2026 Southeastern Residency Conference

Evaluation of Atrial Fibrillation Post-Left Ventricular Assist Device Authors: Glennessa Hodge, Anna Crider, Brian Tran, Kiara Patino, Eleanor Schoen, Mahmoud H. Abdou Background  
Atrial fibrillation is a common postoperative complication following ventricular assist device (VAD) placement contributing to significant morbidity. Beta-blockers are considered first-line agents for the prevention of postoperative atrial fibrillation (POAF); however, many VAD patients are not on beta-blockers preoperatively due to relying on inotropes. Amiodarone is commonly used to reduce the incidence of POAF; however, its role and optimal duration in VAD recipients remains unclear. Current guidelines offer varying recommendations on dosing and therapy duration, ranging from 4 to 12 weeks. This study aims to identify the incidence of POAF, assess the use of amiodarone in management, and explore the impact on patient outcomes.    
Methods  
This study was a retrospective chart review of patients who received a left ventricular assist device (LVAD) at Emory University Hospital (EUH) or Emory Saint Joseph’s Hospital (ESJH) from October 1st, 2022, to March 31st, 2025. 
The primary outcome of this study was to identify the incidence of atrial fibrillation from LVAD implant to post-operative day 90. Secondary objectives included 90-day rehospitalization, postoperative total length of stay, thromboembolic or stroke incidence, and death. Amiodarone dosing was up to provider discretion as there were no protocols in place during this study time frame. Descriptive statistics include mean, median, or mode as appropriate to assess the primary and secondary outcomes. A multivariate regression analysis was conducted to compare POAF and non-POAF patients to identify associated risk factors.   
Results
POAF was seen in 71 out of 119 (37.5%) participants. 90-day all cause and arrhythmia related rehospitalization were 69 (36%) and 12 (6%) out of 190 respectively. The median postoperative length of stay was 22 in the total population, 24 (IQR 19-38.5) amongst POAF patients, and 20 (IQR 16-25.5) for non-POAF patients. 90-Day mortality was 18 (9%) out of 190, 6 from the POAF group and 12 from the non-POAF group. Antiarrhythmics used to treat POAF were amiodarone (35.3%), digoxin (4.2%), and dofetilide (0.5%). Majority of POAF patients were treated with intravenous amiodarone and then transitioned to oral amiodarone 200 mg daily at discharge. 43% of patients discharged on amiodarone did not have their amiodarone plan readdressed within 90 days post discharge. Of the 73 out of patients who received amiodarone prophylaxis for a median of 14 days, 44 (60.2%) did not experience POAF (OR = 1.77, 95% CI = 0.64 - 2.14).  
Conclusion
The POAF incidence is consistent with the 15-50% reported from previous studies. The patient population had common risk factors for POAF, including older age (≥ 60, 42.8%), male sex (77%), hypertension (76%), diabetes (46%), and non-ischemic cardiomyopathy (75%). Amongst patients that experienced POAF, approximately 89.5% were managed with amiodarone at a variety of doses. 56% of patients were on amiodarone prior to admission and 7% were intentionally placed or continued amiodarone for prophylaxis per documentation. 
Analysis of all patients that received amiodarone within the 72 hours immediately prior to LVAD implantation, revealed a greater percentage of patients who got prophylaxis did not experience POAF, however this study was not adequately powered to detect significant difference in outcomes. 90-day outcomes between patients that did and did not experience POAF were similar, however, the POAF group had a longer length of stay at 24 days and a higher percentage of cardiac related readmission compared to the non-POAF group.  
This study suggests a possible benefit in amiodarone prophylaxis in patients with high risk features or history of atrial fibrillation prior to LVAD. The exact dosing regimen and duration is still to be explored, but our study found that higher cumulative amounts of amiodarone did not show better outcomes.