Treatment of congestion is one of the main goals in patients hospitalized for acute heart failure. Nevertheless, current evidence shows that decongestion is often not achieved and that residual congestion at discharge is strongly associated with poor outcomes. While this association has been demonstrated, previous studies have primarily focused on single parameters of congestion (physical examination, biomarkers, or imaging features). The aim of the study is to assess residual congestion at discharge using a multiparametric approach and to compare the prognostic value of each evaluation strategy. Additionally, the analysis will be supported by artificial intelligence to develop a multiparametric prognostic algorithm that can provide an improved predictive model compared to standard statistical approaches.

Inclusion Criteria:

• Age ≥ 18 years.

• Patients hospitalized for new onset heart failure or worsening heart failure defined by:

  1. Symptoms classified as New York Heart Association class III or IV.
  2. Clinical or instrumental signs of volume overload (e.g., dyspnea with evidence of pulmonary congestion on X-rays or lung ultrasound, pitting edema, and jugular venous distension).
  3. Elevated NT-proBNP levels within the first 24 hours of admission (cutoff values: 450 ng/L for patients < 50 years; >900 ng/L for patients aged 50-75 years; >1800 ng/L for patients >75 years).

• Ability to provide informed consent.

Exclusion Criteria:

• Known diagnosis of septicemia.
• Glomerular filtration rate < 15 ml/min.
• Life expectancy < 6 months.
• Active myocarditis.
• Heart transplant recipients.
• Patients with ventricular assist devices.
• Congenital heart diseases.
• Moderate-to-severe liver disease (Child-Pugh B-C).
• Patients that will not be followed up by the Heart Failure Unit.

Patients hospitalized for acute heart failure