Predictors of Frequent Readmissions in Patients With Heart Failure

Area of Science:

• Cardiology
• Health Services Research
• Clinical Epidemiology

Background:

• Heart failure (HF) patients experience high hospital readmission rates.
• Existing risk models often focus on single readmissions, potentially missing high-cost, frequently readmitted patient groups.
• Frequent readmissions significantly contribute to morbidity and healthcare expenses.

Purpose of the Study:

• To develop and validate a risk prediction model for frequent hospital readmissions (≥2) or death within one year in heart failure patients.
• To identify key clinical and demographic predictors associated with recurrent hospitalizations in HF.
• To enable better allocation of healthcare resources by identifying high-risk HF patients.

Main Methods:

• Utilized data from 6252 heart failure patients in the Management of Cardiac Failure Program (MACARF).
• Randomly divided patients into derivation and validation cohorts for model development and testing.
• Employed multivariate logistic regression to identify predictors of ≥2 readmissions or death within one year.

Main Results:

• Identified predictors of frequent readmissions or death: ischemic heart disease, chronic kidney disease, unmarried status, anemia, low serum albumin, elevated creatinine, prolonged hospital stay (>7 days), and lack of beta-blocker discharge.
• A higher risk score correlated with increased event rates (p<0.001).
• The model demonstrated similar predictive performance in both derivation and validation cohorts (C-statistic=0.65).

Conclusions:

• Developed a risk score to identify heart failure patients at high risk for frequent readmissions or death.
• This score can help focus healthcare resources and interventions on a vulnerable subset of HF patients.
• Aims to reduce the disproportionate burden of disease and associated costs in heart failure management.