Early and long-term mortality in 536 patients after theCox-maze III procedure: a national registry-based study.

Background: The cut-and-sew Cox-maze III procedure is the gold standard for surgical treatment of atrial fibrillation. The aim was to study early and long-term mortality based on registry analyses in Swedish Cox-maze III patients.

Methods: Preoperative and early postoperative data were analyzed in 536 patients (male/female (425/111), mean age 57 ± 8.6 years), operated from 1994 to 2009 in 4 centers; 422 (79%) underwent stand-alone Cox-maze III. Atrial fibrillation was paroxysmal in 38% and non-paroxysmal in 62%, mean duration was 7.8 ± 6.3 years. Patients were followed for survival or death in a validated national Cause-of-Death registry. Risk factors associated with observed survival were identified in univariable and multivariable analyses in a standard Cox proportional hazards model.

Results: Four early deaths (0.7%) occurred due to technical complications. At follow-up, 41 of 536 (7.6%) patients had died. Cause of death was cardiovascular in 19 of 536 (3.5%). No ischemic stroke-related death was registered. Univariable risk factors for all-cause mortality included hypertension (hazard ratio [HR] 2.8, confidence interval [CI] 1.5 to 5.3), heart failure (HR 2.4, CI 1.3 to 4.3), concomitant surgery (HR 2.2, CI 1.1 to 4.1), and postoperative complications (HR 2.5, CI 1.3 to 4.8). Gender, non-paroxysmal atrial fibrillation and long arrhythmia duration did not confer increased risk of death. Multivariable risk factors were hypertension (HR 2.9, CI 1.5 to 5.5) and postoperative complications (HR 2.4, CI 1.2 to 4.6). Survival for cardiovascular death at 5, 10, and 15 years was 98%, 96%, and 93%, respectively.

Conclusions: Registry-based follow-up showed low early and long-term cardiovascular mortality and no stroke-related mortality. This is important baseline information when evaluating current surgical and nonsurgical treatment of atrial fibrillation.