Atrial fibrillation: Cardioversion

INTRODUCTION — The restoration (cardioversion) to sinus rhythm (SR) from atrial fibrillation (AF) is performed primarily to improve symptoms, but it may also prevent tachycardia-induced cardiomyopathy, facilitate management of heart failure, and reduce the risk of inappropriate shocks in those with implanted defibrillators. This topic will focus on our approach to cardioversion and the efficacy and safety of the two most widely used approaches: electrical (direct current) and pharmacologic cardioversion. Other related topics include:

●(See "Antiarrhythmic drugs to maintain sinus rhythm in patients with atrial fibrillation: Clinical trials".)

●(See "Antiarrhythmic drugs to maintain sinus rhythm in patients with atrial fibrillation: Recommendations".)

●(See "Prevention of embolization prior to and after restoration of sinus rhythm in atrial fibrillation".)

●(See "Restoration of sinus rhythm in atrial flutter".)

INDICATIONS — In this topic, we are generally referring to cardioversion in stable patients with recently diagnosed symptomatic AF that does not terminate spontaneously.

Cardioversion is indicated to improve symptoms and hemodynamic status from AF. Among patients with early AF and who are at high risk for cardiovascular complications, cardioversion with accompanying rhythm control strategy of AF management can reduce cardiovascular death or stroke. Patients who may benefit from rhythm versus rate control are discussed separately. (See "Management of atrial fibrillation: Rhythm control versus rate control" and "Management of atrial fibrillation: Rhythm control versus rate control", section on 'Indications for initial rhythm control'.)

Cardioversion is most commonly performed in patients who are expected to have long-term maintenance of sinus rhythm, who are expected to convert to normal sinus rhythm, and are at low risk for cardioversion-associated complications (eg, thromboembolism). These may be patients with persistent AF (ie, lasting >7 days) or even paroxysmal AF, if AF is highly symptomatic and cannot be otherwise controlled. It is also preferred in other patient groups.

●Long-term rhythm control – Patients who will be placed on long-term antiarrhythmic drugs or who will undergo catheter ablation will have SR restored as the initial part of that process. (See "Maintenance of sinus rhythm in atrial fibrillation: Catheter ablation versus antiarrhythmic drug therapy" and "Antiarrhythmic drugs to maintain sinus rhythm in patients with atrial fibrillation: Recommendations" and "Atrial fibrillation: Catheter ablation".)

●Long-standing persistent AF – It may be reasonable to attempt to restore SR in patients with long-standing, persistent AF who are symptomatic (or occasionally in patients with presumed permanent AF). An example of the latter is a patient with long-duration AF who now has developed increasing heart failure or worsening cardiomyopathy felt to be related to poorly controlled ventricular rate. However, maintenance of sinus rhythm even for a short time period is not assured. (See "Arrhythmia-induced cardiomyopathy" and "The management of atrial fibrillation in patients with heart failure", section on 'Rhythm control' and "The management of atrial fibrillation in patients with heart failure", section on 'Rhythm control'.)

Frequently, the only way to determine if subtle, nonspecific symptoms are attributable to AF is to restore SR to observe for symptom improvement. Symptom improvement may be delayed for several weeks due to markedly delayed recovery of atrial mechanical function.

●Cardioversion may be used in patients who have very infrequent, persistent episodes who do not respond to a pill-in-the-pocket approach. These individuals may cardioverted as needed from time to time. (See 'Pill-in-the-pocket' below.)

Unstable patients — In some hemodynamically unstable patients who manifest with signs or symptoms such as hypotension, altered mental status, or heart failure, if there is time to attempt ventricular rate slowing or to wait for possible spontaneous reversion to SR, cardioversion may be deferred. Emergency cardioversion should be performed if the patient is hemodynamically compromised due to an uncontrolled rapid ventricular rate or the lack of atrial contraction is thought to impair their cardiac output. If angina is felt to be related to the hemodynamic compromise or lack of atrial contraction with AF, we perform emergency cardioversion; however, the risk for a thromboembolic event needs to be considered. This includes patients with severe acute heart failure, ongoing myocardial ischemia, or hypotension. However, if hypotension occurs with a slow or moderate ventricular response (<110 beats/min), other causes of hypotension should be sought, such as myocardial infarction, pulmonary embolism, sepsis, pericardial effusion/tamponade, or hypovolemia. (See "Hemodynamic consequences of atrial fibrillation and cardioversion to sinus rhythm", section on 'Adverse hemodynamics in AF'.)

REASONS NOT TO PERFORM CARDIOVERSION — Patients in whom it is reasonable to avoid early cardioversion include:

●Those who are asymptomatic or minimally symptomatic, particularly those with multiple comorbidities, advanced age, or poor overall prognosis, where the risks of undergoing cardioversion and/or pharmacologic rhythm control may outweigh the benefits of restoring SR. Not performing cardioversion may also be appropriate in those with a low likelihood of long-term maintenance of SR, such as those with marked left atrial enlargement/dilatation, significant mitral regurgitation or stenosis, or those with florid hyperthyroidism.

●Those who cannot receive anticoagulation, if indicated, are generally not candidates for cardioversion unless the duration of AF is less than 48 hours. (See 'Anticoagulation' below and "Prevention of embolization prior to and after restoration of sinus rhythm in atrial fibrillation", section on 'AF duration uncertain or 48 or more hours'.)

●Those with left atrial thrombus identified during a transesophageal echocardiogram or who have presented with a thrombotic event (eg, stroke or transient ischemia). (See "Prevention of embolization prior to and after restoration of sinus rhythm in atrial fibrillation", section on 'Transesophageal echocardiography-based approach'.)

●Some experts would not pursue cardioversion in a person who has previously failed the procedure (ie, had only a brief period of sustained sinus rhythm following a prior cardioversion).

●Other reasons to consider not cardioverting include:

•Those where AF has been continuously present for more than one year [1-6].

•Those with a left atrium that is markedly enlarged (atrial dimension >6.0 cm, transthoracic echocardiographic biplane atrial volume index >48 mL/m²) [6-11].

•Those with AF recurrence while taking adequate doses of appropriate antiarrhythmic drug therapy and who have recently undergone cardioversion. Drug refractory patients may have successful conversion to SR but are less likely to maintain SR long term. Cather ablation for AF may be a solution to maintaining SR for some patients in this situation.

•Those who do not or no longer respond to more than one antiarrhythmic drug are less likely to maintain SR with other drugs.

•Those for whom cardioversion with long-term maintenance of SR is likely to be unsuccessful if the underlying precipitant (eg, thyrotoxicosis, pericarditis, pneumonia, or mitral valve disease) has not been corrected prior to cardioversion.

PRECARDIOVERSION ISSUES — Prior to cardioversion with either electrical energy or antiarrhythmic drug therapy in stable patients, decisions regarding rate control, timing, and anticoagulation need to be made.

Ventricular rate control — Patients with a rapid resting ventricular response (>100 beats per minute) usually need control of their ventricular rate to improve symptoms. This can be achieved by oral (and occasionally intravenous) administration of short-acting beta blockers or nondihydropyridine calcium channel antagonists, such as diltiazem or verapamil [1]. In the acute setting, the target resting ventricular rate should usually be 80 to 100 beats per minute. (See "Control of ventricular rate in patients with atrial fibrillation who do not have heart failure: Pharmacologic therapy".)

Anticoagulation — Most patients who will undergo cardioversion should be anticoagulated as soon as the decision is made to cardiovert or after assessment of their clinical thromboembolic risk based on their CHA₂DS₂-VASc score. Issues related to anticoagulation around the time of cardioversion are discussed in detail separately. (See "Prevention of embolization prior to and after restoration of sinus rhythm in atrial fibrillation", section on 'AF duration less than 48 hours'.)

Timing — Occasionally, hemodynamically unstable patients may need to be cardioverted urgently, without time to institute optimal oral anticoagulation (see 'Anticoagulation' above). For these patients who are not chronically anticoagulated but are candidates, intravenous heparin or low molecular weight heparin should be started as soon as possible, possibly with a loading dose of heparin prior to cardioversion due to more advanced transient impairment of atrial appendage mechanical function after cardioversion.

For stable patients, the timing of elective cardioversion is determined in large part by the anticoagulant strategy chosen. If possible, patients with valvular and nonvalvular AF duration longer than 48 hours or of unknown duration should be therapeutically anticoagulated for at least four weeks or receive short-term anticoagulation followed by screening transesophageal echocardiography to exclude an atrial thrombus prior to cardioversion [12]. Some of our authors use this approach if the duration is longer than 24 hours. Patients with a history of neurologic event, diabetes, and heart failure appear to be at very high risk for thromboembolism post-cardioversion, and four weeks of anticoagulation or short-term anticoagulation with transesophageal echocardiography should be considered [13].

Evaluation for underlying cause — It is useful to consider the potential for an underlying or precipitating cause, such as hyperthyroidism, but also acute pulmonary embolism, myopericarditis, pneumonia, pericarditis, or sepsis, though atrial fibrillation is rarely the sole manifestation of any of these other than hyperthyroidism. (See 'Reasons not to perform cardioversion' above and "Atrial fibrillation: Overview and management of new-onset atrial fibrillation", section on 'Triggers'.)

ELECTRICAL VERSUS PHARMACOLOGIC CARDIOVERSION — This section compares an electrical with a pharmacologic approach for cardioversion. However, patients who fail pharmacologic cardioversion are generally referred for electrical cardioversion. Thus, the comparison is really an electrical compared with a pharmacologic/electrical approach.

For a first episode, electrical cardioversion is preferred in most cases. This is particularly true for younger patients (<65 years) even if they have no symptoms. For other patients (not the first episode) who need to be cardioverted from AF to SR, either electrical or pharmacologic cardioversion (potentially followed by electrical cardioversion) is a reasonable approach. The choice between the two should take into account the strengths and weaknesses of each approach as well as patient preference. For most patients, we proceed directly to electrical cardioversion to avoid the potential of drug side effect and prolonged monitoring.

Reasons to prefer electrical rather than pharmacologic cardioversion include:

●Avoiding drug side effects, including but not limited to transient hypotension or prolongation of the QTc interval.

●Avoiding need for prolonged telemetric monitoring to screen for a proarrhythmic response. This can result in longer observation periods in the emergency department or procedural area.

●Some antiarrhythmic agents have the potential to convert AF to atrial flutter.

●Superior efficacy as compared with pharmacologic cardioversion.

In contrast, potential benefits from a pharmacologic (rather than an electrical) approach include:

●Avoiding the risk(s) of sedation (required at the time of electrical cardioversion) (see "Cardioversion for specific arrhythmias", section on 'Preparation and personnel')

●Testing for drug tolerance, in the occasional patient (eg, someone with long-standing AF) in whom it has been decided to continue oral antiarrhythmic drug therapy (see 'Preprocedural antiarrhythmic drugs' below)

Studies comparing the two approaches are somewhat limited.

Mortality outcomes following electrical and pharmacologic cardioversion were similar in an observational cohort of 7175 patients from a large anticoagulation registry [14]. There were 2427 (34 percent) patients who received pharmacological cardioversion and 4748 (66 percent) who received electrical cardioversion. During one-year follow-up, event rates (per 100 patient years) for mortality in patients who received electrical and pharmacological cardioversion were 1.36 (1.13 to 1.64) and 1.70 (1.35 to 2.14), respectively.

In the RAFF2 trial, 396 patients with acute AF were randomly assigned to pharmacologic cardioversion with intravenous procainamide or placebo followed by electrical cardioversion, if necessary [15]. The primary outcome of conversion to normal SR for at least 30 minutes at any time after randomization and up to a point immediately after three shocks occurred with equal frequency in the two groups (96 versus 92 percent, respectively; p = 0.07). After procainamide infusion, 52 percent of patients converted (median time of 23 minutes), therefore not requiring subsequent electrical cardioversion. RAFF2 demonstrates that it is feasible for patients to be rapidly cardioverted with medical therapy in the emergency department, resolving their acute symptoms and enabling discharge home.

However, we do not anticipate adoption of intravenous procainamide in this setting due to the additional time required for chemical cardioversion compared with electrical cardioversion and the general lack of familiarity and experience with procainamide and its potential side effects by emergency department clinicians and general cardiologists. Further, the drug has a substantial half-life, making the risk of torsades de pointes due to early discharge problematic. Ibutilide is an antiarrhythmic drug that is more familiar to most clinicians caring for these patients and more likely to be chosen in this setting if an initial pharmacologic approach is attempted. (See 'Specific antiarrhythmic drugs' below.)

ELECTRICAL CARDIOVERSION — Electrical cardioversion of AF is a commonly performed medical procedure with a high success rate and a low complication rate when performed in the setting of anticoagulation. (See 'Arrhythmic complications' below and "Prevention of embolization prior to and after restoration of sinus rhythm in atrial fibrillation", section on 'Rationale for anticoagulation'.)

The overall immediate success rate of electrical cardioversion is greater than 90 percent [4,5,7-10,16,17]. However, the overall success rate falls as the AF duration increases and early recurrences are possible.

Preprocedural antiarrhythmic drugs — Most patients who undergo their first direct current cardioversion do not receive pretreatment with antiarrhythmic drugs. This is particularly true for unstable and other patients who need immediate restoration of SR. Pretreatment with antiarrhythmic drugs may also be omitted in patients in whom the potential for drug-induced arrhythmias is a concern.

The following are reasons to consider initiation of antiarrhythmic drug therapy prior to electrical cardioversion:

●To increase the likelihood of successful cardioversion [1]. Sotalol, ibutilide, and dofetilide seem to decrease the cardioversion energy requirement and may be helpful in refractory patients. (See 'Electrical versus pharmacologic cardioversion' above.)

●To prevent recurrent episodes of AF soon after cardioversion. Ibutilide or verapamil may be useful in preventing early recurrences of AF after cardioversion, particularly if it is of recent onset [18,19].

●To allow for an evaluation of tolerability of one or more drugs. For those patients in whom a decision has been made to attempt the long-term maintenance of SR with antiarrhythmic drug therapy, we believe it is reasonable to preferentially select an agent for cardioversion that can also be used for long-term maintenance of SR (such as amiodarone, flecainide, or propafenone).

●To potentially avoid the need for intravenous sedation and subsequent electrical cardioversion.

●For patients with recurrent AF soon after (<few months) prior direct current cardioversion.

Procedure — A detailed discussion of electrical cardioversion is found elsewhere. (See "Cardioversion for specific arrhythmias" and "Basic principles and technique of external electrical cardioversion and defibrillation".)

The following are a few important points:

●If possible, the patient should be fasting for at least six hours. Oxygen saturation and electrolytes (particularly serum potassium) should be close to normal and anticoagulation status should be reviewed (anticoagulation status is crucial to prevent thromboembolism in the week after conversion; the need for and the type of anticoagulation depends on the patient’s clinical presentation as discussed elsewhere), and drug levels, when measured, should be within the therapeutic range. Digoxin need not be withheld unless digitalis toxicity is suspected [20]. (See "Digitalis (cardiac glycoside) poisoning".)

●Electrical cardioversion, synchronized to the QRS complex, should be performed while the patient is under the influence of procedural sedation and is having blood pressure, heart rate, oxygen saturation, and carbon dioxide capnography monitored. Generally, cardioversion should be done in a situation where airway equipment and airway expertise are present. (See "Procedural sedation in adults in the emergency department: General considerations, preparation, monitoring, and mitigating complications".)

●Electrical cardioversion is very rarely associated with complications such as thromboembolism, respiratory distress due to acute heart failure, myocardial necrosis, skin burns, transient ventricular dysfunction, and transient arrhythmias including bradycardia. These are discussed in detail separately. (See "Cardioversion for specific arrhythmias", section on 'Complications'.)

Postprocedural considerations — Most cardioversions take place without any significant adverse events. However, the following should be kept in mind.

Hemodynamic changes after reversion — Left ventricular systolic function often improves and may normalize after achievement of SR, although it may take months for this to happen. In one study, left ventricular function improved soon after the restoration of SR, a change that was attributed to the reduction in heart rate and return of atrial mechanical contraction [21]. Atrial contractility improves more slowly (atrial "stunning”) when the duration of AF prior to cardioversion is prolonged [22]. (See "Mechanisms of thrombogenesis in atrial fibrillation", section on 'Atrial stunning after cardioversion'.)

In uncommon instances, there can be ventricular "stunning." In these rare cases, there can be acute pulmonary edema after cardioversion. (See "Cardioversion for specific arrhythmias", section on 'Myocardial dysfunction'.)

Recurrence of AF after cardioversion — The recurrence rate of AF after electrical cardioversion can be high, especially in patients with prior recurrences, a large left atrium, underlying structural disease, and a history of heart failure. There are times when it is appropriate to start an antiarrhythmic drug before cardioversion to lower the risk of recurrence soon after cardioversion. For the first attempt, however, especially if the heart is normal and there is no heart failure, a cardioversion without an antiarrhythmic is reasonable, particularly if maintenance antiarrhythmic therapy is not planned [23].

The timing of AF recurrence was evaluated in a review of 61 patients who had daily electrocardiographic recordings using transtelephonic monitoring: 57 percent had recurrent AF during the first month after cardioversion, with a peak incidence during the first five days [24]. This is one of the reasons that post-cardioversion anticoagulation should be used independently of CHADS₂ or CHA₂DS₂-VASc score, as this may obviate the need for a transesophageal echocardiogram or an additional month of anticoagulation prior to another cardioversion.

Electrical cardioversion may be repeated if AF recurs acutely in patients who have not been pretreated with antiarrhythmic therapy. In such patients or those in whom cardioversion fails, the combination of an atrioventricular (AV) nodal blocker plus intravenous loading with amiodarone, ibutilide, or procainamide, or oral dosing with flecainide, sotalol, or propafenone may restore SR pharmacologically.

If the AF persists, electrical cardioversion may be performed and if successful, the patient may be placed on long-term antiarrhythmic therapy. (See 'Pharmacologic cardioversion' below.)

A separate issue is whether to perform repeat cardioversion for later recurrence. One study has shown that two repeat cardioversions increase the likelihood of maintaining SR long-term [25]. However, an important consideration is the time interval between AF episodes. In both of the above reports, a third cardioversion for recurrent AF was performed within seven days of the second cardioversion. Repeated cardioversion may be a reasonable approach for patients with AF that recurs after a longer duration of SR (eg, years). It is most likely to be successful in younger patients with fair exercise tolerance and AF duration of less than three years [26]. Repeat cardioversion or nonpharmacologic therapy (catheter ablation of AF or AV nodal ablation) may be necessary in patients who remain symptomatic when in recurrent AF.

Arrhythmic complications — Bradyarrhythmias are occasionally seen after electrical cardioversion, especially when patients are being treated with drugs to control the rate in AF, in older patients with tachycardia-bradycardia syndrome, and those with known preexisting bradycardia; ventricular tachyarrhythmias are rare but torsades de pointes can occur in patients treated with class III antiarrhythmics, especially in the first 24 hours after cardioversion and especially if there is bradycardia after cardioversion. In a retrospective study of 6906 cardioversions of acute AF in 2868 patients, 63 patients had bradyarrhythmias (51 episodes of asystole >5 seconds and 12 episodes of bradycardia with heart rate <40 bpm) [27]. No episodes of ventricular arrhythmia requiring intervention were reported.

Patients with sinus or AV node dysfunction, as in the tachycardia-bradycardia syndrome, are at higher risk for prolonged sinus pauses and bradycardia if AF is converted without a backup pacemaker. Nevertheless, the hemodynamic benefit from cardioversion may be sufficient to warrant restoration of SR with control of the atrial rate and/or AV conduction with a pacemaker. Physiologic pacing in patients with sinus node dysfunction and SR appears to decrease the likelihood of recurrent AF, especially if the initiation of AF episodes is bradycardia dependent [28]. (See "Sinus node dysfunction: Epidemiology, etiology, and natural history" and "Sinus node dysfunction: Treatment".)

PHARMACOLOGIC CARDIOVERSION — As discussed above, we generally prefer electrical cardioversion to pharmacologic cardioversion in most patients.(See 'Electrical versus pharmacologic cardioversion' above.)

Flecainide, propafenone, ibutilide, dofetilide, procainamide, and, to a lesser degree, amiodarone have efficacy for pharmacologic conversion of AF. Of these, we prefer flecainide or propafenone unless there is evidence for coronary artery disease, left ventricular systolic dysfunction, or the duration of AF is greater than seven days, in which case dofetilide, or to a lesser degree, amiodarone or ibutilide have some role for medical conversion. None of these drugs is as efficacious as electrical cardioversion. In order to avoid side effects and proarrhythmia, we do not use two or more antiarrhythmic drugs at the same time.

Structural heart disease is a contraindication to the use of some of the antiarrhythmic drugs. For the purpose of this topic, we define it as any condition in which there is a deviation in the size, shape, function, or structure of the atria or ventricles (such as left ventricular hypertrophy or dilated cardiomyopathy). This also includes coronary artery disease in the absence of left ventricular dysfunction. The definition does not include processes that alter the electrical properties of the heart, such as electrical activation, conduction, automaticity, or refractoriness.

Specific antiarrhythmic drugs — The following antiarrhythmic drugs have been used to cardiovert patients in AF:

●Flecainide is a very effective antiarrhythmic drug for the pharmacologic conversion of a patient with AF of short (<24 hours) duration. Flecainide should not be used in patients with structural heart disease, particularly those with left ventricular systolic dysfunction or with coronary artery disease. (See "Major side effects of class I antiarrhythmic drugs".)

Intravenous flecainide (2 mg/kg over 10 minutes) acutely reverts recent onset AF in 67 to 92 percent of patients within six hours and is more effective than procainamide, sotalol, propafenone, and amiodarone. The intravenous preparation is not available in the United States. An oral dose of 200 to 300 mg has also been used.

A single, large oral dose of flecainide (100 to 400 mg) is effective for AF reversion [29,30]. One study randomly assigned 79 patients to intravenous or large oral dose of flecainide. The rate of reversion to SR was similar at two hours (64 versus 68 percent for oral drug) and eight hours after treatment (72 versus 75 percent); however, the mean time to reversion was shorter with intravenous flecainide (52 versus 110 minutes) [31].

●Propafenone is significantly more effective in paroxysmal as opposed to persistent AF, with rates likely approaching those seen with flecainide. As with flecainide, we do not recommend propafenone in patients with structural heart disease, particularly those with left ventricular systolic dysfunction or coronary artery disease. (See "Major side effects of class I antiarrhythmic drugs".)

Intravenous propafenone (2 mg/kg over 10 to 20 minutes) is used in Europe for the acute termination of AF. Conversion rates of 23 to 54 percent in patients with AF of variable duration have been reported [32,33].

Oral propafenone can be given as a large dose of 450 to 600 mg. A single oral loading dose reverted AF in 56 to 83 percent of patients, depending upon the duration of AF, in one study [34,35]. The administration of propafenone prior to electrical cardioversion does not alter the energy requirements for, or the success rate of, cardioversion [36].

●Dofetilide has primarily been studied for the medical conversion of persistent AF. Dofetilide has not been directly compared with amiodarone or vernakalant. Its principal use is to help maintain SR, rather than to be used for cardioversion.

It has been used successfully in patients with structural heart disease. As an oral agent, it is rarely used solely for the purpose of cardioversion. It can be started prior to anticipated electrical cardioversion to help maintain SR shortly after either electrical or spontaneous cardioversion, or in those patients in whom it is chosen for the long-term maintenance of sinus rhythm. (See "Clinical use of dofetilide".)

Dofetilide is more effective than placebo for the conversion of persistent AF to SR, particularly at a dose of 500 mcg twice daily [37-39]. As an example, the SAFIRE-D study randomized 325 patients (including 60 percent with structural heart disease) with AF (n = 277) or atrial flutter (n = 48) to oral therapy with 125, 250, and 500 mcg twice daily [37]. Overall conversion rates were 6, 9.8, and 30 percent, respectively. Among patients who convert with dofetilide, successful conversion occurred in 70 percent within 24 hours, while 91 percent converted within 36 hours. Postmarket uncontrolled trials have suggested that oral dofetilide is useful in the medical conversion of persistent AF in up to 60 percent of patients [39]. Dofetilide also appears safe and efficacious in the setting of heart failure (including patients with AF); this is discussed separately. (See "Clinical use of dofetilide", section on 'Heart failure'.)

Dofetilide has the disadvantage of requiring the patient to be hospitalized, as it can cause nonsustained ventricular tachycardia (VT), torsades de pointes, or sudden death [40,41]. In the SAFIRE-D trial, the incidence of torsades de pointes was 1.2 percent [37]. This risk is minimized by dosing based on the creatinine clearance and avoidance of other drugs that may cause torsades de pointes. Initiation of dofetilide is required to be in-hospital, under telemetry conditions by a doctor certified in its use, with a total of at least six dosages given before discharge. (See "Clinical use of dofetilide", section on 'Protocol for administration'.)

●Amiodarone (either oral or intravenous) is not particularly effective for cardioversion (though is effective for maintenance of SR postcardioversion). If reversion to SR occurs it does so several hours later than with flecainide, propafenone, ibutilide, and vernakalant [42-52]. Intravenous amiodarone may be more effective in converting AF after it has been given for hours and days. Oral amiodarone requires long-term loading and is effective in converting about 25 percent of patients with persistent AF to SR after six weeks of loading. Thus, we do not recommend it solely for the purpose of cardioversion. It is not approved by the US Food and Drug Administration for the treatment of AF. (See "Amiodarone: Clinical uses" and "Amiodarone: Adverse effects, potential toxicities, and approach to monitoring".)

However, amiodarone may occasionally have value before cardioversion in patients who will receive the drug long term for maintenance of SR and may be considered as adjunctive therapy to increase the likelihood of successful cardioversion in patients who are known to be refractory to electrical cardioversion or in those in whom there is a concern about early relapse. (See "Antiarrhythmic drugs to maintain sinus rhythm in patients with atrial fibrillation: Recommendations", section on 'Summary and recommendations'.)

Intravenous amiodarone is given at a dose of 150 mg over 10 minutes, with a subsequent infusion of 1 mg/minute for six hours, then 0.5 mg/minute for 18 hours or change to oral maintenance dosing (eg, 100 to 200 mg once daily) [53]. It should be kept in mind that the drug will likely have a significant rate-slowing effect, which may be beneficial in some patients [54]. A table on monitoring for adverse effects is available (table 1).

●Vernakalant is available in Europe and Canada in intravenous forms for the rapid conversion (50 percent conversion within 10 minutes) of recent onset AF (less than eight days’ duration for patients not undergoing surgery, and less than four days’ duration for post-cardiac surgery patients) to SR. This drug is not available in the United States. (See "Antiarrhythmic drugs to maintain sinus rhythm in patients with atrial fibrillation: Clinical trials", section on 'Vernakalant'.)

In a 2019 systematic review and meta-analysis of nine trials (n = 1358) comparing vernakalant with placebo, amiodarone, or ibutilide [55], significant methodological bias was found in four trials. The following was also found:

•Vernakalant was superior to placebo for conversion within 90 minutes (50 percent conversion; risk ratio 5.15, 95% CI 2.24-11.84).

•No significant difference in the rate of conversion was found comparing vernakalant with active drug (56 versus 24 percent; risk ratio 2.40, 95% CI 0.76-7.58).

●Ibutilide, due to its propensity to prolong repolarization and the QT interval, has the potential to provoke torsades de pointes. Ibutilide is available only as an intravenous preparation (1 mg over 10 minutes and potentially repeated once after 20 minutes) and is useful for the acute reversion of AF [56-58]. It has been used in patients with structural heart disease (but without heart failure). We use ibutilide very rarely in this setting. (See "Therapeutic use of ibutilide".)

In trials, the acute AF conversion rate is 28 to 51 percent. Ibutilide is more effective at converting atrial flutter to SR, with conversion rates of 50 to 75 percent. Ibutilide can work in acute situations of persistent AF and can be used (with caution) in patients with structural heart disease. However, while these rates seem low compared with flecainide or propafenone, ibutilide has not been directly compared with these other drugs. In addition, the average conversion times with ibutilide seem shorter than with flecainide and propafenone.

Arrhythmia conversion occurred within a mean of 27 to 33 minutes after the start of the infusion [57,58]. In comparative studies, ibutilide has been more effective for AF reversion than procainamide (51 versus 21 percent and 32 versus 5 percent) [59,60] or intravenous sotalol (44 versus 11 percent) [61].

In four large series, the rate of torsades de pointes with ibutilide ranged between 3.6 and 8.3 percent [56-58,60]. Sustained episodes of torsades de pointes, requiring cardioversion, were seen in 1.7 to 2.4 percent. In addition to polymorphic VT, nonsustained monomorphic VT occurred in an additional 3.2 to 3.6 percent of patients [57,58]. Because of the risk of ventricular proarrhythmia, patients treated with ibutilide should be observed with continuous rhythm monitoring for at least four hours after the infusion or until the QTc interval has returned to baseline. Pretreatment with intravenous magnesium appears to minimize the risk of ibutilide-induced torsades de pointes without affecting the efficacy of conversion [62]. Risk factors for the development of torsades de pointes with ibutilide are heart failure, baseline increase in QTc interval, and low potassium or magnesium. Electrolytes should be checked and normalized before cardioversion with ibutilide.

We occasionally use ibutilide in patients resistant to electrical cardioversion and in patients for whom anesthesia, in conjunction with direct current cardioversion, is not readily available.

The recommended dose varies with patient size. For patients weighing less than 60 kg, the recommended dose is 0.01 mg/kg infused over 10 minutes. If the arrhythmia does not terminate 10 minutes after the end of the infusion, a second bolus (same dose over 10 minutes) can be given. For patients weighing more than 60 kg, the recommended starting intravenous dose is 1 mg over 10 minutes. If AF does not terminate 10 minutes after the end of the infusion, a second bolus of 1 mg over 10 minutes can be given. Most patients studied have had arrhythmia for less than 90 days. Efficacy has not been proven with an AF duration exceeding 90 days. We do not use ibutilide in combination with other antiarrhythmic medications.

Less effective or ineffective drugs — Several medications such as quinidine and procainamide are no longer used for cardioversion as most of the agents presented above have greater efficacy or fewer side effects or both.

The following antiarrhythmic drugs are not particularly effective for the restoration of SR:

●Sotalol – Oral sotalol is less effective than quinidine [63] and equally effective to amiodarone for chemical cardioversion of AF (27 percent at 28 days) [64]. Intravenous sotalol is less effective than intravenous flecainide or ibutilide for reversion of AF [61,65]. Thus, we do not recommend either form for chemical cardioversion of AF. (See "Clinical uses of sotalol".)

●Dronedarone – We do not recommend dronedarone for the restoration of SR in patients with AF since conversion rates are very low [66]. (See "Antiarrhythmic drugs to maintain sinus rhythm in patients with atrial fibrillation: Recommendations".)

●Rate-control drugs such as digoxin, nondihydropyridine calcium channel blockers (eg, diltiazem or verapamil), and beta blockers have not been effective in restoring SR in placebo-controlled studies. Many clinicians overestimate the effectiveness (for cardioversion) of these agents [1,67-72].

Pill-in-the-pocket — A "pill-in-the-pocket" approach with either flecainide (<70 kg: 200 mg; may not repeat in ≤24 hours; ≥70 kg: 300 mg; may not repeat in ≤24 hours) or propafenone can be used to terminate out-of-hospital paroxysmal AF of short duration after these drugs have been shown to be efficacious and safe in a monitored setting. In this approach, we have the patient take a diltiazem or a beta blocker 30 minutes or more (if they are not on chronic AV nodal blocker) before the oral antiarrhythmic drug to prevent a rapid ventricular rate should conversion to atrial flutter occur. Some of our contributors prefer the patient to be observed in the emergency department (or potentially during inpatient hospitalization) the first time the "pill-in-the-pocket" approach is taken so that monitoring of safety and efficacy occurs.

Some authors also instruct their patients to begin a direct-acting oral anticoagulants (DOAC; also referred to as non-vitamin K oral anticoagulants [NOAC]) at the time of the beta blocker/diltiazem so as to decrease the risk of thrombus formation should the patient not convert to SR in the subsequent 48 hours and to prophylaxis for thrombus formation during the transient periconversion period of atrial appendage mechanical dysfunction.

MAINTENANCE ANTIARRHYTHMIC DRUG THERAPY — After successful electrical cardioversion, antiarrhythmic drugs increase the likelihood of long-term maintenance of SR [1]. The use of antiarrhythmic drugs to maintain SR is discussed elsewhere. (See "Antiarrhythmic drugs to maintain sinus rhythm in patients with atrial fibrillation: Recommendations".)

We do not generally recommend maintenance antiarrhythmic drugs after electrical cardioversion in patients with their first episode of non-valvular AF, particularly those at low risk for recurrence (eg, short-duration AF, normal or only mildly increased left atrial size, normal left ventricular systolic function, absence of valvular dysfunction) or those with a transient cause (eg, pericarditis, pulmonary embolism, and corrected or treated hyperthyroidism) [73]. (See 'Recurrence of AF after cardioversion' above.)

RECOMMENDATIONS OF OTHERS — Societal guidelines are available [53,74-76].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Atrial fibrillation" and "Society guideline links: Arrhythmias in adults".)

SUMMARY AND RECOMMENDATIONS

●Unstable patients – For the uncommon patient who is hemodynamically unstable or has angina due to AF and is at low risk for thromboembolism, we recommend urgent electrical cardioversion rather than no cardioversion (Grade 1C). Precardioversion intravenous heparin bolus may be beneficial for prevention of thromboembolism due to postcardioversion atrial appendage dysfunction, but its efficacy has not been studied. (See 'Indications' above.)

●Stable patients – Most patients with AF do not need emergency or even urgent cardioversion, as rate slowing will often improve symptoms. We also attempt to defer cardioversion to allow for initiation of anticoagulation.

For select, stable patients with nonvalvular AF, the restoration of sinus rhythm (SR) with either electrical or pharmacologic cardioversion is necessary or reasonable (see 'Ventricular rate control' above and 'Indications' above):

•Symptomatic, first episode of AF – For most symptomatic patients with new onset/first episode of AF, we suggest an attempt at cardioversion, as opposed to no attempt (Grade 2C).

In patients who do have factors that predict a high likelihood of success of electrical cardioversion, and in whom long-term maintenance antiarrhythmic drug therapy is not planned, we suggest electrical instead of pharmacologic cardioversion (Grade 2B). (See 'Reasons not to perform cardioversion' above and 'Electrical versus pharmacologic cardioversion' above.)

In such patients who are older or have multiple medical comorbidities, it is reasonable to avoid cardioversion if the symptoms can be minimized and pharmacologic ventricular rate control is achieved; the approach needs to be individualized.

•Patients who fail long-term rate control – For patients in whom a long-term rate control strategy has failed due to persistent symptoms, a rhythm control strategy involving cardioversion is reasonable.

•Infrequent episodes of AF – Periodic electrical cardioversion is an option for patients with infrequent episodes of AF that do not spontaneously convert, including those being managed with a rhythm control strategy.

●Roles of rate control and anticoagulation – Most patients in whom cardioversion is chosen will need the ventricular rate controlled and the need for anticoagulation assessed prior to cardioversion. (See 'Ventricular rate control' above and 'Anticoagulation' above.)

●Antiarrhythmic drugs – These may be initiated prior to electrical cardioversion to increase the likelihood of a successful electrical cardioversion, to increase the likelihood of maintaining SR in the hours after cardioversion, or as the first step in a plan for long-term antiarrhythmic therapy. (See 'Preprocedural antiarrhythmic drugs' above.)

For patients with AF and no structural heart disease (including no evidence of coronary artery disease) for whom pharmacologic cardioversion is chosen, we suggest either flecainide or propafenone rather than other antiarrhythmic drugs (Grade 2B). The choice between these two drugs should be influenced by the practitioner’s experience and the clinical situation. (See 'Pharmacologic cardioversion' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Leonard Ganz, MD, FHRS, FACC, who contributed to an earlier version of this topic review.