VIRTUE: An Innovative Ejection Fraction-Based Index for Predicting Outcomes in Acute Heart Failure

Background/Objectives: Acute heart failure (AHF) is a heterogeneous clinical syndrome, and prognosis varies depending on the ejection fraction phenotype. Although the N-terminal pro–B-type natriuretic peptide (NT-proBNP) remains the benchmark biomarker, conventional echocardiographic measures such as the tricuspid annular plane systolic excursion (TAPSE), the right ventricular to right atrial pressure gradient (RV–RA gradient), and the left ventricular outflow tract velocity–time integral (LVOT VTI) provide only partial prognostic information.We previously proposed the Virtue Index, defined as the ratio between the RV–RA gradient and the product of TAPSE and LVOT VTI, reflecting the interaction between right and left ventricular performance. The present study aimed to assess the clinical and prognostic behavior of the index in a cohort of patients with AHF and to determine whether its performance differs between reduced and preserved ejection fraction phenotypes. Methods: We retrospectively analyzed 222 patients admitted with a diagnosis of AHF. Complete echocardiographic data for Virtue Index calculation were available in 168 patients (99 [59%] with heart failure with reduced ejection fraction [HFrEF] and 69 [41%] with heart failure with preserved ejection fraction [HFpEF]). Patients with mid-range ejection fraction (40–49%) or incomplete echocardiographic parameters were excluded from subgroup analyses but retained for descriptive statistics where applicable. Correlation with NT-proBNP was evaluated using Spearman rank testing with bootstrap confidence intervals. Prognostic performance for in-hospital mortality was analyzed using ROC curves, area under the curve (AUC) with bootstrap intervals, and pairwise DeLong comparisons. Results: In the HFpEF subgroup, the index correlated moderately with NT-proBNP (ρ = 0.38, 95% CI 0.13–0.58, p = 0.002) and demonstrated fair prognostic discrimination (AUC 0.704, 95% CI 0.53–0.85), comparable to the RV–RA gradient (AUC 0.724) and higher than TAPSE (AUC 0.637) or LVOT VTI (AUC 0.669). In contrast, in HFrEF, the index showed a weak, non-significant correlation with NT-proBNP (ρ = 0.19, p = 0.06) and modest predictive accuracy (AUC 0.584, 95% CI 0.36–0.79), while LVOT VTI achieved the best discrimination (AUC 0.700). NT-proBNP remained superior in both subgroups (AUC 0.744 in HFrEF, 0.838 in HFpEF). Conclusions: The Virtue Index reflects integrated haemodynamic function and may demonstrate a phenotype-dependent prognostic role in AHF. In our study, its value was more evident in HFpEF, whereas in HFrEF traditional parameters, especially LVOT VTI, remained stronger predictors. Although NT-proBNP consistently outperformed Virtue, the index may complement biomarker assessment by providing rapid, bedside risk stratification of short-term mortality.