Double external direct-current shocks for refractory atrial fibrillation

M of atrial fibrillation (AF) consists of restoring and maintaining sinus rhythm or, if this is not possible, controlling heart rate. Lower risk of systemic embolization and hemodynamic improvement are 2 of the obvious benefits of sinus rhythm. External electrical cardioversion introduced in the late 1950s and popularized by Lown1 in the 1960s is a well-documented and accepted method for cardioversion that successfully restores sinus rhythm in most patients. Lately, intracardiac cardioversion has emerged as an alternative method used especially in cases when external cardioversion has failed.2 Other investigators3 have used transesophageal cardioversion. In similar situations in 1994 we started to use 2 defibrillators.4 Our initial success was encouraging and prompted us to perform a prospective study of the efficacy and safety of double external direct-current (DC) shocks in patients with AF refractory to conventional single shocks. • • • All patients with chronic AF of .1-month duration scheduled for cardioversion between July 1996 and July 1997 in the electrophysiology laboratory at St. Louis University Hospital were considered for the study. No patients were excluded. Each patient was characterized by age, gender, weight, body surface area, underlying heart disease, and antiarrhythmic medication. Information about left atrial size was available from echocardiography in 14 patients. All were euthyroid with normal serum electrolyte levels. The procedure was performed after the patient fasted for at least 6 hours. Conscious sedation was achieved using midazolam hydrochloride and analgesia by fentanyl. The 12-lead electrocardiogram was monitored and stored. The oxygen saturation was monitored transcutaneously and intravenous access was obtained. After shaving, when appropriate, and cleaning the skin with alcohol, 4 self-adhesive, pregelled, polymer-type skin electrodes (13-cm diameter) were positioned and paired for anterior-posterior and apex-anterior cardioversion, respectively. Each pair of electrodes was connected to a Zoll PD 1200 Pacemaker/Defibrillator with a damped sinusoid waveform able to deliver up to 360 J into a 50-V impedance. Operational checks of delivered energy and the discharge buttons were performed daily on all of the defibrillators, and no failures were observed during the study period. For sensing, 3 skin electrodes, unique for each defibrillator, were positioned at the right and left shoulder and lower left abdomen, simulating extremity leads I, II, and III. No attempt was made to choose the same lead for both defibrillators, but the main concern was to obtain a signal without much artifact and with good QRS morphology for triggering both defibrillators independently. No shock electrode was positioned over the breasts in women. The following sequence of shock strength and electrode positioning was used until sinus rhythm was obtained or all 4 shocks used: (1) shock, 200 J anterior-posterior; (2) shock, 360 J anterior-posterior; (3) shock, 360 J apex-anterior; and (4) shocks, 360 J apex-anterior and 360 J anteriorposterior simultaneously. There were at least 3-minute intervals between consecutive shocks, and before each shock transthoracic impedance was measured by ohmmeter. The double shocks were performed by the same person simultaneously pressing the discharge buttons on both defibrillators. Four hours after the procedure, creatine kinase and creatine kinase-MB serum levels were measured. Patients were followed for up to 8 months with arrhythmia status confirmed by electrocardiography. The study was approved by the St. Louis University Institutional Review Board. A total of 24 patients took part in the study. There were 17 men and 7 women, with a mean body weight of 219 lb (range 151 to 371). Six had lone AF and only 5 patients were taking prophylactic antiarrhythmic medication (4 amiodarone and 1 sotalol). Only 9 patients (38%) responded to conventional single shocks with sinus rhythm (Figure 1). Of the remaining 15 patients, 10 (67%) had conversion to sinus rhythm on double shocks, giving a total success rate of 79%. The 5 patients in whom DC cardioversion was unsuccessful did not have any unique features in terms of underlying heart disease, atrial size, duration of AF, skin impedance, or antiarrhythmic medication. Body surface area was no different for the 15 patients whose cardioversion failed, with only 1 defibrillation shock compared with that of patients for whom shock stages 1, 2, and 3 were able to restore normal sinus rhythm (p 5 0.28 using t test for independent samples). Transthoracic impedance varied considerably (75 to 3,000 V) but was no different in those who responded to single shocks compared with those who required double shocks for cardioversion. There was a gradual decrease in impedance with an increase in the number of shocks delivered. In patients who received all 4 shocks, impedance before the first shock was an average of 615 V (range 60 to 3,000) compared with 242 V (range 45 to 750) before the fourth shock. From the Saint Louis University Health Sciences Center, Saint Louis, Missouri. Dr. Bjerregaard’s address is: Electrophysiology and Pacemaker Service, Saint Louis University Hospital, 3635 Vista Avenue at Grand Boulevard, Saint Louis, Missouri 63110-0250. Manuscript received May 19, 1998; revised manuscript received and accepted October 22, 1998.