Predicting the 1-Year All-Cause Mortality After Hospitalisation for an Acute Heart Failure Event: A Real-World Derivation Cohort for the Development of the S2PLiT-UG Score.

ABSTRACT

BACKGROUND: Acute heart failure (AHF) is a complex syndrome associated with high morbidity and mortality. This study aimed to derive a simple risk score with which to identify AHF patients at high risk for an all-cause death event during the first year after hospital discharge. METHODS: Three hundred AHF patients from the Heart Failure registry were included in the analysis. Cox regression with a forward-conditional algorithm and bootstrapping procedure was used to build the prognostic score, while c-statistic was used to assess the prognostic performance of the score. RESULTS: Seven variables were independently associated with an all-cause mortality event during the 1-year follow-up (FU) estimated glomerular filtration rate of 40-60; estimated glomerular filtration rate <40 mL/min/1.73 m2; uric acid >450 µmol/L; left-ventricular ejection fraction <45%; sodium <136 mmol/L; systolic blood pressure <115 mmHg; and a positive history of previous heart failure-related decompensation event(s). The score derived from significant variables enabled classification of patients into three risk categories low (0-2 points), intermediate (3 points), and high (4-6 points). Observed all-cause mortality rates during the 1-year FU were 6.1%, 30.5%, and 80.9% across the three risk categories, respectively. The score demonstrated a high level of discrimination for an all-cause death event in the derivation cohort with the c-statistic value of 0.907 (95% CI, 0.867-0.939; p < 0.0001) and adequate calibration. CONCLUSIONS: The S2PLiT-UG score is a simple tool with potential for facilitating risk stratification and therapeutic decision-making during the first year after hospitalisation for an AHF event. Future external validation studies are required to confirm its prognostic performance.