-dialytic CSmax (dashed columns) was significantly reduced in quartiles 2-to-4 of HD-extracted fluid volume. Likewise, the post-dialytic IVCmax was flattened to significantly lower values. * p < 0.05, ** p < 0.01, *** p < 0.001 vs reference pre-dialytic quartile I; ## p < 0.01, ### p < 0.001 vs corresponding pre-dialytic value. (For interpretation of the references to colour in this figure legend, the…

• 

  Fig. 2. ROC curve analysis for the accuracy of CS and IVC for the identification of volume excess. The values of the area under the curve (AUROC), which estimates the overall accuracy, are presented with 95 % confidence intervals. CSmax was not inferior to IVCmax and IVCC% for identification of congestion (defined as pre-HD state) in patients with HD fluid extraction > 1.5L (n = 49); for all three, p < 0.0001 vs random assessment (i.e. area under the identity line = 0.5); ns = not significant for comparison of AUROC values by Hanley method.

• 

  Fig. 3. Probability of survival according to post-dialysis CS maximum diameter. The Kaplan-Meier curves show that residual congestion, estimated by a post-dialysis CSmax larger than 9 mm, i.e. the optimal cut-off identified above for discrimination of congestion vs decongestion, was associated with an excess mortality risk (Log-rank Chi square = 11.49, p < 0.001; n. at risk: 44 for CSmax ≤ 9 mm [black] and 16 for CSmax > 9 mm [red; univariate Hazard Ratio (HR) at Cox Regression is presented with 95 % confidence interval, p = 0.004]; two patients were censored in the first group at the time of renal transplantation). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)