12-month risk prediction model for heart failure hospitalization after myocardial revascularization

Background

Although prognosis in heart failure has improved with guideline-directed medical therapy (GDMT), heart failure (HF) hospitalizations remain a major contributor to the overall disease burden.

Study aim: To identify risk factors and develop a prediction model for HF-related hospitalizations within 12 months after myocardial revascularization.

Methods

The study included 275 patients with ischemic HF undergoing myocardial revascularization: 54.5% by coronary artery bypass grafting and 45.5% by percutaneous coronary intervention. Mean age was 63.12±0.54 years; 217 men (78.9%) and 58 women (21.1%) were enrolled. Patients were evaluated at baseline and at 12 months after myocardial revascularization using laboratory tests (hemoglobin, creatinine, estimated glomerular filtration rate (eGFR) calculated by the CKD-EPI formula) and echocardiography. Statistical analysis included the Student’s t-test, Pearson chi-square test, odds ratio with 95% confidence intervals, discriminant analysis; model performance was assessed by the area under the ROC curve.

Results

The HF hospitalization rate in the cohort was 18.1%. Patients requiring HF hospitalizations were older, had a longer duration of coronary artery disease (CAD), more frequently underwent surgical myocardial revascularization. eGFR and hemoglobin were significantly lower. A more severe HF profile was observed, with higher NT-proBNP levels and lower left ventricular ejection fraction (LVEF). Echocardiographic signs of pulmonary hypertension were more prevalent (86.0% vs 52.9%, p<0.001), while tricuspid regurgitation velocity (TRV), right ventricular outflow tract acceleration time, right atrium area, the TAPSE/PASP ratio, pulmonary artery and inferior vena cava diameters showed unfavorable values (Table). Right ventricular (RV) dysfunction was more frequent (70.0% vs 32.4%, p<0.001), with increased RV end-diastolic diameter (RVEDD) and impaired RV function reflected by TAPSE, RV fractional area change, RV systolic velocity (Table). Seven parameters demonstrated prognostic relevance for HF hospitalizations: CAD duration (OR=1.173, 95%CI=1.06–1.29), hemoglobin (OR=0.948, 95%CI=0.92–0.97), eGFR (OR=0.980, 95%CI=0.96–0.99), LVEF (OR=0.912, 95%CI=0.88–0.94), TRV (OR=11.129, 95%CI=3.53–34.99), RVEDD (OR=1.134, 95%CI=1.06–1.21), TAPSE/PASP (OR=0.005, 95%CI=0.001–0.041). These variables formed the basis of the prediction model for HF hospitalizations within 12 months after myocardial revascularization, which demonstrated a positive predictive value of 75.5%, a negative predictive value of 76.8%, and a good discriminatory performance (AUROC = 0.827, 95%CI=0.759–0.895, p<0.001).

Conclusion

The 12-month risk prediction model for heart failure hospitalization after myocardial revascularization demonstrates a good predictive accuracy, enabling the identification of patients at risk for adverse HF outcomes who require closer monitoring and timely escalation of GDMT.For image description, please refer to the figure legend and surrounding text.For image description, please refer to the figure legend and surrounding text.