Clinical Factors Associated with Amiodarone Prescription at Hospital Discharge in Patients with Atrial Fibrillation: Insights from the CRAFT Registry

Background and Objectives: Amiodarone remains an important option for rhythm control in atrial fibrillation (AF), particularly when heart failure or structural heart disease limits alternative antiarrhythmic therapies. This study aimed to characterize amiodarone prescription at hospital discharge and identify associated clinical characteristics among patients hospitalized with AF. Materials and Methods: This retrospective analysis of the CRAFT registry included hospitalizations of adults with AF between January 2017 and July 2026, with no data available for 2023. Multivariable logistic regression with patient-clustered robust standard errors accounted for repeated hospitalizations. Sensitivity analyses included first-hospitalization and subgroup analyses and multiple imputation of missing left ventricular ejection fraction (LVEF) values. Results: The study included 3670 hospitalizations involving 2828 patients, with amiodarone prescribed at discharge in 791 (21.6%). The primary complete-case model included 2523 hospitalizations involving 2021 patients. Amiodarone prescription was most strongly associated with paroxysmal or persistent AF (adjusted odds ratio [OR] 24.47, 95% confidence interval [CI] 15.77–37.97). Heart failure with reduced ejection fraction (HFrEF; OR 2.14, 95% CI 1.63–2.80) and lower estimated glomerular filtration rate (eGFR; OR 1.14 per 10 mL/min/1.73 m2 decrease, 95% CI 1.05–1.24) were also independently associated with amiodarone prescription. These associations were generally consistent across sensitivity analyses, whereas those involving direct oral anticoagulant therapy and three-class heart failure pharmacotherapy were sensitive to missing-data handling. Conclusions: Amiodarone prescription at hospital discharge was closely associated with non-permanent AF, HFrEF, and impaired renal function. These findings describe real-world prescribing patterns but do not establish causal determinants of treatment selection, guideline adherence, treatment effectiveness, or comparative safety.