Incidence and predictive factors of atrial fibrillation in paced patients

We have designed a prospective observational study to analyze the incidence and predictive factors of atrial fibrillation (AF) during a long follow-up, in a large population. Atrial fibrillation episodes were documented by the fallback mode switch (FMS) provided by implanted pacemakers. We have included 377 patients (61% men). The pacing indications were atrioventricular (AV) block (49%), sinus node disease (SND, 16%), bradycardia-tachycardia syndrome (BTS, 5%), AV block + SND (19%), AV block + BTS (6%), and BTS + SND (5%). The mean age at implant was 75 +/- 12 (range 28-95). Atrial fibrillation before inclusion was documented in 10% of patients. Drug therapy at first follow-up included beta-adrenergic blockers (17% of the patients), amiodarone (13%), and others (16%). The mean follow-up was 30 +/- 24 weeks. At least one AF episode was stored during follow-up in the memory of 169 pacemakers (45%). Among patients without history of AF at implant, 46% had documented FMS during follow-up. Patients with AF received more antiplatelet medications than patients without AF (P = 0.03). In patients with AF, New York Heart Association functional class was slightly higher, amiodarone and sotalol were more often prescribed, and the proportion of hypertension was higher than in patients without AF. However, these trends were not statistically significant. A significant higher incidence of premature atrial beats was observed in patients with AF than patients without AF (P < 0.0002). Patients with AF had a lower atrial percentage of paced events (55%) than patients without AF (63%, P < 0.02). These preliminary results confirm the high incidence of AF in paced patients and suggest a preventive effect of atrial pacing. The effects of other clinical variables may be confirmed with a longer follow-up in a larger population.