Figure 3.
Figure 3A. Quantitative Relationship Between Drug/Device Effects on End-Diastolic Volume and Mortality. Each data point represents a placebo-corrected change in EDV from an individual remodeling trial plotted against the mortality OR for the specific therapy. Color-coded mortality effect based on data from mortality trials listed in Table 1. Definition of mortality effect for a given intervention as described in Figure 2A legend. Remodeling data derived from analysis of 50 RCTs of 19 interventions, including 8,499 total patients. RCT = Randomized Controlled Trial; EDV = End-Diastolic Volume
Figure 3B. Predicted Probability of a Categorical Mortality Outcome Based on Drug/Device Effect on End-Diastolic Volume. The solid lines represent the likelihood of a categorical mortality outcome based on an intervention’s trial-level effect on end-diastolic volume compared with placebo. The dashed lines signify the corresponding 95% confidence interval. Color-coded mortality effect based on data from mortality trials listed in Table 1. Definition of mortality effect for a given intervention as described in Figure 2A legend. EDV = End-Diastolic Volume