Performance of the LEVO-D score in patients with advanced heart failure at a referral hospital

Background
Advanced heart failure is a prevalent and burdensome condition, and effective therapies are needed. Ambulatory intermittent levosimendan infusion has been used to reduce cardiovascular events, although response rates vary. This study assessed the LEVO-D score as a predictive tool in patients receiving levosimendan infusions under standardised conditions.

Methods
We conducted a prospective observational study at the Heart Failure Clinic of a tertiary-level teaching hospital between September 2020 and March 2024. A total of 67 patients received a median of 10 intermittent levosimendan infusions and were followed for a mean duration of 18 months. The performance of the LEVO-D score was assessed using receiver operating characteristic (ROC) curve analysis and descriptive statistics.

Results
The patients were elderly, with a mean age of 69.5 years, a high comorbidity burden, and an advanced functional class. All patients received beta-blockers, and most had a high rate of guideline-directed therapy. At 18 months, the mortality rate was 36.2% and the hospitalisation rate was 42%. The LEVO-D score showed good predictive value for hospitalisation (AUC 0.81), moderate predictive value for clinical response (AUC 0.64), and poor predictive value for mortality (AUC 0.58). A cut-off value of ≥1.0 provided the optimal balance between sensitivity and specificity for identifying responders.

Conclusions
The LEVO-D score showed consistent discriminatory capacity in our cohort despite a high rate of optimal medical therapy and a low rate of edge-to-edge transcatheter repair, thus supporting its utility and encouraging further validation in diverse clinical settings.