Clinical profiles to guide decongestion in acute heart failure

Background Residual congestion (RC) at discharge is associated with poor outcomes in acute heart failure, yet actionable decongestion targets beyond the first 48 hours are poorly defined. Our objective is to integrate diuretic resistance (DR) with RC at 72 hours to profile risk and guide decongestion. Methods Post hoc pooled analysis of DOSE-AHF and ROSE-AHF with serial congestion assessment and NT-proBNP at baseline and 72 hours. RC was defined using clinical (Orthodema Score [OES]), biomarker-based (NT-proBNP reduction ≤30% and absolute levels), and combined criteria (OES>0 with ≤30% reduction). DR was defined as BAN-ADHF >12. Four 72-hour profiles were prespecified: RC–/low DR, RC–/high DR, RC+/low DR, and RC+/high DR. The primary outcome was a 90-day composite of death, rehospitalization, or Emergency Department visit. Results Among 636 patients (median age 69 [IQR: 60-79] years; 73% male), the primary outcome occurred in 36% over a median follow-up of 56 (34–62) days. Congestion was common at admission (OES>0 in 93%) but decreased to 57% at 72 hours, whereas DR remained stable. OES alone showed limited risk discrimination (OES>0: HR 1.19, 95% CI 0.90–1.59), whereas NT-proBNP–based definitions identified higher-risk patients (≤30% reduction associated with a 1.7-fold higher risk). Using RC−/low DR as reference, risk increased stepwise: RC+/low DR (1.5-fold), RC−/high DR (1.8-fold), and RC+/high DR (1.9-fold higher risk). Conclusions Integrating DR with RC at 72 hours identified four clinical profiles associated with 90-day risk. In the absence of a universally accepted decongestion target, these findings may help contextualize early treatment response and should be prospectively validated before guiding therapy. Condensed Abstract Residual congestion (RC) in acute heart failure (AHF) is linked to poor outcomes, yet optimal decongestion targets beyond 48 hours remain unclear. In a post-hoc pooled analysis of DOSE-AHF and ROSE-AHF (n=636), we evaluated 72-hour RC using clinical (Orthodema Score [OES]), biomarker-based (NT-proBNP reduction ≤30% and absolute levels), and combined definitions (OES>0 with ≤30% reduction), alongside diuretic resistance (DR; BAN-ADHF >12). The primary outcome (90-day death, rehospitalization, or ED visit) occurred in 34%. Clinical congestion alone showed limited discrimination (OES>0: HR 1.19), whereas NT-proBNP–based definitions identified higher risk (≤30% reduction: 1.7-fold higher risk). Compared with RC−/low DR, risk increased stepwise across profiles: RC+/low DR (1.5-fold), RC−/high DR (1.8-fold), and RC+/high DR (1.9-fold). Integrating DR with RC at 72 hours identifies actionable profiles that improve risk stratification and may guide decongestion.