Should paroxysmal atrial fibrillation be treated during cardiac surgery?

Objectives Randomized controlled trials of permanent atrial fibrillation ablation surgery have shown improved outcomes compared with control patients undergoing concomitant cardiac surgery. Little has been reported regarding patients with paroxysmal atrial fibrillation. We hypothesized that treating paroxysmal atrial fibrillation during cardiac surgery would not adversely affect the perioperative risk and would improve the midterm outcomes. Methods From April 2004 to June 30 2012, 4947 patients (excluding those with transcatheter aortic valve implants, left ventricular assist devices, trauma, transplantation, and isolated atrial fibrillation surgery) underwent cardiac surgery, and 1150 (23%) had preoperative atrial fibrillation. Of these, 552 (48%) had paroxysmal atrial fibrillation. Three groups were compared using propensity score matching: treated (n = 423, 77%), untreated (n = 129, 23%), and no atrial fibrillation (n = 3797). Results The treated patients had 30-day mortality similar to that of the untreated patients and those without atrial fibrillation. They had fewer perioperative complications (26% vs 46%, P = .001), greater freedom from atrial fibrillation at the last follow-up visit (81% vs 60%, P = .007), and lower mortality (hazard ratio 0.47, P = .007) compared with the untreated patients. Compared with those without atrial fibrillation, the treated patients had fewer perioperative complications (25% vs 48%, P