Incidence of post-procedural atrial fibrillation after multivessel percutaneous coronary intervention versus coronary artery bypass grafting: a nationwide observational study

Figures

• 

  Figure 1. Flowchart of the study population. Schematic representation of patient inclusion (MV-PCI or CABG), application of exclusion criteria and the final sample size. Missing baseline covariates were handled using a complete-case approach and are not shown in the flowchart. AF, atrial fibrillation; BMS, bare-metal stent; CABG, coronary artery bypass grafting; CCS, chronic coronary syndrome; OAC, oral anticoagulant; MV, multivessel; PCI, percutaneous coronary intervention.

• 

  Figure 2. Primary outcome: MV-PCI versus CABG. Kaplan-Meier curves illustrating the 90-day cumulative incidence of clinically significant post-procedural AF, defined as a newly dispensed OAC prescription within 90 days after revascularisation, confirmed by a subsequent dispensation between 270 and 450 days after the index procedure. Time from the index procedure to the first OAC dispensation is shown among patients fulfilling both criteria, stratified by MV-PCI and CABG. The outcome was analysed using Poisson regression in univariable and multivariable models, adjusted for inclusion year, age, sex, diabetes mellitus, hypertension, hyperlipidaemia, prior major bleeding, heart failure, chronic kidney disease, angiographic findings, smoking status, Canadian Cardiovascular Society score and history of myocardial infarction, stroke, PCI and CABG. AF, atrial fibrillation; aHR, adjusted HR; aRR, adjusted risk ratio; CABG, coronary artery bypass grafting; MV-PCI, multivessel percutaneous coronary intervention; OAC, oral anticoagulant; RR, risk ratio.

• 

  Figure 3. Secondary outcomes: MV-PCI versus CABG. Kaplan-Meier curves illustrating the 90-day event rates of (A) stroke, (B) bleeding events, (C) myocardial infarction and (D) all-cause mortality after MV-PCI and CABG. Outcomes were analysed with Kaplan-Meier estimates and both univariable and multivariable Cox regression. The multivariable model was adjusted for inclusion year, age, sex, diabetes mellitus, hypertension, hyperlipidaemia, prior major bleeding, heart failure, chronic kidney disease, angiographic findings, smoking status, Canadian Cardiovascular Society score and history of myocardial infarction, stroke, PCI and CABG. aHR, adjusted HR; CABG, coronary artery bypass grafting; MV-PCI, multivessel percutaneous coronary intervention; OAC, oral anticoagulant.

• 

  Figure 4. Subgroup analysis of post-procedural atrial fibrillation incidence. Forest plot illustrating adjusted risk ratios for post-procedural atrial fibrillation across predefined subgroups (sex, age, diabetes mellitus, CHA₂DS₂-VASc score and angiographic extent of disease) with p values for interaction. All models were adjusted for inclusion year, age, sex, diabetes mellitus, hypertension, hyperlipidaemia, prior major bleeding, heart failure, chronic kidney disease, angiographic findings, smoking status, Canadian Cardiovascular Society score and history of myocardial infarction, stroke, PCI and CABG. CABG, coronary artery bypass grafting; MV-PCI, multivessel percutaneous coronary intervention.