Improving Emergency Medicine Patient Care
For Questions Or To Order, Call
Mon-Fri 8am-5pm EST
View Full Site
Home Browse Articles CME Tests Login
Subscribe Visit Store View Cart

Home > Browse Emergency Issues

<< Atrial Fibrillation In The ED: Cardioversion, Rate Control, Anticoagulation, And More

Rate Control

Once the patient with rapid AF has been stabilized, the next goal is ventricular rate control. It is the ventricular rate that mostly governs symptoms. Rate control will likely improve the patient’s overall hemodynamic status, resulting in better ventricular filling, less myocardial oxygen demand, and improved cardiac output. Many options for rate control are available, and emergency physicians must understand the indications, contraindications, and nuances of each of these interventions. Sometimes patients with chronic AF will present acutely with a rapid ventricular response due to a physiologic stress, such as fever, dehydration, pain, or anemia. Treatment in such cases would be acetaminophen, fluids, opioid analgesics, and blood transfusion, respectively—not cardioactive drugs.

The most common choices for rate control agents can be found in Table 5. Calcium-channel blockers, ß-blockers, and digoxin are not conversion agents and are not intended for such purposes. While 25%-60% of patients with NOAF may spontaneously convert while on these agents, patients who receive no intervention or placebo demonstrate similar conversion rates.70-73 Amiodarone possesses antiarrhythmic properties in addition to its rate control effects. When choosing a rate control agent, strongly consider whether or not the patient could have an accessory pathway; the use of an AV nodal blocking agent would be contraindicated in that situation. (See the “Special Circumstances” section later in this article.)


DCC is the quickest way to obtain rate control, even in a stable patient. However, it requires procedural sedation, with all of its accompanying risks. Acute DCC is also associated with some potential complications, particularly embolic events and cardiac arrhythmias, which should be minimized prior to “elective” cardioversion of a stable patient (see below). However, in an unstable patient, DCC may save a life. It has been shown to be safe and efficacious in the hands of emergency physicians when managing stable23 as well as unstable patients with AF.

Calcium-Channel Blockers

Calcium-channel blockers (CCBs) are the first-line treatment for rate control in rapid AF.74 They are efficacious, act quickly, and are relatively safe. The two most commonly cited agents used in the ED are diltiazem and verapamil; much controversy exists in the literature as to which agent is better.75-79 They both act on the sinoatrial and AV nodes but do not prolong the refractoriness of accessory pathways. Both agents reduce systemic vascular resistance, and the most frequent complication is hypotension.80 The risk of hypotension can be minimized by the use of IV calcium as a pretreatment prior to giving these drugs, or as a rescue agent if hypotension occurs.81-83 CCBs are effective in controlling ventricular response during exercise in chronic AF.84

Pretreatment With Calcium

If a patient develops hypotension in the setting of receiving a CCB, it can often be reversed or attenuated with IV calcium. Life-threatening tetany has also been reported after IV diltiazem that responded to IV calcium chloride.85 Emergency physicians treating AF/AFl patients with CCBs should have quick access to IV calcium.

Alternatively, calcium as a pretreatment should be considered in patients with marginal blood pressure prior to receiving a CCB.81-83 Calcium has been shown to be efficacious in preventing the negative inotropic effects of the CCBs but does not affect the drugs’ chronotropic effects. Side effects of calcium pretreatment are rare and mild, including flushing and a feeling of “generalized heat” lasting 2-3 minutes.81 While the optimal dose for pretreatment is not known, it seems reasonable to use 1-3 cc of calcium chloride or 5-10 cc of calcium gluconate.

Verapamil vs. Diltiazem

Comparison trials of CCBs (mainly verapamil and diltiazem) with placebo in rapid AF show that the CCBs are more than 90% successful in reducing heart rate.80,84,86-89 This holds true for patients with AFl as well.80,87 CCBs quickly control ventricular response—often in less than five minutes—when given intravenously.87,90 CCBs are more effective for acute rate control than digoxin.90-92 In a comparison study, Schreck et al showed a statistically significant reduction in heart rate for diltiazem-treated patients at five minutes vs. 180 minutes in digoxin-treated patients.90 The combination of IV diltiazem with IV digoxin may result in more efficacious ventricular rate control with fewer fluctuations than that of IV diltiazem alone;93 however, some feel the addition of digoxin is unnecessary. In critical care patients, diltiazem slows the heart rate more profoundly than IV amiodarone, although it produces significantly more hypotension.94

So which agent should we use—diltiazem or verapamil? Advocates of diltiazem state that it has less risk of hypotension and that patients can be maintained on a drip after the initial bolus. Verapamil users state that neither agent has been shown to be definitively better and that verapamil is cheaper. There is some truth to both sides of this debate. Most of the debate regarding which agent produces less hypotension stems from animal data and studies on diseased human heart tissue performed by Bohm et al.75,77 They suggested that verapamil caused greater reduction in contractility than diltiazem, but these findings have been challenged.76,80,84,88,90

The only “head-to-head” human trial did not settle the debate.78 This randomized, double-blind, crossover trial compared the effects of IV diltiazem and verapamil in 17 adults with AF or AFl and a baseline heart rate of 120 bpm or higher. No pretreatment with calcium was used in the trial. Both agents showed similar efficacy for heart rate control at two minutes; however, three of the 17 patients initially entered into the study developed symptomatichypotension after verapamil, which prompted withdrawal from the study. Hypotension was considered life-threatening in one patient who was found to have an ejection fraction of 20%. This study has led many to declare that clearly diltiazem is better than verapamil because of the hypotension. However, the trial was small, and the dosing protocol of the study still leaves this question open to debate. In the patient who has a borderline blood pressure, it may be prudent to use diltiazem. However, the choice of drug may be less important than how much and how quickly it is given. Titrate CCBs over several minutes.

Another reason that diltiazem is commonly quoted as being better than verapamil is that diltiazem can be followed by a drip. A diltiazem bolus followed by a drip did show better efficacy compared with a diltiazem bolus followed by placebo over 24 hours, which is not surprising given the half-life of the drug. However, a verapamil drip has been described and is effective.95 (However, since not many people are familiar with it, you might get some blank stares if you order one.) Some evidence suggests that patients may be quickly switched to oral forms of the CCBs to maintain adequate rate control, possibly avoiding the issue of a drip altogether.96 While verapamil is cheaper than diltiazem by about $10, this difference is unlikely to start a run on the bank.


Kuhn and Schriger published a retrospective review of patients who received verapamil despite a contraindication, and they reported a higher incidence of adverse outcomes and drug failures in that group compared with those who did not have a contraindication.97 CCBs are contraindicated in patients with an accessory pathway, because these patients may develop a rapid ventricular response and induction of ventricular fibrillation secondary to reflex sympathetic stimulation of the accessory pathway.31,33 Hypotension is the most common significant side effect of CCBs.81-83 (See Table 6.)

Special mention of the use of CCBs in patients with signs of CHF is worthwhile. CCB use in patients with CHF is controversial according to some authors, but the literature suggests overall safety and efficacy.84,98-100 Rapid AF with CHF is not necessarily a contraindication to CCB use; in fact, it may be part of the appropriate treatment, particularly in relatively stable patients, and perhaps in unstable patients in whom cardioversion fails. (Remember that the moribund, severely dyspneic patient with acute CHF and rapid AF should be considered unstable and likely have attempts at cardioversion prior to medical therapy.)

In the setting of rapid AF with signs of CHF, the use of CCBs may improve cardiac function by slowing ventricular rate, decreasing oxygen demand and increasing diastolic filling time. Several small studies in patients with Class 3 or 4 heart failure have established CCB relative safety in this setting.84,99,100 In a study by Goldenberg et al of 37 patients with rapid AF and moderate-to-severe CHF, 97% responded well to one or two doses of diltiazem followed by a drip. Hypotension was the most common adverse event (11%), but in all cases, patients responded with nonpharmacologic treatment. No patient was reported to have an exacerbation of the CHF due to the diltiazem bolus or the drip.84 Heywood et al showed similar results with eight out of nine patients with clinical signs of CHF and rapid AF responding to diltiazem without clinical and objective signs of worsening. 100 However, in the trial by Phillips et al, the three patients who received verapamil and became hypotensive had moderate-to-severe LV dysfunction, leading the authors to recommend that IV verapamil be administered with caution to control the ventricular response in patients with AF or AFl and concomitant LV dysfunction.78


The recommended manufacturer dosing for diltiazem is 0.25 mg/kg IV over two minutes followed by a second bolus of 0.35 mg/kg IV over two minutes if there is inadequate response at 15 minutes.79,80 Due to the occurrence of hypotension, some authors have recommended starting at even lower doses, perhaps 0.15 mg/kg IV,79 although published data with this dose are limited. Others have recommended giving 5 mg increments IV slowly up to a maximum of 50 mg. The initial dose(s) can be followed by a drip of 5-15 mg/h for maintenance therapy.80 Verapamil dosing should be 2.5-5.0 mg over 2-3 minutes followed by 5- 10 mg in 15-30 minutes if necessary. As discussed previously, it can be followed by a drip as well, typically starting at 5 mg/h or up to 0.005 mg/kg/min.79 With either drug, start with lower dosages, particularly in patients with underlying CHF or borderline blood pressure, and perhaps the elderly.


A number of different ß-blockers are available (see Table 5 ) that are highly effective in controlling the rapid ventricular rate associated with NOAF. ß-blockers work by decreasing the conduction through the AV node. They act rapidly after IV administration, and adequate rate control can often be established within two minutes of drug administration.101-103 There is likely some benefit to using these agents as the first line in clinical scenarios associated with high adrenergic tone, such as thyrotoxicosis, acute hypertensive crisis, and MI. Particularly in MI patients, ß-blockers decrease both short-term and long-term morbidity and mortality.

Contraindications to ß-blockers are very similar to CCBs, with the same risks in accessory pathways, VT, and high-degree blocks. In addition, ß-blockers are also contraindicated in patients with active bronchospasm or severe asthma, although the ß1-selective agents may have fewer problems associated with their use. ß-blockers do have negative inotropic effects and thus must be used with caution in patients with LV dysfunction. The most frequently reported significant side effect is hypotension.101-103

Esmolol, a ß1-selective agent, has a rapid onset and an ultra-short half-life (7-9 minutes), which has some benefits in patients with rapid AF. Plattia et al found IV esmolol to have similar efficacy to verapamil for rate control in AF.101 Hypotension was reported in both groups, but the authors preferred esmolol because of its short half-life and noted that the infusion could be stopped if the patient suffered side effects. Treatment with esmolol is somewhat cumbersome, however, as it requires repeated bolus dosing and multiple infusion steps. It is also expensive compared with other agents. Alternatives to esmolol include IV propranolol (a nonselective ß-blocker) and metoprolol (a ß1-selective agent). Both of these agents have the disadvantage of longer half-lives, but the advantages of much lower cost and ease of administration. Propranolol is a non-cardioselective ßblocker and must be used cautiously in patients with a history of bronchospasm.


Digoxin has been used to treat AF for more than two centuries. Until recent years, digoxin was almost mandatory in a patient with NOAF. It has fallen out of favor due to its lack of efficacy compared to other agents, its hazardous side-effect profile, and the arrival of newer agents. It is now considered second (or third) line for rate control.104 Although digoxin has some direct influence on the atria and AV node, it affects the autonomic nervous system as well. In AF, the drug exerts its main electrophysiologic effect indirectly, by modulating vagal tone. In the resting patient on digoxin, vagal influences on the AV node and atrium are enhanced, and, as a result, the ventricular rate is slowed.

As digoxin’s predominant rate control effects are due to increased vagal tone, it loses its effectiveness as a rate control agent when a patient is active or if there is increased autonomic sympathetic stimulation. This loss of vagal stimulation in the setting of increased sympathetic tone would also make digoxin a poor choice in patients with rapid AF associated with fever, thyrotoxicosis, hypoxia, or acute blood loss. Some authors suggest that the physician should ask the question, “If this patient were in sinus rhythm, would he or she have a sinus tachycardia?”104 If the answer is yes, then digoxin alone is unlikely to control the ventricular rate, and other agents should be considered. Digoxin does have some modest positive inotropic effects likely due to effects on calcium, and it might still have a role in the management of AF patients with CHF or low cardiac output. One study showed digoxin’s use with esmolol may have helped prevent hypotension, perhaps due to its positive inotropic effects.102 Another study showed some possible benefit using a combination of diltiazem and digoxin over digoxin alone in 52 patients with rapid AF.93 Some of the other limitations of digoxin in rapid AF are its slow onset time and some potential toxicity, particularly with long-term use. It takes 30-180 minutes to slow the ventricular rate significantly,70,71,105 compared with CCBs and ß-blockers, which often work within minutes.90,101 Digoxin blocks the AV node but not accessory pathways and thus is contraindicated in patients with WPW. It does not convert AF, and there is no evidence that it has primary antiarrhythmic actions. This has been shown consistently in three randomized clinical trials in which digoxin, in the absence of heart failure, was no better at conversion than placebo. 70,71,105 Furthermore, based on animal data, digoxin may increase the chance of a recurrence of AF after restoration of sinus rhythm by attenuating the recovery from electrical remodeling of the atria.106 A small upside of digoxin is that it costs about as much as water (unless you buy the ultrapure, spring water in designer bottles). In summary, digoxin is a drug that will likely continue to fall in and out of favor. However, except maybe in rare circumstances (e.g., CHF), it is a drug that has been replaced with better, safer, and more efficacious agents.


Although amiodarone has been used to control AF, it is not approved in North America for any supraventricular arrhythmia. Despite this, IV amiodarone has been evaluated in the treatment of AF for its rate control properties and antiarrhythmic effects. Amiodarone is classified as a Class III antiarrhythmic agent but also does have effects similar to Class I, II, and IV agents. Most studies with amiodarone show that it slows ventricular rate in rapid AF typically with significant rate control effects within 20 minutes of a bolus administration.72,94,107,108 Conversion rates are highly variable (4%-100%) depending on study characteristics. Amiodarone is slower in onset and less effective than diltiazem for rate control but may have less associated hypotension in critically ill patients.94 Several dosing regimens of amiodarone have been used, and no large trial comparing the efficacy and safety of the different dosages in rapid AF has been performed. If you choose to use amiodarone for rapid AF, give an initial loading bolus followed by an infusion.94

Despite causing a prolongation of the QT, amiodarone appears to have a relatively low frequency of pro-arrhythmic effects,109 and the incidence of torsades de pointes is less than 1% according to two evidence-based reviews.107,110 Amiodarone in the acute setting can result in hypotension, which is usually transient and well-tolerated. Hypotension generally responds to IV fluids and stopping the drug.72,110-112
Occasional, seemingly rare episodes of LV failure have been reported,108,113 as have exacerbations of bradyarrhythmias.114 However, amiodarone appears to have few or no negative inotropic effects and appears relatively safe in patients with CHF and LV dysfunction.107 Superficial thrombophlebitis can occur when the drug is administered through a peripheral vein.108,111,113 Because it works on accessory pathways as well as on AV nodal conduction, amiodarone may be useful in WPW and other pre-excitation syndromes. However, case reports of VF have been reported with its IV administration in WPW patients,115 and the drug’s peculiar combination of different electrophysiologic actions might cause differing effects when given intravenously vs. orally. Disadvantages include cost, variable effectiveness, and side effects that can be significant if given long term. The drug is generally welltolerated in the short term.

The current literature does not support the routine use of amiodarone in AF in the ED for treatment of most cases of AF/AFl. An emergency physician might consider its use in the patient with an irregular wide complex tachycardia.


Magnesium, which some advocates declare a “wonder drug,” is useful in the management of ventricular
dysrhythmias, and it has been studied in atrial arrhythmias as well. Magnesium likely exerts its effects in AF by decreasing conduction within the AV node. Several small studies have suggested its effectiveness as a rate control agent in rapid AF;116-119 however, its use is most often recommended as adjunctive therapy, commonly with digoxin. Both Brodsky et al120 and Hays et al116 have reported synergistic effects of digoxin with magnesium. Usually magnesium is given as a bolus of 2 g over 10-20 minutes, sometimes followed by an infusion (studies use different regimens). Rate control is relatively rapid, reported to be within five minutes.116 It is well-tolerated, with mild flushing, warmth, and tingling accounting for the majority of side effects, which are likely related to the rate of the infusion.116 Occasionally, with rapid infusion or large doses (8-10 g), respiratory muscle fatigue, hypotension, and cardiac pauses can be seen.120 The drug can accumulate rapidly in patients with renal failure and should generally be avoided in these patients. In addition to its good safety profile, magnesium has the benefits of being very inexpensive, readily available, and easily administered.

Studies with magnesium show that it controls rate when given alone116 and can convert 50%-60% of patients with recent-onset atrial tachyarrhythmias.117,119,120 A study comparing magnesium with amiodarone in critical care patients with atrial tachyarrhythmias (most of which were AF) found that magnesium controlled ventricular rate just as well as amiodarone.118 In addition, at 24 hours, the magnesium group had significantly more patients convert to sinus rhythm compared with the amiodarone group. Similarly, in a study of 46 patients who presented with a prolonged episode of paroxysmal AF, more patients (57%) converted with magnesium at six hours compared to diltiazem (22%).119 Although largely unproven, magnesium therapy might be particularly useful in alcoholics, who typically suffer from total body magnesium depletion.121 In summary, magnesium is efficacious as a rate control agent in rapid AF, and it might have some antiarrhythmic effects compared to standard agents.