Atrial Fibrillation After Cardiac Surgery: Incidence and Predictors in a New Zealand Cohort

Background: New-onset atrial fibrillation (AF) is common after cardiac surgery and is associated with short- and long-term adverse outcomes, including stroke and mortality. Little is known about postoperative AF in the New Zealand (NZ) population, including how incidence compares internationally, whether there are ethnic differences in incidence, and what factors place patients at highest risk. Methods: A systematic search of literature was conducted in five databases to locate studies reporting the incidence and independent predictors of new AF after coronary artery bypass grafting (CABG) and/or valve surgery. Incidence was calculated using meta-analysis methods. Vote counting methods identified preoperative and surgical variables that are most commonly associated with the development of AF. Identified predictors were used in a secondary analysis of registry data, involving 1630 adults after CABG and/or valve surgery at Auckland City Hospital with no history of AF. Incidence of AF was examined both by surgery type and by ethnic subgroup. Logistic regression was used to identify independent predictors. Results: Analysis of 37 international studies identified that postoperative AF occurs in almost one in three patients but varies by type of procedure. Risk factors for postoperative AF are inconsistent, except for advanced age and valve surgery. In a NZ hospital, the incidence of AF was 29%, with the lowest incidence after isolated CABG (22%), and highest after combined CABG+valve surgery (42%). Māori, followed by Pākehā had the highest incidence of AF, and Māori and Pasifika with AF were 10-years younger compared to Pākehā. Independent predictors of AF were age, higher body mass index, history of congestive heart failure, and valve procedures. However, these factors were modestly predictive at best. Conclusion: Postoperative AF occurs in a NZ population at comparable rates to international settings. Risk stratification using routinely measured factors is unlikely to inform AF prevention due to limited accuracy, but the risk factors identified improve our understanding of which patients are most affected in a NZ setting and who may benefit from strategies to reduce the burden on patients and communities.