Prognostic value of clinical judgment versus the ELAN-HF and BCN-Bio-HF risk scores after hospitalization for acute heart failure

INTRODUCTION AND OBJECTIVES Heart failure (HF) is a highly prevalent condition, particularly among the elderly, and is associated with high rates of readmission and mortality. To improve prognosis, risk assessment tools such as ELAN-HF and BCN-Bio-HF have been developed to estimate mortality after a hospitalization for HF. However, in clinical practice, risk evaluation is often based on the physician's subjective perception. This study aims to compare that perception with the mortality predictions generated by these tools. METHODS A retrospective, observational study was conducted in a secondary-level hospital between 2019 and 2021, including patients recently discharged after hospitalization for HF. At the first follow-up visit, physicians subjectively classified the patients' risk into three levels without access to the results of the risk scores. Afterwards, the ELAN-HF and BCN-Bio-HF scores were calculated and compared with clinical outcomes at 12 months. The predictive capacity of each method was evaluated. RESULTS A total of 125 patients were included (median age: 76 years 71,5-83, 56% male). The overall mortality rate was 34.4%. The ELAN-HF and BCN-Bio-HF scores better identified high-risk patients than the clinical assessment, which underestimated mortality in the low-risk group. ROC curve analysis showed good predictive ability for mortality, particularly for the BCN-Bio-HF score (AUC: 74.6%). CONCLUSIONS The results support that risk scores are more accurate than physicians' subjective evaluation in predicting mortality in patients with HF. Their integration into clinical practice is recommended, given their prognostic value and usefulness in patient follow-up.