Document resource
Background Postoperative atrial fibrillation (AF) is a common complication after coronary artery bypass grafting (CABG) but is less well characterised after multivessel percutaneous coronary intervention (MV-PCI). With the increasing use of MV-PCI, we compared the incidence and clinical implications of clinically significant post-procedural AF following MV-PCI versus CABG.Methods In this nationwide observational study, we analysed data from the Swedish Coronary Angiography and Angioplasty Registry and the Swedish Prescribed Drug Register. We included all patients in Sweden with chronic coronary syndrome who underwent MV-PCI or isolated CABG between August 2013 and October 2020, excluding those with prior AF or oral anticoagulant (OAC) use in the preceding year. Clinically significant post-procedural AF was defined using a treatment-based proxy: a new OAC dispensation within 3 months after the procedure, with persistence at 9 months. The primary outcome was post-procedural AF incidence at 90 days; secondary outcomes were stroke, myocardial infarction, major bleeding and all-cause mortality.Results A total of 4471 patients undergoing MV-PCI and 2232 undergoing CABG were included. Post-procedural AF occurred in 0.6% after MV-PCI and 5.7% after CABG (adjusted risk ratio 15.02; 95% CI 8.72 to 25.87; p<0.001). MV-PCI patients were older and had more comorbidities, whereas CABG patients had higher rates of diabetes and more extensive coronary disease. At 90 days, all-cause mortality and major bleeding did not differ significantly. Myocardial infarction was more common after MV-PCI (1.1 vs 0.4%), whereas stroke was more frequent after CABG (0.6 vs 0.3%).Conclusions Clinically significant post-procedural AF, defined by initiation and persistence of OAC therapy, was nearly tenfold more common after CABG than after MV-PCI. This association remained robust after multivariable adjustment for demographics, comorbidities and angiographic/severity-related factors. Differences in clinical outcomes were modest and should be interpreted cautiously.