Clinical Course of Postoperative Atrial Fibrillation After Cardiac Surgery and Long-term Outcome

Abstract

BACKGROUND: New-onset postoperative atrial fibrillation (POAF) after cardiac surgery is associated with worse short- and long-term outcomes. Although the clinical presentation of POAF varies substantially, almost all studies model it with a dichotomous yes or no variable. We explored potential associations between the clinical course of POAF and long-term outcome. METHODS: This retrospective, observational, single-center study included 6435 coronary artery bypass grafting and/or valve patients between 2010 and 2018. POAF patients were grouped into spontaneous/pharmacologic conversion to sinus rhythm, sinus rhythm after electrical cardioversion, and sustained AF at discharge. Multivariable Cox regression models adjusted for age, sex, type of surgery, comorbidities, and early-initiated oral anticoagulation were used to study associations between the clinical course of POAF and long-term risk for mortality, ischemic stroke, thromboembolic events, heart failure hospitalization, and major bleeding. Median follow-up time was 3.8 years (range, 0-8.3). RESULTS: POAF occurred in 2172 patients (33.8%), 94.9% of whom converted to sinus rhythm before discharge. Of these, 73.6% converted spontaneously or with pharmacologic treatment and 26.4% after electrical cardioversion. Both sustained AF and electrical cardioversion were independently associated with an increased long-term risk for heart failure (adjusted hazard ratio for sustained AF at discharge, 2.55 [95% confidence interval, 1.65-3.93; P < .001]; adjusted hazard ratio for electrical cardioversion, 1.28 [95% confidence interval, 1.00-1.65; P = .049]) but not with increased long-term risk for death, thromboembolic complications, or bleeding. CONCLUSIONS: A more complicated POAF course is associated with increased long-term risk for heart failure hospitalization but not for all-cause mortality or thromboembolic complications.