AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Procedural Complexity, Posterior Pericardiotomy, and Sodium-Glucose Cotransporter-2 Inhibitor Use as Determinants of Postoperative Atrial Fibrillation After Cardiac Surgery
By Nandakumar S, Malik A, Biswas MS
Background: Postoperative atrial fibrillation (POAF) is the most common arrhythmic complication after cardiac surgery and is associated with increased morbidity and resource utilization. Contemporary risk stratification remains imperfect, limiting targeted prevention. Hypothesis: Procedural complexity is a key modifiable determinant of POAF risk, and posterior pericardiotomy and preoperative sodium-glucose cotransporter-2 inhibitor (SGLT2i) use are associated with lower POAF incidence. Methods: We performed a retrospective cohort study of 2,136 adults undergoing cardiac surgery at a single center from 2018-2024. Patients with prior atrial fibrillation were excluded. The primary outcome was new-onset POAF during the index hospitalization. Multivariable logistic regression with Firth penalization was used for adjusted analyses, and false discovery rate correction was applied. Results: POAF occurred in 326 patients (15.3%). On multivariable analysis, coronary artery disease (n = 1,728) was strongly associated with POAF (OR 2.57, 95% CI 1.72-4.01), reflecting its close linkage with coronary artery bypass grafting (CABG) exposure. Compared with isolated CABG (n = 1,407), combined CABG plus valve surgery was associated with higher odds of POAF (OR 1.82, 95% CI 1.07-3.00), whereas isolated aortic valve surgery (n = 560) was associated with substantially lower odds (OR 0.18, 95% CI 0.10-0.31). Notably, traditional risk factors including obesity, hyperthyroidism, diabetes mellitus, hypothyroidism, congestive heart failure, and alcohol use were not independently associated with POAF. In exploratory analyses, posterior pericardiotomy (n = 223) was associated with markedly lower odds of POAF (OR 0.01, 95% CI 0.00-0.08), and preoperative SGLT2i use (n = 93) was also associated with lower odds (OR 0.37, 95% CI 0.13-0.82). Patients with POAF had a longer median hospital length of stay (10 vs 8 days, p<0.001) and ICU length of stay (98 vs 76 hours, p<0.001). Conclusions: In this contemporary cardiac surgical cohort, procedural complexity rather than traditional comorbidities was the dominant determinant of POAF risk. Patients undergoing combined CABG plus valve surgery represent a high-risk subgroup warranting intensified prophylaxis, whereas those undergoing isolated aortic valve surgery appear at comparatively lower risk. Posterior pericardiotomy showed a strong protective association consistent with prior randomized data, and preoperative SGLT2i use warrants prospective investigation as a pharmacologic prevention strategy.