Atrial fibrillation is a reported complication of automatic defibrilla
tor implantation. Its incidence, risk factors, time-course, and compli
cations have not been well-defined. Accordingly, data from 117 patient
s who underwent defibrillator implantation via a thoracotomy (n = 71)
or nonthoracotomy (n = 46) approach were reviewed. To identify risk fa
ctors, 15 variables of potential predictive value were chosen and anal
yzed. Atrial fibrillation developed in 26/117 patients (22%) during th
e early postoperative period and all but one of these 26 patients had
undergone thoracotomy (P < 0.001). Patients who developed atrial fibri
llation tended to be older than those who did not (63 +/- 2 vs 58 +/-
2 years, P = 0.04) and more frequently had a prior history of paroxysm
al atrial fibrillation (31% vs 10%, P = 0.02). The were also less like
ly to be taking Class I or III antiarrhythmic drugs (1/26 vs 24/91, P
= 0.01). By multivariate analysis, operative approach (P < 0.001), the
absence of antiarrhythmic drug therapy (P = 0.006), and a prior histo
ry of atrial fibrillation (P = 0.003) were significant independent var
iables. digoxin neither prevented the occurrence of atrial fibrillatio
n nor slowed the maximal ventricular response. The mortality and compl
ication rates did not differ between the two groups. The major adverse
effect of postimplant atrial fibrillation was automatic defibrillator
discharge; six patients received between 1 and 11 discharges for atri
al fibrillation with rapid ventricular rates.