Acute direct current cardioversion: how 'dangerous' is it?

This editorial refers to ‘Arrhythmic complications after electrical cardioversion of acute atrial fibrillation: TheFinCV study’ by T. Grönberg et al., on page 1432. Atrial fibrillation (AF) of <48 h can, according to the ESC Guidelines, be acutely cardioverted, if the physician is convinced that the patient is certain about the time of arrhythmia onset.1 Cardioversion can be performed either by a direct current (DC) shock or by various pharmacological means. Within Europe, there are large variations regarding which method is used.2 In Scandinavian countries, DC cardioversion is the dominant (∼90%) method, whereas pharmacological conversion is mostly used in southern Europe. There are no prospective head-to-head studies comparing efficacy, safety, and long-term risks of these two methods. In this issue of the Journal, Grönberg et al.3 present data from a large retrospective cohort of almost 7000 acute DC cardioversions in patients with AF. Although this study is retrospective, which always carries its limitations, the database is large and well-described. The authors correctly point out that the retrospective nature of the study might underestimate the incidence of minor adverse events. The manuscript deals specifically with acute arrhythmic complications. The main finding is a low (0.9%) incidence of significant arrhythmias associated with DC cardioversion. These arrhythmias were all various types of sinus node dysfunction; no ventricular arrhythmias needing intervention have occurred. Approximately 45% of patients with significant bradycardias, received a permanent pacemaker 2 months after cardioversion. Old age, female gender, and unsuccessful cardioversion were independent predictors of bradycardia. One patient died within 1 h due to a ruptured aortic aneurysm. No strokes occurred in the acute setting. Other complications, such as laryngeal spasm during anaesthesia,4 were not reported. It is also of interest that ongoing medication with antiarrhythmic drugs including digitalis did not predict an increase in the occurrence of bradycardia. The success rate defined as sinus rhythm at discharge was, as expected, high (94%). This is in agreement with the findings in the Euro Heart Survey on AF.5 These data underline the common clinical observation that DC cardioversion can be performed in a safe and effective way even in the acute setting. The main disadvantage is the need to be fasting and the need to find an anaesthesiologist who is willing to help out. This disadvantage may be overcome by using modern short-acting anaesthetics, such as propofol, by an experienced cardiologist without the need for an anaesthesiologist, thus facilitating this procedure.6 In the emergency room, however, there is usually a lack of cardiologists with that experience. In this study, DC cardioversion was performed by a stepwise increase in energy starting as low as 70 J and ending at 360 J. Such a protocol would lead to an increased number of shocks with a higher accumulated energy level compared with giving the first shock at maximum energy. Whether such a strategy would influence the incidence of significant arrhythmias can only be speculated. Gallagher et al.7 showed that the incidence of arrhythmias associated with DC cardioversion was not related to the total amount of energy. It seems reasonable to assume that one high-energy shock given as the first shock would shorten the procedure time and would be as effective without being proarrhythmic or causing myocardial damage.8 This has been observed in a prospective randomized study.9 Another interesting finding is the fact that >40% of patients with significant bradycardia ended up with a permanent pacemaker within 2 months after the conversion. Why these patients received a pacemaker, is not clear from the study. It is evident, however, that DC cardioversion is an effective but ‘rough’ method to disclose underlying sinus node disease in patients with AF. A positive finding is that there were no ventricular arrhythmias requiring intervention, reported in this large group, irrespective of whether patients were on digitalis and/or antiarrhythmic drugs. Most probably, this is a sign that patients were well-controlled with respect to electrolytes and drug concentrations prior to the procedure. Finally, some important questions which are not answered in this study. The incidence of thrombo-embolic complications after acute DC cardioversion has not been reported. This is a pity as we have no reliable or current data in the literature, regarding this important complication. Cardiac emboli after DC cardioversion usually appear within a timeframe of 30 days and as these patients usually are discharged within 24 h, thrombo-embolic events associated with cardioversion may be overlooked. In addition, it would have been interesting to have information on the CHA2DS2VASc score1 and how many patients were on oral anticoagulants at discharge. One could speculate that assessment of the need for oral anticoagulation might be overlooked after successful cardioversion in the often chaotic environment of the emergency department. It would have been interesting to evaluate if there were any differences in thrombo-embolic events among patients with and without oral anticoagulation after acute cardioversion. The lack of these important data represents a white spot in this manuscript and I would strongly encourage the authors to add this important information in future publications. Conflict of interest: none declared.