Residual Congestion Defined by HFA-ESC Multiparametric Criteria and 90-day Outcomes in Worsening Heart Failure

Residual congestion at discharge remains a major contributor to adverse short-term outcomes in patients hospitalized for acute heart failure (AHF).1 Despite achieving apparent clinical stability, up to 30% of patients are readmitted within the first 90 days post-discharge.2 This has prompted growing interest in objective tools to detect and quantify residual or subclinical congestion prior to discharge. The 2023 expert consensus from the Heart Failure Association of the European Society of Cardiology (HFA-ESC) proposed a multiparametric approach to guide decongestion strategies.3 This includes inferior vena cava (IVC) ultrasound, lung ultrasound (LUS), and changes in natriuretic peptides such as N-terminal pro-B-type natriuretic peptide (NT-proBNP). While each of these modalities has shown prognostic value individually, their combined use as a pre-discharge congestion profile has not been prospectively validated. We conducted a pre-specified subanalysis of the CAVAL US-AHF trial (NCT04549701), a single-centre, randomized study assessing a protocolized decongestion strategy guided by daily ultrasound in patients hospitalized for AHF.4 The trial included adults with new or worsening HF. Exclusion criteria included advanced renal disease, low-output state, severe tricuspid regurgitation, or other non-cardiac causes of congestion. Detailed methodology of the trial has been previously published.5 All patients, regardless of randomization group, underwent daily lung and IVC ultrasound from 24 h after admission through discharge. Studies were performed by trained operators blinded to patient status and centrally adjudicated by a core laboratory composed of three cardiologists with expertise in ultrasound. NT-proBNP levels were measured at admission and discharge. Based on HFA-ESC decongestion criteria, patients were classified into two groups. The ‘optimal’ group met all three of the following: (1) IVC diameter 50% collapsibility; (2) 30% reduction in NT-proBNP. Patients failing to meet any of these were considered ‘non-optimal’. The primary endpoint was a composite of AHF readmission, urgent visit for worsening heart failure, or all-cause mortality at 90 days post-discharge. This study was approved by the Ethics Committee of Instituto Cardiovascular de Buenos Aires and registered in the PRIISA.BA platform of the Buenos Aires Ministry of Health. Written informed consent was obtained from all participants. A total of 60 patients were included in this substudy. The mean ± standard deviation age of the patients was 76.7 ± 13 years, 68.3% were male and the mean left ventricular ejection fraction was 44 ± 15%. Patients had a high prevalence of comorbidities such as arterial hypertension (75.8%), atrial fibrillation (53%), anaemia (34.8%), and chronic kidney disease (42.4%). Overall, 54.5% had a previous diagnosis of heart failure and 21% had been hospitalized for heart failure in the previous year. The most common aetiology of heart failure was related to coronary artery disease in 30.3% of patients, and prior to admission, 45.5% and 27.3% of patients were in New York Heart Association functional class II and III, respectively. Both groups were well balanced with respect to baseline characteristics. Baseline demographics and discharge pharmacologic therapies were well balanced between optimal and non-optimal groups. Twenty-seven patients (45%) met all three optimal decongestion criteria. In this group, no adverse events occurred during the 90-day follow-up. In contrast, among the 33 patients in the non-optimal group, 14 (42.4%) met the primary endpoint (log-rank p < 0.001). The Kaplan–Meier curve is shown in Figure 1. Further exploratory analysis showed no significant differences in discharge furosemide dose, left ventricular ejection fraction, or use of guideline-directed medical therapy between the optimal and non-optimal groups. This suggests that residual congestion may persist independently of clinical assessment or treatment intensity. The present analysis provides prospective evidence supporting the prognostic utility of a multiparametric subclinical congestion profile, as recently proposed by the HFA-ESC.3 Our findings align with prior observational studies demonstrating that elevated B-lines or non-collapsible IVC at discharge are independently associated with poor outcomes. However, the novelty of this analysis lies in the combined application of these tools, along with biomarker trends, to stratify post-discharge risk. Of note, this profile identified a subset of patients (45%) with near-zero risk of events over 90 days. This group may be considered truly euvolaemic at discharge, despite advanced age and high comorbidity burden. Conversely, failure to meet any of the three criteria identified patients at markedly increased risk, despite similar clinical appearance and guideline-directed medical therapy prescription. This approach has several advantages. It uses widely available, bedside ultrasound techniques, requires minimal additional resources, and provides real-time feedback to guide discharge readiness. In busy hospital environments, where early discharge pressures are high, such a profile may offer a standardized and objective way to optimize timing. Several limitations warrant discussion. This was a single-centre analysis with a modest sample size and short follow-up. Furthermore, this was a substudy not powered for outcome comparisons, and validation in larger multicentre cohorts is required. In conclusion, the use of a multiparametric profile incorporating IVC ultrasound, LUS, and NT-proBNP reduction—aligned with HFA-ESC recommendations—identified AHF patients at low risk of 90-day adverse events. These findings suggest that objective discharge criteria may complement clinical judgment by informing the optimal timing of discharge and tailoring the intensity of post-discharge follow-up. Larger studies should explore the role of this strategy in improving outcomes and resource utilization. L.M.B. received research grants from Women As One Research grant and Fiorini Foundation. Conflict of interest: none declared.