Abstract
Background: Systemic venous congestion contributes to organ dysfunction and adverse outcomes in acute heart failure (AHF). The Venous Excess Ultrasound Score (VExUS) evaluates venous congestion, but its prognostic role and the relative contribution of its individual components in ward-admitted patients remains uncertain. We aimed to evaluate the association between global and individual VExUS-derived Doppler findings and clinical outcomes and to explore descriptive congestion patterns using unsupervised clustering. Methods: In this prospective observational study, 100 patients admitted with AHF were screened, and 80 with adequate ultrasound windows underwent standardized VExUS assessment within 24 hours of admission. Associations with 6-month all-cause mortality, readmissions, and renal deterioration were evaluated using multiplicity-adjusted statistical models. Hierarchical clustering was performed as an exploratory secondary analysis to characterize underlying profiles. Results: Individual VExUS components were significantly associated with mortality: hepatic venous flow abnormalities (OR 5.45, 95% CI 1.01–29.40; p=0.04), increased portal vein pulsatility (OR 26.1, 95% CI 1.74–389.59; p=0.01), and abnormal intrarenal venous flow (OR 6.0, 95% CI 1.01–35.62; p=0.049). Conversely, the aggregate global VExUS grade was not independently associated with mortality or other clinical outcomes. All multivariable models included a single doppler predictor adjusted for discharge creatinine. Exploratory hierarchical clustering delineated two descriptive groups with differing patterns of systemic congestion, renal function, and clinical presentation. The group with more severe systemic congestion exhibited lower estimated glomerular filtration rates and a higher proportion of observed deaths (71.4% vs 28.6% of total events; p=0.029). Conclusions: In this single-center cohort, selected individual venous Doppler abnormalities, rather than the aggregate VExUS grade, showed statistically significant associations with adverse outcomes; however, these associations were derived from models with very few events per variable and therefore should be regarded as hypothesis-generating rather than confirmatory. Exploratory clustering suggested the existence of distinct congestion patterns based on VExUS grade and clinical pattern. Larger multicenter studies with adequate event numbers are warranted to validate these exploratory findings and to determine whether they can meaningfully guide decongestive therapy.